proceedings of the lake malawi fisheries management symposium

TECHNICAL COOPERATION REPUBLIC OF MALAWI 

FEDERAL REPUBLIC OF GERMANY 

PROCEEDINGS OF THE 

LAKE MALAWI FISHERIES 

MANAGEMENT SYMPOSIUM 

4 TH – 9 TH JUNE 2001 CAPITAL HOTEL, 

Department of Fisheries 

LILONGWE 

E DITED BY 

OLAF L.F. WEYL & MICHELLE V. WEYL 

December 2001 

NATIONAL AQUATIC RESOURCE 

MANAGEMENT PROGRAMME 

(NARMAP) 

Deutsche Gesellschaft für 

Technische Zusammenarbeit

Table of contents 

Speech by the Minister of Natural Resources and Environmental Affairs, Hon, Harry I. Thomson, MP, at the 

opening of the Lake Malawi Fisheries Management Symposium…………………….………………………... 1 

Speech by the Director of Fisheries, S.A. Mapila, at the opening of the Lake Malawi Fisheries Management 

Symposium…………………….…………………….…………………….……………………. 3 

The Role of the National Research Council of Malawi in protection , conservation and management of 

fisheries resources in Malawi 

Frade K. Nyondo…………………….…………………….…………………….………… 

A general overview of fisheries research and development in Malawi 

O. M. Kachinjika…………………….…………………….………………………………. 11 

Economic security and sustainable programmes for the African Great Lakes 

Benjamin Ngatunga & Anthony Ribbink…………………….…………………….……………………………… 20 

The need to maintain maximum biodiversity in Lake Nyasa 

Benjamin P. Ngatunga………..……………….…………………………….……………… 27 

Fisheries management, biodiversity conservation and genetic stock structure 

George F. Turner…………………….…………………….…………………….………… 31 

Fisheries development, management, and the role of government 

Tony Seymour……………….…………………….…………………….………………… 42 

Seeking sustainability: strengthening stakeholder involvement in fisheries management in Malawi 

Tracy A. Dobson & Aaron J. M. Russell…………………….…………………….………… 55 

Fisheries management and uncertainty: the causes and consequences of variability in inland fisheries in 

Africa, with special reference to Malawi 

Edward H. Allison, Frank Ellis, Peter M. Mvula & Laurence F. Mathieu………………………….. 66 

An overview of indigenous knowledge as applied to natural resources management 

E.Y. Sambo & R. Woytek…………………….…………………….…………………….… 80 

Decentralised environmental management and the implications for fisheries co-management in Lake Malawi 

John D. Balarin…………………….…………………….………………………..…………………….…………… 85 

Status of the small scale fishery in Malawi. 

Mackson J.R. Ngochera …………………….…………………….……………..……….…………………….…… 95 

Effects of overfishing on reproductive potential of major cichlid fish species in southern Lake Malombe 

(Malawi): Need for “Closed Area” strategy as a complementary management option? 

Collins Jambo & Tom Hecht…………………….…………………….……………………. 105 

Management recommendations for the nkacha net fishery of Lake Malombe 

Kissa R. Mwakiyongo & Olaf L.F. Weyl …………………….……..………………….……… 114 

Drifting long line, a potential fishing method for the northern part of Lake Nyasa/Malawi/Niassa 

K.J. Kihedu, M.K.L.Mlay, J.A. Mwambungu & B.P. Ngatunga……………….………………… 121 

6

Gear and species selectivity of the chilimira kauni fishery in Lake Malawi 

Jacqueline Chisambo………….…………………….…………………….……………… 127 

Gear and species selectivity of the gill net fishery in Lake Malawi. 

Richard Dawson Sipawe…………………….…………………….……………………. 133 

Fisheries activities in northern Lake Nyasa (Kyela District) 

John A. Mwambungu…………….…………………….…………………….…………… 142 

Hard choices for chambo management in area A of the southeast arm of Lake Malawi 

Olaf L.F. Weyl…………………….…………………….…………………….…………… 146 

The nkacha fishery in Lake Malombe and the southeast arm of Lake Malawi 

Moffat M. Manase…………………….…………………….………….………………… 156 

The state of the large scale commercial fisheries on Lake Malawi 

Moses Banda………………….……………………………………….………………... 163 

Spatial and temporal distribution of some commercially important fish species in the southeast arm of Lake 

Malawi: A Geostatistical Analysis. 

Geoffrey Z. Kanyerere & Anthony J. Booth…………………….…………………….………... 173 

Preliminary investigations of community level responses to benthic trawling in the demersal fish fauna of 

Lake Malawi/Niassa, Africa. 

Will Darwall…………………….…………………….…………………….…………… 189 

Resource use overlaps between the pair trawl and small scale fisheries in the southeast arm of Lake Malawi 

Thomas E. Nyasulu…………………….…………………….…………………….…………………….…………... 195 

Population biology of the catfish Bagrus meridionalis from the southern part of Lake Malawi. 

Moses Banda…………………………………….…………………….………………… 200 

Feeding habit and development of feeding-related morphological characters in Oreochromis shiranus 

(Boulenger, 1896) larvae and juveniles in Malawi 

Shinsuke Morioka and Daniel Sikawa…………………….…………………….…………………….…………… 215 

Otolith growth increments in three cyprinid species in Lake Malawi and information of their early growth 

Shinsuke Morioka & Emanuel Kaunda……………………………………………………………………………. 220 

Development of African catfish Clarias gariepinus larvae during the transitional phase between endogenous 

and exogenous energy intake. 

Seiji Matsumoto, Shinsuke Morioka & Sigeru Kumagai………………………………………………………… 227 

Effect of temperature on oocyte development of Oreochromis karongae (TREWAVAS) 

L.J. Kamanga, E. Kaunda, J.P. Mtimuni , A.O. Maluwa & M.W. Mfitilodz…………………………………. 233 

Effect of feeding density on survival of African catfish Clarias gariepinus. 

H.K. Zidana & A.O. Maluwa……………………………………………………………………………………….. 242 

Determination of biological reference points for Lake Malawi cichlids 

Anthony J. Booth…………………………………………………………………………………………………… 249 

Appendix I – Abstracts of papers not submitted for publication (with comments from floor)…………………... 260 

Appendix II – List of participants & Acknowledgements………………………………………………………... 270

1 

Lake Malawi Fisheries Management Symposium - Proceedings 

Speech by the Minister of Natural Resources and Environmental Affairs, Hon, 

Harry I. Thomson, MP, at the opening of the opening of the Lake Malawi Fisheries 

Management Symposium 

Read by the Deputy Secretary of the Ministry of Natural Resources and Environmental Affairs 

Mr. E.E. Lodzeni. 

The Director of Administration in the Ministry, The Director of Fisheries and all other fisheries staff, 

Representatives of Donor Agencies, Delegates to the Symposium, Representatives of Government Ministries and 

Departments, Members of the Press, Ladies and Gentlemen. 

It is a great honour and privilege for me to be with you today to officially open this International Symposium on the 

Management of Lake Malawi and Lake Malombe fisheries resources. This is a very important symposium for 

Malawi, since it derives a variety of social and economic benefits from the resources these lakes provide. 

First and foremost, I wish to extend a warm welcome to all participants to this symposium, in particular our 

international guests, who, despite their commitments at home have felt if very important that they come and 

contribute towards the rational management of the fisheries in Malawi. To them I would like to say ‘feel at home’. 

I have noted that the programme does not provide for a field trip. However, I would like to encourage participants 

who will have spare time after the conference, to take a tour of the countryside, especially along Lake Malawi, 

where in addition to having a field observation of the issues to be discussed here, will also be able to admire the 

scenic beauty of some spots in the country, and have a feel of ‘Malawi the Warm Heart of Africa’. 

Ladies and Gentleman, I have been informed that this has been organised with a view to amalgamate all research 

results of previous work done on lake Malawi and Lake Malombe for purpose of formulating management 

recommendations that would ensure sustainability of the fisheries resources. My reaction to this, is that this 

gathering has come at an opportune time when the Malawi Government is looking for ways and means that would 

provide for the sustainable management of its natural resources including fish stocks. Recent trends in the status of 

our fisheries show that there is need for more management guidelines to be put in place, if the fish stocks are to be 

sustained. Hence Government, through the Department of Fisheries, has recommended a number of strategies 

aimed at nurturing and promoting efficient management of the fish resources for the benefit of the present and future 

generations. 

Ladies and gentlemen, these responsibilities should not be taken lightly, as fish is a vital source of animal protein in 

Malawi, contributing about 60-70% of the nation’s animal protein supply. In addition, the fisheries sector employs 

directly well over 50 000 full-time artisanal fishers and about 1 000 commercial fishers. Indirectly, it offers 

employment to 250 000 individuals working in fisheries related activities such as processing, marketing and boat 

building. Therefore, the social and economic significance of the fisheries sector in Malawi is well recognised by 

Government. 

Furthermore, Malawi possesses rich aquatic resources notable of which are the diverse fish species that abound in 

the various water bodies, especially Lake Malawi. Malawi’s fish biodiversity is unparalleled anywhere else in the 

World. It is documented that lake Malawi harbours more that 700 species of fish, which is more than any other lake 

in the world. However, less than half are scientifically described. 

It is in recognition of this that the Fisheries Research Unit (FRU) was set up and has been carrying out research 

studies on Lake Malawi and other lakes and rivers. These are aimed at generating management guidelines for the 

management and sustainable exploitation of our fisheries resources. Previous activities have centred on: 

1. Study of biological characteristics of fish with the aim of determining growth parameters, breeding behaviour, 

spawning behaviour, mortalities, feeding habits and nursery grounds; 

2. Stock assessment involving exploratory survey in shallow and deep waters with a view to identifying underutilised 

fish stocks, and monitoring of the status of fish stocks under exploitation especially in the southern part 

of Lake Malawi. 

3. Studies on productivity of water and its influence on fish production; and

2 


4. Assessment of pollutants and habitat degrading activities, and their effects on aquatic ecosystem productivity; 

and bio-diversity surveys. 

The results generated from the various research activities have assisted the department in coming up with fisheries 

resource management plans, which have contributed to the formulation of various fisheries regulations, such as 

closed seasons, gear restrictions and controlled allocation of fishing licences to the commercial trawl fishery 

Similar efforts have also been made by other organisations such as Chancellor College of the University of Malawi, 

which has recently concluded a project on the ecology of some Lake Malawi fishes. 

However, despite all the past efforts undertaken to control and manage the fish stocks, the fisheries sector is facing 

serious challenges of maintaining a balance between exploitation and conservation despite a number of research 

projects undertaken in the country. 

This is partly due to the fact that the majority of such projects were undertaken at different times without 

considering long-term sustainability. It has, therefore, been difficult for the Department of Fisheries to derive 

maximum benefit from the various recommendations made by the projects. The holding of this symposium is 

therefore, expected to rectify this anomaly by ensuring that all previous studies are presented, discussed and 

integrated into the various management guidelines that have been produced for the Lake Malawi and Malombe 

fisheries. 

Ladies and Gentlemen, I am further informed that in addition to providing an opportunity to amalgamate recent 

scientific information on lake Malawi/Malombe fisheries complex, there will also be discussions with various 

stakeholders in order to formulate management plans/recommendations. The holding of discussions with 

stakeholders is in line with the government’s policy of involving end users in the management of our natural 

resources such as fish. I therefore, wish to request all of you to actively participate in the discussion in an open and 

transparent manner. 

Distinguished Guests, Ladies and Gentlemen, the fact that the presentations to be made at this symposium include 

biodiversity, fish biology and ecology, fisheries assessment and management, social dimensions and economics of 

fisheries management and fisheries co-management assures the Malawi Government that the symposium is of great 

benefit to the nation at large. Such a diversity of issues to be discussed in this symposium will definitely contribute 

towards the determination of management objectives and strategies for lake Malawi/ Malombe fisheries complex. 

Lastly, I would like to urge this forum to look at a positive way forward for the Malawi fisheries by adopting a 

holistic and multi-disciplinary research approach that would always be relevant in the formulation of management 

guidelines and evolve with environment, equity and economics. Cognisance should also be taken of the fact that for 

fisheries research to be an effective tool for management it should have a reliable source of funding. Therefore, it is 

also my hope that the forum will identify alternative sources of funding for the research programmes being 

proposed. 

At this juncture, I would like to thank GTZ for sponsoring this Symposium. My special thanks also go to the 

organisers and management of Capital Hotel for the excellent arrangements. 

Distinguished Delegates, Ladies and Gentlemen, now, it is my honour to declare the International Symposium on 

Sustainable Management of the Fisheries in Lake Malawi/Malombe officially open. 

I thank you for your attention.

3 


Speech by the Director of Fisheries, S.A. Mapila, at the opening of the Lake 

Malawi Fisheries Management Symposium 

The African Great Lakes are considered dynamically fragile ecosystems that are relatively resistant to changes with 

which they have co-evolved for a million years. Lake Malawi is one of the largest fresh water lakes in the world of 

its kind in the SADC region. The lake is rich in biodiversity, of which, fisheries are a major resource for the riparian 

communities. During the last two decades intensive non-selective fisheries in shallow waters, extreme changes in the 

drainage basin and lake vegetation industrialisation, and agricultural developments, have impacted upon the lake. 

There is a need for sound fisheries management plan to be developed and implemented, through participatory 

approaches, a need for sustainable utilisation of the fisheries resource through proper resource monitoring, extension 

and enforcement; and to increase fish production through the availability of a credit scheme to the private sector, the 

introduction of new technologies and the building of infrastructure. 

The environment is the foundation of economic activity for the majority of Malawians. The unsustainable 

exploitation, and underdeveloped potentials, of some of the natural resources in the Lake Malawi basin have 

continued unabated and are challenged by a complex interaction of several factors including population growth, 

widespread poverty, high dependency on smallholder rain fed agriculture, inconsistencies in macro economic 

policies and institutional weaknesses with national capacity. The impacts of competing and unsustainable uses of 

land and water resources are partially understood, and widely manifested throughout the Lake Malawi basin. The 

annual costs of land related environmental degradation in Malawi have been estimated at US$ 243m or 12% of 

GDP. 

Though guidelines are and will continue to be the backbone of Malawi’s economy, since the country is not well 

endowed with mineral resources, the fishing industry is set to take centre stage due to major shifts in the agricultural 

industry. This will thus involve boosting production from inland water resources, and a commitment on the part of 

the authorities. 

Fishing is one of man’s oldest activities, and it remains one of the most important, especially in Malawi. Fish is a 

major source of high-grade protein, and fishing is the main employment in communities along the coast and around 

small lakes and rivers. It is obvious that the use of these natural resources is bound to raise a range of environmental 

issues. Chief amongst them is the conservation and management of the resources themselves, but equally important 

is the problem of allocating resources amongst competing individuals and social groups. This is important because 

without proper allocation there is usually an unrestricted activity, which can undermine management and 

conservation standards. The nature of the various environmental considerations will have to depend on the nature of 

the biological resources, the fishing operations, and their social, economic and administrative contexts. 

Most of the fish resources in Malawi are shared with neighbouring countries. Regional collaboration on these shared 

water resources is going to be essential if biodiversity and production from these water resources is going to be 

sustainable. Within this context, smallholder fishermen form some of the largest constituency of the roughly 100 

million inhabitants of southern Africa. For these fisherman, there is a direct and immediate cause and effect 

relationship between supply of fish and food availability. 

The results are widely demonstrated by the amount of fish consumption and trade in the region. The supply of fish 

suffers seriously from the cyclic fluctuations in catch mainly due to climatic conditions, although 40% of the 

population and labour force relies on fisheries for subsistence as well as employment and income. 

Food security cannot be divorced from poverty. Lack of sustainable income generation, vulnerability to food 

variability and unequal accesses to resources are issues that trap the poverty prone in food insecure situations. 

Increased fish production and availability can significantly affect the national nutrition. 

Statistics indicate that aquatic production in terms of biomass on Lake Malawi has gone up, but fishermen’s catches 

have remained static or even gone down over the last 10 years due to a lack of innovative technology. The fishing 

industry has gone back rather than forwards due to amongst other things, structural measures. While a lot of 

information is known about the fisheries potential of Lake Malawi, a lot still needs to be known about their 

limnology, biology, physio-chemistry, hydrology and socio-economic setting. It is however known that the lake 

supports large biomass of fish in offshore waters, which is not being exploited.

4 


Early aquatic resources development projects on Lake Malawi did not meet with the expected results. During the 

early years after independence, our government invested heavily on infrastructure using foreign assistance. 

However, these facilities were poorly maintained and functioned well below capacity. Some of the reasons were bad 

site selection; lack of government capacity to maintain operations; lack of planning and unclear objectives. Early 

projects were also oriented towards donor driven research projects and transfer of untested technology. Sustainable 

support systems, extension services and credit facilities were not established. 

Following the disappointing results of this donor driven period, the Government of Malawi would like to change its 

project priorities to a production-oriented approach. Priority will be given to the private sector targeting the 

commercial and small-scale fisherman. The overall approach would be to promote fish production for the potential 

benefits of the population. 

Pilot projects undertaken during previous projects have identified an effective approach to aquatic resources 

development on Lake Malawi. Considerable information on aquatic resources use and user’s priorities exist. 

Research for the future should focus on meeting food production needs in Malawi in ways that are beneficial to the 

populace and do not degrade the natural resource base. This will involve identifying the most appropriate 

technological and institutional changes and policies for sustainable and equitable fisheries production. Research 

attempts to identify appropriate policies for stressed fisheries should be given special emphasis. Nevertheless, most 

importantly, the research should also emphasise on extension and on property rights and collective action. 

It is important to note that the most important issue in capture fisheries is the need to match the catch and marketing 

capacity of the fishery to the productivity of the resource. But in most donor driven projects this issue has been 

ignored, and the reason why it has been ignored is that usually the people drawing up national policies, either within 

government or aid agencies, are seldom themselves fisheries specialists. 

Resource management issues dominate ecological concerns in most fisheries, except small bodies of water under 

sole ownership. There are in principle two sets of issues: maintaining the biological productivity, and making the 

optimum use of it, in social and economic terms. The latter issues are often given higher priorities. The main aim of 

the Department of Fisheries is to maintain biological productivity by making sure that the resource is allocated 

wisely between different resource users; i.e. pair trawlers, gill netters, and purse seiners. This will mean taking 

explicit decisions on allocation of resources. Difficulties do arise when the resource users are in conflict as to who 

owns what resource. 

The fishes of Lake Malawi are one of the most remarkably diverse and abundant groups in the world. Some changes 

in species composition on the Lake Malawi fisheries have been reported. Correct identifications are necessary if 

ecological changes, whether or not the result of human impact, are to be taken into account in management plans. 

However, the systematic knowledge of these projects is very poor thereby obstructing further biological research. 

Rich as it is in flora and fauna, Lake Malawi plays host to some of the poorest people in the region, an unfortunate 

paradox. From the Chitipa hills to its delta in Mozambique where it empties into the Zambezi River, it has made 

limited economic contribution to the lives of millions of Africans who live on its banks. Only subsistence survival 

from wetland and floodplain food resources, and building materials offer a meagre existence. 

The lakeshore communities depend heavily on natural resources for a living, the result of which is a high rate of 

environmental degradation. Firewood is a major source of energy for the communities, contributing to increasing 

levels of deforestation. Its waters generate hundreds of megawatts of electricity downstream annually but none of 

this power find its way into the homes of the local communities. The few social amenities like schools and clinics, 

which are thinly spread in the fishing villages, do not benefit from the hydroelectric power schemes through either 

piped water or electricity. Most fishing communities do not have secondary schools or decent clinics. The people do 

not have a good gravel road, let alone a tarred one, or a fixed telephone network. 

Lake Malawi is the largest water body in the SADC region, and yet the water is inaccessible and most often not 

potable and the communities who inhabit it’s shore face severe water shortages each year owing to non-existent 

infrastructure. There has never been an attempt to draw water from this massive body for the local community’s 

consumption.

5 


The major economic activity on Lake Malawi is tourism, which is not fully exploited to reduce the poverty. Tourism 

generates substantial income for most of the countries in the SADC and is the economic mainstay of the Zambezi 

River basin. Top attractions include the lake itself, a unique blend of ecosystems that supports large animals, 

including elephant, hippo, lion, and hundreds of bird species and beautiful flora, and the colourful ornamental fishes. 

Communities in the Lake Malawi catchment are rated as poor in terms of monetary incomes but social scientists say 

they could raise their economic status if they fully exploited the fish resources and tourist dollars through sales of 

local arts and crafts, and game viewing. And yet, no incentives are forthcoming to help Malawians realise this 

potential. 

The beautiful sandy beaches and impressive mixes of biodiversity are, in places, dotted with shanty dwellings, 

which are skilfully and shamefully tucked away amongst mansions, posh hotels and beautiful, but environmentally 

insensitive, leisure facilities. The large amounts of money that have been spent studying the science of the lake have 

not yielded much for the ordinary person who remains buried in deep poverty. The dilapidated lakeshore towns 

exude a sense of lost opportunities. The lake slumbers with its mighty potential, impressive geography and endless 

variety. The lakeshore situation can best be described as a mirror image of the rest of the region where many 

resource endowed communities wallow in poverty while there is plenty on their doorstep. 

It is my contention that, science, strategic plans and management plans do have their place to play in the fishing 

industry, but it is now time that we should put more of our concern on the exploitation of the fisheries resources. We 

as a Department have done a lot in trying to promote this agenda, what we need now is a commitment. Government 

needs to come up and raise the portfolio of the fishing industry as a priority in Malawi. 

It is my hope that this symposium will go a long way in trying to resolve some of these problems. 

With these words I would like to declare this symposium open. 

Thank you

6 


The role of the National Research Council of Malawi in protection, conservation 

and management of fisheries resources in Malawi 

Frade K. Nyondo 

National Research Council Of Malawi, P. O. Box 30745, Lilongwe 3. Malawi 

Abstract 

The Malawi population directly depends on the genetic resources of fauna and flora for its survival in terms of food, medicines, 

aesthetic values and tourism. The fisheries resources contribute a significant proportion of Malawi’s genetic resources. The water 

ecosystem of Malawi comprises between 500 to 1 000 fish species most of which are endemic to Malawi making it an important 

living aquatic resource to the country. Fish contributes 4% to the GNP, 60-70% to the consumption of the nation’s animal protein 

and employs a considerable proportion of the population that is engaged in fishing, fish processing, marketing, boat building and 

repair, engine repair, and gear supply. In addition, a number of companies are involved in the export of aquarium fish. It is 

disheartening to note that despite the important role the fisheries resources play in the economy of Malawi, some unscrupulous 

individuals have collected and exported Malawi’s fisheries resource without following proper procedures. This has led to Malawian 

fish being bred outside the country, making Malawi loose economic benefits and much needed foreign exchange. This situation has 

called for the establishment of a mechanism that allows for sustainable utilisation of fisheries resources while at the same time 

ensuring equitable benefit sharing from such use. 

Malawi is a signatory to various protocols and agreements related to fisheries resources including Convention on Biological Diversity 

(CBD). Malawi does not currently have a coherent policy for the sustainable utilisation of genetic resources. Furthermore, following 

the decentralisation of research clearance for foreign scientists, the system has lent itself to such abuse that it has become rather 

open to over exploitation. To overcome these problems, the National Research Council of Malawi (NCRM) through its Genetic 

Resources and Biotechnology Committee (GRBC) has developed procedures and guidelines for access and collection of genetic 

resources in Malawi which were officially launched in February, 2001. These guidelines were developed based on the notion of 

controlled access which means promoting access to genetic materials and traditional knowledge while setting terms of access 

based upon development priorities. This leads to low to high utilisation and compensation. NCRM has also developed draft 

procedures and guidelines for the conduct of research in Malawi, contractual agreement forms and has formed a taskforce to draft a 

benefit sharing formula. 

Introduction 

The Malawi population directly depends on the genetic resources of fauna and flora for its survival in terms of food, 

medicines, aesthetic values and tourism. The fisheries resources contribute a significant proportion of Malawi’s 

genetic resources. The water ecosystem of Malawi comprises between 500 to 1 000 fish species most of which are 

endemic to Malawi making it an important living aquatic resource to the country. Fish contributes 4% to the GNP, 

60-70% to the consumption of the nation’s animal protein, 40% to the total protein consumption of the population. 

In addition, fisheries employ a considerable proportion of the population that is engaged in fishing, fish processing, 

marketing, boat building and repair, engine repair, and gear supply. A number of companies are involved in the 

export of aquarium fish. It is disheartening to note that despite the important role the fisheries resources play in the 

economy of Malawi, some unscrupulous individuals have collected and exported Malawi’s fisheries resource 

without following proper procedures. This has led to Malawian fish being bred outside the country, making Malawi 

loose economic benefits and much needed foreign exchange. 

To safeguard the over exploitation of the country’s fisheries resources and to ensure regulated access, conservation 

of the fisheries resources, sustainable use of their components, and fair and equitable sharing of benefits arising from 

their use, the Government through various Ministries and Departments have established various committees and 

working groups. These committees and working groups are looking at the policy, legislative and strategic 

frameworks that can ensure that the fisheries resources of Malawi are protected. Realising that most of the fisheries 

resources collected from Malawi end up being used for research purposes in other countries, the Government has 

signed and ratified a number of international agreements that are aimed at protecting the fisheries resources from 

erosion. 

The paper aims at (1) itemising the national initiatives that have been undertaken in order to protect the country’s 

fisheries resources, (2) the paper highlights the international agreement that Malawi has signed and some of their 

provisions, (3) the paper looks at the contents of the procedures and guidelines for access and collection of genetic 

resources in Malawi, (4) the paper outlines the summary of the procedures and guidelines for the conduct of research 

in Malawi, (5) the paper looks at NCRM as a co-ordinator of research, science and technology, (6) it looks at the 

way forward for Malawi.

7 


National initiatives towards the protection of genetic resources 

Several national initiatives have been undertaken in order to protect the fisheries resources of Malawi. There is the 

Genetic Resources and Biotechnology Committee (GRBC) which falls under the National Research Council of 

Malawi (NRCM). The Committee has the following Terms of Reference (TORs): 

• To institute measures harmonious with the relevant guidelines available in the country . 

• To ensure that collection of Malawi’s genetic materials does not lead to loss of biological diversity and/or 

Government Revenue. 

• To ensure that the importation of genetic resources (including genetically modified organisms) and 

germplasm does not adversely affect the conservation and sustainable use of biological diversity. 

• To ensure that exchange of genetic resources and germplasm is done in such a way that Malawi benefits 

economically from whatever is exported. 

• To encourage the establishment of gene banks and genetic data banks (in-situ and ex-situ) and formation of 

strong linkages with banks including the SADC gene bank. 

• To advise the Government on which of the country’s genetic materials should be protected against 

detrimental use by researchers, collectors and traders. 

• To foster the dissemination of information on trends in biotechnology. 

• To keep abreast with the national, regional and global trends in intellectual property rights and trade. 

• To ensure that expatriate researchers work closely with competent Malawian researchers. 

• To encourage and promote endogenous development of biotechnology in areas where Malawi has 

comparative advantage. 

Secondly, the Government of Malawi developed a National Environmental Action Plan (NEAP) which was 

launched in December, 1994. In the NEAP, several factors that affect Fisheries resources were identified. Notable 

among them were, water resource degradation, high population growth, threat to biological diversity, human habitat 

degradation, air pollution, climate change, and depletion of fish resources. 

The NEAP clearly spells out strategies and actions that need to be put in place in order to conserve, sustainably 

utilise and manage the fisheries resources of Malawi. Along with the NEAP, a number of policies have been put in 

place. Such policies include those that deal with population, fisheries, and tourism. The various policies have been 

enhanced through legislation to ensure that the genetic resources are protected by law. The all embracing legislation 

is the Environmental Management Act (EMA) which provides for the: 

• Identification of biological diversity. 

• Determination of threatened biodiversity. 

• Preparation and maintenance of inventory of biological diversity. 

• Putting in place measures for better protection and conservation of rare and endemic species of fauna and 

flora. 

• Determination of actual and potential threats to biological diversity of Malawi and devising measures that 

are necessary for preventing their loss. 

• Integration of conservation and sustainable use of biological diversity into sectoral or cross-sectoral plans, 

programmes and policies of Government, the private sector and the local communities. 

• Re-introduction of ex-situ species in native habitats and ecosystem provided that they do not pose threat to 

in-situ species and habitats. 

• Establishment of regulations and guidelines to control or restrict access by any person to the genetic 

resources of Malawi including. 

• Prohibiting the exportation of germplasm except with government approval. 

• Providing for the sharing of benefits arising from the exploitation of germplasm originating from the 

technology owner or the Government. 

• Providing for payment of fees and charges for access and export licence in respect of germplasm. 

• Promotion of such land use practices that are compatible with the conservation of biological diversity of 

Malawi. 

• Selection and management of environmental protection areas for the conservation of various terrestrial and 

aquatic ecological systems of Malawi. 

• Establishment and management of buffer zones near environmental protection areas. 

• Prohibition and control of introduction of alien species.

8 


• Establishment and management of germplasm banks, botanical gardens, zoos, animal orphanages, and such 

other facilities as may be prescribed. 

• Identification and integration of traditional knowledge into the conservation and sustainable utilisation of 

biological diversity. 

In addition to the all embracing EMA, there are sectoral legislations that also impinge on the conservation, 

sustainable utilisation and protection of fisheries resources. Such Acts include the Fisheries Conservation and 

Management Act etc. 

The provisions of EMA formed the basis of the objectives of the National Biodiversity Strategy and Action Plan 

(NBSAP). The objectives include compiling of a complete inventory of Malawi’s biological diversity, building and 

enhancing capacity, creating awareness of the value of conservation and sustainable use, reviewing and harmonising 

existing legislation, documenting indigenous knowledge and setting up of a Malawi Biodiversity Secretariat. 

The Government of the Republic of Malawi, recognising the importance of biological diversity in the socioeconomic 

development, enshrined biodiversity conservation in the new constitution in 1994 (Chapter II), Section 

13(d). The constitution provides for the State to conserve and enhance biodiversity, prevent the degradation of the 

environment, provide a healthy living and working environment for the people, and accord full recognition of rights 

of the future generations by means of environmental protection in order to achieve sustainable development. 

International instruments signed and ratified by Malawi 

Malawi is signatory to several international agreements that have a bearing on fisheries resources. Some of these 

agreements are the Convention on Biological Diversity (CBD) which Malawi signed on 15 th June, 1992 and ratified 

it on 28 th February, 1994. Other agreements Convention of Wetlands of Significant Importance, Convention on 

International Trade in Endangered Species of Wild Fauna and Flora, African Convention on the Conservation of 

Nature and Natural Resources, United Nations Convention on Law of the Sea, Montreal Protocol for the Protection 

of the Ozone Layer, United Nations Framework Convention on Climate Change, and the Convention of 

Desertification and Drought. The emphasis in this section will be on the provisions of the CBD. 

The CBD provides for sovereign rights over fisheries resources and calls for regulating access on mutually agreed 

terms with prior informed consent. Secondly, it allows member countries to create incentive measures that would 

promote conservation and sustainable utilisation of fisheries resources. Thirdly, it calls for fair and equitable sharing 

of benefits arising out of the use of fisheries resources. 

The member countries are under obligation by the CBD to create incentive measures that can promote conservation 

and sustainable use of fisheries resources, and create sovereign rights over fisheries resources. It obliges member 

countries to ensure fair and equitable sharing of benefits arising out of the utilisation of fisheries resources or out of 

utilisation of knowledge, innovations and practices of indigenous and local communities. It requires them to regulate 

access to fisheries resources and traditional knowledge based on mutually-agreed terms and upon prior informed 

consent, and to create a mechanism to facilitate access to technology (including that which is relevant to the 

conservation and sustainable use of fisheries resources as well as biotechnology that makes use of these resources). 

Finally, it requires member countries to encourage co-operation between the government authorities and the private 

sector in developing methods for sustainable use of these resources (NCRM, 2000 and Ntupanyama, undated). 

Procedures and guidelines for access and collection of genetic resources in 

Malawi 

The NCRM through it GRBC has developed procedures and guidelines for access and collection of genetic 

resources in Malawi. The guidelines clearly stipulate the users, set out the objectives to be achieved by adhering to 

them, define the categories of researchers, outline the procedures and requirements for application, assign 

responsibilities to affiliating and certifying institutions, prescribe conditions for research and material transfer 

agreements and circumstances under which the certificates can be withdrawn, and oblige researchers to follow a set 

standard on publications and data. The guidelines also give a list of certifying institutions and contain an application 

form.

9 


Procedures and guidelines for the conduct of research in Malawi 

The NCRM formed a taskforce that drafted the procedures and guidelines for the conduct of research in Malawi. 

The procedures are meant to enhance the quality of research in order to achieve competitiveness and relevance at 

national, regional and international levels. These procedures aim at assisting the NCRM and sectoral institutions to: 

• Appraise proposals for scientific, professional and ethnical merits. 

• Provide mechanisms for monitoring and evaluation of research projects and activities. 

• Provide a framework for collaboration among researchers within the country and with international 

researchers. 

• Promote originality and complementarily in research in order to avoid unnecessary duplication. 

• Promote capacity building and encourage the development of sectoral programmes and research agenda. 

Ensure proper collection, acquisition, dissemination, storage and management of 

research information 

The guidelines stipulate the need for affiliation and capacity building; provide for formats for review of research 

proposals, monitoring and evaluation, and dissemination of research outputs / results. 

NRCM as co-ordinator of research, science and technology 

NCRM has established several technical committees. The fisheries department is a member of committees such as 

the Agricultural Sciences Committee (ASC) and the Genetic Resources and Biotechnology Committee (GRBC). 

Likewise NRCM is a member of some committees that deal with fisheries. The NCRM in its quest to streamline the 

research agenda in the agricultural and natural resources sector formulated the Malawi Agricultural and Natural 

Resources Research Master Plan. One of the taskforces that worked on the Master Plan was on fisheries. Through 

the Agricultural Services Project, the NCRM through the ASC funded research projects on fisheries which yielded 

some usable technologies under the Contract Research Programme. 

The way forward 

For Malawi to effectively conserve, protect and ensure sustainable utilisation of fisheries resources, there a is need 

to do the following: 

• Document all the endemic and exotic fisheries resources that are in the country. 

• Put in place a strict monitoring system that will ensure that the purpose for the collection of the fisheries 

resources collected is known. This means that the NCRM has to speed up the completion of the procedures 

and guidelines for the conduct of research in Malawi. 

• Ensure enforcement of the procedures and guidelines for access and collection of genetic resources in 

Malawi which requires building capacity in the various affiliating institutions, the NCRM and law enforces. 

• Create awareness on the various stakeholders on the value of Malawi’s fisheries resources and the need to 

protect them. This will need to include the importance of adhering to guidelines and procedures as 

stipulated. 

• Involvement of the local communities in the protection, conservation and sustainable utilisation of fisheries 

resources. Where this is already being done as shown in the paper, the system needs to be intensified and 

internalised. 

• Implement the recommendations. 

• Speed up the process of designing a benefit-sharing formula. 

• Adhere to international conventions and protocols that have a bearing on fisheries resources by 

implementing the provisions contained therein. 

• Encourage the establishment of more protected areas for the conservation of fisheries resources. 

• Complete the formulation of contract agreement and prior informed consent forms. 

• Strengthen the technical arm of NRCM so that it can effectively play its co-ordination role. 

• Another donor should be identified so that the Contract Research Programme is revitalised. 

Conclusion 

The paper has highlighted the importance of fisheries resources in Malawi in terms of food, income and tourism. 

Owing to the vital role these fisheries resources play in the well being of the Malawi population, the need to

10 


conserve, protect and use them sustainably cannot be over emphasised. Procedures and guidelines for access and 

collection of genetic resources in Malawi have been developed for use by various stakeholders and need to be 

adhered to in order to ensure that Malawi benefits from the use of its fisheries resources. Enforcement of regulations 

is key to protection of fisheries resources. There is a need for liaison among the various parties involved to ensure 

that all stakeholders understand what is required of them. 

References 

National Research Council of Malawi, 2000. Policy Recommendation, Procedures and Guidelines on Access to Genetic 

Resources in Malawi : Incorporating Sustainable Utilisation of Genetic Resources into Biodiversity Management 

Planning. Report presented at a National Workshop on Plant Genetic Resources Conservation and Understanding 

International Instruments on Biodiversity Related Issues Held at Lilongwe Hotel , 10 th – 13 th January 2000. 

Ntupanyama, Y. M., Undated. Convention on Biological Diversity and Implementation in Malawi.

11 


A general overview of fisheries research and development in Malawi 

O. M. Kachinjika, 

Department of Fisheries, P. O. Box 593, Lilongwe, Malawi 

Abstract 

Historically, fisheries research and development in Malawi is strongly linked to Lake Malawi and the immense interest it has aroused 

to the outside world due to its sheer size (28 780 km²). Consequently, universal interest in the fisheries of Lake Malawi cropped up 

spontaneously, and has led to a number of research projects being commissioned, beginning with the surveys of 1939, although 

earlier attempts at taxonomic studies can be traced to over a century ago. The paper outlines major fisheries research projects that 

have been undertaken in Malawi. An attempt is made to outline how project results have influenced fisheries management 

strategies and policies. In addition, the implications of management policy on research programs being undertaken is also 

highlighted. Current research programs and future plans are outlined and their relevance to fisheries management is evaluated 

Introduction 

Fisheries research and development in Malawi is strongly linked to Lake Malawi and the immense interest it has 

aroused to the outside world due to its sheer size (28 780 km 2 ), manifested by the fact that it is the third and ninth 

largest lake in Africa and the world, respectively. Consequently, universal interest in the fisheries of Lake Malawi 

can be traced to over a century ago when the first fish collections were taken by the early European explorers who 

came to this part of Africa, notably the early expeditions of Dr. David Livingstone (Tweddle, 1991). Specimens of 

fish collected were deposited in natural history museums where early work was restricted to taxonomy. 

This paper outlines major episodes that have characterised the development of fisheries research and the fisheries 

industry in Malawi by chronologically classifying to the present time the major projects that have taken place. 

These are classed as: (1) activities that took place during the period of limited or very low fishing effort; (2) projects 

that were aimed at maximising fish production; and (3) projects aimed at sustainability and conservation of fish and 

fisheries resources. For each research project, the funding agencies, objectives and major results or outputs are 

given in accordance with the fisheries policy and objectives prevailing at the particular time. This paper culminates 

in outlining currently running projects and a newly adopted approach to research project planning and development 

that is being advocated by the Department of Fisheries in Malawi. It then concludes, by emphasising the need for 

continuous research to provide information required for effective and sustainable fisheries management measures. 

Phase 1- Period of limited or very low fishing effort (1939 – 1960) 

This is a time when the population in Malawi was small, the fishing effort was low, fishing gear was relatively 

ineffective and canoes could only fish in near-shore waters. This effectively meant that Malawi had large fish 

reserves in the areas that could not be fished. Fisheries research studies were initiated at this time, but most of it was 

taxonomic in nature, although some fish biology was undertaken. This period marks the first phase in fisheries 

development. 

However, documented research works date just more than 50 years ago when Bertram, Borley and Trewavas 

conducted the first fishery surveys on Lake Malawi in 1939, and a report was produced in 1942 (Bertram et al. 

1942). This was followed by some biological research work on the chambo, Oreochromis spp. of the Southeast Arm 

of Lake Malawi, in 1945 – 1947, reported by Lowe (1952). These chambo studies started as a result of the 

operations of purse seining, which were introduced into Lake Malawi in 1943, solely for the exploitation of this 

species group. 

A major breakthrough in fisheries research and development came in the mid-1950s with the formation of the Joint 

Fisheries Research Organisation (JFRO) in 1954. This was a research body that was shared by Malawi (then 

Nyasaland) and Zambia (then Northern Rhodesia). In Malawi it had the mandate to carry out research work on Lake 

Malawi. The organisation set up its base at Nkhata Bay, where it carried out biological studies on some fish species, 

and major work on the inshore rock-dwelling fish species, locally known as ‘mbuna’. Summarily, the JFRO carried 

out taxonomic studies, descriptions of traditional fishing methods, experimental fishing, and also carried out limited 

limnological investigations. The report, by Jackson et al. (1963), covers information on hydrological data from the 

lake and affluent rivers, invertebrate studies, a check-list on fishes and ecological zonations, biology of 

commercially important species, experimental fishing and the traditional fisheries. The studies on mbuna, reported 

by Fryer (1959), contributed to evolutionary knowledge and formed a basis for subsequent studies and discussions 

on the topic.

12 


Phase 2- Maximisation of fish production (1961 –1996) 

The need to redirect fisheries research to comply to national food demands, and, fisheries development 

requirements, led to the research focus being shifted in 1962, to Monkey Bay, in the south of Lake Malawi where 

new laboratories had been constructed. This was necessitated by a number of reasons, which included the desire to 

understand the highly productive fishery that was being exploited by purse seines, and the ready availability of 

communication and access facilities to the outside world. At first, the new station continued some of the work that 

was initiated in Nkhata Bay, which included physical limnology, fish taxonomy and biology, gradually focussing 

more on the biology of nchila, Labeo mesops and Kampango, Bagrus meridionalis, Tweddle (1991) and gill-netting 

trials. 

Projects 

Various projects were undertaken to respond to the increased demand for protein supplies to the Malawi population. 

The impetus for this phase began with the introduction of experimental trawling in the Southeast Arm of Lake 

Malawi in 1965, which led to the development of commercial trawling in 1968. This trawling targeted virgin stocks 

of cichlid species unknown to science. Consequently, research studies were initiated on the taxonomy and biology 

of these cichlid fishes with the aim of establishing a management and monitoring regime for this new fishery. 

Preliminary data collected at this time was used in detailed stock assessment work which began in 1969, aiming at 

monitoring the effort being applied on the fish stocks and exploring the potential for trawl fisheries in other parts of 

the lake. These surveys produced some punctuated series of biomass estimates for the demersal fish stocks, and are 

regarded as core trawl sampling programs of fisheries research on Lake Malawi. 

However, the first regular trawl surveys, were undertaken by the FAO Project on Integrated Fishery Development 

(FAO, 1976a). The main aim was to undertake studies on stock assessment to determine the optimum sustainable 

yield of the newly introduced mechanised trawl fishery in the southern part of Lake Malawi. It operated on the 

premise that the stocks being exploited by the trawl fisheries were limited and so, effort had to be monitored and 

management measures put in place, if sustainability was to be ensured. In addition to studies on the trawl fisheries, 

it also did analytical work in the artisanal fisheries of the SEA and SWA of the Lake Malawi. Management 

recommendations included restriction of the number of licenses to be issued in each demarcated area of southern 

part of Lake Malawi and the setting of the a 38 mm minimum mesh size restriction for the trawl cod-end. A 

recording system for the traditional fisheries in the country, the Catch Assessment Surveys (CAS), was put in place, 

and is still being used with minor modifications except in the Mangochi Fisheries District area in the southern part 

of Lake Malawi. When the regular surveys were discontinued, changes in exploited stocks were monitored through 

catch and effort data from the trawl fisheries. The results of the analysis of the catch and effort data showed that the 

numerous demersal haplochromine species, that were being exploited, were sensitive to commercial fishing 

operations with the average size of fish being caught decreasing. It was also observed that there was an increase in 

use of illegal mesh sizes and fishing in prohibited inshore areas. 

These observations led to the birth of the Demersal Fisheries Reassessment Project (ODA/FRAMS Project, 1992), 

which re-introduced regular fishing surveys that were carried out on a quarterly basis, using the same boat and 

methodology as before. Partially funded by an FAO Chambo Project, the aim was to repeat the experimental 

trawling surveys of the early 1970s to examine the current status of the stocks with a view to assess changes that had 

taken place in the fishery and to determine the effect of the 38 mm minimum mesh size. These surveys showed that 

the exploited biomass had declined in the heavily fished areas, and that the species composition of the catches had 

changed from being dominated by large cichlids to mainly small haplochromines. Consequently, it was conclude 

that the 38 mm of the cod end was ineffective in arresting this trend. In addition, the very low levels of biomass 

observed in Area A led to the prohibition of trawling in the area 1992. 

The quarterly stock assessment surveys were continued under an ICEIDA-funded project using the new powerful 

research vessel, the RV Ndunduma, Banda and Tomasson (1997). The findings indicated that there was scope for 

expansion of the demersal fisheries, in the deep waters of the southern part of the lake. This has resulted into the 

development of the deepwater fisheries in the southeast arm of Lake Malawi, and these are currently being exploited 

by three fishing vessels, viz: MV Kandwindwi, RV Ndunduma and the fishing vessel Chenga. 

In line with the fisheries development agenda of the 1970s, which focussed on maximising fish production from 

Lake Malawi and other water bodies in the country, a series of other research projects funded by donors were 

implemented. These include the UNDP/FAO Fishery Expansion Project (FAO, 1982a), which explored the 

opportunities for fisheries development in the central and northern parts of Lake Malawi. Recognising that inshore

13 


stocks were over-fished, the project concentrated on the biology of the deepwater fishes of usipa, utaka, ncheni, 

ndunduma, sanjika, and, studies on plankton and the physical environment. Results showed that the usipa stock was 

characterised by extreme differences in the abundance of year-classes, posing a problem for commercial 

exploitation. Acoustic surveys showed fish density to be much less than in the south. Subsequent trawling indicated 

a low potential catch rate and limited trawlable bottom in the North. It was summarily concluded that pelagic fish 

resources were not sufficient to support a mechanised commercial fishery. As a result, the recommendation was that 

long-term studies on all aspects of the pelagic zone limnology and ecology be undertaken. This recommendation 

gave way to the birth of the UK/SADC Project, which conducted comprehensive studies to assess the fishery 

potential of the pelagic zone (Menz, 1995). 

The UK/SADC Lake Malawi Pelagic Fisheries Assessment Project of 1991-1994, was a regional project between 

Malawi, Mozambique and Tanzania, which studied Lake Malawi as one ecosystem. The aim was to assess the 

fishery potential of the pelagic zone, through various studies that included fisheries biology, stock assessment, 

aquatic ecosystem trophic relationships, and quantification and determination of economically sustainable yields 

from new fisheries targeted at pelagic fish species. An assessment of the standing biomass for the pelagic zone 

was made as well as its fishery potential. The lake fly was shown to be a component of the main energy 

pathways. Description of the physical and chemical characteristics of the lake provided baseline data for 

subsequent studies including monitoring of the lake environment. 

This project was followed by the GEF/SADC Lake Malawi/Nyasa Bio-diversity Conservation Project, which had a 

somewhat similar management arrangement to the above. The goal was to create the scientific, educational and 

policy basis for conserving the biological diversity of the lake, and to enable the riparian states to establish higher 

levels of sustainable production from the Lake’s resources consistent with preserving its bio-diversity and unique 

ecosystem. The outputs from the project include taxonomic reviews of certain taxa, documentation of benthic 

macro-invertebrates in the rocky near-shore zone of Lake Malawi/Nyasa, ecological and biological studies on 

selected species (Duponchelle, 2000), description of physical characteristics of the lake, and the development and 

execution of a Lake Malawi bio-diversity awareness program. It was then recommended to continue monitoring 

the parameters that had been studied, and use the subsequent results to constitute an early warning system for the 

changes taking place in the aquatic environment. 

Other Projects 

Several other projects have been implemented, one of which is the Ornamental Fisheries Research Project (1980). 

This was a collaborative project, which was conducted by the Fisheries Research Unit and Rhodes University 

together with other scientific collaborators. The aim was to provide a sound understanding of the techniques of this 

trade, potential, sustainability and threats to the stocks, together with problems of escape and relocation of the 

species in the lake. It resulted in the provision of management guidelines for the government to regulate the trade. 

It was generally observed that the ornamental fish business was self-regulating, because of market limitations. 

Nevertheless, the project increased the knowledge of the ecology of the inshore fish stocks and widened the 

knowledge for management of the numerous mbuna fish species, and also contributed to advancements in 

conservation and tourism. With such information at its disposal, the project assisted in the early planning and 

development of the Lake Malawi National Park, which is the first lake underwater reserve in Africa. It also 

increased the worldwide attention in freshwater science and generated considerable awareness of Lake Malawi 

natural resources. 

Another important study was carried out by the Chambo Fisheries Research Project, which aimed to carry out 

biological, stock assessment and socio-economic research to provide a management plan for the chambo fisheries of 

the Southeast arm of Lake Malawi, the Upper Shire River and Lake Malombe. This was in response to mounting 

fishing pressure in the 1980s that had caused the stocks in each fishery to be fully or over exploited. 

Studies centred on the elucidation of the life histories of the three chambo species in the area. These included 

breeding, feeding, growth, distribution and migration; description of artisanal gears and, commercial and industrial 

fisheries including socio-economics; quantification of contributions of fish stocks other than chambo; and modelling 

of resource exploitation patterns. The management recommendations included licensing, gear use and 

specifications, closed seasons and future-monitoring regimes. 

The project findings brought to light the crash of the chambo stocks in Lake Malombe and their serious depletion in 

the Southeast arm of Lake Malawi (FAO, 1993). Initially, the project aim was to study the Oreochromis spp., the

14 


most economically valuable component in the Malawi fisheries, but ironically it concluded in a situation where 

smaller, haplochromine species were taking over as the important fishery. In Lake Malombe, chambo declined to 

only 6% of the total catch from 90% in 1976. The results of the Chambo Project confirmed the observations of the 

fishing industry, which was reporting declining chambo catches and an increase in small-sized fish species as 

observed in Lake Malombe. These observations confirmed the fact that the fisheries sector was facing serious 

challenges in maintaining a balance between exploitation and conservation to achieve optimal utilisation without 

impairing long-term sustainability. In view of this, it was recognised that there was a need to put in place an 

effective fisheries management regime that would ensure the sustainability of the fish stocks. Hence, it was 

recommended that national core research programmes be formulated to focus on identification of new unexploited 

stocks as well as the management of those stocks that are under heavy exploitation in order to sustain their 

production potentials and, where necessary, lead to a recovery of the stocks. 

In support of the objective of creating an effective management system while maximising production, an FAO 

Technical Consultation Paper (TCP) on data recording systems was commissioned under the Chambo Fisheries 

Research Project. This developed a gear-based, computerised data handling and analysis system, the Malawi 

Traditional Fisheries (MTF) for statistics in the Southeast arm of Lake Malawi, the Upper Shire and Lake Malombe 

(Alimoso et al. 1990). An evaluation of MTF and the comparison with the prevailing data collection system CAS 

led to the recommendation that MTF be adopted throughout the country . However, MTF proved to be costly, so 

that it is only carried out in the south of Lake Malawi in Mangochi District. However, Coulter (1993), referring to 

the need to review data collection systems, noted that statistical data is very important for monitoring purposes, and 

recommended a recording system that is sufficiently sensitive to detect changes in effort and CPUE, while saving 

costs. 

In response to the need to develop the fisheries industry, the WB/IDA-Fisheries Development Project was 

commissioned, and aimed at institutional building. The main objective was to assist Malawi to realise more fully 

the potential contribution of the fishery sector to the economy, while ensuring that off-take did not exceed 

sustainable yields. This was to be achieved through increasing fish production, generating additional off-farm 

employment and income, conservation of the natural resource and prevention of environmental degradation and 

improving institutional capacity for fisheries sector policy formulation, research planning, monitoring and 

control. Work of the research component of the project confirmed the presence of deepwater fish stocks reported 

in earlier studies and led to the development of the deep water fisheries which are now being exploited by newly 

introduced stern trawlers. In addition, determinations and assessments were made of biomass indices and species 

composition of the demersal fish stocks in both heavily and lightly exploited areas, resilience of the demersal fish 

stocks to exploitation, and economic viability of using a big fishing vessel. 

Other major studies that have been undertaken on Lake Malawi include the Project on the Comprehensive Study of 

Lake Malawi Ecology for Sustainable Development, which concentrated on the ecological aspects of fisheries. The 

goal of the project was to develop a research system and accumulate scientific knowledge on Lake Malawi and its 

surroundings for use by researchers, local communities and policy makers. The specific objectives were to: (1) 

strengthen ecological research and human capacity for sustainable utilisation and proper management of natural 

resources on Lake Malawi; (2) develop comprehensive research methods for sustainable utilisation of fish resources 

and; (3) establish bilateral research partnership, between Japan and Malawi (Lake Malawi Ecology Project, 1999). 

The project trained Malawians through postgraduate training, technician level training, and counterpart visits. It 

also established the Molecular Biology and Ecology Research Unit (MBERU) in the Biology Department of the 

University of Malawi, and a field station at Cape Maclear. 

Environmental problems have also been addressed in the fisheries sector, one of which is the elimination of nuisance 

aquatic weeds. The Malawi National Water Hyacinth Control Project was set up with the aim of controlling the 

spread of water hyacinth in Malawi’s water bodies beginning with the problems in the Shire River system. The 

objectives of the project are: (1) identification and development of biological control agents; (2) assessment of the 

effects of water hyacinth on fish and invertebrate abundance and bio-diversity; (3) assessment of the impact on 

riparian communities; (4) establishment of a capacity to carry out limited spot chemical treatment of water 

hyacinth; (5) informing of local communities and the general public of the dangers of the weed and (6) support to 

research activities on the economic utilisation of water hyacinth. So far, the project has been able to establish 

breeding sites for biological control agents at Makhanga and Mangochi and rearing and distribution of the 

biological agents is in progress. Creation of community awareness is ongoing in different districts.

15 


Several other projects have been carried out on Lake Malawi and in various ways have contributed to the knowledge 

and understanding of the status of the fisheries, and recommended various effective and sustainable strategies for 

exploitation and management (Appendix 1). These include the Traditional Fisheries Assessment Project, which 

carried out a revision and development of monitoring and assessment systems for traditional fisheries, and made 

recommendations for modification of the existing fisheries regulations at the time. The Fisheries Research and 

Management Support (FRAMS) Project aimed at strengthening the capacity of the Department of Fisheries in 

development of research and formulation of policy, and this resulted in the review of management and policy 

options of the fisheries sector. Research under the FRAMS Project, investigated the distribution and ecology of the 

principal fish species of economic importance, involving investigations of catches in all parts of the lake in all 

habitats. Consequently, results of the FRAMS Project plus consultations with the fishing communities, led to the 

establishment of the 19 mm mesh size restriction for the Nkacha net which exploits small haplochromines. 

Since 1966, research projects have been undertaken in other smaller water bodies with varying degrees of 

importance in fisheries ranging from significance at national level, to local significance to the communities in the 

riparian areas. The water bodies of significance are Lakes Malombe, Chirwa, Chiuta and the Lower Shire River 

system, which in total contribute about 30% of the total fish catch in Malawi. In general, the aim was to carry out 

occasional studies of salient aspects as opportunity and time permitted. The projects recommended the 

establishment of regular catch monitoring programs, improvement of fish processing techniques and establishment 

of effective communication and access to these fisheries. In perspective, it was visualised that the problems in these 

water bodies would encompass environmental and socio-economic aspects, emanating from the increased use of 

seine nets, which could lead to weed removal and a reduction in habitat diversity and exposure of fish stocks to 

over-fishing (Tweddle, 1983, 1991). Social-economically, it was visualised that an increase in population around 

these water bodies would lead to increased pressure on the stocks, so that, coupled with any large natural 

fluctuations in the fishery, the riparian communities would face dire social and economic problems. 

Research by other organisations 

Various research programs on the fisheries of Malawi are also being undertaken by other organisations, and these 

include the University of Malawi through its constituent colleges of Bunda College of Agriculture, Chancellor 

College and the Centre for Social Research. Some of the work done at Bunda College of Agriculture covers the 

ecology of the catfish, Bathyclarias spp. and cichlid species of Lake Malombe. Chancellor College has recently 

concluded a research project on the ‘Comprehensive Study of the Lake Malawi Ecology for Sustainable Utilisation 

of the Lake Environment’ and it continues the genetic classification of Malawi fishes using the facilities at the 

Molecular Biology and Ecology Research Unit. The Centre for Social Research is carrying out collaborative studies 

with the University of East Anglia, on the socio-economic aspects of fisheries and fishing communities. Rhodes 

University and the JLB Smith Institute of Ichthyology have collaborated with a number of institutions in Malawi and 

have priously participated in studies on the ornamental fisheries of Lake Malawi. In some cases, individual 

scientists have been able to undertake research on Lake Malawi after obtaining grants from research institutions, 

either in their own capacity or as students. 

Phase 3 - Fisheries resource management, sustainability and conservation (1997 

– to the present) 

Despite the implementation of the various research projects in Lake Malawi and other water bodies, fish catches 

have been declining. Notable is the situation in Lake Malombe, where annual fish landings declined to less than 

3 000 tons in the 1990s from an average of more than 10 000 tons in the 1980s (Bulirani et al. 1999). These 

observations were mainly attributed to overfishing although environmental problems are also reported to be taking 

ground. 

However, the ineffectiveness of past management regulations, emanating from past research and development 

efforts has given important lessons to be considered in any future plans. These include: 

• Limited spatial coverage of research programs. 

• Need for a continuous review of fisheries management regulations. 

• Resource exploitation should conform to the objectives of sustainable management and conservation of 

bio-diversity Research on Malawi’s water bodies should be systematic and based on themes. 

• Management guidelines for the management of Lake Malawi to be formulated based on broad stakeholder 

consultation. 

• Management recommendations have to be adhered to if the goal of sustainable utilisation and conservation 

of biodiversity is to be achieved.

16 


In view of these experiences, the Department of Fisheries, with the assistance of the GTZ-funded National Aquatic 

Resource Management Program (NARMAP), has reviewed the fisheries policy and developed the Fisheries and 

Aquaculture Policy (1999), and a new Fisheries Conservation and Management Act (1997). These are now acting as 

a guiding torch for fisheries research and development in Malawi. Consequently, the mission of the Department of 

Fisheries has been reviewed, as ‘to provide optimal framework conditions and excellent services to enable national 

fisheries industry to satisfy local demand for fish and increase incomes of people dependent on fish’. This is with 

the overall objective of enhancing the quality of life for fishing communities by increasing harvests within 

sustainable yields and promoting aquaculture as a source of income and to supplement fish supply from natural 

waters’. 

To achieve this policy objective, the Department of Fisheries recognises the fact that harvesting of fish stocks should 

be based on established sustainable yields of different species/stocks and that opportunities have to be identified to 

expand existing and/or new aquatic resources. Therefore, the research mission is ‘to provide information necessary 

for sustainable exploitation, management, conservation of bio-diversity and investment in the fisheries sector 

through appropriate biological, technological, socio-economic and environmental research programs’. The 

objectives are: 

• Accumulate comprehensive data on all fisheries resources for all major water bodies of Malawi. 

• Provide information on management of fisheries resources and aquatic environment. 

• Provide information on viable investment opportunities in the fisheries sector 

• Provide information necessary for the maintenance of aquatic biodiversity 

• Establish a fully operational taxonomy section by the year 2003 

Current and future research programs of the fisheries research unit 

The Department of Fisheries relies heavily on scientific information for the development of protocols and plans with 

which to manage Malawi’s fisheries. Essentially, sustainable management is dependent on the ability of fisheries 

managers to determine at what levels of fishing effort, and at which gear selectivity scenarios the catch of a target 

species is sustainable and the spawning stock remains adequate. Therefore, the main role of the Malawi’s 

Government’s Fisheries Research Unit (FRU) is to provide the Department of Fisheries with such information based 

on understanding of the biology, life history and distribution of the target species as well as an understanding of the 

harvesting fisheries. To achieve this, the FRU has adopted a holistic approach whereby research activities are 

geared towards the ultimate formulation of a management strategy for Malawi’s fisheries. This is depicted in Figure 

1. 

Therefore, current research programs have been formulated in a way that they are aimed at addressing the immediate 

need of formulating strategies for the management of the fish stocks facing heavy exploitation pressure, while at the 

same time exploring opportunities to expand existing and/or develop new fisheries resources. Such an approach is 

clearly illustrated in the activities that have been planned for the July 2000 – June 2001 financial year (Table 1), 

where they have also been given a priority ranking depending on the state of the fisheries and the urgency with 

which, management information should be acquired. In this program, the research activities fall under the five 

categories that have been shown in Figure 1, namely gear selectivity, utilisation trends, biological surveys, 

population dynamics and social and economic research.

17 


Table 1. Research projects planned for implementation in the year 2000/2001 (after Banda et al. 2001). 

Project Title Priority 

Biological management parameters for target species in Lake Malawi 1 

Monitoring of catch and effort in artisenal fisheries 1 

Demersal monitoring surveys 1 

Kauni Fishery selectivity survey. 2 

Chilimira fishery selectivity. 2 

Gillnet selectivity surveys. 2 

Handline catch assessment and gear selectivity. 2 

Traditional gear selectivity surveys (Northern Lake Malawi) 2 

A preliminary study of the effectiveness of monofilament gillnets in Lake Malawi 2 

Lake Malombe assessment programmes 2 

Lake Chiuta assessment programmes 2 

Commercial pair trawl selectivity 3 

Trawl net selectivity survey 3 

The economics of processing and distribution of small scale fishing in Lake 

3 

Malawi 

Demersal exploratory surveys 3 

Pelagic exploratory surveys 3 

Limnological Surveys 3 

Aquatic ecology and fisheries of Lake Chikukutu 4 

The research plans presented can be classified as either discrete or long term monitoring. Discrete surveys such as 

the determination of species and size selectivity of the chilimira, kauni, gill net and trawl fisheries in Lake Malawi, 

the determination of biological parameters for major target species or rapid stock assessment of small lakes are 

surveys with outcomes achievable during short-term research programs (1-5) years. Long-term monitoring 

programs, such as catch and effort surveys and trawl surveys can only yield management information if they are 

sustained over long periods (20 years). The classification of research programs on these criteria has major 

implications on sources of funding. This is because while the funding for discrete surveys could be secured on an ad 

hoc basis from a variety of sources, the sustainability of long-term monitoring programs can only be ensured if they 

are budgeted from within the Department of Fisheries. As a result, all the long-term research programs are being 

sponsored by a Fisheries Research Fund (FRF) that was set up under the Fisheries Conservation and Management 

Act (1997), while the majority of the discrete surveys are currently financed by NARMAP. 

Species biology 

• Size at maturity. 

• Reproductive periodicity. 

• Reproductive behaviour 

• Fecundity. 

• Distribution. 

• Age & growth. 

Population dynamics 

• Stock size. 

• Population structure. 

• Mortality rates. 

Determine target 

species 

Identify potential 

over-utilisation 

Predictive 

modelling 

Identify biological 

management 

targets 

Identify 

implementation 

constraints 

MANAGEMENT STRATEGY 

Gear selectivity 

• Species selectivity 

• Size selectivity 

Utilisation trends 

• Catch rates 

• Effort levels 

Social & Economic inputs 

• Fishing economics. 

• Marketing structures. 

• Traditional management systems. 

• Stakeholders & ownership. 

• Migration. 

• Social dimensions. 

Figure 1. Input requirements for the formulation of a management strategy for Malawi’s fisheries (after 

Banda et al. 2001).

18 


Conclusion 

Summarily, Fisheries Research and Management in Malawi has undergone an evolution process dictated by the 

prevailing social and economic demands of each period in time. The first fish specimens were collected by 

enthusiasts based on interest during the early years of contact with Europeans. At that time the population in 

Malawi was small, the fishing effort was low, fishing gear was relatively ineffective, and canoes could only fish in 

near-shore waters. This effectively meant that Malawi had large fish reserves in the areas that were unable to be 

fished. Most of the biological research at this time was taxonomic in nature, although some fish biology was 

undertaken. 

The next phase focussed on expanding the fishery, increasing fish production, and attempts to improve resource 

management to ensure sustainability. This period saw the introduction of motorisation and the establishment of the 

commercial sector, which meant that fishing was now possible in previously unfished areas. The research moved 

towards exploratory fishing for unutilised stocks, predicting effects of intensified effort, monitoring fish stocks, 

refining management regulations and seeking to understand the biological basis of sustainability. Consequently, 

research has tended to be event-driven and reactive and has followed the occurrence of certain fisheries production 

efforts, which might have seriously affected the fish in a sudden manner. These events have usually been due to 

sudden increase and unsustainable pressure of exploitation of a finite resource. The approach was always postmortem 

and focussed on finding out what had gone wrong and then working out corrective measures. 

In view of the failure of previous efforts to fishery management approaches such as mesh size restrictions, closed 

seasons etc. the Department of Fisheries is now aiming at providing information necessary for sustainable 

exploitation, management, conservation of bio-diversity and investment in the fisheries sector through appropriate 

research studies. It is also strongly promoting the participation of the fishing communities in resource management. 

This means that there is a dire need for research to be undertaken to address the following: 

• A review of old fisheries management regulations. 

• While fisheries research should respond to the immediate needs of the fishing industry, due regard should 

be given to sustainable management and conservation of bio-diversity. 

• Research on Malawi’s water bodies should be systematic and based on themes. 

• Management guidelines for the management of Lake Malawi to be formulated based on broad stakeholder 

consultation. 

• Management recommendations have to be taken up by responsible authorities if our fisheries resources are 

to be sustained through time. 

To ensure orderly and systematic conduct of research in Malawi, that ensures that the country benefits from its 

natural resources, the Malawi Government has, through the National Research Council of Malawi, launched the 

‘Procedures and Guidelines for Access and Collection of Genetic Resources in Malawi’. This is to be followed by 

all those intending to carry out research on genetic resources of Malawi. 

All in all, it should be concluded that fisheries research should lead to the establishment of appropriate fisheries 

management regimes that ensure sustainable exploitation of fish stocks, protection of aquatic ecosystem integrity’ 

and conservation of bio-diversity. 

References 

Lake Malawi Ecology Project, 1999. First Annual Report, May, 1998 – April, 1999. Department of Biology, University of 

Malawi, 50p. 

Menz, A. (ed.) (1995). The Fishery Potential and Productivity of the Pelagic Zone Of Lake Malawi/Niassa. Chatham, UK: 

Natural Resources Institute. 

Bulirani, A.E., Banda, M.C., Palsson, O.K., Weyl, O.L.F., Kanyerere, G.Z., Manase, M.M. & Sipawe, R.D., 1999. Fish Stocks 

and Fisheries of Malawian Waters: Resource Report. Government of Malawi, Fisheries Department, Fisheries Research 

Unit.54pp. 

Ribbink, A.J. (1980). The ornamental fish project: report and recommendations 

Ricardo Bertram, C.K., H.J.H. Borley and E. Trewavas (1942). Report on the fish and fisheries of Lake Nyasa. London, Crown 

Agents, 181 p. 

Tweddle, D. (1983). The fish and fisheries of Lake Chiuta. Luso:J.Sci.Tech.(Malawi), 4(2):55-83. 

Tweddle, D. (1991). Twenty years of fisheries research in Malawi. Fisheries Bulletin No. 7, Fisheries Department, Ministry of 

Forestry and Natural Resources, Malawi. 

Tweddle, D. and B.J. Mkoko (1986). A limnological bibliography of Malawi. CIFA Occ.Pap., 13:75 p. 

Duponchelle, F. (ed.) (2000). Fish Ecology Report: SADC /GEF Lake Malawi/Nyasa Biodiversity Conservation Project; 

Ministry of Natural Resources and Environmental Affairs, Lilongwe, Malawi.

19 


FAO (1982a). Fishery Expansion Project: findings and recommendations. Rome, FAO, FI:DP/MLW/75/019 Technical Report, 

21 p. 

Resource Report (1999). Fish stocks and Fisheries of Malawian Waters. Fisheries Research Unit, Fisheries Department, Malawi 

Fryer, G. (1959). The tropic interrelationships and ecology of some littoral communities of Lake Nyasa with especial reference 

to the fishes, and a discussion of the evolution of a group of rock-frequenting Cichlidae. Proc.Zool.Soc.Lond., 

132(2):153-281 

Coulter, G. W. (1993). Report on Fisheries Research and Management Strategies. ODA/FRAMS Project, Malawi, 58p 

Banda , M. C. and T. Tomason, 1997. Demersal fish stocks in southern Lake Malawi: Stock assessment and exploitation. 

Fisheries Bulletin No. 35, Fisheries Department, Malawi. 39pp. 

Banda, M.C., Chisambo, J., Sipawe, R.D., Mwakiyongo, K.R. & Weyl, O.L.F. 2001. Fisheries Research Unit, research plan: 

2000 & 2001. Fisheries Bulletin No. 44. 

FAO (1976a). Promotion of Integrated Fishery Development, Malawi: Project findings and recommendations. Rome, FAO, 

Terminal Report, 36 p. 

Jackson, P.B.N., T.D. Iles, D. Harding and G. fryer (1963). Report on the survey of Northern Lake Nyasa 1954-55. Zomba, 

Government, 171 p. 

ODA/FRAMS Project (1992). Demersal Fisheries Reassessment Project, Preliminary Project Report (November, 1987 – May, 

1992) 

Lowe, R.H. (1952). Report on the Tilapia and other fish and fisheries of Lake Nyasa 1945-47. Part 2 

Fish.Publ.Colon.Off.Lond., 1(2): 126 p. 

Alimoso, S. B., Seisay, M. D. B. & Van Zalinge, N. P. 1990. An efficient method for catch-sampling of the artisanal chambo 

fisheries of the Southeast Arm of Lake Malawi, the Upper Shire River and Lake Malombe. FI: DP/MLW/86/013 Field 

Document 6. 

FAO (1993). Fisheries management in the southeast arm of Lake Malawi, the Upper Shire River and Lake Malombe, with 

particular reference to the fisheries on chambo (Oreochromis spp). CIFA Technical Paper, No. 21. Rome, FAO, 113p.

20 


Economic Security and Sustainable Programmes for the African Great Lakes 

Benjamin P. Ngatunga 1 & Anthony Ribbink 2 

1 Senior Research Officer (Ichthyology), Tanzania Fisheries Research Institute, Kyela Centre, P.O. Box 98, Kyela, Mbeya Region, 

East Africa, [bpngatunga@hotmail.com; tafiriki@africaonline.co.tz]. 

2 Senior Specialist Scientist, JLB Smith Institute of Ichthyology, Somerset Street, Private Bag 1015, Grahamstown 6140, South 

Africa, [A.Ribbink@ru.ac.za]. 

Abstract 

Unquestionably, the responsibility for research, management and conservation of the aquatic resources of any nation reside with 

that nation. If resources are shared, such as is the case where different countries border on a single lake or river, then the 

responsibility is shared and co-management is required. In much of Africa, full responsibility for national resources has not been 

achieved. During post-colonial eras work on aquatic resources remained disproportionately in the hands of foreign workers. Initially 

this was due largely to both the lack of skilled capacity in Africa and to the economic inability of the new governments to meet the 

costs of this work. While the capacity in African countries has grown, despite attrition through the “brain drain” of marketable 

graduates, the economies have not improved to the extent where the countries can be independent of donor participation. A series 

of donor-driven, short-term projects, often delivered in stop-start cycles, have led to situations where sustainable programmes 

cannot be implemented, and long-term goals cannot be set nor achieved. The most pressing step is to provide financial security to 

research stations or conservation units to enable them to develop and progress towards the achievement of long-term visions. This 

step requires independent funding that so reduces reliance on their own and donor governments that financial security is obtained. 

Offshore investments that earn interest provide vehicles for such independence and could take the form of endowment trusts, 

section 21 companies or similar secure endeavours. Such funding, in addition to that of the government, would provide the firm 

foundation on which to launch sustainable programmes and around which to organise other, separately or jointly funded projects. 

Jointly funded projects are envisaged as partnerships between the national facility and donors or research organisations. The 

mechanisms for establishing endowment funds will be discussed and several examples of successful schemes will be given. Most 

donors and aid organisations are not favourably disposed towards making contributions to such funds, preferring rather to maintain 

the stop-start cycle of donor managed projects. It is questioned whether this approach is in the long-term interest of the countries 

receiving the support or whether it might be both advantageous and effective to persuade donors to think afresh. These issues are 

debated. It is concluded that long-term research, monitoring programmes and the achievement of conservation objectives are so 

dependent upon programmes that are driven by the nationals of the countries concerned, and funded in a secure manner for as long 

as funding is required, that it is necessary for those donors that are opposed to the establishment of investment funds to re-examine 

the situation and reconsider their policies. Ideally, to understand and manage the lakes, the programmes should have perspectives 

that embody the phrase “in perpetuity”. And this phrase should be applied to both the nature of the work and the financial resources 

that support it. 

Introduction 

As long ago as 1962, Rachel Carson pointed out that: “It is our responsibility to safeguard our genetic heritage, 

which we received as a fit (in trust) from the past … to hold and to manage (in trust) for the next generation.” 

Conservation, therefore, is a matter of trust; previous generations entrusted us with a responsibility to future 

generations. Although this concept has been used in numerous publications, it has not really taken hold in practical 

sense, partly because our role as trustees is too loose. As trusts or endowments or similar do focus attention on 

responsibilities and duties, and as they do provide one of the solutions to solving problems of achieving financial 

security for long-term programmes, we shall explore with you the value of trusts. 

Although we shall refer to trusts and endowments in this paper, we do so in the knowledge that they are not the only 

way to go, but also in the knowledge that there is a history of success and a legal foundation for their use. Indeed, 

the legal basis for trusts originated in Roman times and have since evolved to be a potentially powerful tool in 

environmental management. 

We shall be dealing with trust in two forms: 

Our responsibilities or duties as trustees charged with caring for the environment (intergenerational moral trust), and 

The manner in which trustees can have the financial resources to meet their responsibilities (financial or endowment 

trust). 

Intergenerational trusts 

In the intergenerational “moral” trust it is the duty of the present generation to care for biodiversity and 

environmental processes so responsibly that they do not compromise the capacity of the next generation to enjoy the 

benefits of natural resources. This or similar statements appear regularly in communications regarding conservation 

of resources, but have little impact in terms of the way the general public live, except perhaps to tweak the 

conscience of a few. Part of the reason for this is that there is no formal trust relationship that individuals have with 

the environment, despite general laws or commitments in many countries that are purported to serve these ends.

21 


However, the concepts and legal framework do come into play with regard to the use of resources that are governed 

by governments or other authoritative bodies. The essence of the intergenerational trust is that the present generation 

should harvest the resource so that the capital or principal is not compromised to the point where the next generation 

would be compromised. For example in a fishery, the off-take or catch should not be so great that the remaining 

population will be unable to recover to meet the needs of the future generations. In general humans are in 

competition with each other to acquire a greater proportion of the environment for themselves, paying little regard to 

the needs of others. If altruism between individuals is not a feature of the present generation, it is unlikely that 

individuals of this generation are likely to meet moral responsibilities to the next generation. Therefore, until 

educational programmes and the poverty issues of current generations are such that individuals can afford to be 

altruistic, formal bodies have to assume responsibility for maintaining the environmental capital in trust for the next 

generation. At present this responsibility resides with central government in much of Africa, but there is a growing 

movement towards decentralisation, in which departments and organisations and even communities are taking 

responsibility for managing the resources. Each of these therefore has the moral responsibility of caring for the 

resources of this generation with the eye on providing for the next generation. In Africa, many of the organisations 

or institutes are unable to fulfil their obligations, as they do not have the financial resources or human capacity to 

do so. One solution is for endowment trusts to be developed that do have finances so that institutions have longterm, 

sustainable support. These can include village trusts where communities are charged with the responsibility of 

managing resources with an intergenerational perspective in mind. 

The legal structures and a wealth of experience exists to provide trusts that have the responsibility to provide for 

future generations. Interestingly, the legal structures also exist in which subsequent generations can sue previous 

generations for stealing their heritage and resources if later generations are of the opinion that the resources were 

inadequately cared for and that they are suffering as a consequence. This underscores the need to guard the capital 

carefully. 

Trusts, endowments and similar 

The problem 

As indicated in an outline of the theme of this meeting, poverty in Africa has led to donor driven research and 

conservation initiatives that are subject to stop-start cycles of sporadic funding and are consequently short-term in 

nature. It impedes the career paths of individuals and promotes the brain drain from countries that should retain 

trained capacity. This has led to situations where sustainable programmes cannot be implemented, and long-term 

goals cannot be set nor achieved. 

Needs to solve problems 

The most pressing step is to provide financial security to research stations or conservation units and to communities 

to enable them to develop and progress towards the achievement of long-term visions. This step requires 

independent funding that so reduces reliance on their own and donor governments that financial security is obtained. 

Endowment trusts, or similar secure endeavours offer possible solutions. Such funding, in addition to that of 

governments, could provide the firm foundation on which to launch sustainable programmes and around which to 

organise other, separately or jointly funded projects. Jointly funded projects are envisaged as partnerships between 

the national facility and donors or research organisations or communities. 

Trust funds offer one solution 

The Trust Fund mechanism offers one solution. Trust funds can guarantee a degree of permanent income, and have 

the potential to contribute significantly to the country’s development through supporting institutional development 

and community initiatives, as well as assisting in the planning and development of income generating activities such 

as eco-tourism. 

What is a Trust Fund? 

As outlined by Barry Spergel (WWF, 1993), a “Trust Fund” is a fairly loose and general term to indicate a sum of 

money that is legally stricted to being used for a specified purpose, separate from other funds such as a 

government’s general budget, and is managed by a board of trustees, which holds legal title to the funds. The legal 

form that a trust may take varies. Most of the trust funds that WWF has worked on have been “endowments”, 

meaning that only the interest income is spent each year and not the principal, which remains invested. However, 

there are other forms that a trust fund may take: the entire principal may be used over a fixed number of years, or it 

may be set up as a “revolving fund”, so that resources come into the trust as existing funds are spent .

ENDOWMENT TRUST 

INTEREST 

CAPITAL 

OR 

PRINCIPAL 

CAPITAL 

OR 

PRINCIPAL 

May be used by present generation 

(biodiversity or financial) 

To be conserved/harvested to generate useable 

off-take/interest for future generations. 

In perpetuity if well managed by 

(board of) trustees. This is sustainable 

DIMINISHING TRUST 

A grant invested in which interest plus 

capital or principal are used by trustees for 

as long as the diminishing resource lasts. 

This is not sustainable 

LIMITED LIFE EXPECTANCY 

22 


An endowment trust can have the specific goals of serving the developmental, research, conservation, educational 

needs of a particular institution, in which case it is rather like a long-term project. This is an Institutional Trust. 

Alternatively it may provide the financial basis for a foundation that supports a variety of environmental projects of 

other organisations, acting rather like a donor . For the parks and research stations around the lake, the institutional 

endowment trust is favoured, but a foundational approach may also be desirable to ‘encourage communities’ 

participation and provide them with a sense of “ownership” of the responsibility to conservation. Thus, in certain 

circumstances we would encourage a mixture of both the institutional and foundational approach. 

Use of trust funds for environmental programmes is not entirely new 

National level trust funds for environmental or conservation purposes are an innovative financing mechanism being 

used by more than thirty countries. Also important are international transborder Trust initiatives (e.g. Transfrontier 

Natural Resource Management Areas, Trans Frontier Conservation Areas). 

According to a recent report commissioned by the UNDP/GEF, between 1990 and 1995 over US$50 million has 

been committed to environmental trust funds. A number of donors, including USAID consider an endowed 

foundation as potentially a better way to provide development assistance, WWF has been instrumental in the 

development of conservation trust funds in more than 10 countries world-wide, and in part of an international forum 

to promote such initiatives in a wider range of countries. 

Environmental Trusts are used mainly to cover recurrent costs of parks and protected areas and research institutions 

to support the overall goal of conserving/managing biological diversity, to reduce air and water pollution, to 

strengthen local environmental institutions, to promote sustainable development, to promote education, community 

participation and public awareness, to promote projects that address problems of poverty alleviation. The emphasis 

has been on supporting integration of conservation and human development needs. 

Most such funds are independent of government in terms of their management (though government may be the main 

source of funds), and most include on their governing board, often in a major way, representatives of nongovernmental 

organisations (NGOs) as well as government representatives. 

Goals are to provide in the long-term, preferably in perpetuity intergenerational, secure financial capital that 

provides sufficient interest to enable the conservation of environmental capital to be achieved by nationals of the 

country. There are numerous subsidiary goals, often specific to the circumstances of the area being conserved and 

the needs of the institution managing the trust. 

Advantages 

Environmental trust funds and foundations hold the potential to meet many needs of both donors and recipients. 

They are attractive in that they provide stable financing for many activities that could not be met by national budgets 

or by short-term projects using donor funding. This long-term stability in funding provides a degree of insulation by

23 


ensuring a dependable flow of money not affected by fluctuations in international donor money or government 

allocations. 

When they involve domestic management of assets, these funds instil a sense of ownership and responsibility for 

environmental conservation, build local capacity for financial management, and reduce the dependence on outside 

experts and international organisations. 

By providing assurance of funds in perpetuity or over a long period of time, they enable long-term planning. And 

they give individuals building careers in such fields some assurance that jobs will be available. When tied to longterm 

environmental action plans, funds help donors identify priority areas for funding and avoid inadvertent 

duplication of efforts. By including a range of stakeholders in the governance and management, funds promote 

democracy, strong civil societies, accountability, and consensus building, and they build national capacity for 

moving towards sustainable development. 

Importantly too, in an intergenerational sense, all trustees are forced to focus on long-term conservation goals, are 

driven by the need to achieve these and by the fact that trustees are accountable for the success or failure of such 

programmes. In essence, trusts of this nature powerfully underscore this generation’s commitment to meeting the 

needs of subsequent generations by imposing duties on trustees. 

There are several additional reasons too why local stakeholders can further benefit from nationally managed 

endowment trusts. 

Disadvantages 

A concern of some donors, especially when setting up an endowment is the need to release a sizeable amount of 

money at the start of a project rather then disbursing over time. A series of smaller appropriations could be a 

solution. However, making a large grant at the beginning could be viewed positively by donors as a way of reducing 

administrative costs and assuring programme continuity. In some respects, making a series of appropriations defeats 

the purpose of the endowment as a large sum needs to be invested for at least a year to earn the interest that drives 

the programme. 

Other disadvantages can arise if the government departments see the trusts as competition rather than support. 

Similarly, if the communities are not included they may be opposed to the trust that could be seen as governmental 

not theirs. Trusts can be caught between the two: government perceives them to be too independent: communities 

perceive them to be too governmental. It is essential, therefore, to have adequate buy-in from all stakeholders. 

One of the biggest problems facing Africa is that many donors do not believe that certain countries at least have the 

capacity to manage trusts. Rather than help develop that capacity and to learn from experience elsewhere, such 

donors simply close the door on trusts and continue to opt for short-term projects that they dominate. There are 

lessons from elsewhere from which everyone can benefit. 

Experience from elsewhere 

If trusts offer attractive alternatives to donor driven, stop-start project cycles for the African Great Lakes then it 

would be prudent to examine experience from elsewhere and be guided by the lessons others learned. Several case 

histories indicate that all are different from one another, some have had problems in gaining momentum and 

acceptance, but all seem to be on successful trajectories and as far as can be ascertained from the sample available, 

none has failed. 

Belize’s Protected Areas Conservation Trust (PACT) uses an innovative method of raising funds for the 

conservation and management of protected areas: a conservation fee charged to all foreign tourists as they leave the 

country. This type of free can serve as a model for other countries, because it has been successful in generating 

revenue even at a time of sever budgetary constraints and declining international aid. Since tourism is now the 

world’s largest industry, charging tourists even a modest conservation fee has the potential to generate large sums 

for management of protected areas. 

In addition to innovation, an equally noteworthy feature of PACT is that money generated by the conservation fee, 

which is essentially a tax, goes to an independent legal entity outside of Government (i.e. the TRUST). PACT is

24 


managed by a Board of Directors (TRUSTEES) with an equal number of members from Government and from nongovernmental 

organisations. This represents a new kind of partnership between Government and civil society. 

Ecuador is regarded as one of the richest reserves of biodiversity on the planet, and is internationally known for 

being host to some of the world’s most precious ecosystems, located in its tropical, Andean, and coastal regions, and 

the Galapagos natural park and marine reserve. There are currently no active specialised facilities to finance the 

tremendous need for environmental management in the country. A national environmental fund, still idle, has 

recently been established. The creation of this fund, the Fondo Ambiental Nacional (FAN), presents an interesting 

case study for two reasons. 

One is the participatory nature of the fund’s conception process, intimately linked to innovative ideas for 

environmental financing. The second is those novel sources of financing themselves. The FAN is the result of 

initiative and work by key private and public-sector agents to grasp a historical opportunity for mobilisation of 

significant amounts of domestic resources toward environmental management. 

Given Ecuador’s unique political and economic conditions in the period from 1994 to mid-1996, including 

economic adjustment, armed conflict with Peru, political scandals, and presidential and congressional elections, the 

future of the fund is still uncertain, but its bases have been strongly established. This case study describes the fund’s 

various stages of creation, and the hurdles faced by its initiators, with the intention of placing the FAN’s experience 

in a real context. 

The idea of mobilising funds for the environment from the proceeds of State de-investment and privatisation 

occurred to FAN’s proponents as something almost obvious. Similarly, the idea of mobilising for the environment 

royalties and other proceeds from the exploitation of non-renewable natural resources (oil) – particularly when the 

exploitation takes place within fragile ecosystems – also seemed rather obvious to a broad community of concerned 

individuals in academia, NGOs, the productive sector, and Government. Implementing those “obvious” ideas, 

putting them into practice, is a different story. The critical mass of individuals who propelled FAN into being is 

aware of the resistance and difficulties that the idea will still need to overcome. However, they felt confident that the 

ample endorsement secured by the participatory process of creation will render both FAN and its proposed funding 

scheme viable for implementation under any in-coming Government. 

Jamaica has two National Environmental Funds. The Environmental Foundation of Jamaica (EFJ) was founded in 

1993. This fund began under the U.S. Government’s Enterprise for the Americans Initiative. The fund is supported 

exclusively by payments from the Jamaican Government, representing interest on rescheduled and reduced bilateral 

debt owed by Jamaica to the United States. In addition to receiving direct payment from the Government of Jamaica 

(GoJ), the fund is able to take advantage of the high interest rate regime in Jamaica to earn substantial interest on its 

capital. The EFJ’s resources, resulting from payments on reduced PL-480 and USAID debt, as well as investment 

earnings, are used for grants to NGOs for projects in environment and child welfare. 

The Jamaica National Parks Trust Fund (JNPT) was legally registered in January 1991 and capitalised with the 

proceeds of Jamaica’s first (and the Caribbean’s second) debt-for-nature swap, on Earth Day 1992. The fund is 

managed primarily as an endowment and makes grants from investment income, its principal remaining untouched. 

The Government’s Green Paper on National Parks for Jamaica indicates that the JNPT should be the major vehicle 

for channelling eligible funds (of public of private origin) to the part system. To this end, the JNPT allows for 

donations to be tied to specific projects, parks, or geographical areas. The Government of Jamaica is expected to 

make annual contributions to the Fund. The Government will also help to develop proposals to bilateral and 

multilateral funding sources. 

Both funds were established through the close co-operation of the Governments of Jamaica and the United States 

and the Jamaica Conservation and Development Trust (JCDT), the island’s leading conservation NGO. This case 

study focuses on the National Parks Trust Fund, although occasionally making comparisons with the EFJ. 

In 1992, the Government of Peru passed legislation creating the National Trust Fund for Protected Areas 

(FONANPE) to assure long-term funding for the country’s protected areas. At the same time, the Government 

created PROFONANPE as a private, non-profit organization to administer the fund. Because of the timing of its 

creation, the new national environmental fund faced a complex panorama in terms of its potential to capture largescale 

national resources. 

On the one hand, the possibility of direct Government disbursements into FONANPE was and continues to be 

almost non-existent. Since the election of President Alberto Fujimori in 1990, the Government has followed a policy

25 


of strict fiscal austerity and has cut spending in almost every branch of Government. Today the policy is being 

reaffirmed in a soon-to-be signed IMF Letter of Intent in which the Government of Peru (GOP) commits to cut 

spending in order to increase its primary fiscal surplus, increase domestic saving, and fend off a potential balance of 

payments crisis. The Government has also committed to increased tax collection, but this additional income will go 

to covering the lean state budget and servicing the $26 billion foreign debt. 

As for non-governmental sources of national funding, Peruvians have no tradition of charitable giving to 

environmental or other non-profit organisations and, despite increased economic stability, most Peruvians still 

endure great economic hardship and have little cash to spare for donations. 

But alongside this grim scenario, there is a bright spot in terms of national resource mobilisation. Since 1990, the 

GOP has been engaged in an all-out effort to “reinsert” Peru into the international financial community. A key 

component of this effort has been the negotiation of Peru’s foreign debt, heavily in arrears since the early 1980’s, 

with multilateral, bilateral, and commercial lenders. The Government’s determination to restore Peru’s 

creditworthiness through these negotiations has provided one of the most important sources of funds to FONANPE, 

which has received more than $10 million in debt donations to date. 

Negotiating debt swaps, while not PROFONANPE’s only line of effort is and will continue to be the organisation’s 

primary focus and best hope for reaching its goal of a trust fund totalling $80 million. Other countries that face tight 

budgets and extensive debt negotiations may find in Peru’s experience a useful example of how a bleak economic 

panorama may in fact hide a more hopeful scenario for the creation of a national environmental fund. 

Developing trusts 

Given that the advantages of environmental endowment trust out-weigh the disadvantages, and, given that there are 

precedents of successful environmental trusts from which one can learn, there seem to be compelling reasons to 

develop trust for lake conservation and research, particularly in developing countries. One of the first steps is to 

persuade donors and governments, which are reluctant to follow this route, of its great benefits. To do this, it is 

essential to be very well prepared in terms of defining visions, goals, modus operandi and, in every respect, 

complete thoroughly that the homework has been done. Also buy-in by stakeholders, particularly the village 

communities and sceptical governments, is an essential prerequisite to success. 

How to fund environmental endowments 

Funding can be derived from a variety of different sources. In many cases in which trusts have been formed the 

national government has shown strong commitment by funding directly or indirectly part or all of the trust. 

Probably, it is worthwhile in all cases for national governments to make a partial contribution as this is an indication 

of sincerity and commitment that would provide additional leverage for fundraisers to use when encouraging others 

to support the initiative. Even the poorest of governments can make contributions as they certainly represent an 

investment in the future of its people and the sustainability of the resources on which they depend. 

There is a broad spectrum of donors that might fund trusts (major bilateral and multilateral donor agencies), USAID 

and the World Bank/Global Environmental Facility, philanthropic donors such as the MacArthur Foundation and 

international conservation organisations including The Nature Conservancy and WWF. Though some of those that 

did so in the past are no longer keen on this route, favouring instead a return to stop-start projects over which they 

can exercise greater donor control. 

How to manage environmental endowments: Investing capital 

The financial rationale of an endowment is to protect the value of assets while generating income from investments. 

Private donor organizations and bilateral and multilateral aid agencies tend to focus on short-term projects whose 

impact often dwindles after the money runs out. For these agencies, endowments offer a way to sustain commitment 

to long-term goals insulated from budgetary and political fluctuations. Operating an endowment, however, 

necessitates acquiring and managing assets in a way that maintains their value and produces income. Ideally, both 

the capital and the interest should grow over time if well managed, so that inflationary trends and growing needs are 

met. Many trusts have grown substantially since their initiation . 

Management in perpetuity 

Regardless of conditions in local financial markets, building a financial management capacity will pay off when the 

financial markets develop to the point that an endowment can be managed in perpetuity. An endowment’s financial

26 


managers should anticipate the various constraints on the handling of an endowment’s assets, including gift 

restriction imposed by donors, reporting requirements of national governments and donors, and the need for 

transparency and for avoiding conflicts of interest in asset management decisions. An endowment that proves able to 

deal with these constraints will be attractive to potential donors. 

The issue of investing within the nation in which the trust is to be launched, or offshore, or in both, has to be 

assessed for each case. Returns on interest are often better in African nations, but the devaluation of currencies and 

the potential for political instability, even in what are apparently strong economies (vide recent events in Zimbabwe) 

has to be carefully analysed by the investment house employed to manage the finances so that the trustees are wisely 

informed. 

It is an advantage for the fund to be independent and outside of government. This increases donor confidence that 

the funds will not be used inefficiently or redirected to other government concerns. It also makes the fund 

independent of political changes as when an election has taken place. Some donors, for example USAID, can only 

give funds to independent non-governmental organisations, or to funds where the management board has a majority 

of NGO representatives. However, the Trust needs to have a good relationship with government and be in line with 

government policy in order to solicit the political good will to make the fund functional. It is also clear that the fund 

cannot replace the government’s normal budgetary commitment to environmental conservation activities, but must 

support them. 

Long-term investments for nature, research, education 

Endowments are defined as assets that are invested to earn income to be used in support of a defined purpose. 

Funds can take several different forms (a trust set up by special legislation, a foundation, a common law trust, a nonprofit 

corporation) but nearly all take the form of an endowment. 

Answerability and accountability 

Appointment of the board of trustees should follow the legally established procedures and should benefit from the 

wealth of experience that exists already (see case histories). Importantly, the trustees have to be fully accountable for 

their performance in both the achievement of environmental goals and in financial management of the capital and 

interest, preferably showing an annual growth of both. 

Conclusion 

Long-term research and monitoring programmes and the achievement of conservation objectives are so dependent 

upon programmes that are driven by the nationals of the countries concerned, and funded in a secure manner for as 

long as funding is required, that the establishment of trust funds or similar forms of sustainable secure funding 

presents an attractive option. Perhaps is it is necessary for those donors that are opposed to the establishment of 

investment funds to re-examine the situation and reconsider their policies. Ideally, to understand and manage the 

lakes, the programmes should have perspectives that embody the phrase “in perpetuity”. And this phrase should be 

applied to both the nature of the work and the financial resources that support it. Whether in the context of lake 

Malawi/Niassa/Nyasa each station or park, such as a Kyela, Metangula, the lake Malawi National Park, Senga Bay 

and others should have independent trusts, each managed by their own trustees, or whether there should be a multinational 

Trust for the entire lake are issues that need not be debated in this paper. What is clear is that trusts 

managed in perpetuity do offer a compelling alternative to the stop-start cycles of donor projects.

27 


The need to maintain Maximum Biodiversity in Lake Nyasa 

Benjamin Peter Ngatunga 

Senior Research Officer (Ichthyology), Tanzania Fisheries Research Institute, Kyela Centre, P.O. Box 98, Kyela, Mbeya, Tanzania, 

East Africa 

Abstract 

Lake Nyasa fish fauna comprises more than 1 000 species, mainly cichlids, and is thus unique among freshwater lakes. But tension 

exists between motives either to conserve or exploit the fish resource. In Lake Nyasa small scale fishery exploits the inshore 

resources, principal fish taken are mainly cichlids which comprise up to 90% of the total landings. Cichlids have low fecundity and 

are k-selected therefore have low rates of recovery in face of stock collapse. Thus what is happening today in Lake Nyasa has great 

consequences for tomorrow. Such a scenario provides the motive and desire for conservation of the Lake Nyasa fish fauna. We 

care about biodiversity and we wish to maintain or conserve it because we see value in biodiversity. The term, biodiversity has 

become so much of a buzzword dominated global environmental debate, that it is often difficult to focus on the critical issues that 

face society. The six categories of biodiversity values relevant to Lake Nyasa are presented and discussed and the need to maintain 

biodiversity in Lake Nyasa is deliberated using a poverty degradation feed-back cycle. 

Introduction 

Biodiversity – a new gimmick? 

According to Ghilarov (1996), the term biodiversity appeared in ecological literature in the middle of the 1980s and 

in less than 10 years it became so popular that for the external observer the whole situation looks like a successful 

breakthrough in some new field of science. Today biodiversity has become an everyday word not just for ecologists 

but for politicians, the media and the general public too. It seems, however, that the real reason for this burst in 

interest in biodiversity (or rather this rapid increase in the usage of the term ‘biodiversity’) lies outside the sphere of 

science. To many professional ecologists, perhaps it sounds offensive, yet I shall dare to claim that the term 

‘biodiversity’ became so popular, primarily because it’s usage facilitates greater access to research funding. So we 

might consider ‘biodiversity’ as a useful fad that helps many ecologists to survive doing science. 

The term ‘Biodiversity’ has become so much part of the buzzword dominated global environmental debate that it is 

often difficult to focus on critical issues that face the society. Biodiversity can be defined in its narrow sense as: 

“The variety and variability among living organisms and the ecological complexities in which they occur” (FAO, 

1994). Biodiversity can be measured in terms of ecosystems, species and genes (Whittaker, 1972, Harper & 

Hawksworth, 1994). 

But a more relevant view of biodiversity, especially to African societies comes from its larger resource oriented 

sense. “In Africa biodiversity is a matter of survival, it is critical for life at grass root level. It is a total variety of 

living matter on which society depends. It provides ecosystem resilience, to allow both people and natural 

communities to cope with periodic environmental stress” (FAO, 1994). 

Why care about biodiversity? 

We care about biodiversity and we wish to maintain or conserve biodiversity because we see value in biodiversity. 

This brings us back to the starting point of value to which part of the society. The debate on biodiversity revolves 

around two opposite poles of the society. 

At one end is the developed western world interested in global values and the options values of biodiversity. It was 

the western world, which devised the funding mechanism such as the GEF to address these global values. At the 

other end are the local communities in the developing world, who depend on the biological resources of natural and 

man modified ecosystems for their livelihood. 

We thus have polarisation of North versus South, Central government versus local government and western 

protectionism versus local utilisation. 

Conservation International stresses that there are mainly five categories of biodiversity values. 

1. Major ecosystem functions (catchment, soil formation, nutrient cycling). 

2. International export values (timber, beeswax, tourist spectacles, and medicines). 

3. Regional and local market values (food, poles and medicine). 

4. Household use values (food, cultural, construction, and fuel).

5. Global intangible values (carbon sequestration, science). 

28 


These values are, however, not all mutually exclusive, for example ‘mbuna’ stocks maintained for international 

export value will also provide global benefit. For developing countries like Malawi, Mozambique and Tanzania, 

emphasis for conserving the stocks of ‘mbuna’ is based on their utilitarian export values, not their long lists of Latin 

names of endemic species. 

Lake Nyasa, a biodiversity ‘hot-spot’ 

The global water reserves comprise of 97.5% salt water and 2.5% fresh water. Surprisingly, only a tiny percentage 

(less than 0.01%) of the global total water reserve is comprised of rivers and lakes. The remainder is ground water, 

glacier and permanent snow and soil moisture, permafrost and swamp water. Yet this tiny percentage holds a 

tremendous diversity of life. Of the 20 000+ species of fishes described so far by scientists (and many more remain 

to be described), 40% live in freshwater. In other words, almost half of the fish biodiversity is contained within a 

very vulnerable 0.01% of the earth’s water. 

GLOBAL WATER 

RESERVES 

FRESH 

(2.5%) 

SALT 

(97.5%) 

2.5% 

0.9% 

Soil moisture, 

permafrost, swamp 

water 

69% 

Glaciers and 

permanent snow 

cover 

0.3% 

Rivers and lakes 

(

29 


The importance of Lake Malawi cichlids as a source of protein cannot be over emphasised; they are a crucial 

resource both as natural populations and subjects of fish aquaculture. Most of Lake Malawi cichlids are small and 

therefore can easily be dried in the sun for upcountry markets. 

Under the circumstances of hunger and poverty, fish conservation is both exhilarating and controversial. 

Exhilarating because those fishers bear the collective wisdom, wit and will to control their resource they depend on 

for survival. Controversial because the need to survive drives them to fish to the last fish! Thus what is happening 

today in Lake Nyasa has great consequences for tomorrow. 

Threats to Lake Nyasa fish biodiversity 

Habitat variability within the lake must have played an important role in sustaining the diversity of fish and 

therefore, preserving the fish biodiversity of Lake Nyasa is not possible without habitat protection of those areas of 

the lake where most species depends on. The most important and also most delicate habitats are the littoral zones of 

the rocky habitats, the shallow water zones associated with river mouths – swamps, submerged macrophyte beds, 

lagoons are important for spawning and nursery grounds of many fish species. 

The main threat to the existence of high diversity of life in Lake Nyasa comes from human activities on the adjacent 

land. The root causes of the lake’s environmental damage that can result in biodiversity loss are of two categories: 

1. Land based causes/sources and activities affecting the quality and uses of the lake (poverty, poorly 

managed social and economic development programs and unsustainable consumption patterns). 

2. Of the land based activities I consider the effect of fishing, both direct and indirect to be of highest priority 

comparable to that of the most serious land-based threats. 

ENVIRONMENTAL 

DEGRADATION 

LONG-TERM EFFECTS 

Loss of: 

Biodiversity 

Habitat 

Ecological 

Functioning 

Agricultural Land 

Fishing Potential 

Tourism Potential 

Opportunity: 

REDUCED 

CAPACITY 

FACILITATE DIVERSIFICATION: 

e.g.: TOURISM 

Poverty / Degradation Feedback Cycle 

INCREASED 

POVERTY 

SHORT-TERM EFFECTS 

Deforestation 

Overgrazing 

Overhunting 

Overfishing 

Pollution 

Erosion 

River Siltation 

Increased Flooding 

FEEDBACK 

ROOT CAUSES 

Alternative Cycle 

REDUCE 

INCREASE DIVERSIFICATION 

& RESOURCE RETURNS 

Poverty 

& 

hardship 

POVERTY 

BEGIN 

HERE 

Not enough 

Capacity 

Reduced 

sustainability 

Increased impacts on 

environment 

Opportunity: 

DEVELOP CAPACITY 

IMPROVED RESOURCE 

MANAGEMENT 

Strategy 

Indeed the need to maintain maximum biodiversity is surely greatest as the rates of environmental change increases. 

The strategy to maintain maximum biodiversity of this fish resource that the local population mainly depends on

30 


should be advocated together with the strategy to alleviate poverty and hunger. An example in this case is the 

provision of an alternative source of cheap protein. 

References 

FAO, 1994. Biodiversity. News letter, East African Biodiversity Project: UNO/RAF/006/GEF. 

Stiassny, M.L.J., 1993. What is a Cichlid? Tropical Fish Hobbyist (TFH, November, 1993), 141-146. 

Ghilarov, A. 1996. What does ‘biodiversity’ mean – Scientific problem or convenient myth? Trends in Ecology and Evolution 

Vol. II, 304-306. 

Harper, J.L & Hawsworth, D.L. 1994. Biodiversity: measurement and estimation. Philos. Trans. R. Soc. London. Ser. B 345, 5- 

12. 

Marcel, A., Brechignac, F. & Thibault, P. 1994. Biodiversity in model ecosystems. Nature 371, 565. 

Ngatunga, B.P. 2000. A taxonomic revision of the shallow-water species of the genera Lethrinops, Tramitichromis and 

Taeniolethrinops (Teleostei, Cichlidae) from Lake Malawi/Nyasa/Niassa (East Africa). PhD thesis, University of 

Rhodes, South Africa, 293pg. 

Stiassny, M.L.J. 1993. What is a Cichlid? Tropical Fish Hobbyist (TFH, November, 1993), 141-146. 

Whittaker, R.H. 1972. Evolution and measurement of species diversity. Taxon 21, 213-251.

31 


Fisheries Management, Biodiversity Conservation and Genetic Stock Structure 

George F. Turner 

Department of Biological Sciences, University of Hull, HU6 7RX, United Kingdom, E-mail: g.f.turner@biosci.hull.ac.uk 

Abstract 

Given the immense difficulties of assessing and managing fisheries on Lake Malawi/ Nyasa, it seems likely that the present 

approach of using intuition and rules of thumb to derive management approaches is likely to continue. It is important that the rules of 

thumb employed are as accurate as possible. At present, almost nothing is known of the migration patterns and movements of the 

major fish stocks in the lake. Both the collection of data and the setting of effort limits are presently based on rather arbitrary fishing 

areas. Local fisheries management will be largely irrelevant if carried out on wide-ranging stocks that are exploited elsewhere. 

Analysis of population structure provides an effective tool for the assessment of fish movement patterns. Tagging is likely to be 

impractical for most lake fish stocks. Species identification is important, but on its own, provides limited geographical resolution of 

stock structure. Molecular approaches using microsatellite DNA are powerful, but are critically dependent on accurate species 

identification. Examples are given of studies implemented on Lake Malawi. A critical advantage of this kind of study is that the 

information it provides need not be updated by regular monitoring, but can be used to inform management decisions indefinitely. 

Introduction 

Assessment and management of fishery resources on Lake Malawi/Nyasa is not easy, and there have been few 

success stories to date. Difficulties arise from (among other things) the sheer size of the lake, the huge number of 

species exploited, the lack of knowledge of fish biology and identification, the diversity and decentralisation of 

fishing activities, the lack of knowledge about the factors influencing fisher behaviour and lack of funding. Little is 

presently known about the movements of fish within the lake, which makes it difficult to define the geographical 

extent of stocks for the purposes of assessment and management. This raises major practical problems in the light of 

decentralisation of responsibility for management of fish stocks, with political authority devolving to district levels 

and with increasing emphasis on local co-management through beach village committees. A further problem lies in 

the assessment of gear and fleet interactions. This is likely to become a more pressing problem, as there are pressure 

to grant more trawl licenses at the same time as artisanal fishers are increasing the use of chilimira nets to target 

offshore benthic species such as chambo (‘kauni’ fishery). 

A ‘stock’ is what fisheries managers try to manage. A group of organisms can be treated as a stock if possible 

differences within the group and interchanges with other groups can ignored without making the conclusions 

reached invalid (Sparre et al. 1989). Why should fishery managers care about stock definition? If a stock is 

overexploited, managers will wish to decrease fishing effort on that stock. This would be of limited use if another 

fishing fleet continues to increase its rate of exploitation of the same stock. Is there any point in the allocation of 

property rights to a local community to enable participatory management of a stock in one area, if rampant 

overexploitation of the same stock is going on somewhere else, or going on in the same area using a different 

method? Equally, if a stock is underexploited, this may be because it is difficult to exploit economically in part of its 

range. If such a stock shows a high degree of migration over its whole range, it might be worth permitting extremely 

heavy exploitation in the areas where it can be economically exploited, in the knowledge that local depletion of the 

population can be compensated by immigration from the rest of the range. Knowledge of stock limits and gear 

interactions is also critical in stock assessment, particular using catch and effort data. For example, there is no point 

in carrying out separate assessments for Salima and Nkhotakota districts if all the main fish species move freely 

between the two areas. If this were the case, the catch and effort data could be pooled and this might not only allow 

more accurate assessment, but might even allow for a reduction in data gathering effort, allowing valuable resources 

to be reallocated elsewhere. Thus, definition of the major fish stocks in Lake Malawi is an essential stage in the 

development of a rationally-based sustainable fisheries management programme. 

Studies of population structure can also provide critical information for conservation biologists. If a local population 

in a reserve area exchanges few migrants with neighbouring populations, its protection is more important for two 

reasons. Firstly, it is likely that if the local population is reduced or goes extinct, immigration from elsewhere will 

do little to help the population recover. Further, it is likely that if migration rates are extremely low, the local 

population may contain unique genetic variants that will not exist in other populations. These variants may include 

genotypes specifically adapted to the local habitat, and if so, immigrants from elsewhere may never become as 

successfully established as the former residents. This can also be critical if restocking is attempted using source 

populations from other localities.

32 


The present paper is aimed at summarising the uses of molecular methods in defining exploited stocks for fisheries 

management. I will also briefly discuss the uses of molecular methods in conservation biology, and discuss other 

methods which might be used to study population structure. 

The basics of molecular ecology 

DNA 

Living organisms are constructed from recipes coded in their genes, made from DNA molecules, passed on from 

their parents. DNA is made up of various subsidiary molecules linked together, but the ‘message’ is spelled out in 

the sequence of 4 different kinds of molecules called ‘bases’, labelled A,C,T and G. An important feature of DNA is 

that it can make copies of itself, with the help of the enzyme DNA polymerase, which is itself made from the DNA. 

The messages in DNA are translated into RNA, which is in turn translated into proteins, which build bodies and 

transmit information around the body. In the course of sequencing the human genome (the genome is the total 

complement of DNA in an organism), it has been discovered that more than 95% of our DNA is not in fact used to 

make proteins, and indeed is not used for anything at all (International Human Genome Sequencing Consortium 

2001). Much of this ‘junk’ DNA is mainly made up of bits that are good at making copies of themselves and getting 

them stuck into other parts of the genome. As long as they do not interfere with the workings of the useful bits of 

DNA, they tend to hang around as a sort of harmless parasite. This is not really as surprising at it sounds, because 

organisms are not ‘designed’ to be efficient, they have merely proved better than their rivals at surviving and 

reproducing over billions of years of evolution by natural selection. 

The discipline of Molecular Ecology, of which Molecular Fish Stock Discrimination is a subdiscipline, is largely 

based on the analysis of this ‘junk’ DNA. Most commonly, molecular ecologists try to look at the same bit of DNA 

(a ‘locus’, plural ‘loci’, meaning ‘place’) in a number of different individuals. There will often be slight differences 

between individuals in the sequences of the DNA at the same locus- these different forms are called ‘alleles’. And 

because each individual has two equivalent chromosomes each containing the same set of loci (ignoring exceptions 

like the sex chromosomes and mitochondrial DNA), then each individual can have two different alleles (in which 

case it is said to be ‘heterozygous’) or two copies of the same allele (‘homozygous’). The trick is to find the same 

locus in each individual in the study. To do this, special ‘primers’ are used to latch on to unique sequences of DNA 

on either side of the region of interest. From these, chemical reactions (PCR or polymerase chain reactions) are 

carried out to make lots of copies of the bit of DNA that lies in between these primers, until there is enough material 

to analyse. The next step is often to work out the sequence of letters in the target locus, by smashing up the molecule 

into lots of little bits and then adding a different fluorescent marker onto the end according to which ‘letter’ is the 

last one. By working out the size of all the bits you can work out the whole DNA sequence. Alternatively, the target 

locus might consist of a very repetitive message (such as GTGTGTGTGTGT… and so on), in which case it is better 

just to see how long the message is. These procedures are increasingly carried out using automated DNA 

sequencers. 

Genetic divergence of populations 

If any groups of sexually reproducing organisms stop interbreeding with each other, they will begin to diverge 

genetically. There are three possible causes for this: selection, mutation and genetic drift (Avise 1994). In recently 

separated populations, drift is likely to be the major force, particularly when populations are small. Genetic drift 

results from the random process of chance differences in survival and reproductive success of individuals with 

different genotypes. This can happen even if these genotypes are equally good at surviving and breeding. Molecular 

estimates of migration rate assume that differences between populations are not due to natural selection. This seems 

a robust assumption for most populations which have become separated fairly recently (hundreds or thousands of 

years), but to be more sure of this, it is best to examine several different molecular markers which are selectively 

neutral and unlinked. By selectively neutral, it is meant that the locus is not involved in making anything useful, like 

a protein molecule. But, there is always the risk that the locus is next door (‘linked’) to something that is useful and 

has recently been subjected to strong natural selection. The risk of this can be minimised by looking at several 

different loci for each individual, which carries the extra advantage of allowing you to get statistically independent 

estimates of migration that can be averaged over to make the results more accurate. Thus, estimates of migration and 

population structure are based on statistical estimates of allele frequencies in different loci in many individuals from 

different populations.

33 


Microsatellites 

Many different molecular marker types exist, such as allozymes, mitochondrial DNA, RAPDs, AFLPs, 

minisatellites etc. Most have been used in studies of population structure. The advantages and disadvantages of 

different kinds of molecular markers are discussed in many excellent review articles and books (e.g. Avise 1994; 

Jarne & Lagoda 1996; Carvalho & Hauser 1998; Mueller & Wolfenbarger 1999; Sunnucks 2000). For studies of 

population structure the general consensus is that the method of choice is analysis of microsatellite DNA. 

Microsatellites are also sometimes called SSRs or Simple Sequence Repeats (e.g. Markert et al 1999; 2001) or, 

along with other larger repeats such as ‘minisatellites’ they are classed as VNTRs or Variable Number Tandem 

Repeats (Majerus et al. 1996). Microsatellites include the monotonous 2-letter repeats already mentioned (GTGTGT 

etc), although sometimes a motif can be as long as 6 letters (ACTGATACTGAT etc), and others can be a bit more 

complex with interruptions of the repeat by non-repetitive sequences (ATATATCGATATATAT etc). It seems that 

the short repeat motifs are difficult to copy accurately, and extra repeats are frequently inserted. Thus, 

microsatellites tend to evolve rapidly (and generally grow in length). This rapid mutation rate, coupled with their 

selective neutrality means that microsatellites tend to be incredibly variable and thus can provide good material for 

statistical comparisons between populations. 

Statistical assessment of population structure 

Most estimates of population structure employ variants of the F-statistics first developed in the 1950s (Wright 1951, 

see, for example Avise, 1994, for an explanation). For population structure, the key measure is FST, which is the 

variance in allele frequencies among populations, standardised as a proportion of the maximum possible for the 

given range of allele frequencies. Basically, it is a measure of the differences between populations in comparison to 

the overall amount of differences between individuals, including the differences between individuals within the 

same population. FST can range from 0 to 1. A high FST indicates a high degree of genetic differentiation between 

populations, and thus a low level of migration. Taking several assumptions into account, it is possible to estimate the 

number of successfully breeding migrants moving between populations per generation (Nm). This figure can also be 

estimated from the number of unique alleles in each population, adjusted for sample size (the ‘Private Alleles 

Method’: Slatkin 1985). Rather surprisingly at first sight, these methods give an estimate of the number of individual 

migrants, not the proportion migrating. There are a number of software packages to calculate FST, Nm and other 

parameters from raw data on microsatellite allele frequencies: most are readily available over the internet. A good 

starting point is the molecular ecology web site at the University of Hull. 

http://www.hull.ac.uk/molecol/nonmol.htm#software 

Molecular stock structure analysis and species identification 

Genetic analysis of stock structure is incredibly sensitive to correct species identification. This is rarely mentioned in 

the literature on the subject, but then in most of the world apart from Lake Malawi and Lake Victoria, it is quite easy 

to identify species of fish, so the question rarely arises. Confusion of closely-related species is just as dangerous as 

confusion of distantly-related ones. No studies of population structure using molecular methods are of any value if 

there is a risk that species might have been confused. The most likely result is that different proportions of different 

species will have been collected in different populations. The consequence of this would be to inflate the estimates 

of genetic differentiation between populations, and to give the impression that migrations levels were lower than 

they actually are. Recent studies have indicated that Lake Malawi cichlid fish cannot be told to the level of species 

on the basis of unique genetic markers (Parker & Kornfield 1997; Turner et al. 2000). So, there is no point in hoping 

that genetic studies in the laboratory can sort out mistaken species identification during sample collection. Properly 

labelled voucher specimens are thus a vital for any molecular ecology study where there might be the slightest doubt 

over species identification. 

Protocols for sample collection for microsatellite studies 

Table 1 gives a suggested protocol for sample collection for microsatellite studies of Lake Malawi fishes. It is 

assumed that there may be some uncertainty over species identification. In practice, it would be advisable to consult 

with leading experts in species identification prior to undertaking the study. This might be helpful in deciding 

whether to focus the collection of ripe males, or whether any stage of fish would do. 

Making labels 

A major cause of problems can lie in labelling of whole fish specimens. Probably the best method is make labels 

from Dymo tape (www.dymo.com). From the manufacturers, a basic manual dymo printer costs $6.80 with spare 

12-foot long tapes costing $6.29 for 3. A single tape is enough to label dozens or even hundreds of individual fish. 

Dymo tapes and printers are sometimes available in Malawi. The more sophisticated electronic dymo printers are

34 


inferior for this purpose, as the labels are not so strong. A greatly inferior substitute is to use relatively rigid card, 

such as sold for file card boxes, and write labels using alcohol-resistant ink, such as is used in the Rotring 

Rapidograph ink drawing pen, http://www.rotring.de/www.rotring.com/index.html. The next problem is to 

attach the label to the specimen. The best method is to use a purpose-made fish-tagging gun loaded with T-bar tags. 

Information on fish tagging materials can be found at http://www.hafro.is/catag/. The best known supplier is 

Floytag, at http://www.halcyon.com/floytag/. The labels can be individually numbered or coded, but this is a lot 

more expensive, and it is probably better to make Dymo labels. Fish tagging guns will have a sharp hollow spike, 

designed to be pushed though fish muscle. To attach a Dymo label to a tag, the spike is simply pushed through the 

label before being pushed through the fish. A click on the trigger, and a strong plastic tag is inserted through the 

fish, held in place by the T-bar, which springs open on the underside. Tags are probably best inserted through the 

caudal peduncle above the lateral line (in case counts of lateral line scales are a useful taxonomic feature). In smaller 

fish, tags might have to be inserted through the main body of the fish above the lateral lines, to avoid breaking the 

caudal peduncle. If a tagging gun is not available, string or fine monofilament twine may be used to tie labels firmly 

round the caudal peduncle, or it can be threaded through the operculum and out the mouth with both ends tied 

together. Neither solution is ideal, as string tends to break and monofilament knots often come undone. Either can 

get caught among the fins and opercula of other fish. This can cause string to break, or the label to get sheared off, 

as fish are put in and out of the storage barrel. If the label has to tied on, it is wise to insert a duplicate label into the 

fish’s mouth and perhaps even another under one of the opercula. The alternative to trying to attach labels is to place 

each fish into a separate plastic bag or jar, which also contains a label. Use of jars is usually impractical in the field. 

If plastic bags are used they should be well filled with preservative, for otherwise the fish may not get enough 

contact with the formalin for proper preservation. 

Table 1. Suggested protocol for sample collection for microsatellite DNA. 

Step Procedure Notes 

1. Select populations for study. The sample area at each site 

should be as small as 

possible, given the need to 

collect sufficient samples 

2. Aim to collect a minimum of 50 live or recently killed individuals from 

each site. 

3. Cut off a piece of a fin and wash off any excess scales and blood (which 

might be from different fish). Place in a vial of 90-100% ethanol. Wash 

scissors before taking next sample. 

4. Label the vial, using alcohol-resistant ink. Best of all, as well as labelling 

the vial, put them into vial boxes where the rows and column are 

already labelled on the box. 

5. Take a colour photograph of each specimen. Before taking the picture, 

write the vial label in large bold characters on a piece of paper and 

place this next to the fish, so that it appears in the photograph. 

6. Label the fish specimen using the same label code as for the vial and 

photo and place in 10% formalin. 

7. Send duplicate copies of all of the photographs to one or more leading 

experts in identification of the fish species concerned, explaining the 

purpose for which the specimens are required. 

8. If required, send whole fish specimens to the identification expert, or 

wait until (s)he has the opportunity to visit your lab to examine the 

material. 

9. Remove from the analysis any sample for which there is the slightest 

doubt over the species identification. 

100 is better. If samples from 

most sites are over 50, the 

occasional population sample 

of 30-40 could be OK. 

If the whole specimen is to be 

preserved, it is best to remove 

the right pectoral fin, allowing 

taxonomic measurements to 

be taken from the left side. 

NB if you rely solely on row 

and column designations on 

the box, and don’t label the 

vials, it is a major disaster if 

you drop the box while the lid 

is open! 

A coarse black marker pen is 

ideal for labelling, and stiff 

card, as used for file card 

boxes. 

Various methods are 

discussed below. 

It would be useful to include 

information on the collecting 

locality for each photograph 

when seeking advice.

35 


Preserving Fish 

A frequent problem with collections of fish specimens made for ID purposes is that they are often bent and twisted, 

and thus very difficult to identify. The main cause of this is packing fish in tightly while they are still soft and fresh. 

It is always best if the fish are placed into a barrel with enough room for them to float freely in the liquid. Ideally, 

the fish’s fins should be open, rather than closed tight against the body. The best way to do this, after killing fish 

with an anaesthetic overdose, is to place them on a sheet of expanded polystyrene (balsa wood would also do, or a 

metal tray filled with about 1cm of candle wax), open out the fins, and pin them in place. Strong formalin is then 

drizzled from a syringe over the fin bases to fix them in place. Fish are best left like this for 10-15 minutes before 

placing them in the formalin bucket. If the fish have asphyxiated (often the case on deck of a trawler), the mouth 

may be fully extended. This also makes identification difficult. When pinning the fins, it is best to also pin back the 

mouth so that it is only slightly open. It is difficult to inspect the teeth, which can be an important character, if the 

mouth is fully closed. With fish brought up quickly from deep water, it is useful to puncture the stomach and 

swimbladder and squeeze out excess air, to try to combat the bloating of the fish resulting from swimbladder 

expansion. A final problem is with decomposition. Large fish should have formalin injected into the gut cavity, 

using a large syringe. The quantity should not be so much that the fish becomes bloated, a normal-looking shape is 

the aim. Very big chambo or ncheni can also benefit from a few injections into the dorsal musculature, carefully 

inserting the needle under a scale, because if you drive it right through the scale, it will come off as you remove the 

needle. It is worth remembering that formalin is diluted by fish as well as by water. The aim is get a 10% formalin 

solution in contact with the fish tissues. If there is more fish than water, then concentration of a 10% solution will 

soon be less than 5%, and fish will start to rot. Another problem with overpacking of fish is that parts of the bodies 

can end up sticking out of the liquid. These can quickly rot. Containers of preserved fish should be inspected 

regularly to maintain levels of liquid, which can drop through evaporation or spillage. Formalin is poisonous, and it 

is extremely unpleasant to inhale the fumes. Before any detailed examination of the specimens is carried out, the 

formalin should be washed out under a running tap (left outdoors for a couple of days) and then the specimens 

should be kept in 70% alcohol. But it is worth waiting until you have checked the quality of your photographs before 

you do this (see below). 

Photographing fish specimens 

I have often been surprised at the poor quality of photographs of fish I have been asked to identify. A common 

problem is that the fish is photographed from an oblique angle, usually because it is high up on a table at the time. 

There is no substitute for getting the specimen onto the floor! If the fish is very big, I often stand on top of a stool to 

get a good aerial view from far enough away. In Africa, the bright light often casts dramatic black shadows over the 

fish. I prefer to photograph fish in the shade where possible. There is usually enough light that a flash is not needed. 

In the open, I usually manoeuvre my own body to cast a full shadow over the fish. An umbrella might come in 

handy for big specimens. Inevitably, there is a risk that photographs will not come out or be over- or under-exposed. 

A digital camera allows you to view your photos immediately and see if they are of acceptable standard. Generally, 

they do not store many images, so it is useful to take backup disks, if it is not feasible to download them directly into 

a computer. If a digital camera is not available, a fallback position in the event of your photos being useless, is to 

photograph the preserved fish. Some element of the live colour is usually still visible if the fish has been kept in 

formalin, in the dark. However, alcohol rapidly bleaches specimens, especially if they are kept in sunlight. So, it is 

advisable to keep specimens in formalin until the quality of photographs has been determined. Live colours often 

fade quickly after death, so it is often useful to take a quick snap of the live fish first, and another picture once the 

fins have been pinned out in order to get a better impression of the shape and finnage. 

Case studies of stock structure analysis in Lake Malawi 

A potential problem with microsatellites is that new primers usually have to be developed for each species being 

studied. This can be a laborious and expensive process. For once, the fact that most of the cichlids are so closely 

related proves to be a plus rather than a minus, since it seems that most primers work across all the haplochromines. 

About half of the primers developed for tilapiine cichlids work on mbuna. Fortunately, Prof. Kocher at the 

University of New Hampshire, USA, has been engaged from several years in developing hundreds of sets of 

microsatellite loci to map the genomes of Nile Tilapia for aquaculture research (e.g. Lee & Kocher 1996, Kocher et 

al. 1998, see also http://tilapia.unh.edu/Default.html) and mbuna for evolution research. Consequently, there is no 

shortage of microsatellite primers available.

36 


Case Study I: Population structure in mbuna 

The first study to use microsatellites to investigate the population structure of Lake Malawi cichlids resulted from a 

research project on the evolution of mbuna, carried out by a consortium of researchers in British Universities, in 

collaboration with the University of Malawi. We wanted to test the hypothesis, dating back to Fryer (1959), that 

there is little migration between populations of mbuna separated by deep-water channels or sandy bays. An earlier 

study by McKaye et al. (1984) had looked at this question by comparing Pseudotropheus zebra from Nkhata Bay 

and Chilumba with Mumbo Islands and Domwe Islands in the southern part of the lake. Although this study clearly 

showed differentiation between the northern and southern populations, it employed allozyme electrophoresis, a 

comparatively crude method which we thought unlikely to demonstrate population differentiation at the smaller 

spatial scales we were concerned with. We used 6 microsatellite loci to compare populations of 4 species: 

Pseudotropheus zebra, P. callainos, and two undescribed members of the P. tropheops complex (P. ‘tropheops 

olive’ and P. ‘tropheops mauve’). We looked at 4 populations from headlands around Nkhata Bay- each headland 

was separated from the next one by a sand/mud bay of 700-1,400 metres in width. Since female P. zebra are similar 

in colour to P. ‘gold zebra’ which is also found at Nkhata Bay, we only sampled males of this species, which are 

easily identified at this site. Our analysis yielded FST values of 0.007-0.016, very small, but statistically significant. 

The number of migrants between sites was estimated to range from 6-8 individuals per generation (van Oppen et al. 

1997). This showed that there is very little movement of mbuna across sandy bays- even very small ones. 

Subsequent studies from Kocher’s lab (and collaborators) have broadly confirmed the pattern from mbuna of other 

genera from the southern part of the lake (Arnegard et al. 1999; Danley et al. 2000; Markert et al. 1999). 

Case Study II: Population structure in pelagic cichlids 

The ODA/SADC Pelagic Fish Resources Project (Menz 1995) estimated Lake Malawi’s offshore pelagic fish stock 

at 168,000 tonnes, of which around 88% is comprised of cichlids (Menz et al.1995). The maximum sustainable yield 

was estimated at around 30,000 tonnes per annum (Thompson 1995). With virtually no fishing presently undertaken 

offshore, important tasks remaining were to estimate the extent to which pelagic stocks could be exploited in inshore 

areas, and to estimate the extent to which exploitation in a limited geographical area might run the risk of 

exterminating local stocks. We addressed these questions through molecular stock structure analysis, during the 

DFID Ncheni Project (Turner et al. 2000). We used six microsatellite loci to screen 5 populations (479 individuals) 

of Diplotaxodon limnothrissa, a surface pelagic species also found inhabiting inshore waters, 3 populations (234 

individuals) of Diplotaxodon macrops, more benthic species found near the bottom on deep shelf areas and 3 

populations (234 individuals) of Diplotaxodon ‘offshore’, a deep-water pelagic species. The status of the latter 2 

forms was then uncertain, since they are very difficult to distinguish in the field, and might have proved to be 

populations of the same species. This was later resolved when we discovered that ripe males of D. ‘offshore’ had 

different breeding colours from D. macrops. We had initially intended to look at population structure in 

Rhamphochromis longiceps, but we later decided that our samples (which were not individually labelled) comprised 

a mixture of R. longiceps and R. ferox. Consequently, we did not analyse these samples. 

No evidence of linkage disequilibrium was found between any of the six loci, confirming that they can be 

considered as independent genetic markers. Levels of genetic variability were remarkably consistent among the 

three. Multilocus estimates of genetic differences (FST) indicated that there was no substantial genetic substructuring 

within the populations of any of the three Diplotaxodon species examined (Table 2). 

Table 2. FST (genetic differentiation) values, at 6 microsatellite loci and over all loci combined (P = 

probability of overall FST not > 0), among samples of three Diplotaxodon species. 

Locus Pzeb1 Pzeb2 Pzeb3 Pzeb4 Unh130 Unh154 Overall P 

Species 

D. limnothrissa 0.0007 0.0014 0.0018 0.0012

37 


distance’. In other words, although all the populations are probably exchanging lots of migrants, genes take a long 

time to exchange between populations at the far extremes of the range. These populations are clearly exchanging 

hundreds, if not thousands of migrants per generation. Samples from Cape Maclear and off the Livingstone 

mountains were not genetically different, nor were samples from the SE Arm and Chilumba. Likewise, for D. 

macrops and D. ‘offshore’, all of the different populations of the same species were so similar, the non-significant 

FST indicates that they could be no more genetically differentiated than two samples taken from the same place. The 

D. macrops populations were from Salima and the SE Arm, while the D. ‘offshore’ populations were sampled 

hundreds of kilometres apart, at Karonga and Nkhata Bay. Included in these analyses were comparisons between 

inshore and offshore samples from the same geographic region: at Chilumba and SE Arm/Cape Maclear for D. 

limnothrissa, in the SE Arm for D. macrops and off Karonga for D. ‘offshore’. In no case did we find significant 

genetic differentiation, showing that samples taken inshore are part of the same populations as found further 

offshore in the same region. The implications of this study for the evolution of these fishes were presented by Shaw 

et al. (2000). 

The minimal population structure we found indicates that these three species are not likely to need management as 

independent stocks in different parts of the lake. Artisanal or bottom trawl catches of these species are very likely 

exploiting the same stocks that are found in the offshore pelagic or deep-water zones. Indeed, Malawian, Tanzanian 

and Mozambican fishers may be exploiting the same stocks. We found clear differentiation between the stocks of D. 

macrops and D. ‘offshore’, which is consistent with our view that they are different species. Diplotaxodon macrops 

is caught in reasonable numbers by deep demersal trawls, and before they were realised to be different species, it 

had been hoped that exploitation of this population might provide a ‘backdoor’ route to exploiting the deep-water 

pelagic stock of D. ‘offshore’. The standing biomass of this stock was estimated at around 33,000 tonnes (Menz et 

al. 1995). Comprising 19% of the pelagic biomass, it might be expected to yield around 6,000 tonnes per annum, 

and it is scarcely exploited. New ways must be tried to get at this stock. 

Case Study III: Are colour forms of mbuna really species? 

An important application of molecular stock structure study lies in the definition of species- the most commonly 

used ‘unit of conservation’. Population structure analysis need not only be applied to populations from different 

locations, but can also be used to test whether two forms co-occurring at a single locality are interbreeding, and 

therefore morphs of the same species, or are not interbreeding and thus are different species. Critically, such studies 

need to be carried out on samples taken from exactly the same place, otherwise there can be confusion of the effects 

of restricted migration. 

The haplochromine cichlids of Lake Malawi share a peculiarity with those of Lake Victoria. It is common to find 

two or more completely different coloured kinds of males in the same place, yet these colour forms show only very 

subtle differences in body proportions or other characteristics. The same holds for the chambo species Oreochromis 

karongae and O. squamipinnis. Although this tends to be taken for granted by people who work on Lakes Malawi 

and Victoria, it is actually an extremely unusual phenomenon. Pioneering researchers, such as Fryer (1959) 

considered that these colour forms were morphs of the same species, but with the advent of SCUBA equipment, 

later studies by Holzberg (1972) and Ribbink et al. (1983) showed that subtle differences in habitat preferences and 

behaviour were correlated with colour differences, suggesting that these forms were indeed different species. 

The first studies to test genetic differentiation between sympatric (co-occurring at the same site) species used 

allozyme electrophoresis to investigate three forms of Petrotilapia spp. at Monkey Bay (McKaye et al. 1982) and 

four forms of Pseudotropheus zebra at Nkhata Bay (McKaye et al. 1984). All of the Petrotilapia were found to be 

genetically differentiated, and they were later described as new species (Marsh 1983). The P. zebra colour forms fell 

into two species groups, the blue and black striped (BB) and the orange blotch (OB) forms were genetically 

indistinguishable, and thus considered to be one species, while the plain blue (B) and the plain white (W) forms 

were considered to be a second genetically distinct species, later named Pseudotropheus callainos (Stauffer & Hert 

1992). 

We employed microsatellites to test for differences among three putative species of the P. zebra complex (P. zebra, 

P. callainos, P. ‘gold zebra’) and six of the P. tropheops complex (‘olive’, ‘mauve’, ‘band’, ‘rust’, ‘black’ and 

‘deep’) at Nkhata Bay. In two of these comparison, P. ‘tropheops band’ v P. ‘tropheops rust’ and P. zebra v P. ‘gold 

zebra’, the females were so similar in appearance that we could not trust ourselves to identify them reliably, and we 

had to confine our sampling to sexually mature males in breeding colours. All nine forms showed significant genetic 

differentiation (van Oppen et al. 1998).

38 


All three studies have consistently demonstrated that colour forms of mbuna found in the same location tend to be 

different species. The exception lies in the BB/OB and B/W forms of the P. zebra complex. These colour forms 

seem to represent within-species polymorphisms. It is striking that the rarer colour form (OB and W) are mostly 

found in females- males of these forms are extremely rare are around 100 times less abundant than females. So, the 

pattern seems to be one common male colour form at a site = one species. 

Other applications to conservation biology 

Molecular methods can also be used in the study of the genetic variability of populations that have undergone a drop 

in density. Sometimes, when population size is dramatically reduced, there can be a sharp loss in genetic variation, 

which can persist even after numbers have increased. This has the potential to limit the population’s potential to 

persist in the long run, as there may be a lack of genetic variants that can cope with novel environmental stresses. In 

the most extreme cases, the population can become inbred, meaning that individuals can end up with 2 copies of 

harmful recessive genes, the effects of which are normally masked. This can lead to populations declining into 

extinction. In such circumstances, transplantation of individuals from other populations might be needed to restore 

variability. Loss of genetic diversity can often be a problem in captive-bred stocks, and can lead to changes in stock 

management practises. 

Non-molecular approaches to assessing population structure 

Tagging 

The most obvious way to study fish movements is to catch fish, mark or tag them in some way and then wait until 

they are recaptured and see where they turn up. Lowe (1952) tagged 6,237 chambo in the SE Arm, with the intention 

of using mark and recapture data to determine growth rates. She recovered 2 fish. A tagging experiment with utaka 

around Likoma was carried out in the 1980s, but with little apparent success (Jennings, pers. comm.). There are 

several reasons why tagging is unlikely to prove effective with most Lake Malawi fishes. 

First of all, the fisheries are based on catching very large numbers of very small fish. In the early 1990s, the estimate 

annual catch of Lethrinops ‘pink head’ from Lake Malombe was 850,000,000 fish (Turner 1996). The total 

population of Diplotaxodon limnothrissa in Lake Malawi was estimated at around 1,450,000,000 fish (Turner 1996). 

With such huge populations, an immense number would need to be tagged and a vast number of captured fishes 

examined for tags to have a hope of finding any tagged and released fish at all. 

For most tagging studies, fishery biologists depend on fishermen to notice tags and return them, usually to claim a 

reward. This is likely to be very difficult in a situation such as pertains around Lake Malawi, where thousands of 

illiterate fishermen are scattered around hundreds of separate fishing villages. Perhaps radio announcements might 

be the only way to contact them. 

The most severe problem lies in studies of the cichlid fishes, which make up by far the most significant part of the 

lake’s catch by weight and value. Cichlids have closed swimbladders. When brought to the surface from deep water, 

the gas in the bladder doubles in volume for every 10 metres ascended. Fish with open swimbladders can expel 

excess gas via a duct to the gut, but cichlids have to rely on much slower physiological processes. In a series of 

experiments, Ribbink and co-workers demonstrated that mbuna required many hours to adjust to modest depth 

changes (summarised in Ribbink et al. 1983). In my own experience, few mbuna or chambo can survive being 

brought directly to the surface from depths in excess of 10 metres. It has been suggested that pelagic cichlids such as 

the ‘big-eye’ species of Diplotaxodon migrate from great depths to the surface at night (Thompson et al. 1995), 

which might indicate special adaptations to rapid pressure changes. However, in April 2001, I conducted a pilot 

study to see if it would be possible to collect live Diplotaxodon and Rhamphochromis from chilimira catches at 

Nkhata Bay. Survival of Diplotaxodon was zero out of more than 50 individuals, including some of the big-eye 

species. It seems highly unlikely that any cichlids collected by normal fishing methods, other than a few seined from 

shallow shores or very close to the surface, could possibly survive capture and tagging. 

Thus, it seems that for more species, tagging studies can be ruled out as a practical method for the study of stock 

structure in Lake Malawi.

39 


Parasite studies 

Several studies of marine fish migrations have been able to show that populations of fish do not intermingle because 

they have distinct populations of parasites, which could be transmitted, were they ever in close contact. Equally, 

knowledge of the life histories of particular parasites may indicate that fish must have been infected in a different 

place to where they have presently been collected, showing that they have spent time at the site where the infective 

stages of the parasite can be found. The use of parasites as ‘biological tags’ is reviewed by MacKenzie & Abaunza 

(1998). Unfortunately, almost nothing is known of the taxonomy or life histories of fish parasites in Lake Malawi. It 

seems likely that many cichlid species will probably prove to be so closely related and recently diverged from one 

another that they have not had time to evolve different parasites. 

Otolith microchemistry 

Another recently developed hi-tech method is the analysis of chemical composition of hard tissues, such as otoliths. 

The method has been useful in distinguishing among fish which spend part of the time in the sea from those living 

entirely in freshwater (Limburg et al. 1998), distinguishing those spending all of their time in the sea from those 

spending part of the time in freshwater (Tsukamoto et al. 1998) and in distinguishing marine fish which had 

originated from different estuaries (Milton et al. 1997) or freshwater streams (Kennedy et al. 2000). It is doubtful if 

the method has any use in distinguishing among fishes from different parts of the same lake, although it might be 

worth trying out the method on river-spawning fishes, such as mpasa and sanjika. 

Species identification 

Given the limited amount of information on species distributions and catch compositions, much can still be learned 

from simply recording the species composition of fishery catches. This can give a useful rough estimate of gear 

interactions within a restricted geographical locality, but cannot really delimit the geographic extent of stocks. 

Conclusion 

Molecular studies of stock structure are expensive, and require considerable technical expertise and sophisticated 

equipment. Is the effort worthwhile? I suggest that it is. 

Most management decisions concerning Lake Malawi fisheries are made on the basis of ‘rules of thumb’, rather than 

hard quantitative evidence. In truth, nobody really knows whether most stocks are over- or under-exploited. Beach 

seines are regarded as destructive on the basis of little hard evidence (Sarch & Allison 2000), based on inspection of 

species and sizes caught, with little knowledge of the real levels of juvenile mortality and its link to population sizes 

of exploited species. Closed seasons are not closely linked to any quantitative assessment of breeding seasonality, or 

any detailed knowledge of the effects of recruitment on exploitation during the breeding season (FAO 1993). Trawl 

fishery licences are allocated on the basis of crude biomass estimates from trawl surveys, with possibly unjustified 

assumptions made on the relationship between trawl CPUE and biomass and between biomass and MSY (FAO 

1976). 

None of these statements should be interpreted as criticism of the fisheries institutions of the riparian countries. The 

size and complexity of the lake, its fauna and its fisheries render attempts at data collection and interpretation a 

daunting task. But what I am suggesting is that it is important to prioritise use of research resources, particularly 

manpower. Molecular studies of stock structure can be carried out with the assistance of externally funded projects 

and external expertise, with little cost in staff time or resources from local institutions. A persistent problem with 

short-term externally-funded and largely externally-implemented projects is that of ensuring that the work continues 

after the end of the funding. Fortunately, this is not necessary with work on molecular stock structure. Occasionally, 

it might be useful to repeat such a study to look for temporal changes in migration patterns resulting from major 

changes in environmental conditions or exploitation regimes. However, in general, the work need only be done once 

in order to generate useful rules of thumb, which can inform management recommendations in perpetuity. 

Acknowledgements 

The research on which this paper was based was funded by the UK Department for International Development 

(DFID) Environment Research Programme and by the Natural Environment Research Council (NERC), and carried 

out with the co-operation of the Dept. of Fisheries, Malawi, TAFIRI and the University of Malawi.

40 


References 

Arnegard, M. E., Markert, J.A., Danley, P.D., Stauffer, J.R., Ambali, A.J. & Kocher, T.D. 1999. Population structure and colour 

variation of the cichlid fish Labeotropheus fuelleborni Ahl along a recently formed archipelago of rocky habitat patches 

in southern Lake Malawi. Proc. Roy. Soc. Lond. B 266: 119-130. 

Avise, J.C. 1994. Molecular Markers, Natural History and Evolution. Chapman & Hall, NY. 

Carvalho, G.R. & Hauser, L. 1998. Advances in the molecular analysis of fish population structure. Ital. J. Zool. 65S, 21-33. 

Danley, P.D., Markert, J.A., Arnegard, M.E. & Kocher, T.D. 2000. Divergence with gene flow in the rock-dwelling cichlids of 

Lake Malawi. Evolution 54, 1725-1737. 

FAO. 1976. Promotion of integrated fishery development, Malawi. FAO, Rome. 

FAO. 1993. Fisheries management in south-east Lake Malawi, the Upper Shire River and Lake Malombe. CIFA Tech. Paper 21, 

Rome. 

Fryer, G. 1959. The trophic relationships and ecology of some littoral communities of Lake Nyasa with especial reference to the 

fishes and a discussion of the evolution of a group of rock-frequenting cichlidae. Proc. Zool. Soc. London. 132, 153- 

281. 

Holzberg, S. 1978. A field and laboratory study of the behaviour and ecology of Pseudotropheus zebra (Boulenger), and endemic 

cichlid of Lake Malawi (Pisces: Cichlidae). Zool. Syst. Evol.- Forsch. 16, 171-187. 

Kennedy, B. P., Blum, J.D., Folt, C.L. & Nislow, K.H. 2000. Using natural strontium signatures as fish markers: methodology 

and application. Can. J. Fish. Aq. Sci. 57, 2280-2292. 

Kocher, T.D., Lee, W.J., Sobolewska, H., Penman, D. & McAndrew, B. 1998. A genetic linkage map of a cichlid fish, the tilapia 

(Oreochromis niloticus). Genetics 148, 1225-1232. 

International Human Genome Sequencing Consortium. 2001. Initial sequencing and analysis of the human genome. Nature 409, 

860-921. 

Jarne, P. & Lagoda, P.J.L. 1996. Microsatellites, from molecules to populations and back. Trends Ecol. Evol. 11, 424-429. 

Lee, W.J. & Kocher, T.D. 1996. Microsatellite DNA markers for genetic mapping in Oreochromis niloticus. J. Fish Biol. 49, 

169-171. 

Limburg, K.E. 1998. Anomalous migrations of anadromous herrings revealed with natural chemical tracers. Can. J. Fish. Aq. 

Sci. 55, 431-437. 

Lowe, R.H. 1952. Report on the Tilapia and Other Fish and Fisheries of Lake Nyasa. Colonial Office, Fishery Publications, 

London. 

MacKenzie, K. & Abauza, P. 1998. Parasites as biological tags for stock discrimination of marine fish: a guide to procedures and 

methods. Fisheries Research 38, 45-56. 

Majerus, M., Amos, W. & Hurst, G. 1994. Evolution: the Four Billion Year War. Longman, Halrow, England. 

Markert, J.A., Arnegard, M.E., Danley, P.D. & Kocher, T.D. 1999. Biogeography and population genetics of the Lake Malawi 

cichlid Melanochromis auratus: habitat transience, philopatry and speciation. Mol. Ecol. 8, 1013-1026. 

Marsh, A.C. 1983. A taxonomic study of the fish genus Petrotilapia (Pisces: Cichlidae) from Lake Malawi. Bull. J.L.B. Smith 

Institute of Ichthyology, 48: 1-14. 

McKaye, K.R., Kocher, T., Reinthal, P. & Kornfield, I. 1982. A sympatric sibling species complex of Petrotilapia Trewavas 

from Lake Malawi analysed by enzyme electrophoresis (Pisces: Cichlidae). Zool. J. Linn. Soc. 76, 91-96. 

McKaye, K.R., Kocher, T., Reinthal, P., Harrison, R. & Kornfield, I. 1984. Genetic evidence for allopatric and sympatric 

differentiation among colour morphs of Lake Malawi cichlid fish. Evolution 38, 215-219. 

Menz, A. (ed.) The Fishery Potential and Productivity of the Pelagic Zone of Lake Malawi/Niassa. Natural Resources Institute, 

Chatham, UK. 

Menz, A., Bulirani, A. & Goodwin, C.M. 1995. Acoustic estimates of fish biomass. Pp. 307-349 in Menz (ed.) The Fishery 

Potential and Productivity of the Pelagic Zone of Lake Malawi/Niassa. Natural Resources Institute, Chatham, UK. 

Milton, D.A., Chenery, S.R., Farmer, M.J. & Blaber, S.J.M. 1997. Identifying the spawning estuaries of the tropical shad 

Tenualosa toli, using otolith microchemistry. Mar. Ecol. Prog. Series 153, 283-291. 

Mueller, U.G. & Wolfenbarger, L.L. 1999. AFLP genotyping and fingerprinting. Trends Ecol. Evol. 14, 389-394. 

Ribbink, A.J., Marsh, B.A., Marsh, A.C., Ribbink, A.C. & Sharp, B.J. 1983. A preliminary survey of the cichlid fishes of rocky 

habitats in Lake Malawi. S. Afr. J. Zool. 18, 149-310. 

Sarch, M.T. & Allison, E.H. 2000. Fluctuating fisheries from Africa’s inland waters: well adapted livelihoods, maladapted 

management. Proceedings of the 10 th International Conference of the Institute of Fisheries Economics and Trade, 

Corvallis, Oregon. http://osu.orst.edu/dept/IIFET/2000/papers/sarch/pdf 

Shaw, P.W., Turner, G.F., Idid, M.R., Robinson, R.L. & Carvalho, G.R. 2000. Genetic population structure indicates sympatric 

speciation of Lake Malawi pelagic cichlids. Proc. Roy Soc. London B 267, 2273-2280. 

Slatkin, M. 1985. Rare alleles as indicators of gene flow. Evolution 39, 53-65. 

Sparre, P., Ursin, E. & Venema, S.C. 1989. Introduction to tropical fish stock assessment. Part 1- Manual. FAO Fisheries 

Technical Paper 306/1. 

Stauffer, J.R. & Hert, E. 1992. Pseudotropheus callainos, a new species of mbuna (Cichlidae), with analysis of changes 

associated with two intra-lacustrine transplantations in Lake Malawi, Africa. Ichthyol. Explore. Freshwaters 3, 253- 

264. 

Sunnucks, P. 2000. Efficient genetic markers for population biology. Trends Ecol. Evol. 15, 199-203. 

Thompson, A.B. 1995 Fisheries potential. Pp. 369-386 in Menz (ed.) The Fishery Potential and Productivity of the Pelagic Zone 

of Lake Malawi/Niassa. Natural Resources Institute, Chatham, UK.

41 


Thompson, A.B., Allison, E.H. & Ngatunga, B.P. 1995. Spatial and temporal distribution of fish in pelagic waters. Pp. 279-307 in 

Menz (ed.) The Fishery Potential and Productivity of the Pelagic Zone of Lake Malawi/Niassa. Natural Resources 

Institute, Chatham, UK. 

Tsukamoto, K., Nakai, I. & Tesch, W.V. 1998. Do all freshwater eels migrate? Nature 396, 635-636. 

Turner, G.F., Carvalho, G.R., Robinson, R.L. & Shaw, P.W. 2000. Biodiversity Conservation & Sustainable Utilisation of 

Pelagic Cichlid Fishes of Lake Malawi/Niassa/Nyasa. Ncheni Project Final Report, University of Hull, UK 

van Oppen, M.J.H., Turner, G.F., Rico, C., Deutsch, J.C., Ibrahim, K.M., Robinson, R.L. & Hewitt, G.M. 1997. Unusually finescale 

genetic structuring found in rapidly speciating Malawi cichlid fishes. Proc. Roy. Soc. London B 264, 1803-1812. 

van Oppen, M.J.H., Turner, G.F., Rico, C., Robinson, R.L., Deutsch, J.C., Genner, M.J. & Hewitt, G.M. 1998. Assortative 

mating among rock-dwelling cichlid fishes supports high estimates of species richness from Lake Malawi. Molecular 

Ecology 7, 991-1001 

Wright, S. 1951. The genetical structure of populations. Ann. Eugen. 15, 323-354.

42 


Fisheries Development, Management, and the Role of Government 

Tony Seymour 

Bryn Eryr Uchaf, Llansadwrn, Anglesey LL59 5SA, UK, Telephone +44 (0)1248 712540, Fax +44 (0)1248 717234, 

Email: seymourevans@hotmail.com 

Abstract 

Studies conducted over the past decade have identified pelagic and deep demersal fish stocks in Lake Malawi capable of sustaining 

an annual yield in excess of 40,000 tonnes but only marginally exploited at present. In addition, there are underexploited stocks of 

small demersal cichlids in Lake Malawi in depths of less than 50m, and there are opportunities for increasing production from the 

existing artisanal fisheries of Lakes Malawi and Malombe by managing them better and by paying closer attention to the near-shore 

environment. From the standpoint of sustainable biological productivity there is no reason why annual production should not be 

doubled over current levels. 

Translating biological potential into industrial output is not simply a matter of importing technologies and securing a source of 

investment capital. Development of the fishing industry will require at least as much attention to entrepreneurial development as to 

technological change. It is suggested that neither of the two smaller-scale categories of fishing business on Lakes Malawi and 

Malombe – artisanal and small-scale mechanised – function in the same way as enterprises of comparable scale in the fishing 

industries of developed nations. It will be necessary to break the barriers which currently limit the capacity of individual enterprises 

to expand, and prevent their graduation from one level to another. It will be necessary also to develop a vision for the industry which 

focuses on human resource development, foresees the creation of intermediate stages in business complexity so as to facilitate 

enterprise growth, and guides the acquisition and flow of information. The capture of information via short-term expatriate experts 

may be less efficient than the secondment of Malawian fishermen to overseas fishing industries and, under controlled conditions, 

the encouragement of joint ventures with other fishing nations. To date, Malawi’s efforts in fisheries development appear clumsy and 

poorly researched when compared to the sophistication of scientific studies on the resource base. 

Hitherto, the market has not been a constraint to fisheries development. Nationally and regionally the demand for fish exceeds the 

supply by a substantial margin, and, in keeping with Malawi’s strength as a trading nation, the fish trade has proved vigorous and 

extremely adaptable. That does not mean that it could not be improved, needs no support, or might not become a constraint at 

higher levels of production. Ultimately, the strengthening of producer-trader linkages, the development of community-based 

organisations for business as well as management functions and the exploration of export markets might provide new points of 

departure for fisheries development. 

Service provision has been a major issue in recent years. Experience gained since the early 1990s has shown that the withdrawal of 

government services to the fishing industry did little to stimulate entrepreneurial development but instead led to a discernible 

weakening of the sector, with negative economic and environmental consequences. World-wide, fishing industries tend to comprise 

a patchwork of private sector and state-led elements, and it is considered normal for governments to intervene in areas unattractive 

to the private sector and to use direct cash transfers and subsidies, both as instruments of development (fleet renewal, new 

technology) and management (decommissioning and buy-back). Although private sector participation in the fisheries ancillary trades 

must be retained as a development goal, the Government of Malawi is now faced with the need to rebuild and expand upon its 

former services network if the production opportunities so clearly identified are to be exploited within the foreseeable future. 

The development of new mechanised and artisanal fisheries has important implications for management if it is to avoid a harmful 

accumulation of effort on hard-pressed traditional stocks. There are some signs that, despite the daunting size of the task ahead, 

Malawi is on the right course in implementing its co-management policy. But the handling of interactions between the anticipated 

future generations of mechanised and artisanal fisheries will require management skills of a higher order than those evident in 

recent years. The potential for increased economic rewards will be tempered by increasing risk as the complexity and catching 

power of the industry grows. 

Introduction 

Whoever wishes to understand the fishery sector in Malawi has to negotiate some fundamental anomalies and 

uncertainties. Lake Malawi is over-fished. It is also under-fished. Its fisheries are over-subscribed, and underdeveloped. 

It therefore needs both management and development, responsibility for which would until recently have 

been assigned without hesitation to the government. But in both of these fields the role of government is in a process 

of redefinition, and the lines dividing public and private responsibility have yet to be clearly drawn. This paper is a 

contribution to the public-private debate, and starts by proposing a framework for increasing production. 

• The immediate productive potential of the fishing industry of Lakes Malawi and Malombe is defined by the 

following five factors: 

• the extent and nature of the fish resources; 

• the market for fish, both domestic and export, and the efficiency with which the fish trade is able to match 

landings with markets; 

• the human resource base of the catching subsector, including skills in fishing and financial management, 

and the ability to retain and reinvest earnings; 

• the availability and nature of ancillary industries and services, including boatbuilding, gear and equipment 

supply, ice, financial services and technical advisory support, and 

• the effectiveness of fisheries management regimes, whether operated by the government, the industry itself, 

or jointly.

43 


The following sections of this paper aim to take stock of the current position with regard to each of these factors, 

and to suggest what roles the government might properly assume in facilitating the realisation of production 

opportunities so clearly identified by recent resource assessments. It will be suggested that merely identifying 

underexploited resources and developing appropriate harvesting technologies will not be sufficient to ensure that the 

desired outcome is actually achieved. Instead, a conscious industrial development approach is required, with as 

much attention paid to assisting enterprise growth as to “transferring technology”, as well as government’s 

resumption of many of the industry support functions divested during the 1990s. 

Resources and fisheries 

Over the past decade a succession of scientific studies have brought about a quantum leap in our understanding of 

the abundance, biology and ecology of the fish stocks of the Lake Malawi system. Although some important gaps 

remain, it is now possible to speak with some confidence of the potential yields from currently unexploited fisheries, 

or of fisheries from which better management might lead to improved yields. Fish stocks in Lake Malawi that are 

currently unexploited or only marginally exploited are listed in Table 1. 

Of the stocks identified above, some are already targeted to a limited extent by existing fishing operations. 

MALDECO harvests around 1,000 tonnes annually from the deep demersal stocks in the north of the South East 

Arm (Area C) and the South West Arm, as well as another 1,000 tonnes of pelagics, principally Diplotaxodon and 

Rhamphochromis spp. In addition, the artisanal fisheries take small quantities of Rhamphochromis, and variable – 

and sometimes very large – quantities of Engrualicypris sardella. Overall, Table 1 represents a biologically feasible 

increase in annual yield from Lake Malawi of around 40,000 tonnes. 

Shallow water (

44 


8. Enhancement of the existing artisanal fisheries by improving management, drawing from a wide range of 

potential management measures including entry and effort limitation, closed or limited-fishery areas and 

prohibition or restriction of harmful gears. 

Table 1. Biomass and sustainable yield estimates of marginally exploited stocks in Lake Malawi 

Stock Area Author Biomass Yield as 

% of 

stock 

Estimated 

sustainable 

yield 

Demersal stocks 

Deep demersal South East Arm (C) 1994-96 Banda & Tόmasson 1997 

6,040 30 1,812 

(>50m) 

South West Arm 1994-96 

7,220 30 2,166 

Domira Bay to Chia Lagoon 1994-95 Banda& Tόmasson 1996 

6,865 30 2,060 

Chia to Nkhotakota 1994 850 30 255 

Nkhotakota to Dwangwa 1994 1,050 30 315 

Dwangwa to Sanga 1994 

3,570 30 1,071 

Ngara to Lufira 1981 Tweddle 1981 1,588 45 715 

Total 27,183 8,394 

Offshore pelagic stocks 

Diplotaxodon Lake Malawi Menz & Thompson 1995 119,700 19 22,700 

Rhamphochromis 16,800 17 2,800 

Copadichromis 8,700 16 1,400 

Other cichlids 3,400 18 600 

Engraulicypris 5,100 63 3,200 

Opsaridium 1,300 23 300 

Synodontis 

13,400 17 2,300 

Total 168,400 33,300 

Total 41,694 

Markets 

Hitherto, the market has not been a primary constraint to fisheries development. Nationally and regionally the 

demand for fish exceeds the supply by a substantial margin, and the fish trade has proved vigorous and extremely 

adaptable. That does not mean that it could not be improved, needs no support, or might not become a constraint if 

fish production is significantly increased; but in Malawi the market must, for the present, be considered of secondary 

importance in conditioning the evolution of the sector. 

The processing, distribution and ultimate retailing of fish within Malawi is mostly undertaken by a very large 

number of small-scale operators. Processing infrastructure is generally simple and requires little capital investment: 

it is frequently owned by members of the lakeshore fishing communities, and may be rented out to visiting 

trader/processors. Some distributors own a means of transport (bicycles, pickup trucks), but many rely on public 

transport. Four principal pricing points exist – the beach or landed price, a lakeshore wholesale price (usually for 

processed fish), an inland/urban wholesale price and the final retail or consumer price. Typically the consumer price 

is 2-4 times the beach price, with the major part of the margin and profit concentrated at the final retail stage. 

Although there is some degree of specialization into processors, wholesale distributors and retailers, the high 

profitability of retailing encourages many traders to engage in the entire cycle, each processing, transporting and 

retailing relatively tiny quantities of fish. 

What is perhaps surprising, given that the fish trade is a well-established and profitable business with an annual 

marginal value of some US$25 millions, is that the scale of processing and trading enterprises remains so small. 

Only on rare occasions has any individual venture (outside Maldeco’s distribution system) advanced beyond the 

level of the one-tonne pickup truck. The reasons for this are probably similar to those which limit the size of 

artisanal fishing businesses, a subject discussed below. It is a matter of debate whether large-scale fish 

processing/distribution enterprises would actually benefit the fishery sector – an increase in scale would lower unit 

distribution costs and might make it easier to improve product standards, but it would also cut into the widespread 

rural and urban employment opportunities afforded by this dynamic and highly accessible trade. 

These statements about the fish trade are based on contemporary personal observation and quantitative studies 

dating from the late 1980s (e.g. Seymour 1988, 1989), but the author is not aware of recent research that might 

substantiate them. In view of the potential for increasing fish production from Lake Malawi, a high priority should 

be placed on investigations into the effective domestic and regional demand for fish products, consumer preferences 

and quality/price relationships, and the present and likely future impact of the September 2000 SADC trade

45 


agreement. Ultimately, the strengthening of producer-trader linkages, the development of community-based 

organizations for business as well as management functions (producer associations) and the exploration of export 

markets might provide new points of departure for fisheries development. 

The fishermen 

A much more important determinant of past fishery performance, and a strong indicator of development potential, is 

the human resource base of the catching subsector. In Malawi the fishing industry is usually – and rightly – divided 

into three levels of operation. These are: artisanal (mostly small-scale commercial “traditional” fisheries, with some 

subsistence fishing), small-scale mechanised (pair trawlers and small stern trawlers) and large-scale mechanised 

(MALDECO). The three have quite different human resource characteristics. 

In 1999 the artisanal fisheries of Lakes Malawi and Malombe employed almost 40,000 fishermen, of which 21% 

were gear owners and the remainder crew (Weyl et al., 2000). Only a tiny proportion of the workforce has received 

any kind of formal training in fishing technology. From unsophisticated beginnings based on natural fibre gears and 

dugout canoes, the industry made substantial technical advances during the 1960s, when multifilament netting first 

became widely available, and the 1970s, when improved V-bottom planked boats became popular. During this 

period imported outboard motors were used in the more profitable – mostly chilimira – fisheries, and simple sail 

technology was copied from Mozambique in Likoma and Makanjira. For the past 25 years, however, very little in 

the way of new technology has been made available to the artisanal fishermen. Their poverty effectively isolates 

them from technical progress in other countries, exposure to which might otherwise be gained through travel, trade 

publications, exhibitions. One might have expected it to be the proper role of the Fisheries Department to keep the 

industry abreast of worldwide technical advances, but the fact is that the government was almost as impoverished as 

the fishermen, and overseas training for Fisheries Department officials focused much more on the biology and 

management of fish stocks than on the mechanics of catching them. Although shut off from the outside world and 

modern fishing materials, there is nothing to suggest that Malawi’s artisanal fishermen are any less competent or 

inventive than those elsewhere. A great diversity of fishing techniques has evolved, including floating longlines, 

bottom-set monofilament longlines, towed gillnets, pair-trawling with dugouts, artificial reefs and a variety of lightattraction 

methods, all using various recombinations of the very limited range of vessels and fishing materials 

available. When interviewed, fishermen often express frustration at their backwardness, and eagerness to experiment 

with new materials, fishing technologies and fishing craft. 

It is notable that within the artisanal fisheries, which produce an annual catch worth US$12-15 millions 1 , hardly any 

fishermen have become wealthy, and the largest enterprises amount only to a few small fishing units and perhaps a 

pickup truck. There is more to this than poor catch rates resulting from overfishing – there is almost certainly a 

range of social and structural factors at play that contain the size of individual enterprises. The social and economic 

dynamics of fishing communities have not been studied in Malawi, but it is suggested that some of the following 

may be important: 

• the extended family system, in which personal responsibility increases in proportion to personal wealth, 

and which operates against a background of poverty and increasing dependency resulting from HIV/AIDS; 

• linked to the above, an inability on the part of artisanal fishermen to separate business finances from the 

household economy, leaving business capital exposed to the needs of the extended family; 

• village-level social mechanisms that inhibit obvious personal success; 

• an inheritance system that more or less guarantees the dissolution of a business on the death of its founder, 

preventing inter-generational transfer of successful businesses; 

• the lack of rural banking services in lakeshore areas; 

• high levels of violent crime that discourage cash-based business activity. 

Artisanal fisheries have never evolved to the point at which individual enterprises could amass sufficient wealth or 

confidence to achieve technical development by themselves. By the same token, they have had little opportunity to 

learn or practice the fundamentals of financial management either. It is not therefore enough to provide new 

technologies and support services alone – until at least some artisanal fishing enterprises are assisted to break 

through the socio-economic barriers that prevent their growth, the vision of an upgraded offshore artisanal fishery 

will remain unfulfilled. 

1 30,000 – 37,500 tonnes @ US$400/tonne.

46 


The next level up – the pair trawl fishery – was essentially a development creation, financed initially under the 

FAO/UNDP Project for the Promotion of Integrated Fisheries Development (IFDP – 1972-76). The project 

identified demersal fish stocks in relatively shallow water that were not targeted by the artisanal fisheries, and 

designed appropriate vessels and gear to catch them. The project went on to support the construction of Mpwepwe 

boatyard and the training of boatbuilders and fishing crews, and put in place a government-operated loan scheme to 

finance the new fishery. At the time it was held up as an example of best practice in fisheries development – a 

conclusion supported by early successes and rapid loan repayment. With few exceptions the pair trawler owners are 

not “graduated” artisanal fishermen, but are mostly professionals, civil servants or businessmen who had sufficient 

personal credibility to attract the business loans (initially from government, later from the commercial banks) which 

these enterprises needed for start-up. The trawler owners do not themselves go fishing, but rely on hired crews 

which, although mostly paid on productivity-bonus systems, do not command the catch share normally paid to crew 

members. Some of these businesses have paralleled very closely to the tobacco estates established by civil servants 

and politicians in the 1970s and 1980s, suffering from an extractive management style, a lack of creativity, minimal 

maintenance of equipment, a poorly-motivated workforce and almost no reinvestment. They differ from enterprises 

of similar size in the developed world in that businessmen, not fishermen, run them. The only individuals within 

these businesses that have the personal wealth and confidence to travel and acquire new ideas are not possessed of 

the technical background to make use of them. In consequence today’s pair trawl units are almost identical to the 

blueprint drawn by the IFDP: there has been no technical development within this fishery for almost thirty years. 

The current demise of the pair trawl fleet is a remarkable illustration of adherence to the blueprint all of the pair 

trawlers were originally fitted with Sabb diesel engines, for which the commercial agent in Malawi was the Fisheries 

Department. When government withdrew from engine and spare part supply in the early 1990s, none of the 15 

companies then in operation changed over to engines with local dealer support. As a result, only those with external 

connections were able to import spare parts and stay in business: now only 5 companies remain active and mean 

annual landings have fallen from 3,000 tonnes to less than 1,000 tonnes (Figure 1). This should be viewed against a 

background of healthy stocks, a strong market and potentially high profitability. Where companies have tried to 

innovate – for instance in occasional forays into stern trawling – these ventures have mostly failed for purely 

technical reasons. Sound technical advice to support these brave attempts has simply not been available. 

Annual landings (tonnes) and effort (days) 

4,500 

4,000 

3,500 

3,000 

2,500 

2,000 

1,500 

1,000 

500 

0 

1976 

1978 

pair trawl effort 

pair trawl catch 

pair trawl cpue 

1980 

1982 

1984 

1986 

1988 

Figure 1. The decline of the pair trawl fishery, 1976-99 

The large-scale mechanised fishery – represented in Malawi only by MALDECO – operates on an entirely different 

plane and is to a large extent independent of external support systems. MALDECO started out in the 1950s as 

Yiannakis Fisheries Ltd., a private company set up by Greek fishermen skilled in ring netting and trawling. The 

1990 

1992 

1994 

1996 

1998 

1.80 

1.60 

1.40 

1.20 

1.00 

0.80 

0.60 

0.40 

0.20 

0.00 

Catch per unit effort (tonnes/day)

47 


company invested in vessels, chilled and frozen storage, ice production and mechanical workshops, and trained a 

large locally recruited workforce. It was confiscated by government in the early 1970s and operated for many years 

as part of the parastatal Malawi Development Corporation. Although it continued to perform well in public 

ownership its profits were diverted to subsidise less viable enterprises, and necessary reinvestment in capital 

equipment was discontinued. After a brief period under the management of ADMARC, another parastatal, the 

company was acquired in 1989 by its present owner, the Press Corporation, now a public company listed on the 

Malawi Stock Exchange. In the 1990s the damage done by years of neglected investment was restored through a 

commercial loan financed (indirectly) by the Nordic Development Fund under the World Bank led Fisheries 

Development Project (1991-2000). The loan provided for the supply of a new 17m multi-purpose fishing vessel as 

well as a range of cold storage, ice-making and ancillary equipment. When the company last changed hands it was 

making a loss, but the impact of its new investments was quickly felt: a positive pre-tax profit was achieved in 1995 

and modest growth has been sustained since then. 

Perhaps more significantly, outside of the project-led programme of capital development and technical assistance, 

the company began to make technical advances on its own initiative. In 1993-94 it imported two kapenta rigs from 

Zimbabwe and adapted them to fish for usipa (Engraulicypris sardella), and in 2000 it acquired a pelagic trawler 

specifically to fish for Diplotaxodon and Rhamphochromis. Although small in themselves these are the first 

examples of a locally owned company importing technology from the international arena: they are the start of 

industry-led development. 

Support industries and services 

In order to function at all, fishing industries need a diverse range of subsidiary support trades and services, including 

boatbuilding and repair, gear supply, harbours and roads, ice and fuel. The current status of these support trades and 

services in Malawi leaves a great deal to be desired: 

Fishing gear supply and technology development 

Multifilament nylon netting of indifferent quality is manufactured in Malawi by the Blantyre Netting Company Ltd. 

(BNC) on rather old Japanese machines from twine spun on site from imported yarn. In recent years this has not 

been a very profitable business, and there has been some uncertainty as to whether BNC (whose main business is 

now the manufacture of polypropylene sacks) would continue it. BNC’s nets and twines are distributed through a 

limited network of company outlets on the lakeshore, and a small range of products is also held by a number of retail 

shops. Annual domestic sales have fallen steadily from 80-100 tonnes in the 1980s to only 50 tonnes in 1999. This is 

a remarkable statistic given that the number of fishing craft has increased by 90% over the same period, although the 

volume of illegal imports of sheet netting from Tanzania and Mozambique is not known. Around half of BNC’s 

annual sales are bulk twine from which fishermen braid their own nets: a sure measure of low returns from the 

artisanal fisheries. BNC also manufactures split-film polypropylene ropes, and imported polyethylene ropes are 

available from many retailers. 

During the 1980s the Fisheries Department was a major retail distributor of BNC’s products, with most District 

Fisheries Offices holding a wide range of stock purchased through the (government-owned) Mpwepwe Boatyard 

Treasury Fund. The legality of this business was questionable, since the gear was purchased from BNC on a surtaxfree 

basis “for research purposes”, enabling the Department to retail at the BNC wholesale price and still retain a 

substantial profit. Although this profit was used to the industry’s benefit (by subsidising the price of fishing boats) 

there can be no doubt that this was a somewhat shady enterprise, and was viewed as a block to private sector 

participation in fishing gear retailing. Not surprisingly, the World Bank made the termination of government’s 

involvement in gear sales a condition of the IDA Fisheries Development Project loan agreement signed in 1991. In 

practice, fishing gear retailing is a risky and unprofitable business, and only really works if the margins are quite 

high (say, 50%). BNC’s recommended retail margin is 13%, and a pricing structure that has always been – to some 

extent – unofficially enforced (and is now reinforced by BNC’s increased participation in the retail market) has 

always inhibited the fishing gear trade – that is why the Fisheries Department became involved in the first place. 

Distribution problems aside, the range of fishing gear available to the industry must be considered extremely 

limited, and has barely changed in 30 years: almost no new materials have been provided, or new technologies 

extended. The fact that the artisanal fishery has expanded in this inefficient form poses another kind of development 

problem. If new materials – for instance, monofilament gill nets – were to be made instantly available at an 

affordable price, the impact on catching power would be dramatic, with potentially disastrous consequences for

48 


inshore stocks. Opening up Malawi’s fisheries to international influences will have to be managed extremely 

carefully if it is not to create more problems than it solves. 

Boatbuilding 

Dugout canoes are constructed by village canoe-builders around the shores of Lake Malawi. Although the number of 

canoes operating in Lakes Malawi and Malombe has increased from 6,200 in 1985 to almost 9,000 in 1999, their 

size has progressively diminished with the increasing scarcity of large trees, and their mean lifespan has declined as 

the use of rot-prone exotic timbers has overtaken the use of indigenous hardwoods. Dugouts are inexpensive but 

wasteful of wood and, in their present form, unsuitable for work beyond the near-shore zone. In Botswana about half 

of the traditional Pterocarpus dugouts (mokoros) of the Okavango delta have been successfully replaced by GRP 

canoes of similar design. It took five prototypes before the GRP design was accepted by the Okavango fishermen, 

after which demand quickly outstripped supply. A first prototype GRP canoe has been tested in Malawi. It was not 

“right” but fishermen are willing to work with the designer to improve it. This may be a way at least to replicate the 

large canoes of former years and extend fishing range at the lower end of the technology/cost spectrum. 

Village carpenters in Mangochi District have built flat-bottomed planked boats up to about 6m since at least the 

early 1970s. These craft are unsuitable for open-water fishing, and have a disturbing tendency to fall apart in rough 

seas. Altogether more than 3,000 planked boats are currently in use on Lakes Malawi and Malombe, of which less 

than one fifth are powered by small outboards. A significant (but unknown) number of these are strong V-bottom 

craft built by established boatyards at Salima and Mpwepwe. 

Salima Boatyard was established by a British boatbuilder as a private business in the early 1970s. The yard produced 

V-bottom planked boats to 6m for the artisanal fisheries, and a few carvel-planked round bilge sailing vessels. When 

the owner left Malawi in 1977 the Fisheries Department acquired the yard (in the absence of a private buyer) and 

over the next decade expanded and improved its production. These boats became very popular with fishermen, but 

production levels declined during the late 1980s to 20-30 per year, and the yard had a long backlog of orders. 

Although the sale price did not truly reflect production costs it is notable that fishermen purchased these boats 

without the need for credit. Salima Boatyard was closed in 1993 in accordance with the conditions of the IDA loan 

that financed the Fisheries Development Project (1991-2000), but no attempt was made to privatise the yard. The 

skilled workforce was dismissed, and some months later the tools and equipment were auctioned to local joinery 

firms, while the Fisheries Department retained the site. 

Several former Salima boatyard employees continue to make boats to an acceptable standard and to the popular 

designs developed by the Fisheries Department. However, they operate as small village businesses with very limited 

capital, and since they are unable to access other than locally available materials their products have a short working 

life. 

Mpwepwe Boatyard, in Mangochi District, was a larger establishment whose main line of construction was a 7.5m 

V-bottom boat powered by a 30hp inboard diesel engine, the workhorse of the pair trawl fleet. By the late 1980s 

maintenance of the existing trawl fleet, plus vessels used in non-fishing roles, was estimated to require the 

construction of six new vessels annually, as well as running repairs. But output had fallen to a very low level – one 

or two boats per year – and the yard was in debt and in urgent need of rehabilitation and new equipment. The 

redevelopment of Mpwepwe was therefore included as a priority sub-component of the Fisheries Development 

Project. 

At the time this project was prepared the prospects for privatisation appeared remote: it was therefore intended to 

rehabilitate the yard and to run it on lines as close to commercial as possible within the confines of government 

accounting and service limitations. Design studies were commissioned, and the boatyard staff reduced to a bare 

minimum in order to minimise losses during the reconstruction period. Stringency measures at this time included 

suspension of the import of marine engines and spare parts on which the pair trawlers depended. But by 1993 

several entrepreneurs had expressed interest in the acquisition of the boatyard. Privatisation appeared for the first 

time to be a real possibility, and the rehabilitation was deferred - it was considered more appropriate to make 

available to the new owner, on a loan basis, those funds which had been reserved for redevelopment. The 

privatisation of Mpwepwe took almost six years, and was not completed until May 1999. The option selected by the 

Privatisation Commission was a management buy-out and the yard was sold to the Mpwepwe Boatyard Company 

Ltd., comprising of staff previously employed by the Fisheries Department. However, by the time the transfer of 

ownership took place the yard was little more than a shell, its new owners inexperienced in business and potentially

49 


crippled by inherited debts. ICEIDA made a small grant available to buy minor tools, but there was insufficient time 

to arrange the proposed subsidiary loan prior to project closedown. Even though the yard is now in private 

ownership the new company has no working capital, almost no stock of raw materials, and very possibly a large 

debt. It stands little prospect of growth – perhaps even survival – without swift external intervention. 

The reduced availability of seaworthy boats to the artisanal fishery (and here it is necessary to distinguish between 

V-bottom boats built by the Salima and Mpwepwe yards and inferior flat-bottomed craft made by untrained 

boatbuilders) has contributed to the concentration of fishing effort in the inshore zone, to increasing poverty in the 

artisanal fishing communities and ultimately to the resource management problems which the Fisheries Department 

now has to face. The mechanised fishery has been more visibly affected. The pair trawl fleet was in a poor state 

when the project started, but after a decade without replacement engines or spare parts only 5 of the 15 units 

operational in 1990 remain in business today. 

Ice supply 

During the 1980s flake ice was supplied to fish traders by the Fisheries Department, from ice plants at Namiasi, 

Salima and Nkhotakota, and by MALDECO. The Namiasi plant became unserviceable in the late 1980s and was not 

replaced because a private flake-ice venture had opened in Mangochi. In 1993 the Salima fish handling depot (which 

provided fish washing and weighing facilities in addition to ice) was closed in accordance with the IDA loan 

conditions, and the machinery auctioned and later removed by the new owner to Blantyre. The small Fisheries 

Department ice plant at Nkhotakota was also taken out of service at this time. 

The current situation is that private plants at MALDECO and Mangochi continue to function, and a new enterprise 

produces small quantities of block ice at Salima. Otherwise, traders from Blantyre and Lilongwe purchase block ice 

from the Cold Storage Company Ltd. depots in those cities. A substantial – but unquantified – unsatisfied demand 

for ice is believed to exist. 

Harbours and roads 

Harbours are not used or currently needed in the artisanal fisheries, which use beach-landing boats and canoes. 

Similarly, beach access roads are not of great importance where the main means of transport between the beach and 

the nearest public transport route is the bicycle. Some major beach landings do attract vehicles, however, and 

anecdotal evidence suggests that landings are becoming increasingly aggregated, i.e. that increasing volumes of fish 

are being landed at a diminishing number of sites. This would point towards the need for road construction at 

carefully selected locations. The EU-funded Central Lake Fisheries Development Project (1980-85) constructed a 

number of beach access roads in Salima and Dedza Districts – these had the strongly beneficial effect of 

concentrating landings, processing and trading activity. 

Harbours are a great advantage to the pair trawl fleet, although with some ingenuity the boats may be operated from 

a simple beach site. The principal advantage of landing onto a jetty is that the catch may be discharged quickly, an 

important factor in reducing spoilage. The other main function of a harbour – shelter from rough lake conditions – is 

in practice very difficult to achieve in Malawi. There are few natural harbours, and medium-term variations in lake 

level make investments in traditional permanent harbour facilities extremely risky. The Fisheries Department 

constructed trawler landings with jetties in Lifuwu, Salima; Malembo, in the South West Arm, and Namiasi, in the 

South East Arm. Salima is no longer considered an inshore trawling ground, and the other two landings are in a very 

poor condition. Maldeco Fisheries Ltd. has constructed its own jetty and onshore infrastructure in Mangochi 

District. 

Development and how to achieve it 

Fisheries development is the growth and technical advancement of the fishery sector. Initially this process aims 

towards a condition in which all known fishing opportunities are fully exploited and the contribution of the fishing 

industry to the national economy maximised. Subsequent development tends towards greater catching efficiency and 

increased value-added through processing. In the more advanced fishing nations fisheries development is led by the 

industry. Even in these countries the government usually has a development role, and will undertake costly 

technological research and provide other services to the industry where these are not attractive to the private sector 

(e.g. the construction of harbours). In developing countries the role of the state tends to be more prominent, and has 

often included the introduction of new capture and catch handling technologies and the mechanisation of artisanal 

fisheries. As we have seen, fisheries development in Malawi appears not to be occurring by itself, and this implies a 

major role for the government in bringing it about.

50 


Where development is to be promoted by government and financed by external donors, the approach, nature and 

extent of development programmes will be heavily influenced by prevailing development styles, which are subject 

to periodic change. Development thinking during the 1970s and 1980s appears now to have been decidedly 

paternalistic, and tended to ignore indigenous knowledge and capacity and impose alien technologies that quite often 

failed. Currently, the development philosophy of many donors and international lending institutions draws heavily 

on the principles of market economics. The state is encouraged to provide a liberal and enabling policy environment, 

within which the private sector is expected to provide productivity and growth. In the productive sectors the role of 

the state is expected to shrink, shedding the more recognisable development functions and focusing on research, 

policy formulation and regulation. This new perspective quite properly accords more respect to the ability of the 

target population, but it also makes a rather risky assumption – that its members will react and respond to new 

economic opportunities in the same way as their entrepreneur counterparts in the developed capitalist economies. 

There is ample evidence to show that in Malawi, and especially in the fisheries sector, this assumption is erroneous – 

the pendulum has swung too far. The private sector is small, and the relatively technology-intensive fishing industry 

and its associated trades have to compete with far more attractive business propositions, notably in crop trading and 

transport. There is no better demonstration of this than the enforced withdrawal of the Fisheries Department from 

the provision of industry support services in the early 1990s. Locally the disastrous results of this intervention are 

well known, and are documented in the Fisheries Development Project Implementation Completion Report (World 

Bank, 2000). But the government is now left in a quandary – should it go back to begin rebuilding the services that 

were dismantled, or should it be looking for some alternative means of promoting fisheries development? 

It is worth considering how the developed nations deal with the question of state support to fishing industries. It 

might be expected that, if it is in some sense inappropriate for the Government of Malawi to provide fishery 

services, then they must be entirely absent from those states with large, modern fishing fleets and advanced 

economies. But this is not the case. In almost all developed fishing nations the government invests in fishing 

harbours and associated infrastructure. In addition: 

In the states of the European Union (EU), grants are available to individual businesses for new or experimental 

fishing ventures and for fish processing facilities. All European states waive duty and taxes on fuel sold to the 

fishing industry. 

• In the United Kingdom, many local government authorities operate ice plants and supply fuel to fishing 

vessels. The central government operates a decommissioning grant scheme to buy out excess fishing 

capacity from the industry. 

• The Republic of Ireland, which does not have an over-capacity problem, provides grants (of around 30%) 

towards the construction of new fishing craft. Here, central government operates a national network of ice 

plants, some of which are leased out to private operators. The government also provides grant support to 

exploratory fishing. 

• In the United States federal support to fisheries is provided through nearly thirty different grant and loan 

schemes providing for fleet renewal, fish processing, research and management (including buy-back and 

decommissioning). Many other regional and state-level support mechanisms complement the federal 

services. 

Malawi has been the subject of an experiment in economic purism that is not matched by practice in the outside 

world. Elsewhere, fishing industries tend to comprise a patchwork of private sector and state-led elements, and it is 

considered normal for governments to intervene in areas unattractive to the private sector and to use direct cash 

transfers and subsidies, both as instruments of development (fleet renewal, new technology) and management 

(decommissioning and buy-back). Given the small size of Malawi’s private sector and the dominance of trade as a 

business activity, it would appear that the entry of entrepreneurs into fisheries ancillary trades will be extremely 

slow. The government is now faced, therefore, with the need to rebuild and expand upon its former services network 

if the production opportunities so clearly identified are to be exploited within the foreseeable future. This process of 

reconstruction may be a painful one, but it will afford the opportunity to plan a development strategy with more 

thoroughness than has previously been the case. It is suggested that the strategy should be based on three essential 

themes – technology development, industrial development and fishery services:

51 


Technology development 

Of the eight fishing opportunities listed in the second part of this paper, three require the development and 

evaluation of new technologies (artisanal fisheries for small demersal cichlids, both deep and shallow, and artisanal 

fisheries for offshore pelagics) and two require the improvement of fishing methods already in use (pair trawling, 

pelagic trawling). Tasks of this kind have always been seen as a proper function of government, and it is to be 

expected that the Fisheries Department would commission the necessary work through an externally financed 

technical assistance programme. 

Industrial development 

This term is used in order to encourage a perception of the fishing industry as something much more than a 

collection of individuals who fish for a living, or technologies that impact on resources. It will be necessary to learn 

much more than we now know about the workforce, about financial and information flows and about business 

strategies and how they may be supported to encourage sustained growth. Three targets for action are: 

• Breaking the barriers to growth. The first requirement here is to clarify through appropriate studies exactly 

what are the constraints to enterprise growth, particularly in the artisanal fisheries. If, as has been suggested 

here, they are linked to an inability to isolate business finances from the demands of the extended family, 

then a carefully administered savings and loan scheme might provide a way forward. Whatever the 

determined constraint or proposed solution, it should not be expected that more than a minority will be 

ready or able to make the necessary changes in the short term. For some purposes it may be necessary to 

work selectively with the most progressive elements of fishing communities, however much this may go 

against egalitarian principles. 

• Improving the structure of the industry. The most expensive artisanal fishing unit – an outboard-powered 

chilimira unit – requires a capital outlay of around US$5,000. A pair trawl unit, plus second-hand pickup 

truck and ancillary equipment, costs US$45-50,000. A hypothetical large-scale fishing venture, based on a 

single 17m trawler plus necessary onshore infrastructure, would require an investment of US$1 million 

upwards. The gaps between these levels of investment (an order of magnitude or more) are so large that no 

individual can aspire to the level above without massive outside assistance. When exploring new 

technologies to harvest underexploited resources in Lake Malawi, thought should be given to creating 

business niches at scales intermediate to the existing three levels, so as to make graduation from one level 

to the next easier to achieve. For the same reason, it is also essential to reverse the recent decline in the pair 

trawl fishery. 

• Management of information. It is necessary to promote an inflow of information on fishing and fish 

handling technology from international sources, and to ensure that such information is (a) delivered 

effectively to the fishing industry, and (b) is matched by the availability of materials and services where it 

results in local demand. Although gear technologists may be recruited through a project-type technical 

assistance programme, the best source of expertise is successful fishermen from other countries. These are 

both difficult and expensive to recruit, for obvious reasons. Three different routes to linking Malawian with 

successful foreign fishermen might be feasible: (a) creating an advisory unit staffed by volunteers who are 

recently retired and carefully selected fishermen; (b) seconding small numbers of Malawian fishermen to 

appropriate fishing businesses overseas, and (c) – although this is not without risk – establishing one or 

more joint venture fishing operations. 

Fishery services 

The following service and ancillary trades are legitimate targets for government support or direct intervention, 

remembering always that private sector participation in these activities will be retained as a development goal: 

• Fishing gear supply. BNC should not be exposed to competition from government, but it may be useful for 

the government to encourage BNC or another supplier (by underwriting the commercial risk involved) to 

import technically desirable but non-traditional fishing materials. Renewed investigation of price structures 

might result in recommendations that would increase the attractiveness of fishing gear retailing and enable 

BNC to focus on production/import. 

• Boatbuilding. Every effort should be made to support the Mpwepwe Boatyard. But boatbuilding capacity 

needs to expand faster than Mpwepwe will be able to, even under the most favourable circumstances. 

Development options could include a design development project, pump-priming operations to support the

52 


construction of steel boats by existing general fabrication companies, an Okavango-style GRP canoe 

experiment, and/or another government boatyard in the Central or Northern Region. 

• Ice supply. It is highly desirable to increase the availability of ice through a planned network of ice 

production and sales points that takes into account possible changes in the distribution of fish landings. The 

network could be designed and installed by the Fisheries Department but ultimately handed over to District 

Assemblies, and the possibility of leasing ice-making facilities to entrepreneurs might be a useful first step 

towards ultimate privatisation. The availability of ice may be essential to achieving the full financial 

potential of proposed offshore fisheries. 

• Fisheries infrastructure. Existing trawler landings at Malembo and Namiasi should be rebuilt, and the need 

for additional sites investigated. This might include public investment in onshore infrastructure to support 

mechanised deep-water or pelagic operations in the centre or north of Lake Malawi. The need for access 

roads to major artisanal landings should be investigated with a view to proposing construction projects to 

the Malawi Social Action Fund. 

• Information and advisory services. The use of experienced volunteers has been discussed above. It may be 

feasible to reserve (development) funds to pay for specialist technical and financial advisory services as 

required by the growing mechanised fisheries – what must be avoided at all costs is the wasting of hardwon 

investment capital through bad technical decisions. 

• Financial services. This is a sensitive but important subject that deserves focused attention. The 

government may wish to support or encourage particular types of fishing enterprise by making available 

grants or subsidised loans. This would not risk distorting Malawi’s financial services sector, since the 

commercial banks have made no investments in the fishing industry for some years; even so, business 

development grants may be the preferred option. It should be noted that the high local currency interest 

rates for commercial borrowing look less frightening when viewed in hard currency terms: the 

identification of export markets for some part of the catch would make commercial borrowing an easier pill 

to swallow. The government may also wish to make grant funding available for management purposes, for 

instance for fishing gear buy-back schemes. 

Development, management and risk 

Fisheries management is the control of fishing industries – usually by governments or, in the case of shared stocks, 

by an agency representing more than one government – with the primary objective of ensuring that harvesting 

remains within biological limits and that the sustainability of the resource base is maintained. In practice, 

management objectives are usually more sophisticated than simply maximising sustainable yields, and it is normal 

to aim for maximum economic return, or maximum employment, or some combination of the two. Development and 

management tend to work in opposite directions, one promoting and the other restraining, but they are not opposites: 

a highly managed fishery is not the same as an undeveloped fishery. The most advanced fisheries combine a high 

level of technical development with an intensive management regime. Worldwide, as fish resources have come 

under increasing pressure and new fishing opportunities have become increasingly scarce, the emphasis has shifted 

from fisheries development to fisheries management. There is a simple principle here: the more developed the 

fishery, the higher the risk of resource collapse through over-exploitation and hence the more important the profile 

and effectiveness of fisheries management. 

In Malawi there has been more than one instance of “managing by not developing”, i.e. deferring development 

because it was feared that government capacity would be insufficient to manage the resulting increase in fishing 

power. This has applied for many years to the reluctance to exposing artisanal fishermen to monofilament gill nets, 

and it was also used as the rationale for postponing investments in the pair trawl fleet during the course of the recent 

IDA-funded Fisheries Development Project. Lack of management capacity may therefore be a further and unwanted 

brake on the pace of fisheries development. 

Participatory Fisheries Management 

To a large extent fish are resources of an open-access nature, regardless of management or resource tenure regime. 

This is due to the sometimes-extreme mobility of fish populations, so that resource tenure can never be pinned down 

to the extent possible with land or trees. Even in situations of closed entry, each fisherman is aware that personal 

restraint means personal disadvantage, however much it may benefit the fishery. For this reason fisheries are always

53 


difficult to manage, and successful management regimes are few. Malawi’s approach to inshore artisanal fishery 

management is to maximise local ownership by transferring use rights and control of access to local fisheries 

institutions, on the basis of mutually agreed management plans. There are two reasons for believing this is the right 

course of action: 

• It would appear that a majority of artisanal fishermen support the principle of co-management, even in 

areas where there has been limited extension and no training. The publicity emanating from the former 

Malawi-German Fisheries and Aquaculture (MAGFAD) Project (now the National Aquatic Resources 

Management Project) has caught the imaginations of fishing communities from all corners of Malawi, and 

demand for extension support now outstrips system capacity. 

• Early indications are that inshore demersal stocks may be highly genetically differentiated (Duponchelle 

and Ribbink, 2000), suggesting that physical movements are restricted and that local-level fisheries 

management in Lake Malawi may have a biological validity previously unsuspected. Genetic research into 

demersal stocks is currently very limited, and further studies should be framed with the objective of 

identifying unit stocks for the main commercial species groups. 

There is however an attendant risk, and that is that over the short to medium term a lack of tangible improvements in 

fishery production and incomes could bring about a reaction or backlash against the new management regime. Part 

of the problem is that in their eagerness to improve their lot (and the status of their fish resources), some 

communities are inflicting considerable hardship on themselves by the imposition of extremely strict management 

regimes. These communities are unlikely to be able to maintain their commitment to co-management unless their 

personal sacrifices result in a swift improvement in incomes – an outcome that, given the nature of environmental 

management, is unlikely. A second part of the problem, and one that compounds the first, is that until new entry into 

the fishery is controlled, expansion of fishing effort will simply mop up all the benefits that would have accrued as a 

result of active co-management. 

There may be potential to take co-management into some constructive new directions. Until now it has been seen as 

a mechanism for implementing more or less traditional fishery management measures – mesh size restrictions, close 

seasons, etc. But it may be possible for the Fisheries Department to work with local fishery managers to develop a 

management vision that takes into account the wider impacts of fishing on near-shore ecology. Future management 

prescriptions might include closed or restricted-fishery areas, the phasing-out or reduction of harmful gears (possibly 

with grant support) and the closure of fisheries to new entrants. 

New fisheries and the management of interactions 

Development of the fishing industry will need to contend with two opposing forces. The opportunities for 

incremental production lie mostly offshore and in deeper waters. Yet the intrinsic productivity of Lake Malawi and 

the unit value of the catch both increase inshore, and in shallower waters. There will therefore be a general profitdriven 

pressure shorewards, which must be strongly resisted in order to avoid damage to inshore resources and the 

existing fisheries that depend upon them. This is most obvious in the case of large-scale mechanised fishing, where 

the towing power needed for deep demersal or pelagic trawling could be devastating in the shallows, but it applies 

equally to the artisanal fisheries. Larger boats and better gear are not desirable if they are used to fish the same 

stocks that are already under intense pressure. 

The tendency to “creep inshore” has already been clearly seen in the pair trawl fishery, but although the legal 

mechanisms to control the deployment of the trawlers are entirely adequate the government’s record in this regard is 

conspicuously weak. It is important to note that in this case one cannot count on harnessing the power of resource 

ownership to drive self-regulation by the mechanised fisheries – because the offence in question is poaching on 

someone else’s territory. It is pointless blaming the fishermen for this, because no fisherman anywhere would pass 

up the chance of a little extra illicit income if the prospects for retribution are low. If the mechanised fisheries are to 

be developed, then law enforcement capacity must be developed simultaneously and as part of the same package. 

References 

Banda, M.C. & Tómasson, T. 1996. Survey of trawling grounds and demersal fish stocks in central Lake Malawi, from Domira 

Bay to Nkhata Bay in 1994 and 1995. Fisheries Bulletin No. 33. Fisheries Department, Lilongwe. 

Banda, M.C. & Tómasson, T. 1997. Demersal fish stocks in southern Lake Malawi: stock assessment and exploitation. 

Fisheries Bulletin No. 35. Fisheries Department, Lilongwe.

54 


Duponchelle, F. & Ribbink, A.J. (eds), 2000. Fish Ecology Report. Lake Malawi/Nyasa/Niassa Biodiversity Conservation 

Project. SADC/GEF. 

Fisheries Department. 1999. Fish Stocks and Fisheries of Malawian Waters. Resource Report 1999. Fisheries Research Unit, 

Monkey Bay. 

Government of Malawi. 2000. National Fisheries and Aquaculture Policy. Fisheries Department, Lilongwe. 

Menz, A. & Thompson, A.B. 1995. Management Report of the UK/SADC Pelagic Fish Resource Assessment Project (Lake 

Malawi/Niassa). 

Seymour, A.G. 1988. Likoma and Chizumulu Islands: production, processing and trading in the chilimira fishery. Fisheries 

Department/GTZ Rural Growth Centres Project. 

Seymour, A.G. 1989. Women’s fish processing and trading groups at Makanjira Rural Growth Centre: report on a short field 

visit 31/01/89 to 03/02/89. Working paper: Project for the Promotion of Women at the Rural Growth Centres. Office of 

the President and Cabinet/GTZ Malawi. 

Tweddle, D. 1981. A preliminary assessment of the bottom-trawling potential of northern Lake Malawi. Luso: J. Sci. Technol. 

Malawi, 2(2):3-13. 

Turner, G.F., Carvalho, G.R., Robinson, R.L. & Shaw, P.W. 2000. Ncheni Project: Final Report. Biodiversity Conservation and 

Sustainable Utilisation of Pelagic Cichlid Fishes of Lake Malawi/Niassa/Nyasa. DfID Project R6414, ERP 49. 

Weyl, O.L.F., Banda, M., Sodzapanja, G., Mwenekibombwe, L.H., Mponda, O.C. & Namoto, W. 2000: Annual Frame Survey 

September 1999. Fisheries Bulletin No. 42, Department of Fisheries, Lilongwe. 

World Bank. 2000. Implementation Completion Report on an IDA Credit in the Amount of USD 8 million to the Republic of 

Malawi for a Fisheries Development Project (Credit No. 22250-MAI), December 29, 2000. Report No.: 21560, Rural 

Development Operations, Eastern and Southern Africa, World Bank.

55 


Seeking Sustainability: Strengthening Stakeholder Involvement in Fisheries 

Management in Malawi 

Tracy A. Dobson & Aaron J. M. Russell 

Department of Fisheries & Wildlife, Michigan State University, 13 Natural Resources Building, East Lansing, MI 48824-1222, Tel: 

517/432-1711, Fax: 517/432-1699, E-mail: dobson@msu.edu 

Abstract 

In response to declining fish stocks in Malawi’s waters, the government and a foreign aid donor collaborated to create a new 

fisheries management approach. A co-management regime was introduced in a pilot program at Lake Malombe at the southern tip 

of Lake Malawi. Fishers elected “beach village committees” (BVCs) to work with Fisheries Department (FD) staff in several phases 

of management. Reviews of the program to date are mixed. BVCs are working with the FD but have unmet fishing needs. 

Expansion of co-management to Lake Malawi has been initiated and is reported to be progressing slowly. The co-management 

approach is being more widely employed as greater demands are placed on scarce government resources everywhere and the 

value of stakeholder involvement is recognized. Using insights from international fisheries co-management programs, suggestions 

for improving the successes of Malawian co-management programs are offered. 

Key words: Fisheries co-management; community-based natural resources management 

Introduction 

A significant reduction in near-shore catch in recent years in Malawi has led to the employment of a new approach 

to fisheries management. In this paper, based on the most recent available data, we describe the fishery and 

management historically and currently. In particular, we lay out the new management approach and assess its 

successes and failures to date as well as speculate on its future prospects for enhancing fishery management. 

Malawi is a small country situated in Southeastern Africa along the African rift. At 119,000 km 2 it is similar in size 

to the State of Indiana. Most of the human population is concentrated in the southern and central regions (National 

Statistical Office 1999) where the best fishing and most suitable agricultural land are located. While 90 percent of 

Malawi’s citizens are engaged in agriculture (primarily subsistence), fishing and associated activities are a 

significant part of the economy. Fish provide an estimated 70 percent of dietary animal protein (National website 

2000, Bland and Donda 1995). 

Historically, five water bodies have been the major producers of fish: the Shire River and Lakes Chiuta, Chilwa, 

Malombe, and Malawi (See Figure 1). By far the largest and biggest producer of fish is Lake Malawi, a rift valley 

lake which yields 40 to 60 percent of the country’s annual catch (State of the Environment Report 1998). It is the 

third largest lake in Africa, taking up approximately 20 percent of the country’s geographical area (27,000 km 2 ). 

Moreover, it is believed to be the world’s most diverse lake in terms of fish species. It is estimated that as many as 

1,000 species exist there, most of which have not yet been described (National Environmental Action Plan 1994). 

The majority of species are believed to fall within the cichlid family, others species are also significant in the 

fishery. The main target species of the fishery include chambo (Oreochromis spp.), mpasa (Opsaridium 

microlepsis), utaka (Copadichromis), usipa (Engraulicyprius sardella), and kampango (Bagrus meridionalis). 

According to McCracken (1987), small scale fishing has a long history in Malawi. Traditional practices were left 

largely undisturbed by the British colonial regime that focused its attention on agriculture. Working from shore and 

dugout canoes, fishers used seine nets and gill nets to harvest a variety of species, including chambo, utaka, 

matemba and usipa. Recent estimates gauge total fishing industry participation at more than 230,000 persons 

(National website 2000, Bland and Donda 1994, Ferguson et al.1993), including boat owners, crew members, fish 

processors, and fish traders. The fisher component stands at approximately 40,000 (Bland and Donda 1995). 

Because of relatively easy entry to the fishery, as other fields of subsistence activity have been reduced or 

eliminated, and as the population has grown, a substantial increase in the number of participants has occurred (Bland 

and Donda 1995), causing greater fishing pressure, providing in part the impetus for a new management approach.

N 

Z A M B I A 

Z I M B A B W E 

Harare 

M A L A W I 

Figure 1. Malawi and its Water Bodies 

Lilongwe 

Lake Malawi 

Lake 

Malombe 

Blantyre 

0 

0 

T A N Z A N I A 

Lake Chiuta 

Lake Chilwa 

M O Z A M B I Q U E 

Figure 1. Malawi and its water bodies 

150 km 

100 mi 

56 


Historical approach to fishing regulation 

To understand the current situation a brief description of the historical context will be useful. According to Munthali 

(1994), territorial use rights, taboos and magic were employed to allocate and maintain the fisheries prior to the 

colonial era. Persons desiring to fish needed to secure permission from a chief or village headman to fish from a 

particular beach. Control by traditional authorities over the number of fishers provided protection against 

overfishing as well as greater equity in access in the view of Munthali (1994). Taboos making certain species 

inappropriate for human consumption in addition to taboos regarding inappropriate times to fish formed parts of an 

arrangement that protected the stocks (Munthali 1994). In addition to the reduced impact on the fishery due to the 

smaller populations dependent on fishing, traditional fishing methods (primarily handmade nets deployed either 

from shore or dugout canoes) are believed to have allowed a sustainable exploitation of the fisheries due to their 

relative inefficiency. 

The British colonial regime of the late 1800s and early 1900s focused its attention on agriculture as the most 

important area of economic activity (primarily tea, coffee and tobacco estate farming), leaving fishing to the locals 

for the most part (McCracken 1987). Most fish were consumed within the fisher’s household, but a significant 

indigenous commercial fishery serving local and regional indigenous markets was noted by Livingstone as early as 

1861 (McCracken 1987). 

European commercial competition was initiated in the 1930s. Fears that the use of European gear would dominate 

and significantly reduce the fishery through their higher efficiency caused the colonials to require non-Africans to 

obtain permits in the late 1930s. In 1949, a drought and fears of famine led to the imposition of a fish export ban 

and price controls. Nevertheless, by the 1950s, McCracken (1987) reports that a small number of Africans had 

developed substantial commercial fishing operations. Fishing rules adopted in 1974 and 1977 sought to limit fishing

57 


pressure (and raise funds) through the requirement of commercial licenses, gear restrictions and closed seasons 

(McCracken 1987). In the 1990s, a government parastatal organization, MALDECO, and one private company were 

engaged in commercial fishing, and 22 pair-trawl units engage in semi-commercial activity (Alimoso 1994, Derman 

et al.1994). The commercial and semi-commercial sectors bring in approximately ten percent of the catch (Bland 

and Donda 1995). 

Over time the influences of western economic and religious ideologies have had their impacts on the “informal” 

institutions regulating fishing activities. The gradual reduction in peoples’ adherence to traditional fishing taboos, 

and the marginalization of traditional leaders’ authorities over fishing in some locations, have led to increases in 

fishing pressure. (Munthali 1994). When Malawi gained independence in 1964, the expatriate regulatory scheme 

remained in place, for many years staffed by expatriates. The gradual transition to a predominantly Malawian staff 

did not significantly change the Fisheries Department’s regulatory approach in the short run. 

Stock reductions lead to new approach to management 

During the 20 th Century Malawi’s human population has increased tenfold to nearly 10 million persons (National 

Statistical Office 1999). At the same time, estate holdings, where export crops such as tobacco and tea are 

produced, have grown to cover ten percent of Malawi’s total land area (roughly one million ha), occupying some of 

the country’s best agricultural land (Eschweiler 1993). The average subsistence farmer’s plot has shrunk below one 

hectare, a size considered insufficient to fulfil a family’s needs (National Population Plan 1993). As a result, more 

and more farmers have resorted to fishing for supplementary income (Bland and Donda 1994, Derman 1995, 

Ferguson et al. 1993). The 1998 State of the Environment Report cites a 32 percent increase from 1983 to 1988. In 

light of the extreme poverty of many persons who depend on fishing, a potential strategy of limiting access through 

requiring licenses and limiting their number, has been considered and rejected (Bland and Donda 1995). 

As the number of persons fishing has risen, pressure on fish stocks has caused dramatic reductions in certain key 

species, giving rise to concerns about the ability of near-shore fisheries to continue to produce an adequate yield for 

human dietary purposes as well as concerns about conserving biodiversity. For example, one of the most important 

catch targets in Lake Malombe, the chambo, has nearly disappeared from that lake, and elsewhere stocks have 

dropped as well. A 1995 report shows Lake Malombe’s chambo catch dropping from slightly over 4,000 tons in 

1976 to about 100 tons in 1994 (Bland and Donda 1995). Catch data indicate that the national chambo catch has 

decreased from 17,000 tons in 1984 to 4,000 tons in 1996 (State of the Environment Report 1998). While 

substitution of other species is occurring (State of Environment Report 1998), the data indicate that a high of 88,500 

tons of overall catch in 1987 has fallen and remains around 70,000 tons (State of the Environment Report 1998, 

Bland and Donda 1995). Consequently, diminishing fish stocks is one of the nine most significant environmental 

problems identified in the National Environmental Action Plan (1994). It should be noted that concern centres 

around the near-shore catch targeted by small scale fishers. A recent Icelandic International Development Agency 

(ICEIDA) study of demersal stocks is reported to have revealed a large unexploited biomass far from shore 

(Seymour1997), where gear employed by small scale fishers, who produce 90 percent of the catch (Bland and 

Donda 1995), cannot reach it. 

Plummeting of near-shore stocks in the 1990s (Ministry of Information 2000, Bland and Donda 1995) resulted in an 

increasing employment of fishing gear such as mosquito nets that harvest with almost absolute efficiency (Scholtz et 

al. 1997, Bland and Donda 1995). At the same time that this type of illegal and unsustainable activity took root, a 

weak economy meant that fewer resources were available for the FD to maintain, let alone expand, its management 

and enforcement activities (Bland and Donda 1995). Everyone, including fishers, however, could see that the fishery 

was in jeopardy, creating a situation in which most stakeholders were willing to consider a different management 

strategy. Lack of resources created an environment in which experimentation with a new approach involving the 

resource-using communities appeared to be a viable option, at least on a trial basis. 

Government and donor recognition that action needed to be taken to conserve the fisheries provided the impetus and 

led to funding. In 1992 an experiment to involve local fishers in management was launched at Lake Malombe at the 

southern tip of Lake Malawi. Since stocks of the highly prized chambo had been nearly fished to extirpation there, 

and the lake was relatively small (390 km 2 ), with about 335 fishers (Government of Malawi 1994), it seemed a 

feasible site for a pilot project. While new to Malawi, fisheries co-management is a strategy that has been employed 

successfully in a variety of regions and situations (Pinkerton and Weinstein 1995, Pinkerton 1989). Desperate for a 

solution, the Fisheries Department collaborated with the German Technical Assistance Agency (GTZ) to create the 

Malawi-German Fisheries and Aquaculture Development Project (MAGFAD) Participatory Fisheries Management 

Program (PFMP) (Scholtz et al. 1997, Bland and Donda 1994). They thought that education and direct participation

58 


in management would alter fisher behaviour, causing them to reduce total effort to permit the stocks the time needed 

to recover, one of the PFMP’s goals. 

The strategies of the project were straight forward: provide extension education on fish biology and stock 

management, provide training and support on self-organization for management and leadership, and then involve the 

fishers in the formerly exclusive governmental domain of fisheries management (Scholtz et al. 1997). Increasing the 

difficulty in bringing about change were the vestiges of decades of political oppression (Africa Watch Report 1990, 

Short 1974) that ended just as the PFMP was being launched. In a 1993 election, Malawi left behind 31 years of 

dictatorship and chose multi-party democracy. Colonialism and the dictatorship had suppressed local level initiative 

and leadership and created a deep mistrust of government. These historical realities presented a challenge to PFMP 

organizers (Scholtz et al.1997, Mr. Ben Mtika, National Extension Agent, personal communication, 1996). Thus, an 

intensive educational and partnership-building effort was necessary along with carefully nurturing the development 

of the administrative and leadership skills needed to participate in fisheries co-management. 

Under girding the new management strategy is the theory that shared regulatory responsibility, i.e., comanagement 

1 , will increase management resources and enhance the fishery. Pinkerton (1989, p. 4) defines comanagement 

as “...negotiated agreements and other legal or informal arrangements ... between groups or 

communities of fishermen and various levels of government responsible for fisheries management...” Another useful 

definition supplied by Selin and Chavez (1995, p. 190) emphasizes the operational mechanism of such 

arrangements, “...a joint decision-making approach to problem resolution where power is shared, and stakeholders 

take collective responsibility for their actions and subsequent outcomes from those actions.” This type of strategy 

recognizes the significant role played by fishers, incorporates their needs, and draws upon their knowledge along 

with the often more technical knowledge of FD professional staff. In a co-management arrangement, fishers begin 

to feel more like partners and less like miscreants that the government seeks to control. Fisheries Department staff 

assume new roles as well, shifting from top-down enforcers to collaborators. Concomitantly, through the 

cooperation and collaboration inherent in co-management, the hostility and mistrust that existed historically in 

relationships between fishing communities and the Fisheries Department in Malawi began to dissipate (Scholtz et al. 

1997). 

Co-management arrangements may incorporate to varying degrees notions of shared authority, shared responsibility, 

stakeholder ownership of the resource, and direct benefits accruing to stakeholders. As a result of their involvement 

as managers, stakeholders may be held accountable and accept accountability for management outcomes which they 

participated in creating. Perceiving themselves as owners, and persons with responsibility for stock protection, they 

are more likely to develop a long term view that changes their fishing behaviour (Bland and Donda, 1994). 

In the case of the PFMP at Lake Malombe, the program worked with the existing political framework to organize 

fishing villages. Each community elected a 10 to14 member beach village committee (BVC) to serve as an 

intermediary between the FD and the fishing community. Traditional authorities, village chiefs or village headmen, 

serve on the BVC within their jurisdiction in an ex officio capacity. After initial training provided by the FD and aid 

organizations, each BVC adopted a constitution and decided on which management activities they would undertake. 

According to Scholtz et al. (1997) BVCs became involved in a variety of important activities including: 

1) discussions of fisheries regulations, 

2) licensing and record keeping for fishing gear and boats, 

3) control of their beach and fishing area (primarily gear and license inspection), 

4) organization of extension sessions, and 

5) participation in fisheries enforcement. 

The FD and donor officials believed that an educated community would be more likely to actively support shortterm 

fishing restrictions in the interest of long term gains. Thus, an extensive program of extension messages was 

launched by MAGFAD. The messages spoke of fish biology and appropriate fishing practices. Since most of the 

target population is illiterate, the primary means of reaching out were discussions with BVC members, radio 

1 “Co-management” incorporates a broad range of approaches to public involvement in resource management. Ranging from 

seeking the views of affected groups to involving them in problem identification, planning and implementation, to actual power 

sharing (IIED 1994). It also one of many terms and phrases used to refer to management by government that incorporates 

community members in some way, including “community-based natural resource management” and “collaborative management.”

59 


broadcasts, and performances of the MAGFAD band (Dawson 1997). The fishing villages rarely have live 

entertainment, so appearances by the band, singing songs about fishing, were always well attended and well received 

(Wilson 1993). UNDP supported the position of a full time extension expert who met regularly with the BVCs and 

trained their members in administration. Under his guidance they learned about group dynamics and how to run 

elections, draft a constitution, create and maintain records and organize and run meetings (Scholtz et al. 1997). 

Co-management outcomes 

Positive fisheries management outcomes were reported by Scholtz et al. in their 1997 evaluation of the PFMP. Early 

in its development, in 1992, BVCs proposed regulations on appropriate fishing net mesh size. They advocated a 

19mm mesh while FD researchers argued for 25mm. Acceptance by the Department of the fishers’ position was a 

significant step forward in persuading the fishers to participate in the pilot project. Other fisher-proposed regulations 

banned kambuzi seine nets and maintained the closed season. According to Scholtz et al. (1997) the FD was not in 

agreement with the usefulness of these two rules, but they were nonetheless adopted. The FD wished to be seen as 

recognizing fisher views in a significant way, and it was felt that the rules would do no significant harm to the 

resource. With the adoption of the rules they had advocated, the fishers knew they had been heard and heeded, in 

keeping with the sentiment expressed by one of them, Mr. Staubi Africa, who said in 1994, “It is for us fishermen to 

decide whether we want to be poor for one or two years, or whether we want to be poor forever” (Dr. J.G.M. 

Wilson, personal communication, January, 1996). 

Adoption rates of the new net sizes were reported to be 17 percent in 1994, 60 percent in 1995, and 85 percent in 

1996 (Scholtz et al. 1997). As the new program moved into full swing, Lake Malombe’s chambo catch data show 

an increase from 79 tons in 1994 to195 tons in 1995 (Scholtz et al. 1997), perhaps the beginning of a reversal of the 

trend reported by Bland and Donda (1995). 

The reviews of the PFMP are not all positive. Indeed, a setback for the project occurred when fishers learned that 

promised “sitting fees” for meeting attendance, provision of new nets, and free replacement of nets of illegal mesh 

size were not forthcoming (DeGabriele 1998, Scholtz et al. 1997). An attempt to address the net replacement 

problem was made through a Fisheries Department small loan program. Through this program it was hoped that 

adoption of regulation size nets would happen as quickly as possible as well as salving the feelings of fishers 

concerned about “broken promises,” but fishers remained dissatisfied because the loan program did not fully meet 

their needs nor the original commitment of the FD (DeGabriele 1998, Scholtz et al. 1997). 

Scholtz et al.(1997) observed that FD staff continued to operate in a “top down” manner after the program began, 

undermining messages of shared authority and trust between managing “partners.” In addition, some recent attempts 

by BVCs to penalize violators were thwarted by FD officers (Dr. J.G.M. Wilson, personal communication, March 

28, 2000). Such overt blocking of BVC management authority will continue undermine the BVC-FD relationship as 

well as confidence in the willingness of the FD to share authority and to honour BVC decisions. 

The lack of effective communication, trust, and coordination between the FD and fishing communities was most 

recently highlighted by Mangochi fishermen’s misunderstanding and mistrust of the Department of Fisheries’s 

regulations establishing a closed fishing season that applied to local fishermen while excluding the commercial 

fishing vessels of MALDECO (Chimwaza 2000). The fact that fishermen’s ignorance of the reasoning behind the 

regulations was dealt with using riot police and tear gas speaks volumes to the need for greater education and, more 

importantly, dialogue between the parties before regulations are passed (a need that can be addressed using 

increased participatory management (Donda 2000) . It is apparent from this incident that the reconfiguration of 

government officials’ roles not been fully addressed, evaluations of stakeholder (including agency staff) incentives 

in the co-management process are needed, and that the participatory governance paradigm has not yet been fully 

internalised throughout the FD. 

Legislation authorises co-management 

From the beginning GTZ had been concerned about the absence of mandated institutional support for fisher 

participation in management (Scholtz et al. 1997). With progress to report that could be used to argue for legal 

status for this kind of activity, and with a proposed new fisheries law, “Fisheries Conservation and Management 

Act,” in the parliamentary hopper, GTZ began to pressure the government for new statutory provisions to formalize 

fisheries co-management (Dobson 1997). A consultant was hired to prepare an analysis of the situation and to draft 

language that could be inserted in the proposed statute (Dobson 1996). Her report was presented to the Director of 

Fisheries in May, 1996. The proposed statute was enacted in 1997, and it includes a new section. “Part III,”

60 


containing provisions authorizing “Local Community Participation” (Fisheries Conservation and Management Act, 

1997). Broadly speaking, BVCs are responsible for “conservation and management of fisheries” within their 

jurisdiction (Fisheries Conservation and Management (Local Community Participation Rules, 1998). The more 

specific details setting forth the institutional structures and authority of BVCs (e.g., licensing of small scale fishers, 

vessels and gear, enforcement) appear in the Fisheries Conservation and Management (Local Community 

Participation) Rules, 1998. The National Fisheries and Aquaculture Policy of 1999, has gone on to support comanagement 

in stating that “..participatory fisheries management has proven to be the most appropriate method to 

manage the fish resources in the lakes of Malawi.” 

Co-management at Lake Malawi and beyond 

With legislation in place securing the position of co-management, in 1998, GTZ authorized funding to expand the 

program it had initiated at Lake Malombe to the fishing villages of Lake Malawi. The expansion is a substantial 

undertaking in view of the size of the Lake. According to a consultant who is currently reviewing co-management 

activities in many of Malawi’s natural resources sectors, including the fisheries, his preliminary data indicate that 

progress is slow (Dr. J.G.M. Wilson, personal communication, March 28, 2000). According to Wilson, no comanagement 

agreements between the Fisheries Department and local communities have yet been signed as 

authorized by the 1997 Fisheries Conservation and Management Act (Dr. J.G. M. Wilson, personal communication 

June 27, 2000). Formal agreements between the FD and BVCs would likely strengthen fishers’ sense of ownership 

and responsibility. 

Wilson (June 2000) further reports that initial BVC organization has occurred at the southern end of the Lake, but 

that the regular radio broadcasts on fishing appear to have induced fishers at the more remote, Northern end of the 

lake to organize into BVCs on their own, in a fashion similar to the Lake Chiuta fishers described below. Having 

come into existence only recently, it may be too soon to begin evaluating the effectiveness of the activities of the 

new groups. 

While beyond the scope of this article, it should be noted that the Lake Malombe project gave rise to a successful 

fisher-initiated effort at Lake Chiuta, a small lake that sits astride the Malawi-Mozambique border (see Figure 1). 

Learning of the Lake Malombe approach, in 1994 fishers organized themselves to protect their fishery from 

approximately 300 new entrants from Lake Chilwa which had substantially dried up. With very limited assistance 

from the FD, they employed the model of Lake Malombe’s BVCs. They established nine 14-member Fish Stock 

Management Committees and drove off the interlopers who were using damaging nkacha seine nets and fouling the 

Lake’s waters with human waste (Dawson 1997, Scholtz et al. 1997). That effort appears to be continuing from a 

recent report describing the successful confiscation of two “illegal seine nets” from armed Mozambican fishers by 

Lake Chiuta fishers (Dr. J.G.M.Wilson, personal communication, March 28, 2000). 

Involving members of the user community in management is still in its fledgling stage in Malawi, and additional 

time is required before impacts on the fishery can be clearly seen and fully assessed. These efforts toward increasing 

the participation of local communities in resource management are part of a larger pattern that is being seen in the 

region. An example is Malawi’s northern neighbour, Tanzania, which has recently reacted to dropping fish stocks 

and water quality concerns by implementing a co-management approach for its Lake Victoria fishery. In response 

to increasing population pressures, overfishing, use of illegal gear and the introduction of a pesticide, thiodan, that 

were damaging the fisheries stocks and the lake water quality, 91 Beach Management Units (BMUs) have been 

established in Mwanza Gulf (Hoza and Mahantane 1998). It appears to be similar in form to the Lake Malombe 

program, with responsibility shared between the Fisheries Division of the government and the BMUs who patrol 

their fishing grounds to reduce illegal fishing. 

It should be noted that the co-management approach is being adopted well beyond the developing nations. Policy 

makers and managers are realizing that working with stakeholders in planning and implementation makes great 

practical sense (Selin and Chavez 1995, discussing inter alia “collaboration models” used by the U.S. Forest 

Service). Just as the fishers in Malawi have useful knowledge to contribute to the management of their fishery, as 

one example, the U.S. Fish and Wildlife Service acknowledges the importance of the knowledge and community 

ownership of residents in the Bitterroot Ecosystem as it moves forward with a plan to re-introduce the grizzly bear 

there in the near future to be managed by a citizen board (U.S. Fish and Wildlife Service 1999).

61 


Participatory approaches in resource management agencies 

Descriptions of the factors that are conducive to making co-management effective have been described by many 

authors (ex. Baland and Platteau 1996, Ostrom 1990). And while most authors discuss at length the requirements 

for capacity-building in communities and user-groups, most give short shrift to the institutional changes that are 

needed in the resource management agencies, whose staff are being drastically altered by fiat. Therefore, we will 

discuss some of the findings of the co-management literature on these neglected participants in the co-management 

process: the resource agency personnel. 

While most resource management field staff have traditionally enjoyed relatively high levels of authority and 

prestige over fishing communities, their new roles in a co-management program have stripped them of much 

authority and prestige. Ironically, they are the recipients of upper-agency/upper-governmental directives mandating 

their participation in a participatory management approach, while this same courtesy of participation is not applied 

to them within their department. In a study done by Gronow (1995), it was noted that the Forestry Department field 

staff in Nepal lacked a true understanding of the participatory approach, and failed to appreciate the legitimacy and 

desirability of the approach as their own lack of knowledge, lack of incentives, job tenure insecurity, and lack of 

agency support had not been taken into account in the design of the new program. Their inability to voice these 

concerns and opinions caused them to resist their gradual loss of authority and power. As a result, neither the 

agency’s nor the communities’ goals were met. 

It was only in certain areas where the forestry field agency concerns were addressed that staff became enthusiastic 

supporters of the new programs, thereby contributing to the programs’ successes. The process of reorientation was 

made in several stages (Gronow 1995). To initiate the process, field personnel were invited to take part in a 10-day 

workshop, where they were encouraged to discuss amongst themselves (with the aid of a professional facilitator) the 

issues surrounding forestry co-management and their roles in it. At this same workshop, they were presented with 

the latest information regarding the resource, co-management research, and participatory methods. Most workshops 

culminated with the staff’s decision to adopt their newly understood roles in the co-management process, 

committing themselves to the process. By allowing the field staff to experience a participatory approach within the 

agency where they saw the value of actively addressing stakeholders’ (in this case, themselves) concerns, they 

understood the merits inherent in doing the same with the forest-stakeholders. 

The second element was to provide the field staff with “follow-up support” to consolidate the process that was 

started at the workshops (Gronow 1995). They were provided with short-term moral and practical support, and job 

security while they attempted to innovatively change the ways that their jurisdictions’ forest resources were 

managed. The presence of competent and supportive role models was seen to be highly predictive of field staffs 

becoming effective and enthusiastic facilitators of the co-management processes with their forest user-groups. 

The last element, which has been shown to be the most challenging is achieving the institutional change necessary 

within the agency as a whole “.. to provide field staff with an environment conducive to their new role.”(Gronow 

1995) This requires a consistent application of the participatory approach within the agency. Lacking this 

consistency, different levels of the agency develop highly disparate attitudes toward co-management, from proparticipatory, 

manipulative, incremental, to anti-participatory (Midgley et al. 1986). This can have the effect of 

stifling motivation and innovation at all levels within the agency, and the development of perverse incentives against 

resource conservation and resource-user participation in management. 

In developing the co-management regimes with communities, the incentives of all stakeholders must be addressed. 

As the field staff are the main representatives of the agency in the co-management process, their roles and incentives 

need to receive as close scrutiny as that of the incentives and institutions facilitating community-members’ actions, 

if these new management regimes are to achieve success. When looking at field staffs’ incentives for acceptance of 

the participatory method, the incentives within their working environment need to be analysed and restructured in 

ways that encourage, rather than discourage their investment in the process. Through periodic participatory 

dialogues with all agency staff, the incentives and disincentives that arise can be addressed to improve staff 

members’ productivity and to reduce the development of perverse incentives within the agency. 

A brief discussion of the literature on human development will assist in understanding why and how this might 

work. The hierarchy of human needs was first described by Maslow (1943) to outline the motivations behind 

peoples’ actions (Figure 2). The bottom four levels of a simplified version of Maslow’s concept are called

62 


“deficiency needs,” which must be met before a person is able to move to the next level. The top level is “selfactualisation,” 

which, with Maslow’s later work (1954), came to be expanded into a grouping named “growth 

needs” consisting of (in order of increasing human growth): 

· Cognitive growth: the need to know, to understand, and explore; 

· Aesthetic growth: the need for symmetry, order, and beauty; 

· Self-actualization: to find self-fulfillment and realize one's potential; and 

· Transcendence: to help others find self-fulfillment and realize their potential. 

Closely related to Maslow’s hierarchy is Herzberg’s (Herzberg et al. 1959) “Motivation Hygiene Theory” which 

describes job factors that motivate employees, dividing motivations in a similar fashion to Maslow (Figure 2): 

hygiene maintenance factors (related to “deficiency needs”), and motivational Factors (related to “growth needs”). 

“Hygiene factors” are a requisite avoiding workers’ dissatisfaction in their jobs, while “motivational factors” 

actually produce job satisfaction (and motivate). Furthermore, while hygiene factor improvements may create shortterm 

changes in job performance, long term job performance changes are created only by increased motivational 

factors. 

Extending Maslow’s and Herzberg’s theories to the roles played by resource management staff members in comanagement 

regimes, it seems clear that in order for agency personnel to actively embrace a role in co-management 

that is foreign to them (and may have the potential to endanger their “deficiency/hygiene” needs), their 

“deficiency/hygiene” needs and “growth/motivational” needs must be addressed. These latter needs can only be 

sustained if they are part of an overall evaluation of the institutional infrastructure, which can best be done through 

an incorporation of participatory approaches in the management of the agency itself. 

Agency personnel analyses also should learn from the co-management literature by addressing the same erroneous 

assumptions that have been applied to the identities and capacities of “community members.” “Agency staff” at all 

levels should not be seen as homogeneous, given their unique ideologies and identities stemming from their 

overlapping profiles of gender, religion, ethnicity, socio-economic background, and so forth. To induce them to 

innovate and be pro-active in their jobs, their incentives and disincentives at every stage of the co-management 

process (policy making, planning, evaluation, enforcement) need to be analysed and addressed in the context of the 

effects that they have on the program. One of the best ways for this to be achieved is through active, consistent, and 

productive dialogue within the agency. 

Conclusions 

This assessment indicates that the evolution in the management approach for Malawi’s fishery from controls through 

traditional methods to license-based open access may be responsible for near-crisis conditions in the near-shore 

fishery. Coming at least part way round the circle, Malawi is following the approach employed in other fisheries to 

reincorporate local participation in fisheries management. While not a panacea, given the magnitude and multitude 

of other social issues in Malawi (e.g., human population growth, a weak economy), co-management appears to be a 

management strategy with many positive potentials. 

Seeing themselves as management partners with a significant role in building sustainable fisheries could cause 

fishers to adopt a longer term view, replacing the short term thinking which dictates taking as many fish each day as 

possible. Such a changed mentality is necessary to move Malawi in the direction of fishery sustainability. Care 

should be taken, however, to avoid the mistakes that have been made thus far that have hindered the effectiveness of 

the PFMP co-management arrangement, such as the failure to make good on commitments to fishers, undermining 

fisher enforcement efforts, and direct conflict between FD staff and fishers. 

Also crucial to the success of fisheries co-management are appropriate training and redeployment of management 

staff as well as a shift to more of a participatory approach within the Fisheries Department itself. The mistakes to 

date could be more easily avoided by ensuring complete understanding, unanimity, and institutional support within 

the agency of the reconfigured roles to be played by the agency and its staff in fisheries management.

63 


Figure 2. Maslow’s Hierarchy of Needs and Herzberg’s Motivation Hygiene Theory (from Allen 1998). 

Figure 3. Herzberg’s Motivation Hygiene Theory (expanded) (from Allen 1998).

64 


References 

Alimoso, S.B. 1994. Recent Research and New Management Strategies for the Fisheries of Lakes Malawi and Malombe. 

Fisheries Department Report. 

Allen, G. 1998. Supervision. (Hyperlink-http://ollie.dcccd.edu/mgmt1374/book_intro.html). 

Baland, J.M., & Platteau, J.P. 1996. Halting degradation of natural resources: is there a role for rural communities? New 

York: United Nations Food and Agriculture Organization and Oxford: Clarendon Press. 

Bland, S.J.R., Donda, S.J. 1995. Common Property and Poverty: Fisheries Co-Management in Malawi. Fisheries Bulletin No. 

30. 1-16. 

Bland, S.J.R. & Donda, S.J. 1994. Management Initiatives for the Fisheries of Malawi. Fisheries Department Report, 

Government of Malawi. 

Chimwaza, E. 2000. “Police tear gasses Mangochi fishermen” Malawi News, December 9-15, 2000. 

Dawson, K.R. 1997. Applying Cooperative Management in Small-Scale Fisheries: The Cases of Lakes Malombe and Chiuta, 

Malawi, M.S. thesis, Michigan State Univ., East Lansing, MI. 

DeGabriele, J. 1998. Evaluation of the participatory fisheries management programme for Lake Malombe and the Upper Shire. 

Report to GTZ, 1-32. 

Department of Economic Planning and Development. 1993. National Population Report. Office of the President and Cabinet, 

Government of Malawi. 

Department of Research and Environmental Affairs. 1994. National Environmental Action Plan., Government of Malawi. 

Derman, B., Ferguson, A. & Mkandawire, R. 1994. Socioeconomic Diversity among Lake Malawi’s Small-Scale Fishers: 

Implications for Biological Diversity and Fisheries Management. Presented at the Annual Meeting of the American 

Association for the Advancement of Science. 

Derman, B & Ferguson, A. 1995. Human Rights, Environment and Development: The Dispossession of Fishing Communities on 

Lake Malawi. Human Ecology. 23:125-142 

Dobson, T. 1996. A Report to the Director of Fisheries on Behalf of the Malawi-German Fisheries and Aquaculture Development 

Project. Submitted 24 May 1996. 

Dobson, T. 1997. Radical Restructuring of Environmental Policy to Preserve Biodiversity in Southern Africa: Malawi at the 

Crossroads. J. of Nat. Resources & Env. Law. 13:149-175. 

Dobson, T. 1999. Community Participation in Natural Resources Management in Malawi: Charting a New Course for 

Sustainability. Colo. J. of Intl. Env. Law and Policy, 1998 Yearbook, 153-177. 

Donda, S. 2000. Journey to sustainable fisheries management: organizational and institutional limitations in fisheries comanagement, 

the cases of lakes Malombe and Chiuta in Malawi. IIFET2000: Microbehavior and Macroresults, Oregon 

State University. 

Environmental Affairs Department. 1998. State of Environment Report for Malawi 1998. , Government of Malawi. 

Eschweiler, J.A. 1993. Malawi Land Uses. Consultant’s Report to the World Bank. i-vii & 1-107. 

Ferguson, A., Derman, B. & Mkandawire, R. 1993. The New Development Rhetoric and Lake Malawi. Africa. 63:1-18. 

Food and Agriculture Organization of the United Nations (FAO) Fisheries Department. 1996. Fishery Country Profile: The 

Republic of Malawi. Http://www.fao.org/fi/fcp/malawie.asp. 

Government of Malawi, Food and Agriculture Organization of the United Nations, United Nations Development Programme. 

1994. Fisheries management in the south-east arm of Lake Malawi, the Upper Shire River and Lake Malombe, with 

particular reference to the fisheries on Chambo (Oreochromis spp.). Chambo Fisheries Research Project. 

Gronow, J. 1995. Shifting power, sharing power: issues from user-group forestry in Nepal. In Nelson, N., and S. Wright, eds. 

Power and Participatory Research: Theory and practice. 

Herzberg, F., Mausner, B. & Snyderman, B.B. 1959. The Motivation to Work (2nd ed.). New York: John Wiley & Sons. 

Hoza, R.B. & Mahatane, A.T. 1998. Fisheries Management Component: Co-Management in Mwanza Gulf. Ministry of 

Natural Resources and Tourism. 

Human Rights Watch. 1990. Where Silence Rules: The Suppression of Dissent in Malawi. An Africa Watch Report. 

International Institute for Environment and Development (IIED). 1994. Whose Eden? An Overview of Community Approaches 

to Wildlife Management. A report to the Overseas Development Administration of the British Government. 

Koro, E. 1999. Where Are Malawi’s Fish?, Africa Information Afrique, May 24, 1999. 

Maslow, A. 1943. A theory of human motivation. Psychological Review, 50, 370-396. 

Maslow, A. & Lowery, R. Ed.. 1998. Toward a psychology of being (3rd ed.). New York: Wiley & Sons. 

Ministry of Information. 2000. Agriculture. Http://www.maform.malawi.net/agricutlure.htm 

McCracken, J. 1987. Fishing and the Colonial Economy: The Case of Malawi. J. of African History. 28:413-429. 

Munthali, S.M. 1994. Protected Areas and People: Sustaining Lake Malawi’s Ecology. Report to the Department of National 

Parks and Wildlife, Malawi. 1-36. 

National Statistical Office. 1999. Population of Malawi in 1998, 1987 and 1977 by area and sex. 

Http://www.nso.malawi.net/Comparative_population.htm. 

Ostrom, E. 1990. Governing the commons: the evolution of institutions for collective action. Cambridge: Cambridge University 

Press. 

Pinkerton, E. & Weinstein, M. 1995. Fisheries that Work: Sustainability Through Community-Based Management. A Report to 

the David Suzuki Foundation. 

Pinkerton, E. 1989. Co-Operative Management of Local Fisheries: New Directions for Improved Management and Community 

Development. Vancouver, British Columbia: University of British Columbia Press

65 


Scholtz, U.F., Njaya, F.J., Chimatiro, S., Hummel, M., Donda, S.& Mkoko, B.J. 1997. Status and Prospects of Participatory 

Fisheries Management Programs in Malawi. A paper presented at the FAO/ODA Expert Consultation on Inland 

Fisheries Enhancements, Dhaka, Bangladesh, 7-11 April, 1997. 

Selin, S. &Chavez, D. 1995. Developing a Collaborative Model for Environmental Planning and Management. Environmental 

Management. 19:189-195. 

Seymour, A. 1997. A Fishy Tale. Notes for a presentation to the [Malawi] Donors Environment and Natural Resources meeting, 

Sept. 4, 1997. 

Short, P. 1974. Banda.. London: Cox & Wyman Ltd. 

U.S. Fish and Wildlife Service. 1993. Grizzly Bear Recovery in the Bitterroot Ecosystem. Environmental Impact Statement. 

Wilson, J.G.M. 1993. Lake Malawi and the Upper Shire River: Community Fisheries Management Programme. Southern Africa 

Development Coordination Natural Resources Newsletter, Issue 4, October. 

Governmental Statutes and Policies cited 

Environment Management Act (1996) (Malawi) 

Fisheries Conservation and Management Bill (1996) (Malawi) 

Fisheries Conservation and Management Act (1997) (Malawi) 

Fisheries Conservation and Management (Local Community Participation) Rules, 1998 (Malawi) 

National Fisheries and Aquaculture Policy, October 1999.

66 


Fisheries management and uncertainty: the causes and consequences of 

variability in inland fisheries in Africa, with special reference to Malawi 

Edward H. Allison 1 , Frank Ellis 1 , Peter M. Mvula 2 and Laurence F. Mathieu 1 

1 School of Development Studies, University of East Anglia, Norwich NR4 7TJ, U.K. 

2 Centre for Social Research, Chancellor College, Zomba, University of Malawi. 

* Corresponding author. 

Abstract 

Uncertainty pervades the management of fisheries. Scientific fisheries management over the last 50 years has been based on the 

premise that there exists an equilibrium relationship between fish production and the level of harvest that can be taken without 

depleting the stocks. These equilibrium ‘surplus-production’ and ‘yield-per-recruit’ models have served to establish the principle that 

unregulated fishing will deplete fish stocks and dissipate economic rents from the fishery, but they have been of limited applicability 

for practical fisheries management when their equilibrium assumptions are violated. The influence of equilibrium models has 

extended beyond stock assessment into management, such that many fisheries management measures are based on a ‘steadystate’ 

view of fishery resources even when most stakeholders are aware that the assumptions are untenable. 

This paper makes the case that fish production in many African inland waters is driven by climate variations. For fisheries where 

stocks fluctuate independently of fishing effort, management for traditional sustained-yield type objectives is inappropriate. While 

there have been many studies attempting to elucidate the mechanisms for environmentally-induced fishery fluctuations, there have 

been fewer studies of the consequences of such variability for fisherfolks’ livelihoods, and for the design of appropriate fishery 

management regimes. A study of the livelihood strategies of fisherfolk involved in the important fisheries for small pelagic species in 

Lake Malawi is used to make the case for management that supports opportunistic exploitation of fluctuating resources by enabling 

geographical and occupational mobility. Livelihood sustainability and resource conservation are best served by support for such 

flexible strategies. The interdependence of fishing and other sectors of the rural economy suggests that policies and development 

interventions aimed at raising fishermen’s incomes without addressing the wider context of rural poverty are unlikely to be 

successful or sustainable. Species-based fisheries management and development focused on the fishing enterprise would benefit 

from re-conceptualisation within a broader natural resource management and rural livelihoods framework. 

Introduction 

Many fisheries resources fluctuate dramatically from year to year due to climactic variability (e.g. Glantz, 1992; 

Bakun, 1998). There has long been widespread recognition that constant catch or constant effort approaches to 

management, based on the paradigm of an achievable optimum sustainable yield, are inappropriate for these types of 

fishery (e.g. Beddington & May, 1977; Larkin, 1977). 

The problem of variability in fisheries, and consequent uncertainty in future stock size estimates can be dealt with in 

one of two ways. Variability can be regarded as random ‘noise’ that obscures underlying steady-state dynamics, or 

its causes and patterns can be investigated, and the results of these investigations incorporated into understanding, if 

not predicting, future stock and harvest levels. 

Although different fisheries are known to have very different patterns of catch series, there have been few attempts 

to classify fisheries according to extent and patterns of variability. Caddy & Gulland (1983) classified fisheries as 

steady, cyclical, irregular and occasional. This latter category denotes so called ‘boom and bust’ stocks that sustain 

important fisheries episodically before disappearing for decades or even for centuries. More recently, there have 

been attempts at formal statistical classifications of patterns of variation in catch or biomass time-series (Spencer & 

Collie, 1998) and to link patterns of variation with fish life-history parameters (Kawasaki, 1983). This work has 

interesting implications for fisheries management, as it suggests that different management targets may be 

appropriate for different fishery types. 

Interpreting the different categories of fisheries in terms of the factors that could drive different patterns of 

variability (DeAngelis & Waterhouse, 1987) suggests that steady and cyclical fisheries are likely to be driven 

primarily by biotic interactions (e.g. stock-recruitment relationships and catch-effort relationships). Irregular and 

occasional stocks could be either chaotic systems driven by strong biotic interactions (feedbacks), or systems where 

biotic interactions are relatively unimportant, and abiotic factors are the main influence on stock dynamics. If the 

latter explanation for the dynamic behaviour of fluctuating stocks is accepted, management based on regulation of 

biotic interactions (e.g. the effect of harvesting effort on future stock size) is much less relevant. 

This paper sets out to examine whether certain fisheries in Africa’s Inland waters are strongly influenced by climatic 

fluctuations, such that the use of management based on standard equilibrium fishery models may be problematic. 

The paper then explores the fishery implications of climate-driven stock fluctuations, through analysis of 

fisherfolks’ livelihood strategies in Malawi, derived from both primary fieldwork and published information.

67 


Finally, the paper draws out some preliminary policy and management implications of these observed livelihood 

strategies. These implications are framed against a context where fisheries policy and management in Malawi is 

undergoing a transition from centralised state-led management towards community-based or co-management (Sholtz 

et al., 1998; Allison et al. 2001). 

Africa's inland fisheries 

Fish protein has made up approximately one fifth of the animal protein consumed in Africa since 1961 (FAO, 1996). 

However, the contribution from inland waters has risen from less than 25 per cent in 1951 to 41 per cent of domestic 

fish production in 1994. In absolute terms, inland fisheries production has soared from 250 000 tonnes in 1950 to 

almost 1 500 000 (FAO, 1996). 

In recent years, Africa's inland fisheries have produced the majority of fish consumed in many African countries and 

almost all of that consumed in Mali, Chad and East Africa (Figure 1). Africa's inland fisheries are important not only 

as a source of food, but as a source of employment and income for resource poor families. They are exploited almost 

entirely by artisanal fishing communities in predominantly rural areas. In 1996 FAO estimated that the number of 

canoes operated by artisanal fishers in Africa's inland waters had increased by 40 per cent in the preceding decade 

and that most freshwater fisheries were intensively exploited. 

" As fishing effort continues to respond to the growing demand for fish, proper inland fisheries management is 

becoming more and more urgent." (FAO, 1996: 10-36). 

‘Proper fisheries management’ in this context has usually meant management for equilibrium production targets 

such as maximum sustainable yield, with measures to achieve these targets enforced by the State (e.g., for Lake 

Malawi, Tweddle & Magasa, 1989; FAO, 1993; GOM, 1999). 

While fisheries management in Africa shows an increasing interest in community and co-management strategies 

(e.g. Normann, Nielsen & Sverdrup-Jensen, 1998), these approaches too, are often based on unjustified assumptions 

about static equilibria and livelihoods based entirely on fishing. These assumptions lead to uncritical promotion of 

territorial use rights in undifferentiated and idealised constructs of a ‘community’ united by fishing interests (Allison 

& Ellis, 2001). The assumption in both cases is that fish yields can be both optimised and stabilised by better 

management. This does not allow for the possibility that optimal strategies may be opportunistic and ‘unstable’ in 

the conventional sense. 

Figure 1: Inland fish production as a proportion of fish supply available per caput in sub-Saharan Africa, 

1994 (Adapted from FAO, 1996).

68 


Climate and fishery fluctuations in Africa’s inland waters 

Conventional fisheries management in industrialised countries over the last 40 to 50 years has been based directly, 

or conceptually, on the Gordon-Shaefer bioeconomic equilibrium model and its derivatives (Figure 2). This model 

proposes an equilibrium between catch and fishing effort, so that fishing effort can be regulated to achieve a 

maximum sustainable yield (FMSY), maximum economic yield (FMEY) and related targets. Failure to regulate fishing 

effort is thought to lead to a situation where fishing effort tends towards the point where economic returns from the 

fishery equal the costs of exploiting the resource – the ‘open access equilibrium’ (FOAE). If signals of resource 

scarcity are distorted or masked by subsidies to the fishing industry (in the forms of grants for modernising fishing 

technology, compensation for poor fishing seasons etc), then fishing effort can even exceed the open access 

equilibrium, possibly leading to stock extinction. 

Although the Gordon-Shaefer model provides an elegant and persuasive overview of how a fishery bioeconomic 

system works, it has been extensively criticised for failing to provide the basis for successful fisheries management. 

There are many practical difficulties with the model: it is difficult to identify the target reference points until they 

have been exceeded; it is difficult to dis-aggregate the models in fisheries where one stock is fished by many fleets, 

or one fleet fishing many stocks; and it is based on catch and effort data that are often unreliable (Hilborn & 

Walters, 1992). There are also the theoretical difficulties with the equilibrium assumptions, outlined in Section 2, 

above. All these problems have led some to question whether the model itself, as well as the fishery management 

systems that are built on its basis, may not be appropriate to some fisheries - particularly those that fluctuate 

extensively (e.g. Sarch & Allison, 2000). 

How applicable are bioeconomic equilibrium (or surplus-production) models to African Inland fisheries? Three of 

the most important aquatic production systems in inland Africa are shallow lakes, river floodplains, and the pelagic 

zones of large lakes. It is also the fisheries of these systems that undergo the most pronounced climate-induced 

fluctuations (Kalk, McLachlan & Howard-Williams, 1979; Plisnier, 1997; Sarch & Birkett, 2000). By contrast, 

fisheries based on longer-lived, larger sized fish in demersal ecosystems in many of the larger and deeper African 

lakes seem more likely to fit with prevailing notions of equilibrium dynamics and the conventional fish stock 

management approaches based on them. There is a reasonable body of evidence suggesting that fisheries are 

significantly impacting productivity in these latter systems (reviewed in Pitcher & Hart, 1995), which is not the case 

for the more variable fisheries we consider in this paper. 

Malawi, and the African Lakes region more generally, contain important examples of all these types of fishery 

system, although river floodplain fisheries are poorly documented in the region. In this paper, we use Lake Chilwa 

and the pelagic fisheries of Lake Tanganyika to illustrate the possible influence of climate variability of fisheries in 

shallow lakes and the pelagic zones of the Great Lakes, respectively. 

Figure 2: The Gordon-Shaefer bioeconomic equilibrium model (Gordon, 1954; Shaefer, 1954) as a basis 

for fisheries management.

69 


Lake level fluctuations and fisheries at Lake Chilwa 

Africa’s shallow lakes are among the most productive fishery ecosystems in the tropics (Talling & Lemoalle, 1998). 

They are also prone to periodic lake level fluctuation, even to complete drying out in low-rainfall years. Most inland 

water ecosystems (with the exception of the African Great Lakes) are young, in geological and evolutionary terms, 

with an adaptable, resilient flora and fauna. They are, in a sense, pre-adapted to cope with a degree of humaninduced 

change (Moss, 1992). This resilience is a feature not often emphasised in fisheries analyses, typically preoccupied 

with stability as a management objective (Shepherd, 1991). 

Lake Chilwa is in many ways typical of the shallower African Lakes. The Lake has recently fluctuated around 1850 

km 2 including both open-water and wetland areas, is less than 3 m deep, and is subject to extreme fluctuations, 

including complete desiccation (Lancaster, 1979). In good years, fish catches can be as high as 25 000 tonnes 

(fishery statistics are rather uncertain and vary between sources) and more than 10 000 people are engaged in fishing 

activities. There was a major increase in fishing effort around the early 1970s, as the region became better integrated 

into the market economy. Minor recessions in lake level, sufficient to reduce fishing for one or two years, can be 

expected every six years or so (see Figure 3). Major recessions which will interfere with fishing in the open lake for 

3-5 years can be expected every 60-70 years, with a possibility of an intermediate recession in 30-40 years 

(Lancaster, 1979). The last drying episode covered the period from late 1994 to 1996, when fishing ceased 

altogether. Fishing operations started again in April 1997 (GOM, 1999). 

Catch (metric tonnes) 

25000 

20000 

15000 

10000 

5000 

0 

1960 1965 1970 1975 1980 1985 1990 1995 2000 

Figure 3. Catch fluctuations (shaded bars) and lake level variations in Lake Chilwa, Malawi 1962-1998. 

Note also that the lake gauging system was changed in 1989, and the lake level measurements from this 

period onwards may not be directly comparable with those in previous years (lower apparent amplitude of 

fluctuation). Fisheries data from Department of Fisheries, GOM (1999); Lake Level Data from 

Environmental Affairs Department (2000). 

Climate and fisheries in the pelagic zones of the African Great Lakes 

The fisheries for small pelagic fish in Africa’s Great Lakes are among the most important on the continent, 

supplying dried fish (variously known as kapenta, usipa, dagaa or omena according to species and region) to 

markets throughout much of East and central/southern Africa. Anecdotal evidence, oral histories, ecosystem and 

environment studies and government fishery statistics all support the notion that of these small clupeids and 

cyprinids fluctuates extensively from year to year, in response to climate-driven variations in primary and secondary 

biological productivity (Tweddle & Lewis, 1990; Allison et al., 1995; Plisnier, 1997). 

12.0 

10.0 

8.0 

6.0 

4.0 

2.0 

0.0 

Lake level (m)

70 


An important study by Plisnier (1997) documents the relationship between fluctuations in stock size (measured by 

proxy as variations in commercial purse-seine CPUE) and the Southern Oscillation Index (Figure 4), demonstrating 

the important link between stock size and climate variations in pelagic fisheries. Evidence of climate-productivity 

links in Lake Malawi’s pelagic fisheries are based on less extensive data, but imply a link between wind-stress, 

upwelling, and fish production (Tweddle & Lewis, 1990; Allison et al., 1995; Irvine et al., 2001). 

CPUE Anomaly (Tonnes/boat) 

0.60 

0.50 

0.40 

0.30 

0.20 

0.10 

0.00 

-0.10 

-0.20 

-0.30 

1976 

1977 

1978 

1979 

1980 

1981 

1982 

1983 

1984 

1985 

1986 

1987 

1988 

1989 

1990 

1991 

1992 

1993 

1994 

Figure 4. The relationship between stock abundance anomalies of small pelagic fish (clupeids) in 

Northern Lake Tanganyika (�), measured as the differences from the long-term average in Catch per unit 

of fishing effort by the Bujumbura-based industrial purse-seine fishery in Nov-Jan, and the Southern 

Oscillation Index or ‘El Niño effect’ (�) in the previous Feb-March. The correlation coefficient of 0.62 is 

highly significant. (Redrawn from Plisnier, 1997). 

Given the highly variable rainfall and wind regime in sub-Saharan Africa (Conway, in press) and the evidence for 

existence of strong climate-fish production relationships, there is case to made for fisheries research and 

management agencies to incorporate fish production-climate linkages in their programmes. These could provide 

more relevant scientific information than the current efforts at estimating parameters for use in single-species 

steady-state fishery assessment models. 

While there has been significant recent interest (if not formal research) in understanding the causes of variability in 

these fisheries, there has been little published work on the consequences of that variability for those involved in 

catching, processing, distribution, sale and consumption of fish. It is this gap in management-related research that 

we aim to address through work on the livelihood strategies of fisherfolk dependent on fluctuating resources. 

The livelihoods approach and research methodology 

The origins of the livelihoods approach 

The livelihoods approach has its origins partly in a literature concerned with understanding the differential capability 

of rural families to cope with crises such as droughts, floods, or plant and animal pests and diseases. This literature 

focuses attention on the assets of rural people, and how different patterns of asset holding (land, stock, food stores, 

savings etc.) can make big differences to the ability of families to withstand shocks (Swift, 1989). This set of 

concerns also links to the concept of vulnerability; defined as a high degree of exposure to risk, shocks and stress 

and proneness to food insecurity (Chambers, 1989; Davies, 1996). Vulnerability has the dual aspect of external 

threats to livelihood security due to risk factors such a climate, markets or sudden disaster; and internal coping 

capability determined by assets, food stores, support from kin or community, or government safety net policies. 

4.0 

3.0 

2.0 

1.0 

0.0 

-1.0 

-2.0 

-3.0 

-4.0 

Southern Oscillation Index

71 


The approach also borrows ideas from an ecological literature concerned with the sustainability of ecosystems or 

agroecological systems (Holling, 1973; Conway, 1987). Here, sustainability is defined as “the ability of a system to 

maintain productivity in spite of a major disturbance, such as is caused by intensive stress or a large perturbation” 

(Conway, 1985). The concepts of resilience and sensitivity as livelihood attributes also originate in this context 

(Bayliss-Smith, 1991). Resilience refers to the ability of an ecological or livelihood system to “bounce back” from 

stress or shocks; while sensitivity refers to the magnitude of a system’s response to an external disturbance. It 

follows from these ideas that the most robust livelihood system is one displaying high resilience and low sensitivity; 

while the most vulnerable displays low resilience and high sensitivity. These ideas are relevant to fishery-based 

livelihoods, as will become apparent in due course. 

The concept of ‘a livelihood’ seeks to bring together the critical factors that affect the vulnerability or strength of 

individual or family survival strategies. These are thought to comprise, chiefly, the assets possessed by people, the 

activities in which they engage in order to generate an adequate standard of living and to satisfy other goals such as 

risk reduction, and the factors that facilitate or inhibit different people from gaining access to assets and activities. 

These considerations result in the following definition of a livelihood (Ellis, 2000, p.10): 

“A livelihood comprises the assets (natural, physical, human, financial and social capital), the activities, and the 

access to these (mediated by institutions and social relations) that together determine the living gained by the 

individual or household.” 

The livelihoods approach is typically set out in the form of a framework that brings together the principal 

components that are thought to comply with the livelihoods definition, as well as demonstrating the interactions 

between them. There are many different diagrammatic representations of this framework (e.g. Carney, 1998; 

Scoones, 1998; Reardon & Vosti, 1995). Here, the framework is summarised in tabular form (Table 1). 

The reference social scope of this framework is typically considered to be the extended household, including 

members who are away from home but send remittances back to the resident homestead. The starting point of the 

framework are the assets owned, controlled, claimed, or in some other means accessed by the household (column A 

in Table 1). The livelihoods framework recognises five main asset categories, comprising physical capital 

(sometimes also called produced capital or economic capital); natural capital (land, trees, fish stocks etc); human 

capital (people, education and health); financial capital (savings, credit); and social capital (kinship networks, 

associations). 

Access to both assets and activities is enabled or hindered by the policy and institutional context of livelihoods, 

including social relations, institutions and organisations (column B). It is also affected by external factors, 

sometimes referred to as the vulnerability context, comprising trends and shocks that are outside the control of the 

household (column C). Assets permit livelihood strategies to be constructed, and these are composed of a portfolio 

of activities, some of which may be natural resource based and others not so (column E). Finally, this framework 

points to outcomes of livelihood strategies, distinguished here between livelihood security effects and environmental 

sustainability effects (column F). 

The livelihoods of artisanal fisherfolk are readily described by this type of framework. In this instance, key assets 

are fishing gears (boats and nets), although many artisanal fishers may also possess land and combine fishing with 

farming (Bailey & Pomeroy, 1996). The policy and institutional context of artisanal fishing includes, but is not 

limited to, the role of state regulations and ‘community’ based rules that affect access to resources. Social relations 

can also determine who has access to fishing opportunities (e.g. the ethnicity of fishing families may differ from 

other families in coastal communities, and roles within fishing activities are often strongly gender-differentiated). 

Fishing families are no less prone than other rural dwellers to adverse events and trends, with natural fluctuations in 

fish stocks being especially critical for them. Finally, fishing families often engage in diverse activities in order to 

achieve livelihood security – an important attribute that we will return to in the context of fisheries management.

72 


Table 1: A framework for micro policy analysis of rural livelihoods (modified from Ellis, 2000, p 30). 

A B C D E F 

Livelihood Access In context Resulting Composed With effects 

platform modified by of in of on 

social relations 

gender trends livelihood security 

class population NR based activities income level 

age migration fishing income stability 

assets ethnicity technological change cultivation (food) seasonality 

relative prices cultivation (non-food) degrees of risk 

natural capital macro policy livestock 

national econ trends nonfarm NR 

physical capital institutions world econ trends 

rules & customs 

Livelihood 

human capital 

land and sea tenure 

Strategies 

financial capital markets in practice 

shocks non-NR based 

social capital 

organisations 

storms 

recruitment failures 

rural trade 

other services 

rural manufacture 

env. sustainability 

soils & land quality 

associations 

NGOs 

local admin 

diseases 

civil war 

remittances 

other transfers 

water 

fish stocks 

forests 

state agencies biodiversity 

Fishing livelihoods research in Malawi 

The livelihoods approach is utilised in many different ways, according to the goal of the study or programme. In 

development practice, it is being used as a ‘process’ tool to enable participants in development programmes to 

identify key constraints and opportunities for development intervention (Ashley & Carney, 1999). In this paper, we 

use the livelihoods approach as a conceptual tool to interpret published literature on Lake Chilwa fisheries, and as a 

primary research tool to understand the livelihoods of people engaged in small-scale fishing on the shores of Lake 

Malawi. 

Livelihoods research utilises a range of existing methodologies in the social and economic sciences, and can 

essentially be regarded as a framework to organise these methodologies in such a way as to reduce sectoral and 

disciplinary biases. The research methods used for Lake Malawi include combination of qualitative and quantitative 

techniques. 

Assets were determined at household level by administration of questionnaires drawn from survey methods used in 

agricultural economics. Relatively small sample-sizes were used (typically 40 households per village) to ensure that 

data quality was maintained and good relationships between enumerators and respondents could be fostered. The 

small sample sizes also meant that each questionnaire could be verified by return visits to households if necessary. 

Sample selection was based on stratification following wealth ranking (based on people’s self-defined criteria), to 

ensure that poorer households are included in the research. Based on people's own definition of wealth, three 

categories emerged, namely (a) the well to do (wopeza bwino), (b) the better off (wopeza bwinno pang'ono) and (c) 

the poor (wosauka). The criteria depend a lot on who has what and who does not have what. The well to do, were 

people that had some form of capital that enabled them to engage in some productive activities like fishing and 

small-scale businesses. They were said to have enough food, live in good houses, own some livestock and could 

afford to send their children to school. The poor were said to be lacking in most things and they did not have 

anything to enable them to engage in profitable productive activities. They often did not have enough food and had 

poor housing. 

Three sites were purposely selected along the shores of Lake Malawi. The sites were Msaka in the Southern Region 

district of Mangochi, Lifuu in Salima (Central Region) and Tukombo in Nkhata Bay (Northern Region). 

A range of qualitative tools drawn from Rapid and Participatory Rural Appraisal (RRA/PRA) and institutional 

analysis were used to investigate how access to assets is modified by social relations, institutions and organisations. 

These included wealth ranking, focus groups, key informant interviews, institutional mapping and ranking of 

organisations’ effectiveness. Trends and shocks were analysed by documenting experiences described in focus

73 


group discussions, the use of relevant secondary social and economic data and the analysis of the political and 

macro-economic context (Allison et al., 2001). 

Resultant livelihood strategies were described through analysis of income sources (including gifts, remittances and 

exchanges) from household questionnaires, and the decision-making processes that lead to choice or adoption of 

these strategies were explored at both intra-household and village-level. 

Although the household questionnaires can provide only a ‘snapshot’ of current livelihood strategies, this was 

complemented by qualitative investigation of dynamic change, pursued though semi-structured interviews and focus 

group sessions, and through documentation of individual life-stories. 

Effects of chosen, or enforced, strategies on both livelihood security and environmental sustainability were 

investigated through existing monitoring systems, such as district and national production statistics and indicators of 

poverty or well-being. 

The research aims to investigate institutional factors that block or enhance people’s ability to pursue a sustainable 

livelihood so that policy and development intervention can address these constraints and opportunities. 

Livelihoods analysis in Malawi – results 

Lake Chilwa livelihoods 

Kalk et al. (1979) offer an interesting insight into livelihood responses to fishery fluctuation during the 1960s and 

1970s at a time of gradual transition from quasi-subsistence to a partial cash economy in this area of Malawi. This 

insight does not appear to have been transferred to current fishery management initiatives. The fisheries of Lake 

Chilwa offer an economically unstable environment, determined by the seasonal and long-term fluctuations in lake 

level. Yet, at high production periods, the fisheries permitted readily earned cash, with “a substantial number of 

men gained an income five or more times greater than that prevailing for unskilled or agricultural labour” (Chipeta, 

1972). In good years, Lake Chilwa supplies almost half the total fish production in Malawi, where fish is said to 

supply around 70% of animal protein in the diets of 12 million people (GOM, 1999). 

Fishing in Malawi is largely a business, not a subsistence activity (Ferguson et al., 1993). Management that 

constrains access to fish in productive periods constrains income-generating opportunities, denies people access to 

much-needed protein and serves no conservation purpose in a lake where the sustainable yield concept is obviously 

untenable. And yet, despite wide-spread acceptance that fisheries management, in its traditional guise of stock 

conservation measures, is inappropriate, there have been recent measures to introduce fishery closures to allow 

recovery after drying periods (even though recovery in the past has been rapid). Various gear and mesh-size 

restrictions have also been introduced, apparently at the behest of fishing communities around the lake, who 

participate in an evolving co-management scheme with the Fisheries Department (Sholtz et al., 1998). 

Work reviewed by Agnew (1979) and Agnew & Chipeta (1979) provide a useful baseline from which to review the 

likely choices available to people in the latest drying episode, and how these might have been impacted by new 

management initiatives. These authors summarise the short-term choices of fishermen during the lake-drying period 

of 1967-68 as: 1) fishing on a very much reduced scale in the remaining swamps, streams and lagoons in the Chilwa 

catchment, 2) transfer to nearby Lakes Malombe, Malawi or Chiuta, 3) increasing the cultivation of rice, cotton, 

cassava and vegetables 4) a switch over to commercial handicrafts such as plaiting carpets, 5) spending considerable 

time trapping birds and digging for rodents or 6) seeking employment elsewhere. These responses varied according 

to income status, asset profiles, ethnicity and time of residence in the area. 

In the drying episode of 1968, around 200 fishermen migrated to nearby Lake Malombe, and others moved to Lake 

Malawi. These were among the richest fishermen, whose investment in fishing-related assets meant that they could 

not simply cease fishing, as could those with a lower stake in this source of livelihood. Since the introduction of 

community-based management in Lake Malombe and Southern Lake Malawi (Sholtz et al. 1998, Chirwa, 1998), the 

option to move fishing operations between lakes is constrained. That this may also prevent Malombe’s fishermen 

from migrating to Chilwa in productive years, thereby relieving pressure on this intensely exploited lake, does not 

appear to have been explicitly considered.

74 


The repercussions of recession in Lake Chilwa waters and consequent decline of fishing are much wider than on 

fishing alone. The whole of the Chilwa plains and lake must be seen as an economic network. Not only are there 

links between fishing and various ancillary services, but also complementary flows of income between fishing and 

farming. The successful fishermen had larger gardens and produced more cash crops than other fishermen (Phipps 

1973). Recognition that there is an “integrated small-scale economy of farming, fishing and cattle-rearing” (Kalk, 

1979; p15) does not seem to have led to any specific policy support for these diversified livelihoods. Instead, 

sectoral concerns for the sustainability of individual natural resource systems have prevailed, even when it is known 

that notions of resource sustainability are questionable. “The Chilwa fishes are clearly well fitted to persist in the 

unpredictable Chilwa ecosystem, provided the refugium of swamps and streams is maintained”, according to Moss, 

(1979, p411), with Kalk (1979, p431) adding: “Man must remain as generalised in activity as the lake fauna in order 

to succeed in the Chilwa area”. 

Moss (1979) also cautions that more dangerous than overfishing in this resilient system were threats to the swamps 

through ‘reclamation’ for agriculture or perhaps as irrigation reservoirs, siltation through changes in catchment landmanagement, 

and pesticides. It is these threats that have led to recent interest in environmental management in the 

Chilwa wetland, and its designation as a Ramsar site. (Environmental Affairs Department, 2000). 

The EAD report reiterates the perceived resilience of the system. However, in an analysis of fisheries issues (EAD, 

2000, Table 5.2), the report highlights “Ignorance, Poverty, Corruption, Migratory fishermen and Lack of 

Resources” as barriers to sustainable utilisation of fishery resources, and recommends the implementation of 

“community-based natural resource management for the benefit of the local people”. There is clearly some 

difficulty in accepting that migration may be a legitimate and sustainable strategy to maximise benefits from a 

fluctuating resource, a factor that needs to be taken into account in the design of any community-based management 

scheme. 

The mobility and livelihood flexibility of the fishing families making their living on the shores of Lake Chilwa in 

the 1970s enabled them to respond to the extreme fluctuations observed. These were not mere ‘coping strategies’, 

but represent active opportunism – adaptations aimed at maximising the contribution of fishing to household 

incomes. It is not particularly useful to talk of the fish stocks as sustainable in the context of this level of ‘natural’ 

fluctuation. Around Lake Chilwa, there are large-scale shifts from fishing to farming, pastoralism and other occupations 

when the lake dries out (and back to fishing when it refills). Such strategies highlight the importance of enhancing or 

maintaining the flexibility of lakeshore livelihoods rather than constraining it with fixed fisheries production quotas, 

seasons or areas. 

Livelihoods on the Lake Malawi shoreline 

In Malawi, usipa (Engraulicypris sardella) is found throughout the lake, but only supports substantial fisheries in 

inshore areas (Thompson et al., 1996; Thompson & Allison, 1997). These fisheries are mainly artisanal and carried 

out using chilimira seines set by two canoes or small ‘plank boats’. The fishery takes place mainly at night with 

light attraction. There is also a substantial daytime beach-seine fishery, often for juveniles. Landings statistics are 

thought to be unreliable and to underestimate the true importance of the fishery, which may reach 50 000 tonnes in 

good years (Tweddle & Lewis, 1990). The fisheries are known to fluctuate extensively, with fishers able to identify 

and refer to ‘good’ and ‘bad’ years for usipa. These seem to be linked to interannual differences in productivity 

(Tweddle & Lewis, 1990) which in turn are generated by variations in the strength of upwelling caused by variations 

in wind stress (Allison et al., 1995). 

Usipa are marketed largely in sun-dried form and, together with the small pelagic species of other African lakes, 

contribute significantly to dietary protein throughout Central and Southern Africa. The fishery is quite seasonal, and 

exploitation of ‘usipa’ is likely to form only part of the livelihoods of those catching it. Coupled with its interannual 

variability, this makes this species a useful case-study of a fluctuating fishery. 

Results from the survey show that livelihood diversification and mobility are key factors enabling fishers to ‘track’ 

resource fluctuations in time and space. People along the Lake diversify in a number of activities including fishing, 

farming and different types of small-scale businesses. 

This type of movement has been adapted for sometime by fishers tracking a 'fluctuating resource'. In the case of the 

usipa this has led to the establishment of special relations with people in different beaches. There are two types of 

movements on Lake Malawi, either long-term or short term. Long-term movement refers to fishers that have moved

75 


from their original homes and have established a permanent camp elsewhere. These often do not have access to land 

and thus rely solely on fish and to some extent small-scale business. With short-term movement fishers move in 

search of fish but operate from their original homes, only establishing temporary camps. Maintaining access to land, 

they are able to return to farming during poor fishing periods. The map below shows the movement of fishers 

around Lifuu beach in Salima District. 

These strategies are well established and accepted around the Lake shore villages, where migrant fishers are seldom 

regarded as problematic, but are rather seen to bring benefits in the form of increased trade and economic activity in 

lake shore villages. Reciprocal access to fishing ‘beaches’ for landing catch and mending nets etc., may be more 

important to fishing-dependent communities than claims for territorial exclusivity of the type encouraged by efforts 

to promote community-based management (e.g. Sholtz et al. 1998). Such studies are urgently required around the 

other African Great Lakes, if the laudable move towards co-management is to develop models for fishery 

management that reinforce sustainable livelihood strategies. There is a danger that the idealised concepts of villageowned 

fishing grounds currently being promoted don’t fit with the ecology of the fish or the livelihoods of the 

fishers. 

Discussion 

The evidence for climate-induced fluctuations in the fisheries of shallow lakes, and for small pelagic fish in the 

African Great Lakes indicate that the management of these stocks needs to be informed by an understanding of how 

fishers, distribution chains and markets cope with fluctuating supplies. Most research on fluctuating stocks has been 

targeted at understanding in detail the mechanisms causing fluctuation in stock size. This is the study of fish

76 


recruitment processes and the environmental factors driving them (e.g. Cushing, 1996). There has been much less 

emphasis on the study of the responses of fishers to stock size fluctuations (Allison & Ellis, 2001). 

Review of secondary data from the Lake Chilwa area, and preliminary analysis of primary fieldwork on the shores 

of Lake Malawi both reveal the importance of livelihood diversity and geographical mobility as livelihood strategies 

of artisanal fisherfolk. Mobility and diversity are required to sustain livelihoods when confronted with resource 

variability that is at least partially climate-induced. The Lake Chilwa case demonstrates that livelihood coping and 

optimisation strategies existed prior to introduction of both State-led and co- management systems. More recent 

information on the impact of fisheries management measures on livelihoods is lacking, but is currently being 

pursued by our research team. 

The results of our research in Malawi are in accord with findings in other developing countries, where several 

studies have suggested that fishers cope with fluctuations through geographical and occupational mobility (Bailey, 

1982; Haakonsen, 1992; Geheb & Binns, 1997; Sarch & Allison, 2000; Béné et al. 2000). Fisheries management 

strategies which focus on optimal catch rates ignore both the role which inland fisheries play in the livelihoods of 

many Africans and the inherent stock fluctuations which have shaped such livelihood strategies. 

Proposals to manage fluctuating fisheries need to be based on a better understanding of fisherfolk’s livelihood 

strategies. In fisheries where exploitation has little demonstrable impact on fish stocks, and productivity is closely 

linked with climate, it is not useful to talk about sustainable yields, or of fixed limits for fishing effort. Neither are 

community-based or co-managed territorial use rights, in the form of geographically fixed territories, useful for 

fisheries management in areas where lake or floodplain levels are highly variable, or where fishers have to track 

mobile pelagic resources to sustain their catch rates. 

In Malawi, realisation of the importance of fisherfolk’s mobility is leading to a move away from management based 

on beach village committees (Sholtz et al., 1998), towards larger spatial scales – lake management areas defined in 

terms of movements of range of operations of artisanal fishers, and on ecological criteria (Weyl, personal 

communication, 2001). The mechanisms for governance of these lake spaces are still being discussed. 

It is relatively straightforward to outline what management approaches should not be taken, less easy to identify 

appropriate management support for sustainable livelihoods from fluctuating fisheries. While removing unnecessary 

impediments to sustainable opportunistic exploitation strategies is one important step, it may not be enough, given 

the increasing pressures on resources and livelihoods in Africa. Common property institutions that have evolved 

mechanisms, such as reciprocal access agreements between migrants, should be considered more appropriate than 

territory-based approaches as a way of implementing any effort-limitations deemed necessary. Even when 

embryonic and of limited functionality for resource conservation, such as in Malawi, such institutions can be built 

upon, rather than replaced by externally-conceived ‘perfect’ ones. 

Formal recognition, in national policy and legislation, of the legitimacy of opportunistic livelihood strategies, 

coupled with active removal of barriers to mobility and livelihood diversification would seem to be appropriate 

policy responses at national or district level. Active support for livelihood diversification (not the same as providing 

incentives for people to diversify out of fishing altogether) is another management option. 

The apparently greater importance of climate, relative to fishing, in driving the dynamics of fish stocks in many of 

Africa’s shallower wetlands (Sarch & Allison, 2000) also suggests that effort could be redirected at protecting 

wetland functions and broader ecosystem integrity and away from trying to manage fish stocks for sustainability. 

Management needs to lose its preoccupation with stability and gain an increased appreciation of resilience. 

‘Modern’ fisheries management has often consisted of setting stock conservation objectives, and then finding means 

of modifying fishers’ behaviour or investment to fit these objectives (Mahon, 1997). This has usually meant 

imposing closed seasons, closed areas, size limits, gear restrictions, access or ‘fishing effort’ restrictions. While 

there has been concern for the effects of different regulatory options on fishing communities, there has usually been 

little systematic research on their effects on fishers livelihoods. Fisheries management is becoming more 

consultative, and fishing communities now have greater participation in management, sometimes through comanagement 

arrangements (Pomeroy & Berkes, 1997). There is still little systematic discussion of the effects of 

different management options on livelihoods. There is a requirement for both participatory research to help to

77 


identify acceptable management solutions to fishery problems, and further studies of livelihoods to understand how 

fishers cope with and react to both inherent fluctuations and changing externalities. 


We thank Olaf Weyl and the staff of NARMAP for their help and organisational support, and for allowing us to 

incorporate our dissemination workshop within the National Fisheries Management Symposium. This research is 

funded under the Fisheries Management Science Programme of the U.K. Department for International Development 

(project number R7336). We also thank Terri Sarah, for contributions to a related paper (Sarah & Allison, 2000). 

The opinions presented in this paper are our own, and do not necessarily represent those of DFID. 

References 

Agnes, S. 1979. The People and the Land, pp 309-342. In M. Kalk, A. J. McLachlan and C. Howard-Williams (eds.) Lake Chilwa: 

Studies of Change in a Tropical Ecosystem. The Netherlands: Dr. W. Junk. 

Agnew, S. & Chipeta, C. 1979. Fishing and Fish-trading: Socio-economic Studies, pp 343-368 in Ibid. 

Allison, E. H., Patterson, G., Irvine, K., Thompson, A. B. & Menz, A. 1995. The Pelagic Ecosystem, pp 351-377. In A. Menz 

(ed.), The Fisheries potential and productivity of the pelagic zone of Lake Malawi/Niassa. Chatham, U.K.: Natural 

Resources Institute. 

Allison, E. H. & Ellis, F. 2001. The Livelihoods Approach and Management of Small-scale Fisheries. Submitted to Marine 

Policy. 

Allison, E. H., Mvula, P. M. & Ellis, F. 2001. Fisheries Management: by Whom, for Whom? Competing Agendas in the 

Development and Management of Fisheries in Lake Malawi. In Geheb, K. and Sarch, M-T. (eds.) Broaching the 

Management Impasse: Perspectives on Fisheries and their Management from the Inland Waters of Africa. Oxford: 

Oxford University Press. 

Ashley, C. & Carney, D. 1999. Sustainable Livelihoods: Lessons from Early Experience, London: Department for International 

Development. 

Bailey, C. & Pomeroy, C. 1996. Resource Dependency and Development Options in Coastal Southeast Asia. Society and Natural 

Resources 9: 191-199. 

Bailey, C. 1982. Small-scale Fisheries of San Miguel Bay, Philippines: Occupational and Geographic Mobility. ICLARM 

Technical Reports 10, 57p. 

Bakun, A. 1998. Ocean Triads and Radical Interdecadal Variation: Bane and Boon to Scientific Fisheries Management, pp 331- 

358. In T. J. Pitcher, P. J. B. Hart and D. Pauly (eds.) Reinventing Fisheries Management. Dordrecht, The Netherlands: 

Kluwer. 

Bayliss-Smith, T. 1991. Food Security and Agricultural Sustainability in the New Guinea Highlands: Vulnerable People, 

Vulnerable Places. IDS Bulletin 22(3): 5-11. 

Beddington, J. & May, R. M. 1977. Harvesting Natural Populations in a Randomly Fluctuating Environment. Science 197: 463-465. 

Béné, C., Mindjimba, K., Belal, E. & Jolley, T. 2000. Evaluating Livelihood Strategies and the Role of Inland Fisheries in Rural 

Development and Poverty Alleviation: The Case of the Yaéré Floodplain in North Cameroon. Proceedings of the 10 th 

International Conference of the Institute of Fisheries Economics and Trade, Corvallis, Oregon, July 9-14 th 2000. 

http://osu.orst.edu/dept/IIFET/2000/papers/bene.pdf 

Caddy, J. F., & Gulland, J. A. 1983. Historical Patterns of Fish Stocks. Marine Policy 7(4): 267-278. 

Carney, D. (ed.) 1998. Sustainable Rural Livelihoods: What Contribution Can We Make? London: Department for International 

Development. 

Chambers, R. 1989. Editorial Introduction: Vulnerability, Coping and Policy. IDS Bulletin 20(2): 1-7. 

Chipeta, C. 1972. A note on Lake Chilwa Cyclic Changes. Journal of Interdisciplinary Cyclic Research 3: 87-90. 

Chirwa, W.C. 1998. The Lake Malombe and Upper Shire River Fisheries Co-management Programme: An Assessment, pp 61-77. In 

A. K. Normann, J.R. Nielsen and S. Sverdrup-Jensen (eds.) Fisheries Co-management in Africa. U.K.: Institute for 

Fisheries Management and Coastal Community Development. 

Conway, D. in press. Extreme Rainfall Events and Lake Level Changes in East Africa: Recent Events and Historical Precedents. 

In E.O. Odada (ed.) The Second International Symposium on the Limnology, Climatology and Palaeoclimatology of 

the East African Lakes. In press, Kluwer. 

Conway, G. R. 1985. Agroecosystem Analysis. Agricultural Administration 20: 1-55. 

Conway, G. R. 1987. The Properties of Agroecosystems. Agricultural Systems 24(2): 95-117. 

Cushing, D. H. 1996. Towards a Science of Recruitment in Fish Populations. Oldendorf/Luhe, Germany: Ecology Institute. 

Davies, S. 1996. Adaptable Livelihoods Coping with Food Insecurity in the Malian Sahel. London: Macmillan Press. 

DeAngelis, D. L., & Waterhouse, J. C. 1987. Equilibrium and Nonequilibrium Concepts in Ecological Models. Ecological 

Monographs 57: 1-21. 

Ellis, F. 2000. Rural Livelihoods and Diversity in Developing Countries. Oxford: Oxford University Press. 

Environmental Affairs Department. 2000. Lake Chilwa Wetland – State of the Environment. Environmental Affairs Department, 

Malawi Government, June 2000. 

FAO. 1993. Fisheries Management in the South-east Arm of Lake Malawi, the Upper Shire River and Lake Malombe, with Particular 

Reference to the Fisheries on Chambo (Oreochromis spp). CIFA Technical Paper 21. Rome: FAO.

78 


FAO. 1996. Fisheries and Aquaculture in Sub-Saharan Africa: Situation and Outlook in 1996. FAO Fisheries Circular No. 922 

FIPP/C922. Rome: FAO. 

Ferguson, A., Derman, B. & Mkandawire, M. 1993. The New Development Rhetoric and Lake Malawi. Africa 63(1): 1-18. 

Geheb, K. & Binns, T. 1997. Fishing Farmers or ‘Farming Fishermen’? The Quest for Household Income and Nutritional Security 

on the Kenyan Shores of Lake Victoria. African Affairs 96: 73-93. 

Glantz, M. H. (ed.) 1992. Climate Variability, Climate Change and Fisheries. Cambridge: Cambridge University Press. 

GOM. 1999. Resource Report, 1999. Fisheries Department, Government of Malawi. 

Gordon, H.S. 1954. The Economic Theory of a Common Property Resource: The Fishery. Journal of Political Economy 62: 124- 

142. 

Haakonsen. 1992. A Regional Policy on Fishermen Migration in West Africa - Seminar on Artisanal Fishermen Migrations, 

Cotonou, Benin, 3-5 December 1991. Marine Policy 16: 237-238. 

Hilborn, R., & Walters, C. J. 1992. Quantitative Fisheries Stock Assessment: Choice, Dynamics and Uncertainty. U.K.: Chapman 

& Hall. 

Holling, C. S. 1973. Resilience and Stability of Ecological Systems. Annual Review of Ecology and Systematics 4: 1-23. 

Irvine, K., Patterson, G., Allison, E. H., Thompson, A. B. & Menz, A. 2001. The Pelagic Ecosystem of Lake Malawi, Africa: 

Trophic Structure and Current Threats, pp 3-30. In M. Munawar and R.E. Hecky (eds.) The Great Lakes of the World 

(GLOW): Food-web, Health and Integrity. Leiden, The Netherlands: Backhuys. 

Kalk, M. 1979. Focus on Social Problems, pp 417-432 in ibid. 

Kalk, M., McLachlan, A. J. & Howard-Williams, C. (eds.) 1979. Lake Chilwa: Studies of Change in a Tropical Ecosystem. The 

Netherlands: Dr. W. Junk. 

Kawasaki, T. 1983. Why Do Some Pelagic Fishes Have Wide Fluctuations in Their Numbers? Biological Basis of Fluctuation 

from the Viewpoint of Evolutionary Ecology. FAO Fisheries Report 291: 1065-1080. 

Lancaster, N. 1979. The Changes in the Lake Level, pp 41-58. In M. Kalk, A.J. McLachlan and C. Howard-Williams (eds.) Lake 

Chilwa: Studies of Change in a Tropical Ecosystem. The Netherlands: Dr. W. Junk. 

Larkin, P. A. 1977. An Epitaph for the Concept of Maximum Sustained Yield. Transactions of the American Fisheries Society 

106(1): 1-11. 

Mahon, R. 1997. Does Fisheries Science Serve the Needs of Managers of Small Stocks in Developing Countries? Canadian Journal of 

Fisheries and Aquatic Sciences 54(9): 2207-2213. 

Moss, B. 1979. The Lake Chilwa Ecosystem – A Limnological Overview, pp 399-416. In M. Kalk, A.J. McLachlan and C. Howard- 

Williams (eds.) Lake Chilwa: Studies of Change in a Tropical Ecosystem. The Netherlands: Dr. W. Junk. 

Moss, B. 1992. Uses, Abuses and Management of Lakes and Rivers. Hydrobiologia 243/244: 31-45. 

Normann, A. K., Nielsen, J. R. & Sverdrup-Jensen, S. 1998. Fisheries Co-management in Africa. U.K.: Institute for Fisheries 

Management & Coastal Community Development. 

Phipps, P. 1973. The ‘Big’ Fishermen of Lake Chilwa. A Preliminary Study in Entrepreneurism in Rural Malawi. In M.E. Page (ed.) 

Land and Labour in Rural Malawi. Rural Africana 21: 39-48. 

Pitcher, T. J. & Hart, P. J. B, (eds.) 1995. The Impact of Species Changes in African Lakes. London: Chapman & Hall. 

Plisnier, P.-D. 1997. Climate, Limnology and Fisheries Changes in Lake Tanganyika. FAO/FINNIDA: Research for the 

Management of the Fisheries on Lake Tanganyika. GCP/RAF/271/FIN-TD/72. ftp://ftp.fao.org/fi/ltr/TD72.PDF). 

Pomeroy, R. S. & Berkes, F. 1997. Two to Tango: the Role of Government in Fisheries Co-Management. Marine Policy 21(5): 

465-480. 

Reardon, T. & Vosti, S. A. 1995. Links Between Rural Poverty and the Environment in Developing Countries: Asset Categories 

and Investment Poverty. World Development 23(9): 1495-1506. 

Sarch, M-T. & Allison, E. H. 2000. Fluctuating Fisheries in Africa's Inland Waters: Well-adapted Livelihoods, Maladapted 

Management. Proceedings of the 10th International Conference of the Institute of Fisheries Economics and Trade. 

Corvallis, Oregon, July 9-14th 2000, 11 pp. 

http://osu.orst.edu/dept/IIFET/2000/papers/sarch.pdf 

Sarch M-T & Birkett, C. M. 2000. Fishing and Farming at Lake Chad: Responses to Lake Level Fluctuations. The Geographical 

Journal 166(2): 156-172. 

Scoones, I. 1998. Sustainable Rural Livelihoods: A Framework for Analysis. IDS Working Paper 72. Sussex, U.K.: Institute of 

Development Studies. 

Shaefer, M. B. 1954. Some Aspects of the Dynamics of Populations Important to the Management of Commercial Marine Fisheries. 

Bulletin of the Inter-American Tropical Tuna Commission 1: 25-56. 

Shepherd, J.G. 1991. Stability and the Objectives of Fishery Management. MAFF Laboratory Leaflet 64. Lowestoft, U.K: Directorate 

of Fisheries Research. 

Sholtz, U., Njaya, F. J., Chimatiro, S, Hummel, M., Donda, S. & Mkoko, B. J. 1998. Status and Prospects of Participatory Fisheries 

Management Programs in Malawi, pp 407 - 425. In T. Petr (ed.) Inland Fishery Enhancements. Fisheries Technical Paper 

374. Rome: FAO. 

Spencer, P. D., & Collie, J. S. 1997. Patterns of Population Variability in Marine Fish Stocks. Fisheries Oceanography 6(3): 188- 

204. 

Swift, J. 1989. Why Are Rural People Vulnerable to Famine? IDS Bulletin 20(2): 8-15. 

Talling, J. F. & Lemoalle, J. 1998. Ecological Dynamics of Tropical Inland Waters. Cambridge: Cambridge University Press. 

Thompson, A. B. & Allison, E. H. 1997. Potential Yield Estimates of Unexploited Pelagic Fish Stocks in Lake Malawi. 

Fisheries Management and Ecology 4: 31-48.

79 


Thompson, A. B., Allison, E. H. & Ngatunga, B. P. 1996. Distribution and Breeding Biology of Offshore Pelagic Cyprinidae and 

Catfish in Lake Malawi/Niassa. Environmental Biology of Fishes 47: 27-42. 

Tweddle, D. & Magasa, J. H. 1989. Assessment of Multispecies Cichlid Fisheries of the Southeast Arm of Lake Malawi, Africa. 

Journal du Conseil International pour l’Exploration de la Mer 45: 209-222. 

Tweddle, D. & Lewis D. S. C. 1990. The Biology of Usipa (Engraulicypris sardella) in Relation to Fluctuations in Productivity of 

Lake Malawi and Species Introductions, pp 67-72. In T. J. Pitcher and C. E. Hollingworth (eds.) Collected Reports on 

Fisheries Research in Malawi. Occasional Papers 1. London: Overseas Development Administration.

80 


An overview of indigenous knowledge as applied to natural resources 

management 

E.Y. Sambo 1 and R. Woytek 2 

1 University Office, University of Malawi, P.O. Box 278, Zomba, Malawi 

2 Operational Quality and Knowledge Services, Africa Region, The World Bank, 1818 H Street, NW, Washington, DC 20433, USA 

Abstract 

Historically, rural communities have acquired detailed knowledge, skills and strategies based on their interaction with the local 

environment over long periods of time. Local techniques and practices, which are essentially subsistence-oriented and are distinct to 

a particular social group and culture, have been developed and built up over centuries of experience and adaptation and are 

generally in harmony with the environmental conditions and responsive to constraints. This stock of knowledge is communicated in 

local languages and handed down from generation to generation. It is an expression of African tradition, rich in intricacies. When 

properly understood, this indigenous knowledge (IK) demonstrates how communities appreciate their environment and constantly 

seek to live in harmony with nature. Often this knowledge expresses itself and is communicated through songs, taboos, totem 

animals, custom laws and practices, place names and nicknames, riddles and proverbs. 

Indigenous knowledge permeates the social structure. It may have been influenced by innovations emerging from within itself, from 

other indigenous systems or from external systems but it essentially evolved locally. It is the basis for local-level decision-making in 

agriculture, food security, natural resource management, and a host of other activities in the rural communities. IK, therefore, deals 

with folk beliefs, skills, methods and practices. Some of the practices have led to mismanagement of resources, some are less 

efficient than modern technologies, while IK is generally less precise as measured by international science. IK, however, offers 

insights into possible alternative approaches to interpreting environmental and development change. 

Previously, external solutions offered in natural resource management projects often failed because they did not fit with the local 

knowledge and circumstances. It is now recognised that the local community's decisions to adopt or reject a new development 

project is strongly influenced by its existing skills, values and beliefs which are usually reflected in its indigenous knowledge. 

Although integration of IK in natural resources management may not be the ultimate answer in itself, there is increasing recognition 

of the value of local knowledge in production systems and rural development. Rural people with their detailed, holistic, integrated 

knowledge of local ecosystems, are experts in their own right. Utilising IK in the development process requires identification, 

validation, documentation and integration. 

This paper advocates that the opportunity should be seized to utilise the available IK for the fisheries resource management, the 

stabilisation of land-use activities in the catchment, the integrated monitoring programme of the lake ecosystem, and for the design 

of the institutional arrangements in the Lake Malawi Environmental Management Project. 

Introduction 

Renewable natural resources that are currently the subject of major impacts in Malawi include land, water, fish, 

wildlife and forests. Natural resource management poses major challenges as environmental degradation continues 

to worsen as a result of high population growth rate, poverty, low agricultural productivity, dwindling smallholder 

farmlands and escalating input prices, which collectively build up pressure on resource utilisation and exploitation. 

The discounted economic cost of soil erosion, deforestation, water resources degradation, fisheries depletion and 

biodiversity loss amounted to 10% of the GDP by 1994 and it represented a substantial income loss to the country 

(EAD 1998, p.9). This calls for innovative mechanisms that take into account the participation of all stakeholders in 

the natural resource management in order to arrest the declining situation. 

Historically, rural communities have acquired detailed knowledge, skills and strategies based on their interaction 

with the local environment over long periods of time. This stock of knowledge permeates the social structure and it 

is the basis for local-level decision-making in agriculture, food preparation, natural resource management, and a host 

of other activities in rural communities. It is an expression of African tradition, rich in intricacies. When properly 

understood, this Indigenous Knowledge (IK) demonstrates how communities appreciate their environment and 

constantly seek to live in harmony with nature. Often this knowledge expresses itself and is communicated through 

songs, taboos, totem animals, custom laws and practices, place names and nicknames, riddles and proverbs - all 

these have a story to tell (Matusse, 1995). Indigenous knowledge, therefore, refers to history, cultural heritage, and 

customs as developed in direct response to the physical and social realities...and in the context of Africa, indigenous 

means what is African, or what Africa has learned and adopted from other societies (Chavunduka, 1995). One 

scholar says that a particularly striking "philosophy of the indigens is empathy, selflessness, service to other beings, 

unconditional caring, honesty, accountability, respect for all members of your society ...." (Nxumalo, 1995). 

Such a complex collection of knowledge, described as indigenous knowledge systems (IKS), ethnoscience or 

traditional wisdom (Kipuri 1995) evolves over time, has collective ownership and is communicated orally from one

81 


generation to the next. Before rural communities experienced any institutionalised foreign intervention, it is these 

indigenous knowledge systems that have regulated the level of exploitation and utilisation of environmental 

resources and ensured their survival. The land and its resources belonged to the people, held in trust by the chief, 

and the chief had the management authority as regards access to the resource, permission for usage, period of usage 

and even levels of usage. Resource utilisation was effectively regulated since this system was embedded in 

traditional governance structures of the society that excluded challenging authority of the chief. In the post-colonial 

era, the traditional structures became weakened as a result of partial or complete transfer of authority over the 

resource, to the different sectors of government. 

The United Nations Conference on Environment and Development (UNCED), commonly known as the Earth 

Summit, which convened in Rio de Janeiro in June 1992 adopted a global plan of action. The Agenda 21 is 

considered a blueprint for a global partnership aimed at reconciling the twin requirements of a high quality 

environment and a healthy economy for all peoples of the world. Chapter 26 of Agenda 21 "Strengthening the Role 

of Indigenous People" provides a framework for recognising Indigenous Knowledge Systems (Box 1). 

Box 1. Strengthening the Role of Indigenous People 

“Indigenous people, who represent a significant part of the world’s population, depend on renewable resources and 

ecosystems to maintain their well-being. Over many generations, they have evolved a holistic, traditional scientific 

knowledge of their land, natural resources and environment. The ability of indigenous people to practise sustainable 

development on their lands has been limited by economic, social and historical factors. Governments should recognise 

that indigenous lands need to be protected from environmentally unsound activities, and from activities the people 

consider to be socially and culturally inappropriate. There should be national dispute-resolution procedures to deal with 

concerns about the settlement of land and use of resources. Some indigenous people may require greater control over 

their lands, and self management of their resources. They should also participate in development decisions that affect 

them, and in the creation of protected areas, such as parks. 

Governments should incorporate the rights and responsibilities of indigenous people into national legislation. Countries 

could also adopt laws and policies to preserve customary practices, and protect indigenous property, including ideas 

and knowledge. Indigenous people should be allowed to actively participate in shaping national laws and policies on 

the management of resources or other development processes that affect them. Governments and international 

organisations should recognise the values, traditional knowledge and resource management practices that indigenous 

people use to manage their environments, and apply this knowledge to other areas where development is taking place. 

They should also provide indigenous people with suitable technologies to increase the efficiency of their resource 

management.” 

Source: Keating (1993), Chapter 26 

The significance of traditional knowledge is also incorporated in the International Convention on Biodiversity (Box 

2). Previously, solutions offered by some development projects failed because they did not fit with the local 

knowledge and circumstances. Many organisations such as the World Bank and IUCN (International Union for 

Conservation of Nature and Natural Resources) have in the past decade began to directly address environmental 

issues through promoting environmental sustainability in their development efforts and emphasising country 

ownership (Asibey, 1995). It is now recognised that the local community's decisions to adopt or reject a new 

development project is strongly influenced by its existing skills, values and beliefs which are usually reflected in its 

indigenous knowledge. Thus, involvement of the community at all stages "is likely to improve project design by 

ensuring that full advantage is taken of the local technology and indigenous knowledge, and by ensuring that the 

project is fully adapted to the social organisation and value systems of the community" (Matusse, 1995). 

Such a bottom-up approach empowers the community to take charge of resource management in development activities. 

It is to be admitted, however, that due to cultural hybridisation there has been inevitable loss of IKS. In any case, during 

the colonial era, this knowledge was often regarded as primitive, scorned at, marginalised and dismissed as worthless 

(Matowanyika et al., 1995), so that preference towards being "modernised" or "westernised" has played a major role in 

eroding IKS. In addition, one of the biggest pitfalls is that most of African IKS remains unrecorded. It is, therefore, 

essential that researchers, extension agents and rural development experts study indigenous knowledge. Before imposing 

essentially foreign perspectives and technologies their appropriateness to the local setting should be examined.

82 


This paper advocates the utilisation of available IK in the Lake Malawi Environmental Management Project. 

Various studies undertaken in Malawi over the previous years suggest relevant IK is available for the fisheries 

resource management, the stabilisation of land-use activities in the catchment, the integrated monitoring programme 

of the lake ecosystem, and for the design of the institutional arrangements of the programme. 

Box 2. Convention on Biological Diversity 

“The world’s biological diversity - the variability among living organisms - is valuable for ecological, genetic, social, 

economic, educational, cultural, recreational and aesthetic reasons. The diversity is important for evolution, and for 

maintaining the life-sustaining systems of the biosphere. The conservation and sustainable use of biological diversity 

are of critical importance to meet the food, health and other needs of the growing world population. However, biological 

diversity is being significantly reduced by certain human activities, and it is vital to anticipate, prevent and attack the 

causes of this loss. Substantial investments are required to conserve biological diversity, but they will pay off with a 

broad range of environmental, economic and social benefits. 

The world needs to conserve biological diversity and make sustainable use of its components in a fair and equitable 

way. Sustainable use means use in a way and at a rate that does not lead to the long-term decline of biological 

diversity. This will maintain its potential to meet the needs and aspirations of present and future generations. The uses 

include those of genetic material, which is any plant, animal, microbial or other material containing functional units of 

heredity. We also need to conserve ecosystems, which are groupings of living and non-living material that act as a unit. 

Countries have rights over their biological resources, but they are also responsible for conserving their biological 

diversity and for using their biological resources in a sustainable manner… 

Many indigenous and local communities have a close dependence on biological resources, and nations should make 

use of this traditional knowledge of the conservation and sustainable use of biological diversity. Countries are to 

preserve and maintain such indigenous and local knowledge and promote its wider use. This is to be done with the 

approval and involvement of those who have such knowledge, and these people should benefit from the use of their 

practices….” 

Source: Keating (1993) p.66 

The scope of IK 

Indigenous knowledge encompasses local knowledge that is developed outside the formal systems of science and 

education. It is embedded in community practices and institutions. It is locally managed and unique to a given 

culture or society. IK includes all information, experiences and practices community members have acquired over 

time to maintain or improve their livelihoods, to produce and to engage in social, cultural, and productive activities, 

to cope with crises or to manage natural resources. Rural people with their detailed, holistic, integrated knowledge 

of local ecosystems, are experts in their own right. They are capable of classifying and structuring environmental 

data and knowledge using their own terminology of understanding environmental processes and developing skills 

and strategies. IK is not static but becomes modified as the learning process demands fine adjustments. The 

community uses IK as the basis for decision making in agriculture, health care, food preparation, education, naturalresource 

management and as a coping mechanism in general. IK is primarily tacit knowledge of communities, 

elders, women, farmers, healers, artisans, spiritual leaders, etc. As such, it is not easily codifiable (World Bank, 

1998). 

Traditional ecological knowledge, traditional technical knowledge and traditional practices enable people make 

choices in the use of natural resources. Facets of such knowledge systems and practices include: local soil and 

biodiversity classification, traditional hunting norms and regulations, traditional medicine (both human and ethnoveterinary), 

traditional institutions of chieftainship, community based land allocation (such as for sacred groves), 

agricultural production systems (such as slash and burn), informal savings society, etc. Such a knowledge base 

persists and defines the limits of local culture. Thus in order to maintain the culture, IK must first be shared. 

Various methods of exchanging IK exist, including: apprenticeship; artifacts; demonstration; story telling, myths, 

legends, taboos and songs; training etc. Secondly, IK must be used, developed and preserved. Because IK is 

primarily developed and communicated outside the formal systems of science and education, it is frequently 

overlooked in the development process.

83 


Rationale for integrating IK in development 

The resilience of IK as it gets passed on from generation to generation implies that it is a component of human and 

social capital. This suggests that the community has had the opportunity to decide on what values, practices and 

beliefs to maintain, on the premise that there is general acceptance. Basically, community empowerment is achieved 

when choices are endogenous and any future activities based on such consensus are likely to be supported. In such 

situations, IK will increase efficiency and effectiveness, thus making IK a basis for sustainable development. 

It is important to recognise that communities can be helped to make better use of IK for development. This can only 

happen by first learning from them (i.e. recognising and identifying the IK), validating, recording, documenting, 

storing it in a retrievable form, transfering it back to the communities, then disseminating successes to a wider 

community, and finally promoting the exchange of the successful IK between communities. Learning from the 

community and helping the community learn (a bottom-up approach) can lead to empowerment through enablement. 

The following approaches have proven to help in creating the environment for enablement: 

� exchange knowledge of local practices in order to build local knowledge networks; 

� engage researchers and experts in order to determine research agendas; 

� engage authorities in dialogue in order to enhance good governance; 

� dialogue with development partners in order to shape endogenous development agenda; and 

� leverage local and global knowledge in order to integrate IK in development. 

MAINSTREAMING IK INTO DEVELOPMENT ACTIVITIES 

The Policy Environment 

In the past, the value of harnessing global and local knowledge for development has not received the needed 

attention in developing countries in general and in Africa in particular. With the advent of Information and 

Communication Technologies this appears to be changing, especially since the early 1990’s. In Malawi, the National 

Science and Technology Policy and the draft Science and Technology Bill have been endorsed by stakeholders in 

May 2001. The S&T Policy provides for the identification, development, protection and promotion of IK. This will 

help move IK from the realm of folklore into the development domain. 

IK in Lake Malawi Environment Management Project (LMEMP) 

IK could be the most important resource of small-scale fishing and farming communities in the Lake Malawi basin. 

There is no other resource that is owned, controlled and managed to the same extent as IK. The proposed LMEMP 

offers a unique opportunity to utilise the available IK for the fisheries resource management, the stabilisation of 

land-use activities in the catchment, the integrated monitoring programme of the lake ecosystem, and the design of 

the institutional arrangements. 

IK-related activities would identify options and formulate actions for integrating the IK in the management of 

natural resources. It is a challenge in so far as the approach departs from the conventional practices of project 

implementation. The process will involve: 

making an inventory of actors, resources, publications; 

convening a national exchange involving researchers, NGOs, CBOs, local knowledge bearers, etc; 

determining areas of relevant IK for the LMEMP; and 

agreeing on action plan involving IK research, exchange and institutional arrangements. 

The expected outcomes and impact include: 

increased ownership of the management of the lake basin resource by local communities; 

higher productivity of the lake basin resources; and 

sustainable access to natural resources by communities. 

Conclusion 

There is a global recognition that indigenous knowledge has an important role to play in consonance with modern 

scientific and technological intervention in social and economic development and cultural and political 

transformation. IK is a valuable resource of the local community that needs to be used to the community’s 

advantage. Incorporating IK in the LMEMP will provide not only a testing ground but will also offer an opportunity 

to understand and internalise IK in the development planning.

84 


References 

Asibey, E.O.A. 1995. Indigenous knowledge systems and prudent natural resource management. In: Matowanyika, 

J.Z.Z., Garibaldi,V. and Musimwa, E. (eds.). Indigenous Knowledge Systems and Natural Resource 

Management in Southern African. Report of the Southern African Regional Workshop, Harare, Zimbabwe, 

20-22 April 1994, IUCN-ROSA, Indigenous Knowledge Systems Series No.1. 

Chavunduka, G.L. 1995. Keynote address by Professor G.L. Chavunduka, Vice-Chancellor of the University of 

Zimbabwe. In: Matowanyika, J.Z.Z., Sibanda, H. and Garibaldi, V. (eds.). The Missing Link: Reviving 

indigenous knowledge systems in promoting sustainable natural resource management in southern Africa. 

Proceedings of a Regional Workshop, held in Midmar, Kwazulu-Natal Province, South Africa, 23-28 April, 

1995. IUCN-ROSA Indigenous Knowledge Systems Series No.2. 

EAD (Environmental Affairs Department). 1998. Environmental Management Framework. In: State of Environment 

Report. Malawi Government, Lilongwe. 

Keating, M. 1993. Agenda for Change: A plain language version of Agenda 21 and the other Rio Agreements, 

Chapter 26, Published by the Centre for Our Common Future. 

Kipuri, Naomi. 1995. International perspectives and importance of indigenous knowledge systems and institutions 

in natural resource management: contributions and experiences in the pastoral Maasai of East Africa. 

Paper presented at the Workshop on Indigenous Knowledge Systems in Natural Resource Management. 

Held in Kwazulu, Natal Province, South Africa, 23-28 April 1995. 

Matowanyika, J.Z.Z., Sibanda, H. and Garibaldi, V. (eds.). 1995. The Missing Link: Reviving indigenous 

knowledge systems in promoting sustainable natural resource management in southern Africa. 

Proceedings of a Regional Workshop, held in Midmar, Kwazulu-Natal Province, South Africa, 23-28 April, 

1995. IUCN-ROSA Indigenous Knowledge Systems Series No.2. 

Matusse, R. (1995). Indigenous people and people-centred development: Are we moving South to meet ourselves? 

Paper presented at the IUCN sponsored Workshop on Indigenous Knowledge Systems in Natural Resources 

Management in Southern Africa. Held at Midmar, Kwazulu-Natal Province, South Africa, 23-28 April 1995. 

Nxumalo, O.E.H.M. (1995). Keynote address by Professor O.E.H.M. Nxumalo: Indigenous knowledge systems and 

natural resource management in South Africa. Delivered to the Regional Workshop on the Study and 

Promotion of indigenous Knowledge Systems and natural Resources Management in Southern Africa, at 

Fern Hill - Harwick, Natal Province, South Africa, 23-28 April 1995. 

World Bank. (1998). Indigenous Knowledge for Development: A Framework for Action. Knowledge and Learning 

Center, Africa Region, The World Bank.

85 


Decentralised environmental management and the implications for fisheries comanagement 

in Lake Malawi 

John D. Balarin 

Chief Technical Advisor, Danida Environment Support Program, Pvt Bag 396, Lilongwe, Malawi, (tel 265-836533, e-mail: 

DESPS@malawi.net) 

Abstract 

Malawi has made a political commitment to decentralization through the recent appointment of the District Assemblies. Environment 

Management, has been decentralised since 1996. The Environment Act requires communities to participate actively at district level 

in State of the Environment Reporting (SOER) and Environmental Action Planning (EAP) and management. These processes have 

now become integral to the District Development Planning (DDP) system and are the responsibility of the Assemblies. Community 

consultation and participation is fundamental to the process. The resultant EAPs are implemented at target group level through 

micro-projects. As decentralised environmental management takes hold, communities will become better empowered to assess their 

own needs, to plan and act accordingly. In addition, the empowerment of the Assemblies as custodians of national laws through 

district by-laws will bring communities closer to policing and becoming resource managers. In addition, a degree of security of 

tenure can be secured through the by-law and this may allow a socio-economic change in community behaviour from 

“hunter/gatherer” to introduction of husbandry system options. 

This paper elucidates the SOER, EAP and DDP process, its implications for the BVCs and the future role of community participation 

in the co-management of Lake Malawi. 

Introduction 

With the election of Councilors and the formation of District Assemblies in late 2000, Malawi has taken a great step 

closer towards making community based natural resource management (CBNRM) a reality. CBNRM and 

decentralization are inextricably linked (Balarin et al. 2001) in as much as the latter transfers power for 

administration and management from a central level to a lower, more appropriate, district, or sub-district level. 

Decentralization strives, through consultative processes and popular participation to empower communities to 

become pro-active in the planning and decision making process of development, and advocates for change in policy. 

It also brings them closer to being the custodians of the resource with self-imposed by laws and more secure access 

rights. 

Tools to achieve this are now in place, described in the Local Government Decentralization Strategy (OPC, 

1998)(LGDMP, 1999) and the Strategy for Decentralised Environmental Management (DEM)(EAD, 2000, 2001 a, b 

and c). 

The 1999 Fisheries and Aquaculture Policy of Malawi has “collaborative management” as one of its key objectives. 

Subsequent, subsidiary legislation for the first time in the history of fisheries development in Malawi, empowered 

communities through the Act, to formulate by-laws that can self-govern the management of the fisheries resources in 

their locality. It would be myopic to think that the fisheries legislation alone is adequate. To be successful, any 

fisheries natural resource management strategies would need to be coupled with decentralization, and include crosscutting 

elements from the recent trends in other sectors (notably forestry, wildlife, water, etc). The stage, is set, now 

more-so than before for Malawi to embark on some of the first steps towards fisheries co-management. 

CBNRM Defined 

Guiding Principles of CBNRM 

Some guiding principles for CBNRM are, in brief (adapted from COMPASS, 2000a): 

a. Communities should be prime beneficiaries and take the lead role, actively participating in identifying, 

planning and implementing CBNRM activities. 

b. CBNRM activities should be managed by democratically elected institutions or committees linked to the 

local authority. 

c. Communities must develop clearly defined constitutions for their CBNRM institutions and establish bylaws 

for CBNRM in conformity with national policy. 

d. User groups and boundaries must be clearly defined with clear rights of access, lease or ownership. 

e. Natural resources should be treated as economic goods and any intervention should be seen to have 

tangible value added benefits to communities.

86 


f. CBNRM programs must be gender sensitive, promote equitable sharing of costs and benefits and be 

supportive of community priorities. 

Co-management vs CBNRM 

Most natural resources in Malawi are state owned and have largely been state regulated for the benefit of all present 

and future generations. However, in recognition that management initiatives that do not involve the resource users, 

is not sustainable, CBNRM strategies aim to place responsibility for resource management at the lowest appropriate 

level. A spectrum of co-management options emerge (Figure 1)(Balarin et al. 2001). Essentially, the current 

situation of 100 % state management of resources, should gradually giving way to progressively more resources 

coming under varying degrees of joint management between state and user, until eventually, for some resources, the 

community is 100 % in control. 

Figure 1. Conceptual Diagram of the different stages of co-management 

Husbandry vs Natural Resource Management 

The NEAP (1994) identified that the nexus to environmental degradation in Malawi is poverty and illiteracy. 

Because of poverty, foraging from natural resources has been a fundamental sustainable livelihood strategy, 

especially in times of need. CBNRM strategies should therefore not only aim at the sustainable management of the 

status quo of ecosystems, (ie. cropping according to sustainable maximum yield), it should seek out sustainable 

economic benefit streams that offer incentives to the user to diversify into alternative, value added uses or income 

generating activities. For CBNRM to succeed, extension strategies would have to introduce the resource user to 

knowledge of “best practices” of equal or better economic significance. 

Husbandry of the resource is one option capable of sustainable enhancement of production. To maintain the needs 

and pressures of growing human populations, implementation of CBNRM needs to bring about change in 

community behaviour. Communities need to change from present practices of hunter/gatherer scenario, and move a 

socio-economic step forward to the next stage in evolution of husbandry systems (ie. the active farming of 

resources) (see Figure 2)(Balarin et al. 2001).

87 


Figure 2. Conceptual diagram of evolution trends in natural resource husbandry. 

Institutional Framework 

History of National Environmental Legislative Framework 

The history of some of the more relevant steps in legislative and policy development that have had influence on 

environmental management and fisheries are as follows: 

• 1994: National Environmental Action Plan (NEAP)(Vol I and II) 

• 1996: Environment Management Policy 

• 1996: Environment Management Act 

• 1996: Decentralization Policy 

• 1997: Fisheries Conservation and Management Act 

• 1998: EIA Guidelines 

• 1998: Environment Support Program (ESP) 

• 1998: First National State of the Environment Report (SOER) 

• 1998: Local Government Act 

• 1999: Fisheries and Aquaculture Policy 

• 2000: Subsidiary legislation (Community by-laws) 

• 2000: First Lake Chilwa State of the Environment Report 

• 2000: First District SOERs 

• 2001: First Lake Chilwa Management Plan. 

• 2001: First District EAPs 

• 2001: Poverty Reduction Strategy Paper (PRSP)

88 


The NEAP was Malawi’s response to Rio 1992, and set the stage for the formulation of the Environment Policy and 

Act. The Environment Management Act of 1996 placed responsibility for environmental management at the district 

level. Environmental monitoring, through State of the Environment Reporting (SOER) every 2 years and 

development of mitigation measure through Environmental Action Planning (EAP) every 5 years, became the 

responsibility of the Assemblies assisted by a Environment District Officer (EDO). 

The Environment Act established a clear role for the mainstreaming of in natural resource management (NRM) in 

the local government structures, notably the Assemblies, Area Development Committees (ADC), Village 

Development Committees (VDC), and associated community institutions. Environment Impact Assessment (EIAs) 

were also a legal requirement for all development projects and the EIA guidelines established community 

consultation as an essential part of the process. Despite the Decentralization Policy of 1996, it was however not 

until the Local Government Act of 1998 that decentralization became a national objective, to place administrative 

authority for natural resource management (NRM) at the lowest appropriate level. This became effective in 

November 2000, when local government elections lead to the formation of the District Assemblies. 

The legal and political framework is now in place to empower communities to take a more pro-active role in NRM. 

The Environment Act and Local Government Act have become integrated through the Environmental Affairs 

Department (EAD) strategy for decentralised environmental management (EAD 2000, 2001 a, b and c). SOERs and 

EAPs are now integral parts of the decentralised, District Development Planning (DDP) process (OPC, 1998). 

The District Development Planning Framework and Fisheries Co-management 

In the various stages in the District Development Planning (DDP) process, Figure 3 (OPC, 1998), community 

participation is essential. The DDP cycle from situation analysis, preparation of a district socio-economic profile 

(SEP) culminating in a District Development Planning Framework (DDPF) is synonymous with SOER. SOER is a 

process of information compilation and analysis and a tool for information dissemination for better planning and 

decision making. The process assists communities and planner alike to identify environmental problems, and to 

illustrate trends and spatial distribution of pressure, state and response indicators (EAD, 2001 a). Prioritization of 

environmental concerns or hot spots, leads to the formulation of appropriate development objectives and strategies. 

The next stage of the cycle from identification of mitigating actions, development options and formulation of the 

DDP, for the environment sectors, is synonymous with the DEAP. DEAPs are community decisions for mitigation 

or environmental management of problems. Communities decide for each priority environmental concern, the 

action, by who, where, when and how (EAD, 2001 a.). Thereafter, once approved by the Assembly, DEAPs are 

implemented through community based environmental micro-projects and an element of participatory M&E is 

involved. 

In so much as the DDP process is consultative and community based, so are the preparation of SOERs and EAPs. 

The process is flexible. It can be applied as a “fast track”, a means to identify priority hot spots that need immediate 

intervention even before the SOER and DEAP process is complete. The SOER and DEAPs are divided into chapters 

by sector, and fisheries is dealt with as a separate Chapter. The first district based SOERs and DEAPs are now under 

active preparation. Table 1 summarizes the status for the fisheries sector around Lake Malawi. 

Table 1. Status of decentralised environmental management in Lake Malawi Districts 

District 

Year DESC 

Status of SOER Status of EAPs 

Formed DSOER ADC SOER DEAP ADC EAP 

Karonga 1999 Draft Final Draft Final 

Rumphi 2000 Draft underway Draft underway 

Likoma 2001 not started not started not started not started 

Nkhata Bay 1998 Draft underway Draft underway 

Nkotakota 2000 Draft underway Draft underway 

Salima 2000 Draft underway Draft underway 

Dedza 1998 Draft underway Draft underway 

Mangochi 1998 Draft underway Draft underway 

Key: DESC - District Environment Sub-Committee; DEAP - District Environmental Action Plan; DSOER - 

District State of the Environment Report; ADC - Area Development Committee

Figure 3. SOER and DEAP in relation to District Development Planning System 

89 


National Environment Management Framework and Fisheries Co-management 

The framework document, the National Environment Policy of Malawi recognizes the importance of CBNRM and 

calls for “national, regional and district development plans to integrate environmental concerns, in order to improve 

environmental management and ensure sensitivity to local concerns and needs”. 

The DEM strategy (EAD, 2000 a) allows for community based plans, separated by sector, to feed upwards from 

community to VDC to ADC level and be consolidated into the DEAPs. DEAPs aggregate into the NEAP from 

which emerge sector investment programs (SIPS). All NRM SIPs are compiled into the Environment Support 

Program (ESP) and further, are consolidated into the national Poverty Reduction Support Program (PRSP), the 

umbrella document for all national plans and government budgeting (Fig 4 a and b). 

Likewise, the SOERs from communities, pass to VDC to ADCs and consolidated in the DSOERs, are similarly 

aggregate into the NSOER and become the M&E tool of the NEAP. The SOERs are reported by the Minister to 

parliament every year and thereby provide a powerful opportunity for communities and districts to communicate 

their concerns and needs to the highest policy and decision making level. 

The institutionalization of the SOER DEAP process as part of the tools of fisheries co-management would provide 

the sector with a powerful tool for communities to lobby at the highest level for resources.

Figure 4a. CBNRM and the PRSP (adapted from the PRSP working group). 

90 


Figure 4b. The national framework for integration of EAPs into the national Budget.

91 


National Institutional Framework for Environmental Management 

Strategic objectives to achieve CBNRM would be to ensure that the emerging local government structures at District 

Assembly, Area Development Committee (ADC), Wards and Village Development Committee (VDC), provide a 

logical institutional anchor point for community based organizations such a Beach Village Committees. CBNRM 

should therefore be seen as a mainstreamed role of local government, and fits within the existing institutional 

framework for decentralised environmental management (Figure 5), notably: 

• Parliament is the highest policy making authority and it is advised by a Parliamentary Committee on the 

Environment (PCE) that serve as the watchdog of policy and the state of the environment. 

• The Ministry co-ordinates the implementation of policy, law enforcement and EIA inspections, establishing 

standards and maintains international representation, co-ordinating conventions and donor support. The 

Ministry is advised by the National Council on the Environment. 

• The various NRM departments administer policy directives, formulate national plans and policy, guiding 

the districts, provide capacity building support, facilitate EIAs and compile sector NSOERs and NEAPs. To 

co-ordinate these sectors, national environmental focal points have been established and there is a National 

Steering Committee of these focal points, made up of the directors of each sector. 

• At district level, the Assemblies co-ordinate preparation of the SOERs and DEAPs, and facilitate the 

implementation of micro-projects and awareness campaigns. 

• The Assembly is assisted by a District Environment Sub-Committee (DESC)(Figure 6) made up of focal 

points from each of the Technical Directorates, assisted by the EDO and anchored in the office of the 

District Planning Director. They serve to prepare the SOERs and EAPs, train the ADCs and VDCs, oversee 

micro-projects and generally mainstream environmental management into all planning processes at all 

levels. 

• Environment management should be mainstreamed in the roles of all the ADCs, Wards and VDCs, all the 

way down to the communities. 

Decentralization and Fisheries Management 

Fisheries management in Malawi suffers from a number of concerns (Mapila, 2001), amongst which are: 

• A poor track record of enforcement of the law. 

• Unrestricted access due to problem in allocating security of tenure. 

• Poor access to research knowledge by the user groups. 

• Slow take off of aquaculture. 

• No national or regional coordination mechanism. 

Under the Local Government Act, envisaged is that Assemblies will establish by laws and that these shall be 

extended all the way down through the subsidiary authorities to community level. In fisheries, the fishing 

communities could be governed by district by-laws which take their point of departure from the national law. The 

assemblies could mobilize their legal and administrative powers to: 

• Police the resource, monitor use patterns and enforce regulations through the district police system. 

• Policing by the Assembly, would free up the Fisheries Officer, allowing a change from a policeman role to 

becoming a friend of the community and a technical advisor whose sole purpose would be to impart 

knowledge to change fishermen’s behaviour. 

• District by laws could provide rights of access through allocation of lease rights to VDCs who pay lease 

fees based on limited entry licensing. They in turn through the BVCs, allocate resource use rights and 

police those rights through their constitutions. Security of tenure will ultimately lead to behaviour change 

and see the introduction of husbandry systems. 

• The Local Government Act empowers one or more Assemblies to come together to form an Inter-district 

Committee for common resource management like Lake Malawi. This could be extended to provincial 

administration in neighbouring states. 

In the coming years, the Fisheries Department will have to embark upon a devolution strategy for the sector. Some 

of the above becomes food for thought.

92 


Figure 5. Institutional framework for the coordination of environmental management

Figure 6. Institutional framework for decentralised environmental management. 

93 


Implications for Fisheries Co-management 

In conclusion, in fisheries the implications of decentralization are that the Beach Village Committees (BVC) now 

find an anchor point in the District Assembly for legal recourse to enforce their constitutions through district bylaws. 

In addition, access rights through leasehold becomes possible and this will bring with it a change in behaviour 

favouring a more husbandry type approach to management. 

In addition, the BVCs through participation in the SOER process are able to undertake a consultative situation 

analysis, to better understand their problems and to identify problem areas or environmental “hot spots” and bring 

these to the attention of the authorities. SOERs are published as district reports, consolidated into national reports 

and presented to Parliament every year. This provides communities a powerful tool to lobby at the highest level for 

support. SOERs allow community prioritization of problems to be dealt with through EAPs. 

The participatory process of EAP preparation means that BVCs participate in planning and take ownership of their 

own remedial actions to mitigate against prioritized problems. Generally these are expressed in the form of 

environmental micro-projects and would be included in the DDPs. At an inter-district level, the 8 Districts 

surrounding Lake Malawi will be able to come together and their combined DDPs would make up a coherent and 

comprehensive Lake Malawi Management Plan. A similar initiative has already been completed for Lake Chilwa. 

Under the Local Government Act, the 8 districts could form an inter-district committee to manage Lake Malawi.

94 


The DDPs provide the District Assemblies with a basis for seeking funding from government, donors and other 

agencies. Currently available to the Lake Malawi BVCs are potential environmental micro-project funds from 

Danida, World Bank, Malawi Environment Endowment Trust (MEET), COMPASS/USAID, EU and GTZ. 

Key References 

Balarin, J. D., Jensen, A. and Ndovi, W (2001) Developing CBNRM Planning and Implementation Tools. 

Proceedings, National conference to Develop CBNRM Strategic Plan for Malawi, COMPASS: 14 p. 

COMPASS (2000a) Workshop on Principles and Approaches for CBNRM in Malawi. COMPASS Document 10: 

100p (March 2000). 

COMPASS (2000b) Draft Strategic Plan for CBNRM in Malawi. Compass Document 23: 56p (November 2000). 

EAD (2000) Strategy for Decentralization of Environmental Management. EAD publication: 50p. (Draft October 

2000). 

EAD (2001a) Decentralised Environmental Management Manual. Vol. I: A guide to 

SOER, DEAP and micro-project preparation. EAD: 126p. (Draft January 2001). 

EAD (2001b) Decentralised Environmental Management Manual. Vol. II: Data Capture Tool for SOER. EAD 

publication: 75p. (Draft January 2001). 

EAD (2001c) Decentralised Environmental Management Manual. Vol. III: A guide to environmental micro-projects. 

EAD publication: .. p(in preparation, April 2001). 

LGDMP (1999) Village Action Planning Manual for AEC level facilitators. LGDMP publication:160p. 

Mapila, S. A. (2001) Towards Responsible Development of the Fisheries Sector in Malawi. Department of Fisheries: 

41 p. 

OPC (1998) District Development Planning Handbook. OPC publication:128p. (March 1998).

Status of the small scale fishery in Malawi 

Mackson J.R. Ngochera 

Fisheries Research Unit, P. O. Box 27, Monkey-bay. 

95 


Abstract 

Most of Malawi’s population is dependent on fisheries directly or indirectly as a source of food security, livelihood and income. Total 

catch in the recent years for the major water bodies, Lake Malawi, Lake Malombe, Lake Chiuta, Upper Shire and Lower Shire have 

declined from an average of 60 thousand metric tons in the period 1976-1990 to 49 thousand metric tons in 1991-1999. Over the 

same period, the number of fishermen increased by 27%, similarly the number of fishing gears and fishing crafts has also increased. 

Introduction 

The fish stocks of Malawian waters are, undoubtedly among the most important natural resources of Malawi. Out of 

the 120000 km 2 area covered by Malawi, 20% is water (Weyl, 2001). There is therefore little doubt that a large 

number of Malawi’s population depends directly or indirectly on the fishery as a source of food security, livelihood 

and income. The value of fish does not lie only in their scientific interest, but also in their primordial nutritional 

status. It is a known fact that fish provides some 70% of the animal protein consumed by Malawians (Mkoko 1992, 

cited by Ribbink (1994) and Turner (1994b)). Currently the fishery employs over 48 000 fishermen (Weyl et al. 

2000) and fish also contributes about 4% to the Gross National Product (GNP) of the country. 

However, it has been noticed in the recent years that fish catches in Malawi have declined. The reasons are: the ever 

increasing fishing effort in the shallow waters; the absence of alternative employment; the rapidly growing human 

population, which exerts an increased fish demand, on the overexploited stocks (Turner 1995) and the use of 

destructive fishing gears. 

The aim of this report is to provide an account of the status of the different fisheries and fish stocks of the major 

water bodies of Malawi. In agreement with the current practices in fisheries management an attempt is made to 

apply the principles of the precautionary approach. The FAO Code of Conduct for Responsible Fisheries, adopted in 

1995 stipulates in article 7.5: “States should apply the precautionary approach widely to conservation, management 

and exploitation of living aquatic resources in order to protect them and preserve the aquatic environment. The 

absence of adequate scientific information should not be used as a reason for postponing or failing to take 

conservation and management measures.’’ 

The findings presented in this report are based on catch-effort data collected by the Department of Fisheries during 

the 1976-1999 period. The data was used to calculate CPUE (Catch Per Unit Effort) and relative effort. In the 

analysis of catch and effort data, standard methods have been applied. For details of the methodology, a reference is 

made to Sparre and Vanema’s Introduction to Tropical Fish Stock Assessment (FAO 1992) where treatment of catch 

and effort data are explained in detail. 

Total catch in Lake Malawi 

Lake Malawi, the southern most of the African Rift Valley Great Lakes, supports fisheries yielding some 30,000 

tons of fish annually from waters controlled by Malawi (Tweddle and Magasa, 1989). Over 88% of the total 

landings in Lake Malawi, come from the small-scale fishery although there exists a large-scale commercial fishery. 

Total catches in the small-scale fisheries in Lake Malawi, for the period 1976-1999 is shown in Figure 1. A clear 

increasing trend in the period 1976-1989 is observed and peaks of more than 30 000 metric tons was attained in 

1987, 1989, 1996 and 1997. Catches fluctuated widely during the period 1990 – 1999 and the highest catch of 

40 000 metric tons was recorded in 1996. The reason for this was the good usipa catch (Engraulicypris sardella) in 

that year. 

Figure 2 shows the average contribution to the total Lake Malawi small-scale fishery by the various districts. 

Mangochi and Salima accounted for 45% and 18% respectively. The high fish yields from these two districts are 

mainly due to the suitable limnological conditions. In addition to the suitability of the area for highly productive 

fishing methods such as trawling and seining, the broad shallow shelf of these areas supports high benthic 

productivity. The shape of the lake and the direction of the prevailing winds mean that seasonal upwelling of 

nutrient-rich water occurs mainly in the south. (Eccles, 1974 cited by George F. Turner). Most of the rest of the lake

96 


is often rocky and very deep and the fisheries are mainly dependent on pelagic zooplankton-feeding fish, which in 

turn are supported by low pelagic primary productivity. 

Catch(metric tons) 

40000 

35000 

30000 

25000 

20000 

15000 

10000 

5000 

0 

1976 

1978 

Figure 1. Total traditional catch in Lake Malawi for the period 1976 – 1999. 

Nkhotakota 

13% 

Figure 2. Contribution to the total catch in Lake Malawi by district. 

1980 

Nkhata bay 

12% 

Salima 

18% 

Figure 3. Average species-group contribution to the total Lake Malawi catch 

1982 

O.big 

cicchlids 

3.5% 

Others 

7.2% 

Kambuzi 

8.0% 

Catf ish 

8.0% 

Chambo 

11.7% 

1984 

1986 

1988 

Year 

Likoma 

4% 

Ntchila 

0.3% 

1990 

Usipa 

28.7% 

1992 

Utaka 

32.7% 

1994 

Karonga 

8% 

1996 

1998 

Mangochi 

45%

97 


Catch composition by species group reveals that utaka has been the main dominating group with an average 

contribution of 32.7%, followed by usipa (28.7%) and chambo (11.7%)(Figure 3). These three species groups 

yielded 73.1% of the total average catch. Contribution by the other species groups has been in the range 0.3%-8%. 

Fisheries in Lake Malawi appear to be stable when analysed as a bulk biomass. However, when analysed at species 

or genus level, some of the most important stocks seem to have declined. The most apparent cases are those of 

chambo stocks in Lake Malawi, which are considered to be in the state of decline and precautionary management 

action, are recommended due to the excessive fishing effort (Pálson et al. 1999). 

Trends in fishing units. 

The number of fishermen operating in the waters of Malawi in the period 1990-94 (10 601-10 602) has been 

relatively stable but has increased by 27% in the following years (Figure 4). Also, within the same period, the 

number of assistants/crew increased by 37%. Currently, there are about 48 000 people that are directly employed in 

the fishery (Weyl et al. 1999). 

Dugout canoes operating in the waters of Malawi counted 9671 – 11457 in the period 1990 – 1999 signifying an 

increase of 18% (Figure 5). The number of plank-boats with engines over the same period 1990 – 99 has increased 

by 48% (360 – 534). However figures of 1993 and 1994 (449 393) are probably errors, since a decrease or increase 

of 50 units within a year is very unlikely. Plank-boats without engines have also increased by 38% (2 167 – 2 991) 

over the same period although in 1995 only 1 361 units were counted. This again is obviously an error since a 

change of such a scale is by no means reasonable. 

No/ of fishers 

Figure 4. Number of fishermen and assistants in Malawian waters 1990 – 99. 

16000 

14000 

12000 

10000 

8000 

6000 

4000 

2000 

0 

40000 

35000 

30000 

25000 

20000 

15000 

10000 

5000 

0 

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 

Gear ow ner Crew 

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 

Boat + engine Boat - engine Dugout canoes 

Figure 5. Number of small-scale fishing crafts Malawi for the period 1990-1999. 

The most important gears in Lake Malawi are gillnets, chilimira nets and kambuzi seines. The importance has been 

determined by looking at the amount of catch that each gear contributes in Lake Malawi. (Figure 6). The three gears

98 


contributed over 90% to the total Lake Malawi catch, gillnet (43.1%), chilimira (32.9%) and kambuzi seine (14.5%). 

Long-lines, chambo seine, mosquito net, fish trap and hand-lines are of intermediate importance. They contributed 

less than 10% to the total Lake Malawi catches and ranged from 0.4% - 3.3%. 

Contribution of the total catch by gillnets over the 24-year period has been relatively stable at an average catch of 7 

thousand metric tons. However, chilimira’s contribution has increased into the late 1990s from an average of 10 

thousand metric tons in the 1980s to 16 thousand metric tons in the 1990s. Kambuzi seine catches have fluctuated 

around an average catch of 2 thousand metric tons. The number of gillnets and chilimira have increased two-fold 

over the period 1990-1999 (from 19 699-43 318 for gill nets and from 1 479-2 568 for chilimira nets). The number 

of kambuzi seines however shows a declining trend over the same period, from 654 gears in 1990 to 343 gears in 

1999. 

a 

b 


40000 

35000 

30000 

25000 

20000 

15000 

10000 

5000 

0 

1976 

Kambuzi 

14.5% 

Mosquito 

2.9% 

Longline 

3.3% 

1978 

Gillnet 

43.1% 

1980 

1982 

Handline 

0.4% 

1984 

1986 

Fishtrap 

1.1% 

Chambo 

seine 

1.8% 

Chilimira 

32.9% 

Figure 6. Gear contribution to the (a) total Lake Malawi catch and (b) annual Lake Malawi catch from 

1976 to 1999. 

1988 

1990 

1992 

1994 

1996 

Year 

Chilimira Gillnet Kambuzi Others 

1998

99 


Catch and effort in Lake Malawi. 

Catch, effort and CPUE in the southeast arm small-scale fishery is shown in Figure 7. Unless otherwise stated, effort 

refers to the relative effort (YT(y) /R(y) where YT(y) is the total yield of all gears and R(y) is the sum of relative 

CPUE weighted by the yields in year (y), normalized effort where all the small-scale fishing gears have been 

standardized to a common effort unit. CPUE also refers to weighted, relative CPUE where relative catch per unit 

effort of each gear has been weighted by the catch taken by that gear (For further details see; Sparre, P. & Venema, 

S.C., 1992. Introduction to tropical fish stocks assessment. Part I- Manual. FAO Fisheries Technical Paper, 306/1, 

376 pp). 

Catches have increased into the 1990s from 5 000 metric tons in the 1970s to 10 000 tons and a highest catch of 

about 20000 metric tons recorded in 1996 mainly due to high usipa catches in that year. Effort increased in the mid 

1980s and has been at a relatively high level in the 1990s. CPUE however has declined except the high value 

recorded in 1996. 

Effort and CPUE 

2.50 

2.00 

1.50 

1.00 

0.50 

0.00 

1976 

Figure 7. Catch, effort and CPUE in the South East Arm of Lake Malawi for the period 1976 – 1999. 

To obtain an estimate of maximum sustainable yield (MSY) for the southeast arm of Lake Malawi, relative effort 

and relative CPUE for the period 1976 – 1999 was used. There was fairly good correlation between ln(CPUE) and 

effort. The Fox surplus production model was fitted to the data to yield initial MSY estimates for the southeast arm 

fishery (Figure 8). 


1978 

25000 

20000 

15000 

10000 

5000 

0 

1980 

83 

99 

93 

80 

82 

77 

88 

86 

78 85 

79 76 

95 

90 

0 2000 4000 6000 8000 10000 12000 14000 16000 

Relative effort 

Fox Total Yield Poly. (Fox) 

Figure 8. Fox surplus production function of the SE Arm 1976-99 

1982 

1984 

1986 

1988 

1990 

1992 

96 

87 

1994 

97 

1996 

89 

94 

91 

1998 

Catch Relative CPUE Normalised effort 

92 

98 

25000 

20000 

15000 

10000 

5000 

0 

Catch(metric tons)

100 


The statistical quality of the data is fairly reasonable since a regression of CPUE and effort gave a coefficient of 

determination of 57%. The results of the Fox model are estimated at 10 502 metric tons at an effort level of 16 667 

units. It is however evident that MSY of 10 000 tons was exceeded in the late 1980s to mid 1990s. Although catches 

seem to be increasing, the increase in effort in the recent years might have remarkable effects in the long run. 

Catch, effort and CPUE in the southeast arm traditional fishery is shown in Figure 9. Catches have fluctuated 

throughout the period, around a mean of 6 000 tons. Effort was relatively low until late 1980s when it increased and 

has remained at that high level, but fluctuating in the 1990s. CPUE however has declined in the early 1980s from 

high values in the late 1970s and has since remained at that low level. 


3.50 

3.00 

2.50 

2.00 

1.50 

1.00 

0.50 

0.00 

1976 

1978 

1980 

Figure 9. Catch, effort and CPUE in the South West Arm of Lake Malawi for the period 1976 – 1999. 

To obtain an estimate of MSY in the South West Arm of Lake Malawi, relative effort and relative CPUE data was 

used. There was a good correlation between ln CPUE and relative effort. The Fox surplus production model was 

fitted to the data to yield initial MSY estimates for the SW Arm fishery (Figure 10). The statistical quality of the 

data is also fairly reasonable since a regression of CPUE and effort gave a coefficient of determination of 58%. 

The results of the Fox model are estimated at 6 839 metric tons at an effort level of 10 000 units. It is however 

evident that MSY was attained at relative effort levels that exceeded current effort and it appears that fMSY was 

exceeded in the in the 1980s to 1990s with no stock recovery (Weyl et al, 2001 in press) 

Catch 

12000 

10000 

8000 

6000 

4000 

2000 

0 

Figure 10. Fox surplus production function of the SW Arm 1976-1999. 

1982 

1984 

1986 

1988 

Year 

1990 

1992 

1994 

1996 

1998 

12000 

10000 

8000 

6000 

4000 

2000 

Total Yield Relative CPUE Normalised Effort 

81 

77 

83 

78 

86 

82 

76 

79 

80 84 

84 

87 

88 

91 

96 

89 

98 

95 

97 

85 

93 90 

99 

0 5000 10000 15000 20000 

Relative ef fort 

92 

0 

Catch (metric tons)

101 


Catch, effort and CPUE in Domira-bay for the period 1976-99 are shown in Figure 11. Catches have declined from 

high values in the 1970s and late 1980s. Effort however has remained relatively stable except the high value in 1989. 

CPUE has declined from high values in the early 1980s and has remained at that low level. 

Catch, effort and CPUE in Nkhota-kota for the period 1976-1999 are shown in figure 12. Catches show an 

increasing trend from low values in the 1970s to high values in the 1990s. Effort was relatively low in the 1970s but 

increased substantially in mid 1980s and has remained at that high level. CPUE has remained relatively stable. 

Effort/CPUE 

3.50 

3.00 

2.50 

2.00 

1.50 

1.00 

0.50 

0.00 

Figure 11. Catch, effort and CPUE in Domira-bay from 1976-99. 

Ef f ort/CPUE 

2.50 

2.00 

1.50 

1.00 

0.50 

0.00 

1977 

1976 

1980 

Figure 12. Catch, effort and CPUE in Nkhota-kota from 1976-99. 

1983 

1986 

1989 

Catch, effort and CPUE for Nkhata-bay for the period 1976-1999 are shown in Figure 13. Catches show an 

increasing trend. Effort has also taken a similar trend, increasing from low values in the 1970s to high values in the 

1990s except the low values in 1995 and 1996. CPUE however has remained relatively stable. 

Catch, effort and CPUE in Likoma Island over the period 1976-1999 are shown in Figure 14. Catches have 

fluctuated without a definite trend. Effort has declined in the early 1990s from high levels in the 1970s and 1980s. 

CPUE has also fluctuated greatly without a definite trend. 

1992 

1995 

1998 

Catch 

Year 

Relative CPUE Normalised effort 

1979 

1982 

1985 

1988 

Year 

1991 

1994 

1997 


3500 

3000 

2500 

2000 

1500 

1000 

500 

0 

8000 

6000 

4000 

2000 

0 


10000 

Catch(metric tons)

102 


Catch, effort and CPUE for Karonga for the period 1976-1999 are shown in Figure 15. Catches show an increasing 

trend from low values in the 1980s. Effort has also taken a similar trend, increasing from low values in the 1980s. 

CPUE however has remained relatively stable. 

Effort/CPUE 

3.50 

3.00 

2.50 

2.00 

1.50 

1.00 

0.50 

0.00 

1976 

1979 

Figure 13. Catch, effort and CPUE in Nkhata-bay from 1976-99. 

Effort/CPUE 

Figure 14. Catch, effort and CPUE in Likoma from 1976-99. 

Effort/CPUE 

3.50 

3.00 

2.50 

2.00 

1.50 

1.00 

0.50 

0.00 

2.50 

2.00 

1.50 

1.00 

0.50 

0.00 

1976 

1980 

Figure 15. Catch, effort and CPUE in Karonga from 1976-99. 

1982 

1985 

1988 

1991 

1994 

1997 

Catch 

Year 


1979 

1982 

1985 

1988 

1991 

1994 

1997 

Catch 

Year 


1983 

1986 

1989 

Year 

1992 

1995 

1998 


6000 

5000 

4000 

3000 

2000 

1000 

0 

6000 

5000 

4000 

3000 

2000 

1000 

0 

4000 

3500 

3000 

2500 

2000 

1500 

1000 

500 

0 



Catch(metric tons)

103 


Trends in total catch 

Total catches in Malawian waters fluctuated between 50 and 60 thousand metric tons in the period 1976-86 (Figure 

16). During the period 1987-91, catches were in the range of 60-70 thousand metric tons. Catches declined in the 

early 1990s to a minimum of 30 thousand metric tons in 1995. The decline in total catch as seen from Figure 16 is 

caused by declining catches in other water bodies other than Lake Malawi, in particular in Lake Malombe, Lake 

Chilwa, Lake Chiuta and the Lower Shire. Also the reduced catches in the commercial fishery contributes to the 

decline. 

Traditional catches in Lake Malawi increased into the late 1980s to a peak of 33 thousand metric tons in 1987. After 

which catches have remained relatively stable. The highest catch in Lake Malawi traditional fisheries was recorded 

in 1996 (37 thousand tons). Traditional fisheries in other water bodies other than Lake Malawi were relatively stable 

until early 1990s, fluctuating around 30 thousand metric tons with a peak of 45 thousand metric tons in 1990. 

Catches however declined by 1994 to some 20 thousand metric tons and they dropped to 6-10 thousand metric tons 

in 1995/96. The main reason for the low catches in these two years is primarily due to the low or zero catches in 

Lake Chilwa when it dried up. 


90000 

80000 

70000 

60000 

50000 

40000 

30000 

20000 

10000 

0 

1976 

1978 

Figure 16. Annual catch contribution by water body from 1976-99 (* traditional and commercial). 

Management measures 

Current fisheries management regulations are based on technical restriction of fishing gears, i.e. gear mesh size or 

size of gear (head line length) and restrictions of fishing areas or fishing times- closed areas and seasons for other 

gears such as beach seines. A number of gears are without any restrictions such as longlines except for permissible 

setting times in Lake Malawi, hand lines, scoop nets and cast nets. Some gear types are locally prohibited such as 

kauni for chambo, nkacha in Lake Malawi, beach seines in Lake Chiuta, upper shire and in rivers and dams. Some 

species are subject to minimum size i.e. all species of chambo (Oreochromis) 15.0 cm, other tilapia e.g. 

Oreochromis shiranus shiranus 10.0 cm folk length and mpasa (Opsaridium microleps) 30.0 cm fork length. 

The traditional fisheries are ‘’open entry’’, although a license is formally needed to operate in that fishery. An 

annual basic fee must be paid depending on the type of gear being used. In view of the current status of fisheries and 

fish stocks in Malawian waters, the effectiveness of these measures appears to be limited. 

The fishery in the other smaller water bodies however show declining trends in catch and CPUE and many appear 

currently to be at low level in terms of resource status (Table 1). The prospects for fisheries in Lakes Chilwa, Chiuta, 

Malombe and upper shire seem to be extremely discouraging. However due to heavy limitations of available data for 

Lakes Chilwa and Chiuta, precise recommendations are not given at this stage. 

In summary, recommended precautionary management actions are as follows: 

Lake Malawi traditional fishery: 

1980 

1982 

1984 

1986 

1988 

1990 

1992 

1994 

1996 

1998 

*L.Malaw i L.Malombe 

Year 

L.Chilw a L.Chiuta U+L.Shire

104 


• Total lake wide ban on chambo seines and ban on kauni for chambo in Area A in order to restore the 

chambo stocks. 

Lake Malombe fishery: 

• Total ban on gill nets and chambo seines, in order to restore the chambo stocks. 

• Effort limitations of nkacha and kambuzi seine by denying new entries in order to maintain kambuzi 

stocks. 

Lake Chilwa fishery: 

• Currently no management recommendations since the fishery are seasonal i.e. the lake dries up. 

Lake Chiuta fishery: 

• Reduction in effort in order to restore stocks. 

Upper Shire fishery: 

• Total ban on seines and gill nets, in order to restore stocks. 

Table 1. Summary of the status of fisheries and fish stocks in Malawian waters (after Bulirani et al. 1999). 

Fish Long term trends (1976-1999) 

Stock(s) Fishery/Survey Waterbody Area(s) Catch Effort CPUE 

All Traditional South East Arm All Increasing Increasing Decreasing 

All Traditional South West Arm All Variable Increasing Decreasing 

All Traditional Domira-bay All Decreasing Stable Decreasing 

All Traditional Nkhotakota All Increasing Increasing Stable 

All Traditional Nkhata-bay All Increasing Increasing Stable 

All Traditional Likoma All Variable Decreasing Variable 

All Traditional Karonga All Increasing Increasing Stable 

All Traditional Lake Malombe All Decreasing Increasing Decreasing 

All Traditional Lake Chilwa All Decreasing Increasing Decreasing 

All Traditional Lake Chiuta All Decreasing Increasing Decreasing 

Chambo Traditional Lake Malawi All Decreasing Increasing Decreasing 

Chambo Traditional Lake Malombe All Decreasing Increasing Decreasing 

Chambo Traditional Upper Shire All Decreasing Variable Decreasing 

Kampango Traditional Lake Malawi All Decreasing Increasing Decreasing 

Bombe Traditional Lake Malawi All Decreasing Increasing Decreasing 

Utaka Traditional Lake Malawi All Variable Increasing Decreasing 

Usipa Traditional Lake Malawi All Increasing Increasing Variable 

Kambuzi Traditional Lake Malawi All Increasing Increasing Stable 

Kambuzi Traditional Lake Malombe All Decreasing Stable Stable 

References 

Bulirani A.E., Banda M.C., Pálson Ó.K., Weyl O.L.F., Kanyerere G.Z., Manase M., Sipawe R., 1999. Fish Stocks 

and Fisheries of Malawian Waters: Resource Report. Government of Malawi, Fisheries Department, 

Fisheries Research Unit.54pp. 

Kolding, J. Population ecology and simple sustainable yield estimators in fisheries: a review. University of Bergen, 

High Technology Centre, Norway. 

Pálson Ó.K., Bulirani A., Banda M, 1999. A review of biology, fisheries and population dynamics of chambo 

(Oreochromis spp., CICHLIDAE) in Lakes Malawi and Malombe. Malawi Fisheries Bulletin. 

Sparre, P. and Venema, S.C., 1992. Introduction to tropical fish stock Assessment. Part I – Manual. FAO Fisheries 

Technical Paper, 306/1, 376pp. 

Turner, G.F., 1995. Management, conservation and species changes of exploited Fish stocks in Lake Malawi in The 

Impact of Species Changes in African Lakes. Tweddle, D. and Magasa, J.D. (1989) Assessment of multispecies 

cichlid Fisheries of the Southeast Arm of Lake Malawi, Africa. J. Cons. int. Explor. Mer., 45: 209 – 

222. 

Weyl O.L.F., Banda M, Sodzabanja G., Mwenekibombwe L.H., Namoto W., Mponda O.C., 2000. Annual 

Frame Survey, September 1999. Fisheries Bulletin No.42. Fisheries Department, Lilongwe, Malawi. 

Weyl, O.L.F. 2001, Small-scale fisheries statistics summary. Short communication No. 2

105 


Effects of overfishing on reproductive potential of major cichlid fish species in 

southern Lake Malombe (Malawi): need for “closed area” strategy as a 

complementary management option? 

Collins Jambo 1 and Tom Hecht 2 

1 Fisheries Research Unit, P.O. Box 27, Monkey Bay, Malawi. 

2 Department of Ichthyology and Fisheries Science, Rhodes University, P.O. Box 94, 6140 Grahamstown, South Africa. 

Abstract 

The three major species Lethrinops ‘pinkhead’, Otopharynx argyrosoma ‘red’ and Copadichromis cf. virginalis, which used to 

contribute about 75% to the total catches (by weight) of Lake Malombe in the last decade were investigated. The main aim of the 

investigation was to assess the impact of fishing intensity on reproductive potential of the three species. Fecundity, reproductive 

seasonality, sexual maturity, and sex ratio were related to habitat types of the south western side (heavily fished) and south eastern 

side (lightly fished). The three species have low fecundity and they are synchronous spanners, with a breeding peak during July to 

October period. Females of all three species mature earlier than males while the sex ratio of the three species was not significantly 

different from 1:1 in both sides of the lake. The length-fecundity relationships for L. ’pinkhead’ and O. argyrosoma ‘red’ indicated that 

fecundity was more closely related to length in the south eastern side than in the western side. The frequency occurrence of mature 

females and juveniles was greater in the south eastern side than in the south western side of the lake. Juveniles of Oreochromis 

spp. (chambo) were also abundant in the south eastern side. The south eastern side of the lake is characterised by low fishing 

intensity, muddy substratum and aquatic macrophytes. It is also functioning as a spawning area for Oreochromis spp. and nursery 

area for the three haplochromines. These findings provisionally suggest that the efficacy of closed area method as an additional 

management tool in Lake Malombe. Such a management tool would protect juveniles and breeding stocks of the main species, 

hence meet the criteria of ensuring the sustainability and utilisation of fish stocks in Lake Malombe. 

Key words: reproduction, fishing intensity, closed area, management, Lake Malombe and haplochromines 

Introduction 

The Lake Malombe fishery has been under severe fishing pressure since 1986 (FAO 1992, Coulter 1993). The 

fishery provides the cheapest dietary protein to the entire human population surrounding the lake and directly 

employs 4660 people. During 1970-1986 period, Oreochromis spp. (chambo) used to dominate the catches but 

collapsed in 1988. The collapse of the chambo fishery prompted artisanal fishers to use fine-meshed nets with which 

to target the small haplochromines (kambuzi). As such, small haplochromines became predominant in the catches 

after the collapse of the chambo fishery. During 1986-1993, three haplochromines (Lethrinops ‘pinkhead’, O. 

argyrosoma ‘red’ and Copadichromis cf. virginalis) contributed about 75% to the total annual catch of 8 000 tons 

(FAO 1993). However, five haplochromines currently constitute about 67% of the total annual catch by weight 

(Jambo 1997). 

Recent findings indicate that catches of kambuzi have declined from 20 kg per haul in the late 1980s (Tweddle et al. 

1995) to 3.5 kg per haul (Jambo 1997), indicating that the Lake Malombe fishery has been overexploited. The 

current overall decline of small haplochromines threatens both the livelihood of the rural people who depend on fish 

as their source of animal protein and employment, as well as the nation’s animal protein supply as a whole. 

Considering the fact that many African Great Lakes cichlids are narrowly stenotopic, philopatric and practise mouth 

brooding (Ribbink et al. 1983), in a way to safe guard their offspring within the same habitat, such a drastic decline 

in catch per unit effort threatens the biodiversity of fish species in Lake Malombe. Further increase in fishing effort 

and inappropriate use of fishing gears on the remaining small populations would consequently lead to a total 

collapse of the kambuzi fishery. 

Based on Roberts & Polunin (1991) views it was hypothesised, in this study, that high fishing intensity in the south 

western side of Lake Malombe might have depleted populations of adult fishes and reduced the average size of 

females, resulting in poor recruitment into the fishery. In order to test this hypothesis, aspects of reproductive 

biology of the three principal haplochromines were investigated in the south western side (heavily fished) and south 

eastern side (lightly fished) of the lake. The main aim of the investigation was to assess the impact of fishing 

intensity on reproductive potential of the three principal species, with a view of formulating decisions to do with the 

placing and size of a protected area. 

Materials and methods 

A monthly sample of 25 to 45 fish for each species was collected from each sampling station. A standard nkacha 

offshore seine net (250 metres long and 9 metres deep) was used at all stations on each sampling trip. The nkacha 

offshore seine net is an active gear with a gradation of mesh sizes of 39 mm and 25 mm on the wings and 19 mm on 

the bunt. It involves divers who pull the footrope and tie weights together to ‘purse’ the net. Sampling stations were

106 


located along two transects across the lake. On each transect, two sampling stations were situated in the south 

western side and the other two stations in the south eastern side of the lake (Figure 1). 

Figure 1. Map of Southern Lake Malombe showing the location of major landing sites and sampling 

stations. SS refers to sampling stations on the two transects. 

For each specimen, total length to the nearest millimetre and total body mass to the nearest gram were measured. 

Fish mass was taken as the mass of the fish minus gut contents. The breeding seasons as well as sexual maturity 

were determined by a monthly visual appraisal of the gonad activity stages. The visual assessment of gonadal stages 

was done using the descriptions of Marsh et al. (1986) (Table 1). To eliminate possible error, only those data which 

were collected during the peak period of gonad activity, were used. 

To establish the breeding seasonality, the percentage frequency occurrence of females with active ripe (AR), ripe 

(R) and spent (S) stages for each species, was plotted against sampling time (months). The breeding season was then 

identified by comparing peaks of AR, R and S stages. Only mature females with AR, R and S gonadal stages were 

used in this investigation to avoid the possible masking effect, which immature fish could have had on the results. 

To establish the length at 50% sexual maturity, with minimal possible error, the data collected during the period of 

maximum gonad activity were used. Fish with active ripe, ripe and spent gonads were considered mature. For each 

species and sex, the percentage frequency of such mature fish was plotted against total length. A logistic curve was 

fitted to the percentage of sexually mature individuals by length (L), using King's (1995) equation: 

P = 1/(1 + e -r(L - L m ) 

where r is the slope of the curve and Lm is the mean length which corresponds to size at 50% maturity. The logistic 

curve was used in order to minimise unreasonably high estimation of Lm.

Table 1. Criteria used for categorising gonads to reproductive stage. 

Stage Description of female gonadal stages 

Immature (I) Eggs were elongate, pale yellow / pale orange in colour 

107 


Active Ripe (AR) Eggs were ovoid in shape and they were closely packed and pale orange in 

colour 

Ripe (R) Eggs were irregular in shape and loosely connected to one another. They 

were bright orange in colour. 

Spent (S) Ovaries were thin and opaque wall. Red eggs dominated with few bright 

orange eggs. 

Description of male gonadal stages 

Immature (I) Testes were translucent and almost colourless. 

Ripe (R) Testes were opaque and creamy. 

Spent (S) Testes were thin, flaccid and irregular. 

To determine the absolute fecundity of each species the ripe ovaries for each fish were separately preserved in 

Gilson's solution. After keeping the ovaries in Gilson's solution for one week, the number of eggs were counted 

(Bagenal 1978). The length-fecundity relationship was determined by regression analysis. The best regression fit 

was given by the linear equation: 

F = b L + a 

where F is the absolute number of eggs counted per individual, L is the total length of fish and a, b are linear 

equation constants. Relative fecundity (number of eggs per gram fish weight) was used to compare reproductive 

strategies of the three species. 

Sex ratio for each species was evaluated by recording the number of fish sampled by sex. The numbers of males and 

females were compared using a Chi-square test to determine whether sex ratio differed significantly from unity. This 

test was done to evaluate if size-selective fishing, which has been aggravated by use of small meshed nets, might 

have altered the sex ratios of the three species. 

Results 

Reproductive seasonality 

For all three species, a high frequency of potential breeding females (AR and R) was evident between July and 

October, with prominent peaks in July and October for active ripe female and in August for ripe females. A high 

frequency of potential breeding ripe females was also noted for O. argyrosoma ‘red’ in December (Figure 2). From 

these data it is evident that all three species breed mainly from July to August. 

Sexual maturity 

Gonads of 991 males and 979 females were examined. More than 50 % of the males, irrespective of species and side 

of the lake, were mature at approximately the same size (79 mm), thus corresponding to an age of between two and 

three years. However, they appear to attain 100% sexual maturity at different lengths (TL). In contrast, females of 

the three species attained 50% sexual maturity at different lengths (Table 2)

108 


Table 2. Comparison of length at 50% sexual maturity (Lsm) in millimetres and length-at-age 1 (L-at-1) in 

millimetres for both sexes of the three species studied in the southern Lake Malombe 

Species Male 

Female 

Lsm L-at-1 Lsm L-at-1 

L. ‘pinkhead’ 79 67 66 86 

O. argyrosoma ‘red’ 77 69 72 89 

C. cf. virginalis 79 32 83 45 

% Frequency 

% Frequency 

50 

40 

30 

20 

10 

0 

70 

60 

50 

40 

30 

20 

10 

0 

L. ' pinkhead' ( n=297) 

Active ripe Ripe 

Jun Jul Aug Sep Oct Nov Dec Apr May Jun 

Months 

C. cf. virginalis (n=189) 



Months 

% Frequency 

70 

60 

50 

40 

30 

20 

10 

0 

O. argyrosoma ' red' (n=389) 



Months 

Figure 2. Frequency distribution of female fish with active ripe (AR) and ripe (R) gonads. 

Size distribution and Sex ratio 

The frequency occurrence of smaller mature females was greater in the south eastern side than in the south western 

side. Females of L. ‘pinkhead’ and O. argyrosoma ‘red’ dominated in the smaller size classes (66-75 mm) while 

females of C. cf. virginalis dominated in the larger size classes (78-90 mm) (Figure 3 a,b and c). 

These results suggest that more of the immature females in the L. ‘pinkhead’ and O. argyrosoma ‘red’ populations 

were vulnerable to the standard nkacha net. However, there was no significant difference in size between sexes and 

sides (ANOVA) P>0.05 (Table 3). Similarly, the sex ratio of the three species did not differ significantly from 1:1, 

X² test; P > 0.05 (Table 4).

109 


Table 3. Mean lengths (mm) of potential breeding males and females in the south western (SW) side and 

south eastern (SE) side of Lake Malombe. 

Species SW 

(Heavily fished) 

L. ‘pinkhead’ 78 

67 

O. argyrosoma’red’ 71 

72 

C. cf. virginalis 79 

SE 

(Lightly fished) 

78 

67 

76 

71 

80 

83 83 

Table 4. Sex ratio of the three species sampled from southern Lake Malombe, irrespective of the sides. 

Species Male Female Ratio x² 

L. ‘pinkhead’ 1450 1441 1:1.03 0.017 

O. argyrosoma ‘red’ 801 1014 1:1.23 0.001 

C. cf. virginalis 749 802 1:1.07 0.790 

Discussion 

Breeding seasonality 

It appears that all three species breed throughout the year with one or two distinct peaks occurring from July to 

October for L. ‘pinkhead’ and C. cf. virginalis and in September and December for O. argyrosoma ‘red’. In general, 

these results confirm the findings of Tweddle & Turner (1977), Mwanyama (1993) and Banda et al. (1994) who 

found that the main breeding season for most Malawian cichlids falls between August and September. The high 

frequency occurrence of potential breeding fish, throughout the year, provides further evidence that the three species 

breed throughout the year and this makes the application of a closed season problematic. Nevertheless, high 

frequency of spent ovaries and juveniles after November suggests that a fine-tuned closed season would cover the 

month of November so that juveniles could be protected. 

The breeding peaks of the three species occurred during the phytoplankton biomass peaks. This could be a strategy, 

which synchronises the breeding behaviour with availability of food. Eccles (1974), Lowe-McConnell (1987) and 

Mwanyama (1993) have indicated that more nutrients are available during the period June to September from 

sediments pertubated by the south easterly winds and between February and March during the rainy season. These 

findings provide some evidence that the synchrony in breeding of the three species may reflect a direct response to 

food availability. Similar relationships have also been found for cichlid species elsewhere (McKaye 1984 and Marsh 

et al. 1986). It was also apparent that the main breeding season of the three species occurred during the time when 

CPUE was also highest, indicating that the three species are more vulnerable to fishing during their breeding season. 

From a fishery perspective, the coincidence of greater CPUE and breeding season during the open season, suggested 

that the closed season (1January-31 March) was ineffective in protecting breeding stocks.

(a) L. ‘pinkhead’ 

% Frequency 

(b) O. argyrosoma ‘red’ 

% Frequency 

12 

10 

8 

6 

4 

2 

0 

15 

10 

5 

0 

(c) C. virginalis 

% Frequency 

South western side 

Male 

Female 

55 6 65 69 73 77 8 85 89 93 98 

Total length (mm) 

15 

10 

5 

0 



Male 

Female 

55 62 66 70 74 78 82 86 90 94 99 


Male 

Female 

65 69 73 77 81 85 89 93 97 103 


110 


South eastern side 

Figure 3 a, b and c Length frequency distribution of mature fish of three principal species in the southern 

Lake Malombe. 

% Frequency 

% Frequency 

15 

10 

5 

0 

12 

10 

8 

6 

4 

2 

0 

% Frequency 

10 

8 

6 

4 

2 

0 


Male 

Female 

58 63 67 71 75 79 83 87 91 100 


Male 

Female 

33 60 64 68 72 76 80 84 88 92105 



Male 

Female 

30 62 66 70 74 78 82 86 90 94 100111123 

Total length (mm)

111 


The spawning period of O. argyrosoma ‘red’ coincided with the onset of the rainy season in December. This species 

belongs to a group of small zoobenthic feeders, substrate spawners and it is most abundant in the shallower parts of 

the lake (Eccles & Trewavas 1989). During the rainy season the preferred habitats for O. argyrosoma ‘red’ had high 

turbidity values. Considering the fact that high turbidity does not prohibit breeding of some haplochromines 

(Greenwood 1974), it appears that breeding of O. argyrosoma ‘red’ coincides with high water turbidity. 

Sexual maturity 

Analysis of the data, using non-pooled data, showed that the sizes at 50% maturity of female fish, from both sides of 

the lake, were not significantly different. Results for the three species have shown that the females mature 

approximately six months to one year before the males. The early maturing of females is considered to be 

advantageous to the reproductive potential of the three species because females could breed at least once before they 

are exploited in their second year of life. 

Cichlid females maturing at a small size could also mean a reduction in absolute population fecundity (Fryer & Iles 

1972, Pitcher & Hart 1982) resulting in insufficient recruitment levels from such small and early maturing females. 

However, some scholars have argued that while mouth brooding cichlids may mature at a small size and lay few 

eggs, recruitment is guaranteed due to the fact that the young are painstakingly cared for by the parent. This implies 

that although females of the three species are maturing early, at a small size, they may still have a good recruitment 

potential, which may however need to be conserved by protecting the breeding grounds. 

Studies on sexual maturity of related species such as Haplochromis virginalis (107 mm), H. quadrimaculatus 

(164 mm) and H. pleurostigmoids (130 mm) have shown that they all matured during their third year (Fryer & Iles 

1972). Tweddle & Turner (1977) who investigated the sexual maturity of Lethrinops. parvidens , L. longipinnis, 

and H. anaphyrmus (closely related to the species studied here) also postulated that they breed in their third year. 

According to Fryer & Iles (1972), reduction in size at 50% maturity in all female individuals might have been a 

direct adaptation to heavy fishing. It appears that the reduction in size at sexual maturity for C. cf. virginalis has 

been compensated by a faster growth rate (Jambo 1997). Pitcher & Hart (1982) and Gulland (1988) pointed out that 

a smaller size at sexual maturity coupled with an increasing growth rate offers the potential for compensatory 

adjustment to offset declining spawner stock size. This phenomenon could probably explain why these three species 

have been resilient to heavy exploitation and give an impression that the fishery could be restored if there is a 

substantial curtailment of fishing and proper management. 

Sex ratio 

For all three species the overall sex ratio was approximately 1:1 at 95% confidence interval. This suggests that there 

was an equal removal of both sexes by the standard fishing gear in southern Lake Malombe. However, it was 

apparent that for all three species, males slightly outnumbered females two months prior to the breeding season. This 

may be a reproductive strategy that could increase the reproductive success of the females through the exercise of 

mate choice. Lande (1981) quoted by McKaye (1986) and Turner (1993) emphasised that such preponderance of 

males satisfies the polygynandry of mouth brooding cichlids, whereby females have numerous males from which to 

choose and apportion their eggs and hence achieve greater mating success. Another explanation could be that males, 

as in Tilapia moori (Konings 1988), move about in groups prior to and during the spawning period whilst females 

are solitary in less accessible breeding grounds. Such a breeding behaviour could contribute to males dominating the 

catches during and/or prior to the breeding season. 

Fecundity 

In this study, the linear function best described the significant relationship between absolute fecundity and length. 

However, the relationships for L. ‘pinkhead’ and O. argyrosoma ‘red’ were poorer in the south western side than in 

the south eastern side of the lake. The possible explanation for such a poor length-fecundity relationship in the south 

western side could be that the high fishing intensity has influenced a shift in the size distribution of these species 

resulting in the impairment of reproductive output (Sadovy 1996). 

This study indicates that relative fecundity indicate is variable among the three species and between the sides of the 

lake. Considering the fact that mean weights for each species, in both sides of the lake were not significantly 

different (t-test, P>0.05), it appears that the higher values of relative fecundity for C. cf. virginalis, in the south 

western side, might have been a reproductive adaptation to high fishing pressure. Pitcher & Hart (1982) and Sadovy 

(1996) pointed out that changes in reproductive strategy are expected to involve increases in relative fecundity at a

112 


given size and age when fish populations are subjected to high fishing pressure. The lower value of relative 

fecundity of O. argyrosoma ‘red’, in the south western side, could be an indication of poor habitat breeding 

conditions for this species in the south western side. It could be envisaged that the sandy substratum of the south 

western side of the lake may have contributed to the low values of relative fecundity of O. argyrosoma ‘red’ while 

the muddy substratum of the south eastern side of the lake may have contributed to the better habitat conditions for 

reproductive success of O. argyrosoma ‘red’. 

In overall, absolute fecundity estimates from this investigation showed general agreement with estimates by 

Mwanyama's (1993). Mwanyama (1993) documented a range of 13 to 78 eggs. The slight difference might have 

originated from sampling, differences in population structure in both sides and period of sampling. Comparisons of 

absolute fecundity values of each species with respect to sides of the lake showed that L. ‘pinkhead’ and C. cf. 

virginalis had the highest mean values of absolute fecundities (28 and 51 eggs respectively) in both sides of the lake, 

where as for O. argyrosoma ‘red’ the highest mean value was observed in the south eastern side of the lake. These 

results provide circumstantial evidence that absolute fecundity of L. ‘pinkhead’ and C. cf. virginalis have not been 

affected much by high fishing intensity. 

Conclusions and recommendations 

The three dominant species breed throughout the year with distinct peaks occurring between July and October. From 

a fishery perspective, the high frequency occurrence of breeding females during July - October period suggested that 

the old closed season (January- March) was ineffective in protecting breeding females. As such, a new closed season 

for Lake Malombe (October- December) was adopted. It was also evident that the breeding peaks coincided with 

phytoplankton biomass peaks, suggesting that intraspecific competition for food between adults and juveniles is 

reduced during the period of greater fish abundance. 

Of the three species, females of L. ‘pinkhead’ and O. argyrosoma ‘red’ mature earlier than males irrespective of side 

of the lake while males of the three species attain sexual maturity at approximately the same size (79 mm). The 

relative small size at sexual maturity is of an advantage for the population because females are given a chance to 

spawn at least once before they are caught. However, benefits of early maturation are limited in Lake Malombe due 

to the current high fishing pressure. According to Jennings & Beverton (1991), a reduction in size at 50% sexual 

maturity for females could be a direct consequence of overfishing or response to physical and biological 

characteristics of the environment. These views suggest that in overall, degradation of substratum and high fishing 

intensity may have caused a reduction in size at sexual maturity of L. ’pinkhead’ and O. argyrosoma ‘red’ in Lake 

Malombe. 

In conclusion, the south eastern side of the lake is characterised by low fishing intensity, muddy substratum and 

aquatic macrophytes. These factors appear to have been associated with greater abundance of breeding females and 

higher frequency of juveniles rather than fecundity and sex ratios. Considering these results, it could be suggested 

that the existing sanctuary area, situated in the southern part of Lake Malombe, should be extended. It is appreciated 

that the optimum size of the sanctuary has to be scientifically established. However, given the seriousness of the 

situation, it is recommended that the sanctuary area boundary be increased immediately but in consultation with the 

fishers. It is anticipated that enlarged and well managed sanctuary area will protect the unique aquatic environment 

in the south eastern side (which tends to be the breeding ground), the breeding fish stocks, and juveniles, hence 

prevent the total collapse of the fishery. 


Special thanks are due to Mr O.K. Mhone and A. Phiri for their assistance in identifying fish species, weighing and 

measuring fish, and counting eggs. We sincerely thank Dr. Tony Booth and Mr. Dennis Tweddle who 

enthusiastically provided statistical advice and supervised data processing. Overseas Development Agency (ODA) 

funded this study.

113 


References 

Bagenal, T.B. 1978. Methods for assessment of fish production in freshwaters, IBP Handbook 3. Blackwell Scientific 

Publications Ltd., Oxford. 

Banda, M.C., Tomasson, T. & Tweddle, D. 1994. Assessment of trawl fisheries of the south east arm of Lake Malawi using 

exploratory surveys and commercial catch data. Conference on Inland Fisheries Stock Assessment, Hull England, April 

1994. 

Coulter, G.W. 1993. Report on Fisheries Research Strategies. The Government of Malawi Fisheries Department. ODA-FRAMS 

pp 58. 

Eccles, D.H. 1974. An outline of the physical limnology of Lake Malawi (Lake Nyasa). Limnol. Oceanogr., 19:730-742. 

Eccles, D.H. & Trewavas, E. 1989. Malawian cichlid fishes. The classification of some Haplochromine genera. Herten (W. 

Germany), Lake Fish movies. 

FAO. 1992. Fisheries Management in the South East Arm of Lake Malawi, the Upper Shire and Lake Malombe. CIFA Technical 

paper, Vol. 21 GOM/FAO/UNDP Chambo Fisheries Research Project. 

Fryer, G. & Iles, T.D. 1972. The Cichlid Fishes of the Great Lakes of Africa: Their Biology and Evolution, Oliver and Boyd. 

Edinburgh. 

Greenwood, H. 1974. Cichlid fishes of Lake Victoria, East Africa: the biology and evolution of a species flock. Bull. Br. Mus. 

nat. Hist. (Zool.), suppl., 6:1-134. 

Gulland, J.A. 1988. Fish Population Dynamics: The implication for management. Wiley, London. 

Jambo, C. 1997. 

King, M. 1995. Fisheries Biology, Assessment And Management. Fishing News Books. Pp341. 

Konings, A. 1988. Cichlids Of Lake Malawi. Tfh Publications. Neptune City. 495 Pp. 

Lande. 1981. 

Lowe-Mcconnell, R.H. 1987. Ecological Studies In Tropical Communities. Cambridge Uni. Press, Cambridge, 382 Pp. 

Marsh, B., Marsh, A.C. & Ribbink, A.J. 1986. Reproductive seasonality in a group of rock-frequenting cichlid fishes in Lake 

Malawi. J. Zool., London (A) 209:9-20. 

Mckaye, K.R. 1984. Behavioural aspects of cichlid reproductive strategies: patterns of territoriality and brood defence in Central 

American substratum spawners and African mouthbrooders. In: Fish Reproduction, (eds G.W. Potts and R.J. Wooten). 

Academic Press, London. pp245-273 

Mckaye, K.R. 1986. Mate choice and size assortative pairing by cichlid fishes of Lake Jilo'a, Nicaragua. J. Fish. Biol., 29 

(Supp.A), 135-50). 

Mwanyama, N.C. 1993. Relations entre les ressources alimentaires, I'alimentation et la reproduction des chambos (poissons 

tilapia's di genre Oreochromis dans les lacs Malawi et Malombe (Afrique centrale) Ph.D. thesis, Universite' de 

Provence-Axix-Marseille, 158p. 

Pitcher, T.J.& Hart, P.J.B. 1982. Fisheries Ecology, Croom Helm, Beckenham, 414 pp. 

Ribbink, A.J., Marsh, B.A., Marsh, A.C., Ribbink, A.C. & Sharp, B.J. 1983. A preliminary survey of the cichlid fishes of rocky 

habitats in Lake Malawi. S. Afr. J. Zool., 18:149-310. 

Sadovy, Y.J. 1996. Reproduction of reef fishery species. In: Reef Fisheries (eds N.V.C. Polunin and C.M. Roberts). Chapman & 

Hall. 

Turner, G.F. 1993. Teleost mating behaviour. In: Behaviour of Teleost Fishes (ed. T.J. Pitcher). Chapman & Hall, London. 

pp.307-331. 

Tweddle, D.& Turner, J.L. 1977. Age, growth and mortality rates of some cichlid fishes of Lake Malawi. J. Fish. Biol., 10:385- 

398. 

Tweddle, D., Turner, G.F. & Saisay, M.B.D. 1995. Changes in species composition and abundance as a consequence of fishing in 

Lake Malombe, Malawi. In: The Impact of Changes in African Lakes (eds. T.J. Pitcher and P.J.B. Hart). Chapman & 

Hall. London, pp413-424.

114 


Management Recommendations for the Nkacha Net Fishery of Lake Malombe 

Kissa R. Mwakiyongo 1 & Olaf L.F. Weyl 2 

1 Fisheries Research Unit, P.O. Box 27, Monkey Bay, Malawi, e-mail: fru@malawi.net 

2 National Aquatic Resource Management Programme, P.O. Box 27, Monkey Bay, Malawi, e-mail: narmapbay@malawi.net 

Abstract 

Fish catches from the nkacha net fishery on Lake Malombe were sampled on a monthly basis from March 2000 to January 2001. 

Although more than 60 species were identified, the catch composition was dominated (about 80%) by only five species. These were 

Copadichromis chrysonotus, C. virginalis, Lethrinops sp. ‘pinkhead’, Otopharynx argyrosoma and O. tetrastigma. This species 

composition was similar to that recorded in previous assessments and indicated a relative stability in the species that drive the 

nkacha net fishery. This paper investigates size selectivity for ¼ inch and ¾ inch mesh size nkacha nets from a yield per recruit 

(YPR) perspective for three of the five species: C. virginalis, L. sp. ‘pinkhead’ and O. argyrosoma. Management recommendations 

for the Lake Malombe nkacha net fishery are made using two target reference points (TRPs), F0.1 and Fmax, derived from the YPR 

models. 

Introduction 

The Lake Malombe fishery has undergone dramatic changes both in gear use and in catch composition (Weyl et al. 

2001). Chambo seines and gill nets dominated the fishery up to the mid 1980s when chambo (Oreochromis spp.) 

was the mainstay of the fishery. With the collapse of the chambo fishery in Lake Malombe, kambuzi seines and 

nkacha nets became increasingly important and currently over 90 % of total fish catch is from nkacha nets (Weyl et 

al. 2001). However, even this fishery is considered to be declining in Lake Malombe with catches fluctuating 

between 2000 and 4000 tones from 1995 to 1999 and over the same period the number of kambuzi seines, chambo 

seines and nkacha nets have also declined (Weyl et al. 2000). The latest count (Weyl et al. 2000) shows 165 nkacha 

nets being operated in Lake Malombe and the Upper Shire River. 

Although the legal minimum mesh size for nkacha nets is ¾ inch (GOM 2000), the mesh size in most common use 

in nkacha nets is ¼ inch (pers obs.). This study investigates the species composition in nkacha net catches to 

determine whether there have been major species composition changes over the last eight years and investigates the 

effect of the dominance of the ¼-inch mesh size on the fishery. 

Methods 

Fish catches from the nkacha net fishery on Lake Malombe (14 o 29’–14 o 45’S, 35 o 12’–35 o 20’E), Malawi, were 

sampled (n=107) from ¼ inch and ¾ inch mesh size nkacha nets on a monthly basis from March 2000 to January 

2001. No sampling was carried out between 1 October and 31 December, 2000, as this period is closed to nkacha net 

fishing in Lake Malombe. In most cases the landed catch was sub-sampled and all fish in the sample were identified 

to species level, each species component of the catch was weighed and each fish was measured for total length (TL) 

in millimeters. Since, age and growth data were available for only three species in the lake Malombe fishery, 

Copadichromis virginalis, Lethrinops sp. ‘pinkhead’, and Otopharynx argyrosoma, further analysis was restricted to 

these species. 

Age specific selectivity 

Length-specific selectivity was estimated using cumulative length frequency distributions for Copadichromis 

virginalis, Lethrinops sp. ‘pinkhead’, and Otopharynx argyrosoma in the nkacha net fishery. For age-specific 

selectivity, length frequencies were converted to age-frequencies using the von Bertalanffy growth parameters 

determined by Jambo (1997). Both age- and length-specific selectivity was described by a logistic distribution as 

selectivity was assumed to increases to, and remain at, a maximum. The selectivity pattern of the nkacha net fishery 

was therefore assumed to be temportally invariant and described by the logistic function: 

S 

a 

= 

( ) 1 

−( 

a−φ 

) / σ − 

1+ 

e 

where Sa is the selection of the gear on fish of age or length a, φ is the age- or length-at-50% selection and σ the 

width of the selectivity function.

Per-Recruit analysis 

Yield-per-recruit (YPR) as a function of fishing mortality (F) were determined by: 

where N a 

~ is the relative proportion of fish at age a is defined recursively as: 

~ 

N 

a 

⎧1 

⎪ ~ 

= ⎨N 

⎪ ~ 

⎩N 

a−1 

115 


where Sa is selectivity at age a, F is the instantaneous rate of fishing mortality on fully recruited cohorts, M is the 

instantaneous rate of natural mortality and max is the maximum recorded age. In all per-recruit models the weightat-age 

was described by: 

Wa= α( la ) β 

e 

max−1 

−( 

M + S 

e 

a −1 

−( 

M + S 

F ) 

max −1 

−( 

M + Sa 

F ) 

[ 1 − e ] ( M + S F ) ∆a 

max 

~ 

YPRF = wa 

S a FN 

a 

a 

∑ 

a= 

0 

F ) 

( 1 − e 

−( 

M + S 

max −1 

F ) 

if a = 0 

if 1 ≤ a < max 

if a = max 

where la is the length-at-age determined by the von Bertalanffy growth equation and α and β are parameters 

∆ a of 1/12 th of a year. 

describing the length-weight relationship. All summations were conducted with a step size ( ) 

) 

Input parameters used in YPR models for the three species in the Lake Malombe nkacha fishery are summarised in 

Table 1. Growth and mortality parameters were obtained from data presented in Jambo (1997), mortality estimates 

were derived from Weyl 2001, and all summations were run using 5 years to define the parameter for maximum age 

(max). Gear specific fishing effort was standardised to the effort exerted by one nkacha unit per year. 

Table 1. Input parameters used in per-recruit analysis for three species in the Lake Malombe nkacha 

fishery. 

Parameters C. virginalis L. pinkhead O. argyrosoma 

L ∞ Predicted asymptotic length 

146mm 118.4 101.7 

K Brody growth coefficient 

0.88 0.6 0.52 

Max Maximum age 5 5 5 

M Natural mortality rate 0.97 yr -1 0.62 yr -1 0.55 yr -1 

F Fishing mortality rate 1.59 yr -1 0.81 yr -1 0.71 yr -1 

Q Catchability coefficient 0.0096 0.0049 gear -1 .yr -1 0.0044 gear -1 .yr -1 

α Parameter for length/weight equation 

0.044687 0.02138 0.041687 

β 

,, 

2.33 2.7 2.33 

a50: ¼ Age-at-50%-selectivity 

0.706131 1.302769 1.891944 

bδ50: ¼ Width of the selectivity logistic ogive 

0.091881 0.241381 0.349569 

A50: ¾ Age-at-50%-selectivity 

0.68007 1.426617 2.054715 

δ50: ¾ inch Width of the selectivity logistic ogive 

0.086659 0.250679 0.345974

116 


Target Reference Points 

In order to determine target reference point (TRP) fishing mortality for the ¼ inch and ¾ inch mesh sizes, two 

TRP’s were investigated: 

1) F0.1 or marginal yield strategy which corresponds to the rate of fishing mortality at which the slope of the 

YPR curve falls to 10% of its value at the origin. 

2) Fmax which corresponds with the asymptote of the YPR curve. 

Results 

Species selectivity 

Over sixty species of fish were identified in the nkacha net fishery, however only five dominated the catch. These 

were Copadichromis chrysonotus, C. virginalis, L. sp. ‘pinkhead’, O. argyrosoma and O. tetrastigma, each one of 

which made up >5% of the catch (Table 2). The monthly trends in catch composition for the period March 2000 – 

January 2001 indicated that the five most dominant species together made up between about 60% and 90% of the 

total nkacha net catch (Figure 1). 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Relative abundance (%) O. tetrastigma O. argyrosoma L. sp. "pinkhead" C. virginalis C. chrysonotus Others 

Mar '00 Apr '00 May '00 Jun '00 Jul '00 Aug '00 Sep '00 Jan '01 

Month 

Figure 1. Monthly trends in catch composition (%) from March 2000 to January 2001. 

Size selectivity 

Size selectivity comparisons between ¼ inch and ¾ inch nkacha net mesh sizes for C. virginalis, L. sp. ‘pinkhead’ 

and O. argyrosoma are shown in Figs. 2, 3 and 4, respectively. For C. virginalis, the mean selectivity for ¼ inch 

mesh size was 67 mm versus 66 mm for the ¾ inch mesh size (Figure 2). For L. sp. ‘pinkhead’ the mean 

selectivities for ¼ inch and ¾ inch mesh sizes were 64 mm and 67 mm, respectively (Figure 3). The mean 

selectivities of O. argyrosoma for ¼ inch and ¾ inch mesh sizes were, respectively, 63 mm and 66 mm (Figure 4). 

The very small mean size selectivity differences (1 mm – 3 mm) between ¼ inch and ¾ inch mesh sizes among the 

three species demonstrates that there is little or no influence on fish size selection by the two different mesh sizes. 

Age-at –selectivity is shown in Table 1. 

Per recruit analysis 

The YPR curves for ¼ inch and ¾ inch nkacha net mesh sizes for C. virginalis, L. sp. ‘pinkhead’ and O. argyrosoma 

are shown in Figure 6. The response of YPR to differences in the age at capture using ¼ inch or ¾ inch nkacha nets 

was marginal for the three species. The number of nkacha gear units calculated for the F0.1 and Fmax TRP’s for C. 

viginalis, L. sp. ‘pinkhead’ and O. argyrosoma are presented in Table 3. C. virginalis was the least resilient to

117 


increased fishing effort (F0.1 = 84-85 gears; Fmax = 154-160 gears), while O. argyrosoma was the most resilient (F0.1 

= 161-168 gears; Fmax = 348-395 gears) of the three species. 

(a) 

Frequency 

600 

500 

400 

300 

200 

100 

0 

0 20 40 60 80 100 120 140 


Figure 2. Size selectivity of Copadichromis virginalis caught by (a) ¼ inch and (b) ¾ inch nkacha net 

mesh sizes in Lake Malombe. 

(a) 

Figure 3. Size selectivity of Lethrinops sp. ‘pinkhead’ caught by (a) ¼ inch and (b) ¾ inch nkacha net 


(a) 

Frequency 

Frequency 

800 

600 

400 

200 

0 

600 

400 

200 

0 

0 20 40 60 80 100 120 140 


0 20 40 60 80 100 120 140 


Figure 4. Size selectivity of Otopharynx argyrosoma caught by (a) ¼ inch and (b) ¾ inch nkacha net 


(b) 

(b) 

(b) 

Frequency 

Frequency 

Frequency 

100 

80 

60 

40 

20 

0 

150 

100 

50 

0 

120 

100 

80 

60 

40 

20 

0 

0 20 40 60 80 100 120 140 


0 20 40 60 80 100 120 140 


0 20 40 60 80 100 120 140 

Total length (mm)

118 


Table 2. Catch composition in the nkacha net fishery of Lake Malombe, all months combined. (n = 107 

catches) 

Family Species name (%) Family Species name (%) 

ANGUILLIDAE 

Anguilla nebulosa labiata 0.00 

CICHLIDAE cont. 

Oreochromis shiranus 1.17 

BAGRIDAE Bagrus meridionalis 0.29 

CHARACIDAE Brycinus imberi 0.08 

CICHLIDAE Astatotilapia calliptera 0.12 

Aulonocara guentheri 0.98 

Aulonocara macrochir 0.10 

Buccochromis atritaeniatus 0.85 

Buccochromis nototaenia 0.18 

Copadichromis chrysonotus 17.98 

Copadichromis cyaneus 0.08 

Copadichromis trimaculatus 0.04 

Copadichromis virginalis 11.93 

Corematodus taeniatus 0.01 

Ctenopharynx intermedius 0.31 

Ctenopharynx nitidus 0.00 

Ctenopharynx pictus 0.01 

Dimidiochromis compreceps 0.06 

Dimidiochromis kiwinge 0.02 

Fossochromis rostratus 0.00 

Hemitaeniochromis discorhynchus 0.21 

Hemitilapia oxyrhynchus 0.01 

Lethrinops “deep water” altus 4.44 

Lethrinops lethrinus 1.55 

Lethrinops longipinnis 0.28 

Lethrinops macrochir 0.01 

Lethrinops parvidens 0.38 

CLARIIDAE 

CYPIRINIDAE 

Oreochromis spp. 0.52 

Oreochromis squamipinnis 0.88 

Otopharynx argyrosoma 7.60 

Otopharynx spp. 0.06 

Otopharynx tetraspilus 1.10 

Otopharynx tetrastigma 16.68 

Placidochromis subocularis 0.75 

Protomelas labridens 0.23 

Protomelas similis 0.33 

Protomelas triaenodon 0.02 

Rhamphochromis longiceps 0.53 

Rhamphochromis macrophthalmus 0.15 

Rhamphochromis spp. 0.32 

Sciaenochromis gracilis 0.02 

Stigmatochromis woodi 1.41 

Tilapia rendalli 0.00 

Tramitichromis lituris 0.13 

Trematocranus placodon 1.42 

Clarias gariepinus 0.67 

Bathyclarias longibarbis 0.00 

Bathyclarias spp. 0.00 

Barbus arcislongae 0.05 

Barbus litamba 0.16 

Engraulicypris sardella 0.83 

Opsaridium microcephalus 0.13 

Lethrinops sp. ‘pinkhead’ 23.97 MASTACEMBELIDAE Aethiomastacembelus shiranus 0.00 

Lethrinops spp. 0.03 

Maravichromis anaphyrmus 0.05 

Mylochromis mola 0.05 

Nyassachromis argyrosoma 0.66 

MOCHOKIDAE 

MORMYRIDAE 

Synodontis njassae 0.02 

Hippopotamirus discorhynchus 0.13 

Table 3. Number of gear units calculated for the F0.1 and Fmax TRPs for Copadichromis virginalis, 

Lethrinops sp. ‘pinkhead’ and Otopharynx argyrosoma in Lake Malombe. 

Number of gears 

Species F0.1 Fmax 

Copadichromis virginalis 84-85 154-160 

Lethrinops sp. ‘pinkhead’ 122-131 224-249 

Otopharynx argyrosoma 161-168 348-395

Yield-per-recruit 

2.5 

2.0 

1.5 

1.0 

0.5 

0.0 

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 

2.0 

1.6 

L. "spp" pinkhead 

F0.1 Fmax 

1.2 

0.8 

0.4 

0.0 

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 

1.2 

1.0 

O. argyrosoma 

F0.1 Fmax 

0.8 

0.6 

0.4 

0.2 

0.0 

C. virginalis 

F CUR 

F 0.1 

F CUR 

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 

F 

119 


Fmax 

F CUR 

1/4 inch mesh 

3/4 inch 

Figure 5. Yield per recruit curves for ¼ inch and ¾ inch mesh sizes nkacha nets for Copadichromis 

virginalis , Lethrinops sp. ‘pinkhead’ and Otopharynx argyrosoma in Lake Malombe. 

Discussion 

The use of TRPs for the management of a fishery is common practice (Caddy & Mahon 1995), and the Malawi 

Department of Fisheries recommends TRPs based on catch per unit effort (CPUE) in the small-scale fishery 

(Bulirani et al. 1999). The yield-per-recruit (YPR) approach allows for the determination of at least two commonly 

used TRPs: firstly, the fishing mortality which corresponds to the maximum of the yield-per-recruit curve (Fmax) and 

secondly, the marginal yield or F0.1 strategy (Gulland & Boerema 1973, Deriso 1987). The Fmax approach has 

received some criticism in the past (Punt 1993), and the F0.1 TRP is this is considered to be more robust management 

strategy (King, 1995). 

It has long been recognised that when a single gear catches several different species it is impossible to manage each 

species at its optimum level (Murawski & Finn, 1988). This complicates the management of multi-species fisheries. 

One approach that has been applied in such cases is to manage the fishery according to the least resilient species 

(Weyl 1998). In the Lake Malombe scenario, C. virginalis was the least resilient of the three species and 85 nkacha 

nets are recommended if the species is to be managed at a F0.1 level. Based on surplus production models, Weyl et 

al. (2001) recommended 170 nkacha gear units for Lake Malombe in order to maintain a MSY of 5000 – 7000 tons. 

This study shows that from a YPR perspective, this level of fishing would exceed the F0.1 TRP for all three species 

(Table 2). Management at Fmax, for any of the three species, would allow for 150-160 gears which approximate 

Weyl et al.’s (2001) estimate for MSY. Since difference in selectivity of ¼- and ¾-inch meshes yielded only a 

marginal difference in YPR the failure of the nkacha net fishery to adopt the ¾-inch mesh nets is not considered 

drastic.

120 


However, it should be recognized that the Fmax and F0.1 strategies do not take into account whether sufficient 

spawner biomass is conserved to ensure sufficient recruitment in the future (Deriso 1987, Sissenwine & Shepherd 

1987, Clarke 1991, Punt 1993). It is also recognised that the per-recruit approach has limitations, such as its 

assumption of constant recruitment (Perreiro 1992), and, therefore, its use as a predictive tool is limited to shortterm 

predictions. It is crucial that relevant long-term directed catch-at-age or –length data are collected, in all 

fisheries, to allow for the combination of the per-recruit data with other age-structured models in order to provide 

more accurate, comprehensive and sustainable strategies for long-term management. 


We should like to thank Mr. Davis S. Mandere for help in fish identification. We also extend our gratitude to all the 

research and support staff who assisted in the programme. This study was funded by the Fisheries Department and 

the National Aquatic Resource Management Programme (NARMAP). 

References 

Bulirani, A.E., Banda, M.C., Palsson, O.K., Weyl, O.L.F., Kanyerere, G.Z., Manase, M.M. & Sipawe, R.D. 1999. Fish Stocks 

and Fisheries of Malawian Waters: Resource Report. Government of Malawi, Fisheries Department, Fisheries Research 

Unit.54pp. 

Caddy, J.F. & Mahon, R. 1995. Reference points for fisheries management. FAO Fisheries Technical Paper No. 347, Rome, 

FAO. 83 pp. 

Clark, W.G. 1991. Groundfish exploitation rates based on life history parameters. Can. J. Fish. Aquat. Sci. 48: 734-750. 

Deriso, R.B. 1987. Optimal F0.1 criteria and their relationship to maximum sustainable yield. Can. J. Fish. Aquat. Sci. 44 (Suppl. 

2): 339-348. 

FAO. 1993. Fisheries management in south-east Lake Malawi, the Upper Shire River and Lake Malombe, with particular 

reference to the fisheries on chambo (Oreochromis spp.). CIFA Tech. Pap. 21. FAO, Rome. 113 pp. 

Gulland, J.A. & Boerema, L.K. 1973. Scientific advice on catch levels. Fish. Bull. 71: 325-335. 

Jambo, C.J. 1997. Aspects of the ecology and reproductive biology of three cichlid fish species of southern Lake Malombe 

(Malawi). M.Sc. Thesis. Rhodes University, Grahamstown, South Africa. 126 pp. 

King, M. 1995. Fisheries biology, assessment and management. Fishing News Books, Oxford. 341 pp. 

Murawski, S.A. & Finn, J.T. 1988. Biological bases for mixed-species fisheries: species co-distribution in relation to 

environmental and biotic variables. Can. J. Fish. Aquat. Sci., 45: 1720-1735. 

Pereiro, J.A. 1992. Some conceptual remarks on yield per recruit. Fish. Res. 13: 423-428. 

Punt, A.E. 1993. The use of spawner-biomass-per-recruit in the management of linefisheries. Special Publication of the 

Oceanographic Research Institute, Durban, 2: 80-89. 

Sissenwine, M.P. & Shepherd, J.G. 1987. An alternative perspective on recruitment overfishing and biological reference points. 

Can. J. Fish. Aquat. Sci. 44: 913-918. 

Weyl, O.L.F. 1998, The dynamics of a sub-tropical lake fishery in central Mozambique. PhD Thesis, Rhodes University, 200p. 

Weyl, O.L.F. 2001. Mortality rates and catchability coefficients for three major target species in the nkacha net fishery of Lake 

Malombe. National Aquatic Resource Management Programme (NARMAP) Working Paper No. 3. 7 pp. 

Weyl, O.L.F., Banda, M., Sodzabanja, G., Mwenekibombwe, L.H., Namoto, W. & Mponda, O.C. 2000. Annual frame survey: 

September 1999. Fisheries Bulletin No. 42. Department of Fisheries , Lilongwe, Malawi. 56+23 pp. 

Weyl, O.L.F., Banda, M.C., Manase, M., Namoto, W. & Mwenekibombwe, L.H. 2001. Analysis of catch and effort data for the 

fisheries of Lake Malombe, 1976-1999. Fisheries Bulletin No. 45. Department of Fisheries , Lilongwe, Malawi. 53 pp.

121 


Drifting Long Line, A Potential Fishing Method for the Northern Part of Lake 

Nyasa/Malawi/Niassa 

K.J. Kihedu, M.K.L. Mlay, J.A. Mwambungu and B.P. Ngatunga 

Tanzania Fisheries Research Institute, Kyela Centre, P.O. Box 98, Kyela, Mbeya Region, East Africa 

Abstract 

Fishing gears commonly used in the Tanzanian Waters of Lake Nyasa consist of open water seine nets (Ndaturu/Pajero)) gillnets 

(vilepa) traps (migono) beach seine nets (kokoro) and hook and line (ndoano). These methods of fishing exploit mainly the near 

shore and riverine species. Based on recommendation of the UK/SADC Pelagic Fish Resource Assessment Project on the use of 

drifting long line to exploit offshore fish communities, trial fishing was carried out off Lupingu, Ikombe, Manda and Mbamba-bay 

villages in Tanzania. The results were encouraging and they are presented and discussed. With a mean daily catch rate of 

5.8 kg/100 hooks/2hours of fishing, this method proved to local fishermen to be more profitable and environmentally friendly than 

using the destructive beach seine nets. 

Introduction 

Lake Nyasa is the most southern of the African great lakes situated in the rift valley of East Africa. It lies between 

9° 30'S and 14° 30'S. It is the third largest lake in Africa with respect to surface area and the second most 

voluminous. The lake is long and narrow with an approximate length of 550 km and a mean width of 50-60 km. 

With a surface area of 28 800 km² and a volume of 8 400 km³ it gives an average depth of 292 m. (Menz 1995). The 

maximum depth is given as 785 m by Gonfiatini et al. (1979), with deepening of the basin towards the north. 

Tanzanian shoreline of Lake Nyasa extend 300 km in the north west to north east. Most of the inshore area of the 

lake on the Tanzanian side is very deep with few patches of shallow zones especially around the mouths of rivers 

Songwe, Kiwira, Mbaka, Rufilyo and Ruhuhu. 

The fisheries of the lake are very varied exploiting most of the inshore water communities as well as the river 

running species and some offshore communities. 

Most of the fishing in Tanzanian waters of the Lake is done with dug-out canoes. Fishing gears commonly in use 

consist of open water seine nets, gillnets, beach seine nets, traps and hooks. Longline has been used in Lake Nyasa 

to catch fish mainly catfishes (Bagrus meridionalis, Clarias species and Bathyclarias species). No pelagic longline 

has been reported on Lake Nyasa (ICLARM and GTZ 1991). The UK/SADC pelagic Fish Resource Assessment 

Project had recommended fishing with pelagic handline or longline using usipa or some other fish as bait in catching 

Rhamphochromis spp. especially the larger individuals that occur in 100-200 m of water (UK/SADC interim report, 

1993). 

Acoustic evidence strongly suggests that fish biomass further offshore remained relatively constant (Rufli, 1982). 

The fish in the Pelagic zone formed a simple but highly structured community. Fish biomass of the pelagic zone of 

Lake Nyasa was estimated to be 90 kg/ha (Rufli and Vitullo, 1982). Fish stocks in Lake Nyasa still seem to be high 

enough suggesting that the fish are less heavily exploited in the lake especially in the pelagic zone. 

Materials and Methods 

Fishing was conducted in the Tanzanian pelagic waters of Lake Nyasa in Ludewa district off Lupingu village 

(Figure 1). An 18 footer rubber dingy mounted with a 30Hp Yamaha out-board petrol engine was used during 

fishing operations. Three gangs of longlines each consisting of 100 hooks of sizes No. 7, No 11 and No. 13 were 

constructed. According to the numbering system of hooks number 7 was the largest and number13 the smallest. For 

each gang, hooks were attached to 0.5 mm thick monofilament line and individually hang 4m apart on a 0.8mm 

thick monofilament line 400 m long. During the fishing operations the three gangs were joined together to constitute 

a line 1200 m long. Each line was suspended at each end on a rope 4 x100 m deep attached to a surface buoy with a 

flag at both ends (fig.3). The hooks were baited with Engraulicypris sardella ‘usipa’, a bait known to be preferred 

by most of the predator fish found in the lake. 

The long line was set early in the morning at the depth of 100 m at a point where the water depth was 130 m or more 

targeting for any available predatory fish at that depth. The longline was left in the water for 2hrs before hauling. 

This is considered as the maximum time the fish is allowed to feed and taking into consideration that the quality of 

the bait deteriorates with time (Allison et al. 1995). 

The total catch from each longline was weighed and then sorted into species and the total length of each fish was 

recorded. These data were prerequisite for the information required; the catch rate and catch composition.

122 


The same method was employed when setting drifting longlines at Ikombe, Manda and Mbamba-bay (Fig.1). In 

these demonstrations however hooks No.13, 11 and 9 were employed instead of hook sizes No.13, 11 and 7 set at 

Lupingu. 

Malawi 

Kyela 

Itungi port Matema 

Nkanda 

Ludewa 

Lupingu 

Manda 

Figure 1. Lake Nyasa (Tanzania side). 

Results and Discussion 

TANZANIA 

Mbinga 

Mbamba-bay 

MOZAMBIQUE 

Catch rates 

Catch rates (kg/300 hooks/2 hours) for the three longlines from pelagic waters of Lake Nyasa off Lupingu village 

(Ludewa district) set for three weeks in August/September, 1995 are shown in Table 1. Total daily catch ranged 

between 279kg/300h/2hrs and 13.5kg/300h/2hrs (mean = 5.759kg/300h/2hrs). Catches were higher in hook size 

umber 11 (Average catch rate = 2.63kg/100h/2hrs) and least in hook size number 7 (Average catch rate = 

0.99kg/100h/2hrs). There is no significant difference in catches with respect to varied water depth, weather and 

station (t-test; p.0.05). 

Table 1: Catch rates (kg/100 hooks/2 hours) for three different sizes of hooks from pelagic waters of 

Lake Nyasa off Lupingu village (Ludewa district) caught in August/September, 1995. 

LL 7 LL 11 LL 13 TOTAL 

Date Cruise No. Depth. Weather Station Weight Station Weight Station Weight 

22.8.95AM 1 140m Strong wind 1 3.5 3 7.29 2 2.25 13.50 


24.8.95AM 3 180m Calm 2 1.00 3 0.15 1 1.33 3.48 

25.8.95AM 4 150m Calm 3 0.00 1 2.46 2 0.85 3.41 

26.8.95AM 5 130m Calm 3 0.35 2 4.17 1 2.94 7.46 

27.8.95AM 6 130m Calm 2 0.58 1 1.73 3 1.16 3.47 




31.8.95AM 10 130m Calm 3 0.44 1 1.05 2 1.70 3.19 

1.9.95AM 11 150m Calm 3 0.00 2 2.12 1 1.95 4.07 

2.9.95AM 12 140m Calm 3 0.00 1 0.4 2 3.31 3.71 

2.9.95PM 13 130m Calm 3 0.00 1 2.45 2 0.90 3.35 

3.9.95AM 14 150m Calm 3 0.00 1 7.10 2 1.90 9.00 

4.9.95AM 15 160m Calm 2 0.84 1 0.40 3 2.97 4.21 

5.9.95AM 16 140m Calm 1 6.52 3 3.11 2 2.20 11.83 


TOTAL WEIGHT 16.86 44.73 35.29 100.43 

AVERAGE WEIGHT (KG) 0.99 2.63 2.08 575.90

123 


Catch composition 

Catch composition of fish caught in three hook lines is summarized in Table 2. The species caught contribute five 

families out of the ten considered to exist in Lake Nyasa. The families are cichlidae (Rhamphochromis and 

Lethrinops), Clarridae (Clarias spp.) (Bathyclarias sp.), Bagridae (Bagrus meridionalis), Mochokidae (Synodontis 

njassae) and Cyprinidae (Opsaridium spp.). From all three longlines the catch composition showed that the 

predatory cichlids of the genus Rhamphochromis are more common in the catches followed by clariid catfishes, 

Bathyclarias spp. Other fish in the catch included Bagrus meridionalis, Synodontis njassae, Clarias spp., Lethrinops 

spp. and the predatory cyprinids Opsaridium microlepis and O.microcephalum. Hook number 7 caught more bigger 

fish Bagrus meridionalis, number 11 caught more medium sized fish, Rhamphochromis ferox while small fish- 

Rhamphochromis longiceps was caught mainly on hook number 13. 

Table 2. The size composition of fish caught in three hook lines off Lupingu village in August/September, 

1995 

Family Genus/Species Weight (Kg/haul) and numbers/haul per 100 

hooks 

Line No. 7 Line No. 11 Line No. 13 

Kg. No. Kg. No. Kg. No. 

Claridae Bathyclarias spp. 0.01 0.07 1.19 0.41 0.29 0.24 

Clarias spp. 0.09 0.07 0.00 0.00 0.00 0.00 

Bagridae Bagrus meridionalis 0.53 0.60 0.15 0.24 0.30 0.18 

Mochokidae Syodontis njassae 0.00 0.00 0.01 0.06 0.01 0.06 

Cichlidae Rhamphochromis spp 0.35 0.97 1.30 3.53 1.32 3.52 

Lethrinops spp. 0.11 0.13 0.06 0.41 0.07 0.35 

Cyprinidae Opsaridium microcephalum 0.00 0.00 0.00 0.00 0.02 0.06 

Opsaridium microlepis 0.00 0.00 0.01 0.06 0.00 0.02 

TOTAL 1.08 2.62 2.07 

Length distribution 

Catch rates (Kg /300hooks/2hrs) for three longlines from pelagic waters of Lake Nyasa off Nkanda village (Kyela 

district) set for 12 days in October, 1998. Deductions as Shown in Table 3 show that the total daily catch ranged 

between 2.2 kg/300 h/2 hours and 18.95 kg/300 hooks/2 hours (Mean = 9:40 kg/300 hooks/2 hours). Catches were 

higher in hook size number 13 (Average catch rate = 3.61 kg/100h/2 hours) and least in hook size No. 9 (average 

catch rate = 2.57 kg/100 hooks/2 hours). 

The catch composition of fish caught from the three gangs of long lines from the pelagic waters of Lake Nyasa off 

Nkanda Village (Kyela district) set for two weeks in October 1998 is summarized in Table 4. The species caught 

include Rhamphochromis and Lethrinops (cichlidae), Bathyclarias spp. (claridae), Bagrus meridionalis (bagridae) 

and Synodonts njassae (mochokidae). The catch composition show that Bagrus meridionalis was more common in 

the catches followed by predatory cichlid of the genus Rhamphochromis and by clariid catfishes, Bathyclarias spp. 

Table 3. Catch rates (Kg/100 hooks/2 hours) for three different sizes of hooks from Pelagic Waters of 

Lake Nyasa off Nkanda Village (Kyela District) caught in October, 1998. 

LL 9 LL 11 LL 13 

Date Cruise No. Water Station Weight Station Weight Station Weight Total 

Depth. 

Weight 

8.10.98 1 120m 3 0.50 2 2.30 1 4.90 7.70 

9.10.98 2 120m 2 0.10 1 0.15 3 - 

0.25 

12.10.98 3 119m 3 0.70 1 1.55 2 1.75 4.00 

13.10.98 4 120m 1 2.35 3 8.61 2 4.45 15.41 

14.10.98 5 120m 1 3.65 2 2.55 3 1.05 7.25 

15.10.98 6 120m 2 9.15 1 2.75 3 4.22 16.12 

16.10.98 7 120m 3 0.46 1 2.15 2 2.45 5.06 

17.10.98 8 123m 3 5.10 1 7.85 2 6.00 18.95 

18.10.98 9 120m 2 4.05 1 5.25 3 5.95 15.25 

21.10.98 10 120m 1 4.30 3 1.25 2 5.45 11.00 

21.10.98 11 120m 3 0.45 1 1.50 2 0.25 2.20 

22.10.98 12 120m 3 1.50 2 1.90 3 2.05 5.45 

TOTAL WEIGHT 32.31 37.81 38.52 107.09 

AVERAGE WEIGHT(KG) 2.68 3.02 3.21 8.92

124 


Table 4. The catch composition of fish caught in three hook lines off Nkanda village in October, 1998 

Family Genus/Species Weigh (kg/haul) and numbers/haul per 100 hooks 





Mochokidae Synodontis njassae 0.00 0.00 0.00 0.00 0.02 0.15 

Cichlidae Diplotaxodon spp. 0.00 0.08 0.05 0.31 0.02 0.31 

Rhamphochromis spp. 0.28 0.77 0.13 0.62 0.38 0.92 


TOTAL: 2.57 3.22 3.61 

Catch rates ( Kg/300hooks/2hrs) for three longlines from pelagic waters of Lake Nyasa off Manda village (Ludewa 

district) set for 14 days ( June/July 1999) are shown in Table 5, the daily catch ranged between 2.70 kg/300 hooks/2 

hours and 33.73 kg/300 hooks/2 hours. (Mean = 16.33 kg/300 hooks/2 hours). Catches were higher in hook size 

number 9 (average catch rate = 5.58 kg/100 hooks/2 hours) and least in hook size number 11 (average catch rate = 

5.24 kg/100 hooks/2 hours). 

The catch composition summarized in Table 6 below. Species caught include Rhamphochromis, Diplotaxodon, 

Lethrinops (Cichlidae), Bathyclarias spp. and Clarias (Claridae), Bagrus meridionalis (Bagridae) and Synodontis 

njassae (Mochokidae). The catch composition showed that Bagrus meridionalis was more common in the catches 

followed by Rhamphochromis followed by Diplotaxodon and by clariid catfishes Clarias and Bathyclarias. spp. 

Catch rates (Kg/300hooks/2hrs) for three sets of longlines from pelagic waters of Lake Nyasa off Mhalo village 

(Mbamba-bay) Mbinga district set for 9 days May,2000. Table 7, show the daily catch ranged between 1.41 kg/300 

hooks/2 hours and 6.72 kg/300 hooks/2 hours (Mean = 3.68 kg/300 hooks/2 hours). Catches were higher in hook 

size No. 13 (average catch rate 1.57 kg/100 hooks/2 hours) and least in hook size No. 9 (average catch rate 

0.97 kg/100 hours/2 hours). 

Table 5. Catch rates (kg/100 hooks/2 hours) from three different sizes of hooks from Pelagic Waters of 

Lake Nyasa off Fokland (Manda) caught in June-July, 1999 

Date Cruise Water Long Line 

Long Line 

Long Line 

Total 

No. Depth(m) Hook No. 9 Hook No. 11 Hook No. 13 Weight 

Station Weight Station Weight Station Weight 

No. (kg) No. (kg) No. (kg) 

20.06.99 1 120 3 4.50 2 0.95 1 0.25 6.40 

21.06.99 2 115 3 4.25 2 - 1 - 4.25 

22.06.99 3 115 3 6.03 1 13.57 2 11.93 31.53 

23.06.99 4 115 1 18.28 2 1.90 3 - 20.18 

24.06.99 5 110 2 6.45 1 6.59 3 14.09 27.13 

25.06.99 6 110 2 0.30 1 10.20 3 1.35 11.85 

26.06.99 7 110 1 6.62 2 2.50 3 0.49 9.61 

27.06.99 8 115 1 1.81 2 0.45 3 0.90 3.16 

02.07.99 9 110 1 0.33 2 1.38 3 0.99 2.70 

03.07.99 10 115 1 3.12 2 8.30 3 11.45 22.87 

04.07.99 11 120 1 2.69 2 8.84 3 4.49 16.02 

05.07.99 12 110 2 6.75 1 8.34 3 8.31 23.40 

06.07.99 13 115 2 3.55 1 3.55 3 8.62 15.72 

07.07.99 14 115 1 13.40 2 6.74 3 13.59 33.73

125 


Table 6. The catch composition of fish caught in three hook line off Forkland-Manda in June-July, 1999 

Family Genus/Species Weight (kg/haul) and number/haul per 100 hooks 




Clarias spp. 0.94 0.93 1.14 0.93 0.15 0.21 


Mochokidae Syonodontis njassae - - 0.04 0.07 0.01 0.07 

Cichilidae Rhamphochromis spp. 0.16 0.50 0.65 2.43 1.26 3.64 

Diplotaxodon spp. 0.04 0.57 0.06 0.50 0.13 1.29 


Others 0.09 0.21 0.04 0.14 - - 

The catch composition of fish caught from the three sets of long lines from the pelagic waters of Lake Nyasa off 

Mhalo – Mbamba-bay (Mbinga district) set for 9 days in February 2000 is summarized in Table 8. 

Species caught include: Rhamphochromis, Diplotaqxodon and Lethrinops (Cichlidae) Bathyclarias spp. and Clarias 

(Claridae) and Bagrus meriodionalis (Bagridae). The catch composition showed that Rhamphochromis spp. were 

more common in the catches (actually dominating the catches) followed by Diplotaxodon and by Lethrinops spp. 

As pointed out earlier Rhamphchromis species were considered more abundant for long line fishery in the pelagic 

waters of Lake Nyasa. The setting of these demonstrations thus took into consideration feeding habits of this 

predatory cichlid. The major feeding period of which is in the morning and a minimum peak in the afternoon. 

However as observed from the catch composition Bagrus meridionalis was more common. Important also are the 

clariid catfishes Bathyclarias species and Clarias species. Since they constitute significantly to the long line catches 

in the northern part of the lake, it is imperative to consider their feeding behaviour in future demonstrations. 

Table 7. Catch rates (kg/100 hooks/2 hours from three different sizes of hooks from Pelagic Waters of 

Lake Nyasa off Mhalo (Mbamba-bay) caught in May, 2000 

Date Cruise Water Long Line Hook Long Line Hook Long Line Hook Total 

No. Depth(m) No. 9 

No. 11 

No. 13 

Weight 

Station Weight Station Weight Station Weight 

No. (kg) No. (kg) No. (kg) 

8.02.2000 1 120 1 0.51 2 1.70 3 2.46 4.67 

9.02.2000 2 115 3 1.73 2 0.65 1 0.51 2.89 

10.02.2000 3 120 3 - 2 0.50 1 0.91 1.41 

11.02.2000 4 120 3 1.69 2 0.53 1 2.08 4.30 

12.02.2000 5 120 1 - 2 2.77 3 3.95 6.72 

13.02.2000 6 120 1 0.28 2 0.82 3 1.49 2.59 

16.02.2000 7 120 1 2.27 3 2.50 2 0.95 5.72 

17.02.2000 8 120 2 0.35 1 0.25 3 1.05 1.65 

18.02.2000 9 130 3 1.90 1 0.60 2 0.70 3.20 

Table 8. The catch composition of fish caught in three hook lines off Mhalo-Mbamba-bay in February, 

2000 

Family Genus/Species Weight (kg/haul) and numbers (haul/100 hooks) 



Claridae Bathyclarias spp. 0.16 0.11 - - - - 

Clarias spp. 0.04 0.11 - - - - 

Bagridae Bagrus meridionalis 0.04 0.11 - - - - 

Cichilidae Rhamphochromis spp. 0.64 2.00 0.98 3.33 1.96 4.22 

Diplotaxodon spp. 0.03 0.11 0.06 0.22 0.03 0.22 

Lethrinops spp. 0.06 0.22 - - - -

126 


Table 9. Range and mean of total lengths (cm) of fish caught by long line fishery from Pelagic Waters of 

Lake Nyasa off Nkanda/Ikombe in October, 1998. 

Fish Species Range Mean N. 

Bagrus meridionalis 18.0 – 84.0 37.1 154 

Rhamphochromis spp. 20.0 – 45.0 31.7 33 

Clarias spp. 28.0 – 72.0 59.5 14 

Diplotaxodon spp. 16.0 – 23.0 19.1 9 

Synodontis njassae 13.0 – 18.0 15.0 3 

Haplochromis spp. 16.0 – 28.0 23.5 14 

Table 10. Range and mean of total lengths (cm) of fish caught by long line fishery from Pelagic Waters of 

Lake Nyasa off Fokland (Manda) in June- July, 1999. 


Bagrus meridionalis 19.0 – 66.5 40.7 118 


Clarias sp. 19.0 – 90.0 61.0 23 


Bathyclaria spp. 36.0 – 87.5 59.5 21 

Lethrinops spp. 11.0 – 27.5 23.1 14 

Synodontis njassae 13.0 – 15.0 14.0 2 

Haplochromis spp. 21.0 – 34.5 29.3 5 

Copadichromis spp. 10.0 – 10.0 10.0 1 

Table 11. Size range and mean of total lengths (cm) of fish species caught by long line fishery from 

Pelagic Waters of Lake Nyasa off Mhalo (Mbamba-bay) in February, 2000. 



Bagrus meridionalis 33 33 1 

Bathyclarias spp. 31.5 31.5 1 


Lethrinops spp. 18.5 – 25.0 21.8 2 

Recommendations 

Fishermen should be encouraged to practice drifting longline because: 

• It is cost effective 

• Environmentally friendly 

• Targets for mainly large fish 

• Hooks # 9-13 are recommended 

• Other baits also be investigated 

• Demonstrations should cover the entire Tz coast. 

References 

Gonfiatini, R., Zuppi, G., Eccles, D. H. & Ferro, W. 1979. Isotope investigation of Lake Malawi. In Isotope in lake studies. 

International Atomic Energy Agency, Vienna. 

ICLARM & G.T.Z. 1991. The context of small scale integrated agriculture – Aquaculture systems in Africa. A case study of 

Malawi: ICLARM Stud. Rev. 

Menz A. 1995. The fishery Potential and Productivity of the pelagic zone of Lake Malawi/Niassa. Chathan, U.K. Natural 

Resources Institute. 

Rufli, H. 7 Vitullo, J. A. 1952. Preliminary estimate of the abundance of Pelagic Fish Stocks in Lake Malawi. In biological 

studies on the pelagic ecosystem of Lake Malawi. Rome/FAO/UNDP F< LDP/MLW/75/019 Technical Report I: 

138/153. 

UK/SADC. 1993. Pelagic Fish Resource Assessment Project. Interim Technical Report. 

Walezak, P.S. 1982. Feeding habits and daily food consumption rates of the major Pelagic fish species of Lake Malawi. 

Supplement to: Biological Studies of the Pelagic Ecosystem of Lake Malawi, FAO Technical Report 1, FI: 

DP/MLW/75/019, Field document 15.

127 


Gear and species selectivity of the chilimira kauni fishery in Lake Malawi 

Jacqueline Chisambo 

Fisheries Research Unit, P.O. Box 27, Monkey Bay 

Abstract 

Kauni fishing in Lake Malawi occurs by using a chilimira net (an open water seine net) at night using light from paraffin lamps to 

attract the fish. Kauni fishing occurs in both the south-east (SEA) and south-west arms (SWA) of Lake Malawi. Catches from kauni 

gears landing at Kela beach (SEA) and Msaka beach (SWA) were examined for a one-year period to determine the potential effects 

on the fish stocks. Catch composition and length frequency assessments were performed. The kauni fishery in both areas was 

multi-species with 70 species recorded in the Kela fishery and 65 species in Msaka. At Kela the major genera targeted were 

Rhamphochromis, Oreochromis and Copadichromis, which made up over 75 % of the catch composition. At Msaka the major 

genera were Rhamphochromis, Engraulicypris and Copadichromis, which made up over 85 % of the catch. Length frequencies of 

Oreochromis spp. Rhamphochromis spp. and Copadichromis virginalis are examined and recommendations for managing this 

fishery are made. 

Introduction 

It is a well-known fact for Lake Malawi that there has been a decline in the chambo Oreochromis spp. stocks. 

Catches of chambo in Lake Malawi peaked at 8-9 000 tons in the late 1970s and in the mid-1980s but declined in 

1986-87 (Bulirani et al. 1999). The south-east arm is the most important fishing ground for chambo. Catches were 

stable at 5-6 000 tons during 1986-92 but have declined to 1 800-2 800 tons a year (Bulirani et al. 1999). The bulk 

of the catch has been taken in the traditional fishery, which uses various gears such as gill nets, chambo seines, 

kambuzi seines and chilimira nets. The chambo stocks are now considered threatened and in need of management 

(Bulirani et al. 1999). 

In the south-east arm of Lake Malawi it was observed that, while chambo formed only a small component of the 

chilimira net fishery as a whole, the high effort levels in this fishery result in the fishery contributing between 25% 

and 46% of the total chambo catch (Weyl, 1999). The chambo is being exploited by chilimira nets (open water 

seine nets; FAO, 1993), which are locally known as kauni (a term referring to the light attraction method by using 

lamps at night to attract fish into the nets: Banda unpublished). The chilimira that targets chambo is a derivative of 

the gear used for catching usipa and this derivative gear used in the kauni fishery is known to operate in both the 

southeast and southwest arms of the lake. It has been found that the chilimira fishers in Area A of the south–eastern 

arm directly target chambo (Weyl, 1999, Weyl & Banda, 2001). A preliminary report on the fishery in this area 

showed that over 80% of the chambo caught in the kauni fishery was immature (Banda unpublished). This indicated 

that the fishing method was affecting the recruitment of chambo (Banda, unpublished). Due to this observation 

Banda recommended the closure of the kauni fishery. 

The chambo directed kauni fishery of Area A is of major concern and it has been recommended that there should be 

a closure of this fishery in this area since it is an important fishing ground for chambo (Weyl, 1999; Bulirani et al. 

1999). This recommendation came about in order to restore chambo stocks to previous levels and increase 

production. The fisheries Conservation and Management Regulations 2000 prohibit ‘kauni for chambo’ thus 

catching chambo in the kauni fishery is illegal (Fisheries Management and Conservation Act, 2000). 

This paper is a result of a one-year study of the chilimira/kauni fishery in the southern part of Lake Malawi. The 

objective of this study was to examine and compare the kauni fisheries in the south east arm (SEA) and south west 

arm (SWA) of Lake Malawi to determine the potential effects of these fisheries on the fish stocks. 

Methodology 

Catch assessment of the kauni fisheries operating at Kela (SEA) and Msaka (SWA) landing sites were conducted 

once for three consecutive days a month for a one-year period. At these landing sites, as many gears as possible 

were sampled. The weight of the entire catch was assessed and a sample was randomly selected and weighed. The 

species composition of the catches was assessed as each fish in the sample was sorted to species level. Each species 

group was weighed separately and each fish was measured for total length (TL) in millimetres. 

Results 

The kauni fishermen, landing at Kela, fish in Area A, south of Boadzulu Island, while those landing at Msaka 

usually fish close to Mumbo Island. It was observed that both the fisheries in the SWA and SEA are multi-species 

catching many different species, many of which were cichlids. The number of species recorded at Kela was 70 and

128 


at Msaka 65 (Table 1). The contribution of the different species in the fishery varied from month to month as well 

as by area. 

The fish species identified in the fishery were lumped in major species groups for species composition analysis. 

These major groups were Rhamphochromis spp. Oreochromis spp. Copadichromis spp. as well as usipa 

(Engraulicypris sardella) and ‘other’ group. The ‘other’ group consisted of all other species caught in the fishery 

that were not included in the major species groups. 

At Msaka the main contributing species groups were Rhamphochromis spp, usipa and Copadichromis spp. 

However, the contribution of these various groups varied for the different months of the year. The ‘other’ group 

contributed significantly in November. There was no data collected in June, and in December there was an anomaly 

in the data that lead to all the data in December being discarded (Figure 1). The Rhamphochromis spp group 

contributed 51% of the total catch followed by usipa, which contributed 25%. The ‘other’ group contributed 14%, 

Copadichromis spp. group contributed 10%, and chambo contributed only 0.5% of the total catch (Figure 2). 

In Kela the situation was different. The most important species groups were Rhamphochromis spp. and chambo 

followed by the ‘other’ group and the Copadichromis spp. group. These species groups contribution also varied 

each month (Figure 3). Rhamphochromis spp. contributed 37% of the total catch followed by chambo which 

contributed 34%. The ‘other’ group contributed 21%, Copadichromis spp. 5% and usipa only 3% to the total catch 

(Figure 4). 

Length frequency analyses were conducted for some of the most important species groups i.e. Oreochromis spp. and 

Rhamphochromis spp. For the Copadichromis spp. group, one of the species that contributed significantly from this 

group C. virginalis was used in the analysis. Selectivity analyses for the different species groups were conducted by 

combining all the individuals caught in the different mesh sizes (0.5 to 2 inch) in the fishery. This is due to the fact 

that analyses with the separate mesh sizes showed no significant difference in the sizes of individuals caught. 

100% 

80% 

60% 

40% 

20% 

0% 

Mar 

Apr 

May 

Jun 

Jul 

Figure 1. Monthly trends in catch composition (%) of Msaka kauni fishery. 

Aug 

Sep 

Oct 

Nov 

Dec 

Jan 

Feb 

Other 

Rhamphochromis spp. 

Oreochromis spp. 

Usipa 

Copadichromis spp.

0.5% 

24.5% 

129 


Figure 2. Cumulative species composition by weight (%) of Msaka kauni fishery (2000 Mar–2001 Feb). 

100% 

80% 

60% 

40% 

20% 

0% 

Mar 

Figure 3: monthly trends in catch composition (%) of Kela kauni fishery. 

21% 

14% 

10% 

Apr 

May 

Jun 

Jul 

Aug 

Sep 

34% 5% 

3% 

Oct 

51% 

Nov 

37% 

Rhamphochromis spp 

Copadichromis spp 

Rhamphochromis spp 

Copadichromis spp 

Usipa 

Oreochromis spp 

Figure 4: cumulative species composition by weight (%) of Kela kauni fishery (2000 Mar–2001 Feb). 

Dec 

For the Oreochromis spp. group, the length frequency showed that the majority of individuals caught were bigger 

than 150 mm TL. Few juvenile chambo were caught in this fishery, 3 % of individuals were 150 mm or less in total 

length. This lack of juvenile chambo in the kauni fishery is not the result of the gear selecting for the adults but 

rather because fishing occurs in areas where the juveniles are not found being an open water fishery (juvenile 

chambo are found in shallow water closer to shore). The length at 50% maturity for chambo species is 203 mm total 

Jan 

Feb 

Usipa 





Rhamphochromis spp. 


Usipa 

Copadichromis spp.

130 


length (Palsson et al. 1999). In this fishery 50% selectivity was at 210 mm showing that mature adults are mostly 

caught. 

The length frequency for the Rhamphochromis spp. group showed that 97% of the individuals caught were 110 mm 

or bigger in total length. This group is very diverse with a length at maturity range of 109–232 mm standard length 

(Turner et al. 2000). This range reflects the fact that there are different species with different lengths at maturity 

caught in this fishery. Many of the individuals caught belonged to the species R. longiceps and R. ferox which are 

on the lower side of the maturity range having lengths at maturity of 109-130 mm SL and 138 –143 mm SL 

respectively. In this fishery 50% selectivity was at 183 mm TL. 

The length frequency for C. virginalis, a major species in the Copadichromis spp. group showed that 89 % of the 

individuals caught were 110 mm or bigger in total length. The length at maturity for this species is 106 mm TL 

(Marechal, 1991). Mean (50%) selectivity was at 128 mm indicating that mature adults are mainly caught in this 

fishery. 

Frequency 

500 

450 

400 

350 

300 

250 

200 

150 

100 

50 

0 

Figure 5. Length frequency for Oreochromis spp. sampled from the kauni fishery (n=3015). 

Frequency 

10 

1400 

1200 

1000 

800 

600 

400 

200 

0 

40 

10 

70 

50 

100 

90 

130 

130 

170 

Figure 6. Length frequency for Rhamphochromis spp. sampled from the kauni fishery (n=6943). 

160 

210 

190 

Length at maturity range 

109 mm-232 mm 

Mean selectivity 183 mm 

250 

220 

290 

TL (mm) 

250 

TL (mm) 

Length at 50% 

maturity 203mm 


330 

280 

370 

310 

410 

340 

450 

370 

490 

400 

430

Frequency 

140 

120 

100 

80 

60 

40 

20 

0 

5 

20 

Length at maturity 

106 mm 

35 

50 

131 


Figure 7. Length frequency for C. virginalis sampled from the kauni fishery (n=661). 

65 

80 

95 

TL (mm) 

Discussion 

In Kela and Msaka it was observed that the Rhamphochromis spp. group was very important, contributing 37% and 

51%, respectively. Usipa was more important in Msaka, were it contributed 25% to the catch, than at Kela were it 

only contributed 3% of the total catch. The Oreochromis spp. group was more important at Kela, were it 

contributed 34% of the total catch. At Msaka this group was of minor importance, contributing less than 1% to the 

total catch of the kauni fishery. 

Due to the wide-ranging nature of the Rhamphochromis stocks, they are not considered threatened by local 

overexploitation at the moment and are therefore not in need of any restrictions (Turner et al. 2000). The 

Copadichromis stocks do not form a huge component in the fishery (contributing 5% in Kela and 10% in Msaka) 

and the sizes of individuals harvested are of those that have reached adult size. For this reason, the Copadichromis 

stocks do not seem to be threatened by the kauni fishery. However, it should be kept in mind that other fisheries 

also harvest this species group and these should be closely monitored. For the usipa stocks, no management 

recommendations can be given because the stock experiences large fluctuations that appear to be caused by 

environmental factors (Allison et al. 1995). However, due to the declining nature of the chambo fishery, 

Oreochromis spp. stocks need management intervention if the stocks are expected to recover. 

From an area point of view, none of the species groups dominating the kauni fishery at Msaka seem to be 

threatened, therefore restrictions at Msaka are not necessary. However for the chambo stocks at Kela, management 

is necessary. 

The ban of ‘kauni for chambo’ in the Fisheries Conservation and Management Act (2000) has not been effective 

because the fishermen are still catching chambo. It appears unrealistic to inform the kauni fishermen to fish but not 

catch chambo and if chambo is caught in their gear, they are unlikely to throw it back into the water. However it is 

not known whether they specifically target chambo in their fishing procedure or not. A more realistic management 

procedure would be to ban this fishery in a given area i.e. Area A as recommended by Bulirani et al. (1999) and 

Weyl (1999). Due to the use of lights in the fishery at night, it would be easy to monitor whether or not the 

fishermen are following the regulations. 

From the length frequency analyses of the Oreochromis spp. group it was observed that the adult chambo are mainly 

being harvested in this fishery. The harvesting of immature chambo is not a concern in this fishery. However, since 

adult chambo are mainly caught, the stocks are still being depleted. Therefore, there is a need to protect the adults 

until a recovery of the chambo stocks can be seen. 

With the aim of protecting the chambo stocks, the option of closing the kauni fishery and its consequences should be 

considered. With an annual estimated catch of 1500 tons (Catch statistics, 2000), the kauni fishery in Area A 

contributes approximately 40% to the total landings in the area. In addition, the closure of this fishery would result 

in a decreased yield of other major species groups such as Rhamphochromis spp. and Copadichromis spp. that are 

presently exploited in this fishery. 

110 

125 

140 


155 

170

132 


It has been reported that 40 to 60 boats land at Kela beach each day (Ganter et al.2001). Therefore over 400 people 

are directly involved in catching fish only at this beach. In Area A, there are 119 chilimira gears which are all likely 

involved in kauni fishing and this means that in this fishery over 1 000 people in the area are directly involved in 

catching fish (Weyl et al. 2000). Closing down this fishery would lead to a high number of people losing 

employment. Options of limiting effort in this fishery to levels where enough adults are protected while some 

fishing continues, need to be explored. 


Thanks to the technical assistants and other staff of the Fisheries research unit (FRU) who assisted in the collection 

of data. Thanks to Dr. Olaf Weyl who assisted in the data analysis. Special thanks to the National Aquatic 

Resource Management Programme (NARMAP) whom funded this study. 

References 

Allison, E.H., Patterson, G., Irvine, K., Thompson, A.B. and Menz, A. 1995. The pelagic ecosystem. In: The fishery potential 

and productivity of the pelagic zone of Lake Malawi/ Niassa. Menz, A. (ed). Chatham, UK: Natural Resources Institute 

Banda, M. 1996. (unpublished) A preliminary report on the kauni fishery. Unpublished report. 

Bulirani, A.E., Banda, M.C., Palsson, O.K., Weyl, O.L.F., Kanyerere, G.Z., Manase, M.M. & Sipawe, R.D. 1999. Fish stocks 

and fisheries of Malawian waters: Resource report. Government of Malawi, Fisheries Department, Fisheries Research 

Unit. 54pp. 

Catch statistics, 2000. Unpublished, Fisheries Research Unit, Monkey Bay. 

Fisheries Conservation and Management Regulations. 2000. The Malawi Gazette Supplement, 14 th July, 2000, containing 

Regulations, Rules, etc. Government Notice No. 32. Fisheries Conservation and Management Act, 1997 (No. 25 of 

1997). 

Ganter, E. 2001. Socio economic survey No. 2. Kela village – summary of preliminary results. NARMAP &-Dept of Fisheries, 

Malawi. GTZ publication, March 2001. 

Marechal, C. 1991. Copadichromis. In: Check-list of the freshwater fishes of Africa (CLOFFA). J. Daget, J.P. Gosse, G.G. 

Teugels and D.F.E. Thys van den Audenaerde (Eds). ISNB, Brussels. Vol 4, p51-58. 

Palsson, O.K., Bulirani, A. and Banda, M. 1999. A review of the biology, fisheries and population dynamics of chambo 

(Oreochromis spp., CICHLIDAE) in Lakes Malawi and Malombe. Fisheries Bulletin No. 38. 

Turner, G.F., Carvalho, G.R., Robinson, R.L. and Shaw P.W. 2000. Biodiversity Conservation and sustainable utilisation of 

pelagic cichlid fishes of Lake Malawi/Niassa/Nyasa. Ncheni Project Final report. University of Southampton & Hull, 

U.K. 

Weyl, O.L.F. 1999. Artisanal fishery catch-assessment for the south-east arm of lake Malawi; 1994-1998. NARMAP Technical 

Report No.2. 41pp. 

Weyl, O.L.F., Banda, M., Sodzapanja, G., Mwenekibombwe, L.H., namoto, W. and Mponda, O.C. 2000. Annual frame Survey, 

September 1999. Government of Malawi, Fisheries Bulletin No. 42. 

Weyl, O.L.F. and Banda, M.C. 2001. Notes on the chilimira net catch composition from catch and effort in Mangochi District. 

NARMAP Technical Report No. 5. 8 pp.

133 


Gear and species selectivity of the gill net fishery in Lake Malawi. 

Richard Dawson Sipawe 

Fisheries Research Unit, P.O. Box 27, Monkey Bay, Malawi. 

Abstract 

Size selectivity of some important fish species caught in gill nets was estimated indirectly. Species composition of the catches from 

gill nets with mesh sizes between 1 and 2 inches showed that the catch was dominated by C. virginalis (74%) and that chambo 

contributed less than 0.2% to the catch of these small meshed gill nets. Chambo was not caught in these meshes because the size 

range of chambo that could be susceptible to this gear inhabits shallow waters where small meshed nets are not set. The small 

meshed gill nets were considered ideal for catching offshore cichlids. Gill nets with mesh sizes between 2 and 3 inches selected 

chambo at sizes corresponding with the size range at which this species migrates to deep water. Since this migration occurs before 

maturity the use of gill nets with mesh size 2-3 inches is not recommended. Gill nets with mesh sizes between 3 and 4 inches were 

considered ideal since they select for mature chambo. However, the study showed that gill nets with the current mesh sizes in use 

do not catch mature catfish. For management purposes, development and adoption of distinct gill net fisheries for specific target 

species is suggested. 

Introduction 

The gill net is one of the most commonly used gears in Lake Malawi. Its use dates back to the early 1940s. Gill nets 

contribute between 15 and 20% to the total annual fish landings from Lake Malawi (Bulirani et al. 1999) and in 

terms of employment, this fishery engages about 13 600 fishers, representing 30-35% of people directly involved in 

the fishing industry both as gear owners as well as crew members (Weyl et al. 2000). 

The number of gill nets being operated on the lake has increased tremendously during the late 1990’s (Weyl et al. 

2000). In 1995 for example, there were over 17 000 gill nets, the number grew to 24 000 in 1997 and in 1999 there 

were approximately 30 000 gill nets (Weyl et al. 2000). In addition, recent frame survey showed that the majority 

of these nets were under-meshed. For instance, 90% of gill nets recorded for Mangochi District were less than 3.0 

inches (Weyl et al. 2000) which was below the 3.75 inch legal mesh size for gill nets operated in the southeast and 

southwest arms of Lake Malawi. 

The increase in the number of gill nets in the late 1990s coupled with the use of under-meshed gill nets in Lake 

Malawi, initiated the execution of this project in-order to determine species composition and size selectivity of the 

catches in gill nets of mesh sizes ranging from 1inch to 3.75 inches. This was done in order to assess potential 

impacts of small meshed nets on the fishery. 

Methods 

Sampling area 

Catches were sampled from local fishermen’s gill nets in the southeast arm of Lake Malawi. In particular, gill net 

samples were acquired from Makawa (Kela beach), Kadango, Nkope, Ng’ombe and Namiasi landing sites in the 

southeast arm of Lake Malawi. Since gill net fishers are highly migratory, the sampling programme was designed in 

such a way as to target beaches where most gill-netters were landing at that time. 

Data collection and analysis 

Fish catches from the gill net fishery of the Southeast arm were sampled from 1inch to 3.75 inch mesh size gill nets 

on a monthly basis from October 1998 to December 2000. The fish were identified to species level, each species 

component in the catch was weighed and then total length (TL) measurements were recorded to the nearest 

millimetre. 

Retention in gill nets occurs when a fish penetrates a mesh beyond its gill covers but does not completely penetrate 

through, this implies that a fish is caught if its head girth is smaller but its maximum girth is larger than the mesh 

perimeter (Hamley, 1975). Since there is a relationship between girth and total length (Konda 1962, Burd 1963 and 

Kipling 1963), total length (TL) is used exclusively in this study. 

The selectivity curves were fitted using a re-parameterised gamma selectivity function as outlined in Punt et al. 

(1966). A typical gill net selectivity curve is bell-shaped falling to zero on both sides of a maximum (Hamley, 

1975). The gamma selection function was chosen because it easily describes all the members of the exponential 

family of distributions more precisely than the other functions (Manase, 2000).

Results and discussion 

134 


1-2 inch gill nets 

Figure 1 shows the percent catch composition of the 7 main fish species that contributed over 90 % of the catch in 

gill nets with mesh size between 1 to 2 inches. Table 1 shows the catch composition and total number of fish species 

caught in gill nets of various mesh sizes. 

In gill nets with mesh sizes between 1 and 2 inches, 120 different fish species were recorded. The majority of the 

species caught in such nets were those that are naturally small in size. The only exception was that 6% and 0.2% of 

Bagrus and chambo respectively (both comprising of juveniles only) were caught in these nets. 

Percent catch contribution of main species in 1-2" nets 

Otopharynx 

2% 

Synodontis 

3% 

Letrinops 

4% 

Aulono 

2% 

Figure 1. Catch composition in 1-2inch gill nets. 


9% 

Bagrus 

6% 

Copadichromis 

74% 

Table 2 gives the mean lengths at capture, total length ranges as well as variances for fish caught in various mesh 

sizes and analysed using gamma distribution. The mean length at capture of chambo in 1.5inch nets was 11cm 

(figure 2) while the mean length at maturity for chambo is 20.3 cm (Bulirani et al. 1999). For the 1.5 inch gill nets 

chambo varied between 10 cm and 13 cm in total length. 

Table 3 gives the mean length-at-capture (lc) divided by length-at-maturity (lmat) (Lc/Lmat) values for species 

analysed with gamma function. When mean length-at-capture (Lc) was expressed as a proportion of the length-atmaturity 

(Lmat), it was found out that all chambo caught in gill nets of 1.5 inch mesh size were immature 

(Lc/Lmat≥1). However, chambo contributed only 0.2% of the total catch in gill nets of this particular mesh sizes. 

This is because small immature chambo inhabits very shallow waters (FAO, 1993) while these nets are set in deep 

water where they are set to target offshore pelagic fish species especially as utaka. 

The length at first capture for C.virginalis in 1 and 1.75 inch gill nets was 13.3 and 12 cm respectively (Figure 3). 

These sizes are well above 10.6 cm, the length at maturity for C. virginalis (Marechal 1991). Though 74 % of the 

catch consisted of C.virginalis, this stock seems not to be threatened because only mature individuals are being 

harvested. Since approximately 90% of the catch in gill nets with mesh sizes ranging between 1-2 inches comprise 

small fish species that mature before capture by this gear, such nets therefore may be used as harvesting strategy for 

utaka.

Figure 2. Chambo 1.5 inch mesh selection curve. 

% Proportion 

Oreochromis spp. 1.5" selectivity 

100 

80 

60 

40 

20 

% Proportion 

% Proportion 

0 

9 10 11 12 13 14 

Length (cm) 

Figure 3. C.virginalis selection curves for 1.0 inch and 1.75 inch gill nets. 

135 


Observed Predicted 

Copadichromis virginalis 1.0" selectivity 

100 

80 

60 

40 

20 

0 

100 

80 

11 12 13 14 15 16 

Length (cm) 


Copadichromis virginalis 1.75" 

selectivity 

60 

40 

20 

0 

7 9 11 13 15 

Length (cm) 

Observed Predicted

136 


Table 1. Percent catch composition in gill nets of mesh size 1-2 inch (25-50mm), 2-3 inch (51-75 mm) 

and 3-4 inch (76-95mm) sampled between October 1998 and December 1999 in the southeast arm of 

Lake Malawi. (n= 422 kg for 1-2 inch nets; n = 23kg for 2-3 inch nets and n = 1 107 kg for 3- 4 inch nets). 

* Species for which Gamma analysis was done. 

Species 1-2 in. 2-3 in. 3-4 Species 1-2 in. 2-3 in. 3-4 in. Species 1-2 in. 2-3 in. 3-4 in. Species 1-2 

in. 

in. 

Alticorpus Coreematodus Maravichromis Rhamphochromis* 

A. geoffreyi 0.07 C. shiranus 0.01 M. ericotaenia 0.03 0.01 R. spp_ 0.26 

A. macrocleithrum 0.05 C. taeniatus 0.04 M. anaphyrmus 0.43 1.08 0.27 Sciaenochromis 

A. mentale 0.13 Ctenopharynx M. formossus 0.06 S. benthicola 0.09 0.04 

Aulonocara C. intermedius 0.86 0.03 Mormyrus S. gracilis 0.19 0.03 

A. copper 0.14 Dimidiochromis M. deliciosis 0.87 S. spilostichus 0.13 0.18 

A. spp_ 0.01 D. compreceps 0.00 Mylochromis Stigmatochromis 

A. blue orange 0.08 D. dimidiatus 0.01 M. mola 0.00 0.11 S. pholidophorus 0.01 

A. brevirostris 0.01 D. kiwinge 0.03 0.43 0.07 M. spilotichus 0.01 S. woodi 0.11 0.00 

A. guentheri 0.15 0.36 D.strigatus 0.05 Nimbochromis Synodontis 

A. macrochir* 0.81 0.09 Diplotaxodon N. livingstonii 0.50 S. njassae* 3.27 0.89 

A. rostratum 0.30 0.95 D. argenteus 0.23 0.02 N. polystigma 0.07 Taeniaochromis 

A.yellow 0.06 D. greenwoodi 0.07 N. venustus 0.06 0.35 T. spp_ 0.02 

Bargrus D.intermedius 0.01 Nyassachromis T. holotaenia 0.16 

B. meridionalis* 5.35 3.12 9.83 D.limnothrissa 0.08 N. leuciscus* 15.16 Taeniolethrinops 

Barbus Fossochromis Opsaridium T.furcicauda 0.59 1.25 0.02 

B. arcislongae 0.94 F. rostratus 0.13 0.02 O. microcephalus* 0.04 T. laticeps 0.01 

B. eurystomus* 1.80 Haplochromis O. microlepis 0.10 T.praeorbitalis* 0.03 

B. Johnstoni* 3.50 H. spilopterus 0.01 Oreochromis Tramitichromis 

B. litamba 0.32 0.15 Hemitaeniochromis O. shiranus 0.03 0.35 0.02 T.lituris* 0.43 

Bathyclarias H. discorhynchus 0.42 0.62 0.00 O. spp*_ 0.14 22.83 68.20 T. spp_ 0.06 

Bathyclarias spp* 7.40 H. insignis 0.01 0.12 Otopharynx Trematocranus 

Brycinus H. urotaenia 0.28 0.52 0.02 O. argyrosoma* 1.40 T. labifer 0.65 

B. imberi 0.04 Hemitilapia O. auromarginatus 0.06 0.48 T. labifer 0.65 

Buccuchromis H. oxyrhynchus 0.00 O. cf_ productus 0.06 T. labifer 0.65 

B. atritaeniatus 0.35 Labeo O. speciosus 0.25 T. placodon* 5.18 0.28 

B. lepturus* 0.55 L. cylindricus 0.01 3.12 O. spp_ 0.00 Tyrannochromis 

B. nototaenia 0.01 0.62 Lethrinops Placidochromis T. macrostoma* 0.08 

B. rhoadesi 0.06 0.03 L. alba 0.12 0.00 P. acuticeps 0.02 

B. spp. 0.49 L. alta* 0.18 P. long 0.00 

Champsochromis 0.03 L. christyi 0.50 0.36 P. longimanus 0.04 

C. spilorhynchus 0.03 L. gossei* 0.02 P. macrognathus 0.01 

Chilotilapia 0.05 0.15 L.lethrinus 0.14 2.31 0.02 P. subocularis 1.09 0.78 0.09 

C. rhoadesii 0.05 0.15 L.long 0.08 Platygnathochromis 

Clarias L.longimanus 0.64 P.kirkii 0.04 0.26 

C. gariepinus 0.51 16.45 1.23 L. longipinnis 0.16 0.14 Protomelas labridens 0.07 

Copadichromis L.macracanthus 0.01 P.similes 0.07 0.69 0.03 

C. chrysonotus 0.02 0.09 0.00 L. macrochir 0.22 0.01 P. spilopterus 0.01 

C. eucinostomus* 0.22 L. matumbai 0.05 P. triaenodon 0.00 

C. inornatus 0.00 L.mylodon 0.00 Pseudotropheus 

C. pleurostigma 0.03 L.parvidens* 0.42 0.00 P. elegans 0.11 0.13 

C. prostoma 0.00 L. pink head* 0.41 P. livingstonii 0.07 

C. quadrimaculatus* 2.03 L. polli 0.18 0.03 P.tropheops 0.04 

C. trimaculatus 0.00 L. stridei 0.28 P. williamsonii 0.19 

C. virginalis* 67.90 14.73 Lichnochromis 

L. acuticeps 0.11 0.26 

2-3 

in. 

3-4 

in.

137 


Table 2. Mean retention lengths (maximum selectivity), total length (TL) ranges and variances for fish 

caught in various mesh sizes and analysed using gamma distribution. 

Species/Species group (n) Mesh size Max. Sel. Range TL (cm) Variance 

Oreochromis spp. 48 1.5 11.00 8.5-13.5 0.79 

95 2.0 13.94 11.0-17.0 0.96 

398 3.0 24.05 17.0-35.0 3.17 

2101 3.5 24.24 17.0-35.0 2.51 

135 3.75 25.30 18.0-35.0 2.28 

Copadichromis virginalis 36 1.0 13.30 11.0-16.0 0.98 

393 1.25 12.07 10.0-16.0 0.73 

1814 1.5 12.48 10.2-15.0 0.96 

209 1.75 12.08 10.5-16.0 0.96 

113 2.5 12.38 10.0-15.0 0.85 

Bagrus meridionalis 71 1.5 22.74 18.0-32.0 2.54 

136 3.0 34.77 36.0-50.0 3.53 

114 3.5 37.54 28.0-50.0 3.10 

83 3.75 37.66 30.0-50.0 3.30 

Otopharynx argyrosoma 69 1.25 10.09 8.0-12.0 0.66 

135 1.5 10.63 7.0-17.0 1.74 

Rhamphochromis spp. 131 3.5 38.28 36.0-47.0 1.80 

102 3.75 38.76 37.0-47.0 1.64 

Synodontis njassae 239 1.5 13.39 11.0-20.0 1.45 

35 2.0 13.58 11.0-19.0 1.19 

C. quadrimaculatus 123 2.0 16.64 13.0-21.0 1.33 

47 3.0 16.95 13.0-21.0 1.30 

Bathyclarias spp. 69 3.5 49.98 41.0-62.0 3.02 

79 3.75 49.45 40.0-62.0 3.19 

Lethrinops pinkhead 43 1.25 9.62 7.50-12.0 0.78 

48 1.5 10.10 8.00-14.0 0.93 

Trematocranus placodon 81 3.0 19.57 16.0-24.0 1.18 

31 3.5 20.51 18.0-24.0 0.95 

Barbus johnstoni 62 3.5 34.00 10.0-38.0 2.23 

Barbus eurystomus 39 3.75 38.00 33.0-44.0 2.72 

Opsaridium microcephalus 36 3.5 47.00 44.0-50.0 0.91 

Tyrannochromis macrostoma 52 3.5 20.00 18.0-23.0 1.11 

Nyasachromis leuciscus 51 2.5 17.00 14.0-20.0 0.82 

Tramitichromis lituris 34 1.5 11.00 8.50-14.5 1.28 

Taeniolethrinops praeorbitalis 76 3.0 23.00 21.0-26.0 1.07 

Lethrinops parvidens 34 1.5 12.00 10.0-14.0 1.50 

Lethrinops longimanus 81 2.5 17.00 14.0-22.0 1.31 

Lehrinops gossae 43 2.5 16.50 14.0-20.0 0.93 

Lethrinops alta 33 1.5 10.80 9.0-14.0 0.80 

Copadichromis eucinostomus 38 2.0 11.30 9.0-13.0 0.74 

Buccochromis lepturus 52 3.5 26.04 23.0-30.0 1.35 

Aulonocar macrochir 33 2.0 11.70 10.0-14.0 0.70

138 


2-3 inch mesh gill nets 

Figure 2 shows the percentage catch composition of the 7 main species that contributed almost 90% of the total 

catch in gill nets with mesh sizes between 2 to 3 inches. 

In gill nets with mesh sizes ranging between 2-3 inches, 99 fish species were recorded. Chambo dominated the 

catch; it comprised 25% of the catch by weight. The importance of small fish species in the catch started to decrease 

because these meshes selected for slightly larger fish species. 

% catch ontribution of main species in 2-3" 

gill nets 

Trematocranus 

6% 

Oreochromis 

25% 

Figure 4. Catch composition of main species in 2-3 inch gill nets 

The mean length at capture for chambo in these nets was 14 cm and the largest chambo caught was around 17 cm 

(figure 5). Therefore, chambo caught in these gill nets, were immature. The vulnerability of the chambo caught in 

gill nets of 2-3 inch mesh sizes increases due to its migratory behaviour, since at this size the chambo moves from 

shallow to the deep water where it meets the nets. 

Oreochromis spp. 2.0" selectivity 

100 

Figure 5. Chambo 2.0 inch gill net selectivity curve. 

% Proportion 

50 

0 


14% 

Bagrus 

3% 

Nyassachromis 

16% 

11 12 13 14 15 16 17 18 19 20 21 

Length (cm) 


Clarias 

17% 

Labeo 

3% 

Copadichromis 

16%

139 


Table 3. Mean length-at-capture (lc) expressed as a proportion of the length-at maturity (Lmat) for the main 

target species analysed with gamma function. Adapted from Weyl et al. 2000. 

Species Lmat Mesh size in inches for sampled species 

1.0 1.25 1.5 1.75 2 2.5 3 3.5 3.75 

Aulonocar macrochir 13.5 cm* 0.9 

Buccochromis lepturus 31.5 cm 2 0.83 

Copadichromis eucinostomus 10 cm* 1.1 

Copadichromis quadrimaculatus 17.7 cm 2 0.9 0.96 

Copadichromis virginalis 10.6 cm 2 1.25 1.14 1.18 1.14 1.2 

Lethrinops alta 11.25 cm* 1.23 

Lethrinops gossie 14.7 cm* 1.12 

Lethrinops longimanus 15.9 cm* 1.14 

Lethrinops parvidens 15.3 cm* 0.71 

Lethrinops pinkhead 9.0 cm* 1.07 1.12 

Nyasachromis leuciscus 12.0 cm 2 1.9 

Oreochromis spp. 20.3 cm 1 0.5 0.7 1.18 1.19 1.25 

Otopharynx argyrosoma 11.4 cm 2 0.89 0.93 

Rhamphochromis spp. 28.71 cm* 1.33 1.35 

Taeniolethrinops praeorbitalis 22.5 cm* 1.05 

Tramitichromis lituris 13.5 cm* 0.87 

Trematocranus placodon 18.8 cm 2 1.04 1.09 

Tyrannochromis macrostoma 18.1 cm* 1.11 

Barbus eurystomus 30.0 cm 2 1.18 1.19 

Barbus johnstoni 30.4 cm 2 1.15 1.15 

Opsaridium microcephalus 35.6 cm* 0.94 

Lc/Lmat 

Bagrus meridionalis 52.2 cm 1 0.43 0.66 0.72 0.72 

Bathyclarias spp. 60.0 cm 1 0.83 0.82 

Synodontis njassae 10.0 cm 1 1.34 1.4 

Lmat was derived from * estimated at 75% of the maximum length for species given in Turner 1996, 

1- Bulirani et al. 1999, 2- Weyl 2000. 

3-4 inch mesh gill nets 

Figure 6 shows the percent catch composition of the 5 main species that contributed 95% of the total catch by 

weight in gill nets with mesh sizes between 3 to 4 inches. In all, 70 fish species were recorded from this category of 

nets (Table 3). Chambo comprised 70% of the total catch in these gill nets and for the 3.5 inch gill nets, the mean 

length at capture for chambo was 24.5 cm, 2.2 cm longer than the length at maturity. Therefore, gill nets with mesh 

sizes between 3-4 inches catch mature chambo. 

The mean length at capture for B. meridionalis in 3.5 inch gill nets was 37.5 cm while the length at maturity is 52.2 

cm (Bulirani et al. 1999). This showed that the length at first capture was less than the length at maturity. 

The mean length at capture for Bathyclarias spp. in 3.5 inches was 49.5 cm while the length at maturity is 60.0 cm. 

This means that individuals caught in this mesh size were immature. The range of capture for both Bagrus 

meridionalis and Bathyclarias spp. was very wide and individuals were caught outside the selection curve in both 

cases reflecting the fact that two methods of capture were involved (Hamley, 1975). These were entanglement and 

wedging. Catfish are vulnerable to entanglement because they have dorsal and pectoral spines.

% catch Contribution of main species in 3-4" nets 


4% 

Oreochromis 

70% 

Figure 6. Catch composition of main species in 3-4 inch gill net. 

Figure 7. Selection curve for chambo in 3.5 inch gill nets 

Figure 8. Bagrus meridionalis 3.5 inch gill net selection curve. 

% Proportion 

% Proportion 

140 


Bagrus 

10% 

O r eochr om i s spp. 3.5" selectivity 

100 

80 

60 

40 

20 

Barbus 

6% 

0 

14 18 22 26 30 34 38 

Length (cm) 

100 

80 

60 

40 

20 


Bathyclarius. 

8% 

Buccochromis. 

2% 

Bagrus meridionalis 3.5" selectivity 

0 

23 27 31 35 39 43 47 51 55 

Length (cm) 

Observed Predicted

% Proportion 

100 

Figure 9. Bathyclarias spp. 3.5 inch selection curve. 

80 

60 

40 

20 

Bathyclarias spp. 3.5" selectivity 

0 

34 38 42 46 50 54 58 62 66 

Length (cm) 


141 


Conclusion. 

Gill nets with mesh sizes between 1-2 inches have the potential to catch immature chambo. However, they do not as 

small chambo inhabit very shallow waters of less than 5 metres (FAO, 1993) while the nets are set in water deeper 

than 40 metres where they are set to target small offshore cichlids especially utaka. 

Gill nets with mesh sizes between 2-3 inches catch immature chambo normally of less than 17 cm TL that migrate 

from the shallow into the deep waters. Gill nets with mesh sizes between 3-3.75 inches catch mature chambo. Gill 

nets with mesh size of 3.75 inches and less, that are currently in use, catch immature catfish. Any mature catfish 

caught in these nets do so by being entangled. 

Therefore, we recommend the development of distinct gill net fisheries for specific target species. This may be 

achieved by identifying the most important fishing grounds or areas in the lake and then by adopting and promoting 

the best harvesting strategy for each target species. We also discourage the use of 2-3 inch gill nets as these catch 

immature chambo in deep waters. 


Funding for this study mostly came from the European Union (EU) and the Department of Fisheries’ Fisheries 

Research Fund (FRF). Some additional data came from the gear selectivity surveys sponsored by the GTZ supported 

National Resource Management Programme (Narmap). Special thanks are due to Dr. Olaf Weyl, the Narmap 

Research Advisor and Dr. M. Banda the officer-in-charge for Fisheries Research station. 

References. 

Bulirani, A.E., Banda. M.C., Palsson, O.K., Weyl, O.L.F., Kanyerere, G.Z., Manase. M.M., Sipawe, R.D., 1999. Fish and 

Fisheries of Malawian waters: Resource report. Government of Malawi, Fisheries Research Unit. 

Burd, A. C., 1963. On selection by the drifter fleets in East Anglia herring fishery. J. Cons. Int. Explor. Mer 28: 91-120. 

Hamley, J. M., 1975. Review of gillnet selectivity. J. Fish. Res. Board can. 32: 1943-1969. 

FAO, Fisheries management in the South-east arm of Lake Malawi, upper Shire River and Lake Malombe, with particular 

reference to the fisheries on chambo (Oreochromis spp). CIFA Technical Paper. No. 21. Rome, FAO 1993. 113p. 

Konda, M., 1962. The relation between the size of the mesh of salmon gill net and the length of salmons in the catches. (in 

Japanese with English summary) Bull. Hokkaido Reg. Res. Lab. 24: 138-147. 

Kipling, C., 1957. The effects of gillnet selection on the estimation of weight-length relationship. J. Cons. Int. Expor. Mer. 23: 

51-63. 

Manase, M. M., 2000. Traditional Gear Selectivity in Southeast arm of Lake Malawi. GTZ-NARMAP Tech. Rep. No. 4. 

Marechal, C. 1991. Copadichromis. In the check-list of the freshwater fishes of Africa (CLOFFA). I Daget, J. –P. Gosse, G. G. 

Teugels and D> F. E. Thys van den Audenaerde (eds.). ISNB, Brussels, MRAC, Tervuren; and ORSTUM, Paris. Vol. 

4: 51-58. 

Weyl, O.F.L., Banda, M.C., Sodzapanja, G. Mwenekiombwe, L.H., Namoto, W. and Mponda, O.C., 1999. Annual Frame survey. 

Malawi Government. Fisheries Bulletin No. 42.

142 


Fisheries activities in northern Lake Nyasa (Kyela District) 

John A. Mwambungu 

Tanzania Fisheries Research Institute, Kyela Centre, P.O. Box 98, Kyela-Tanzania 

Abstract 

When compared with the other East African Great lakes, the fisheries activities of Lake Nyasa are very low. This might be due to 

poor fishing vessels and gears used by the artisanal fishermen and the ultra-Oligotrophic nature of the lake. The present paper 

looks at the fisheries potential of Lake Nyasa in Kyela district and suggests strategies for development. 

Introduction 

Lake Nyasa lies at the Southern end of the East African rift valley. The lake is long and narrow. It is 560 km long 

and has a surface area of 30 800 km² ranking ninth in order of size among the lakes of the world. On a south, north 

transect, the lake shows a remarkable increase in depth, with the north being deeper and the south shallower. It is 

one of the deepest lakes in the World with an average depth of 275 m ranging from 200 m to 700 m (Ngatunga et al. 

1997). The lake is shared by three riparian states Malawi, Mozambique and Tanzania (Figure 1). 

Malawi 

Kyela 

Itungi port Matema 

Nkanda 

Figure 1. Lake Nyasa (Tanzania side). 

Ludewa 

Lupingu 

Manda 

TANZANIA 

Mbinga 

Mbamba-bay 

MOZAMBIQUE 

An area of 5 569 km², which is about 18% of the lake in the north, is within Tanzania jurisdiction. Three 

administrative regions namely Ruvuma, Iringa and Mbeya share the lake. Kyela district, which is a focus of the 

present paper, falls within Mbeya Region 

In Tanzania the shoreline of the lake is 258 km long, and has three main ecological zones: First is the rocky shore of 

the north east shoreline which is characterised by deep slopes from the Livingstone mountains. The mountains 

which reads high as 3 000 m in the north extend along the lake shore from Matema to Manda, curve in land 

eastwards and reappear after Ndumbi, where they become lower and less steep. In the southern part of the range the 

shore is characterised by reefs, inshore rocks and small uninhabited islands; second is the wide sandy beaches which 

occur in few areas such as Matema, Lupingu, Manda, Njambe, Liuli and Mbamba-bay. Third are the river estuaries 

associated with swampy areas covered with reeds. They are found in two areas: Wissman Bay from Songwe river on 

the Malawi border to Matema were four rivers flow into the lake (Songwe, Kiwira, Mbaka and Lufilyo) and in 

Amelia Bay, around the Ruhuhu delta which covers an area of 400 km² (Ng’ang’a 1993) from Manda to Ndumbi 

Reef. The shoreline in these areas is characterised by sand and mud the water is predominantly muddy. 

The lake is of enormous importance to Tanzania as a means of transport, reservoir of fresh water and a source of 

food (i.e. fish). Characteristics of the shoreline are very important in understanding the distribution of the fishes and

143 


hence the fisheries of Lake Nyasa in the Tanzania waters. Fishing activity is mainly artisanal, using traditional crafts 

and gear. However, the lake is characterised by very rough weather conditions, which make the traditional fishing 

with canoes a very dangerous venture. Moreover, because of the depth, nutrient levels available to phytoplankton are 

generally very low and based on nutrient; the lake is considered Oligotrophic (nutrient poor). An exception to the 

above is the shallow south and around river mouths, areas which are considered to be highly productive. The present 

work tries to evaluate the fish resource potential of Lake Nyasa in Kyela district and suggests strategies for 

development. 

Fish Exploitation 

The capture fisheries of Kyela are divided into two main sectors namely food and Ornamental fisheries. The latter is 

useful to the nation as a source of foreign money, however the food capture is the most important sector constituting 

a major source of income and protein for coastal population. 

Fishing on the Tanzania side is under developed and has received very little attention. The common fishing vessel is 

the dugout canoes (3-4 m length). The canoes are small, unstable and propelled by paddles; one or two fishermen 

normally operate them. The fishery is mainly restricted to inshore waters within 2 km of the shoreline. 

The principal fishing gears are gillnets, seines, hook and lines scoopnets and traps. In Kyela, the anadromous 

piscivorous cyprinids Opsaridium microlepis and O.microcephalus are of great importance. Several major affluent 

rivers (Songwe, Kiwira, Mbaka and Lufilyo) provide suitable habitats for spawning. Fisheries around and in the 

river mouths have long been intensive whereby gillnets and sometimes seines are used to exploit these cyprinids and 

other species during spawning period. This type of fishery makes the anadromous species more vulnerable to over 

fishing leading to destruction of stocks. 

Gillnets of stretched mesh size of 2 inch are used for catching cichlids (haplochromine and tilapine sp.), 2½ inch- 

3½ inch for bigger tilapine sp. and Opsaridium microcephalus, 3½-6 inch used for catching bigger fish species 

(Opsaridum microlepis, clarias sp. and Bathyclarias sp.). 

With the introduction of open water seine nets from Malawi, artisanal fishing units have been concentrated at Itungi, 

Mwaya, and Matema, (Wissman Bay). The open water seine known as ndaturu which is similar to the chilimira nets 

of Malawi, use light to attract Engraulicypris sardella (usipa) and small Rhamphochromis sp(Ngelwa) during 

moonless nights. During the day the gears are used to catch Diplotaxodon sp. (mantura) and Copadichromis sp. 

(vituwi) from deep inshore waters, and Tilapiine cichlids (magege) in the shallow waters of the Bay. 

Handline is commonly operated by most of the fishermen for catching tilapine species while bottom-set longlines 

are used for catching bigger demersal species (Clarias and Bagrus sp.). 

Recently fishermen in Kyela district have adopted drifting pelagic longlines after successful demonstrations to 

fishing communities of Itungi, Mwaya and Matema by TAFIRI. The gear target for large predatory fish such as 

large Rhamphochromis sp. (lumbulu/hangu) and Bathyclarias sp. (Ngoshora). The relationship between fish 

production and different components of artisanal fishery in Kyela district between 1990 and 1996 is shown in Figure 

2. The yield has fluctuated between 1 664.7 tons and 2 142.7 tons. Production is observed to drop from 1 428.5 tons 

in 1991 to 264.2 tons in 1992 and increased abruptly in 1992 to 1 423.1 tons. Canoes decreased from 295 in 1990 to 

205 in 1992 and increased from 312 in 1993 to 374 canoes in 1996 while fishermen decrease from 682 in 1990 to 

412 in 1992 and initially increased from 542 in 1993 to 709 in 1996. 

Catch composition of fish landed at beaches of Kyela district in 1996 is summarised in Figure 3. Catch was 

dominated by Engraulicypris sadella registering 28% followed by Rhamphochromis sp. (18%), Tilapines (17%), 

Others (15%) and Haplochromines (11). Bagrus meridionalis and Clarias sp. each contributed 4% while 

Opsaridium sp. and Labeo sp. contributed 2% and 1% respectively. Barbus species contribution was less than I% to 

the total catch.

Catch (tonnes)& Fishing effort 

2500 

2000 

1500 

1000 

500 

0 

1990 1991 1992 1993 1994 1995 1996 

144 


Figure 2. Fishing effort (fishers and canoes) and annual catch for Kyela district (Fisheries data 1990- 

1996). 


Barbus15% 

0% 

Usipa 

29% 

Opsaridium 

2% 

Figure 3. Catch composition by weight of fish landed at Kyela beaches (Fisheries data). 

Years 

Tilapiines 

18% 

Fish processing 

The production of fish from Lake Nyasa does not meet the demand and thus most of the fish from the lake is 

consumed fresh. Processing activities are in most cases done for immediate consumption. Frying is the most 

common method. A big fish like Opsaridium microlepis is first cut into small pieces. The pieces are fried in oil and 

sold on the rural market,"pombe" (local brew beer) shop and bars, occasionally, when there is a bumper catch or 

when fish cannot fetch immediate market fish is hot smoked in the evenings. 

Fish Marketing 

As mentioned earlier most of the fish is consumed fresh. At the landing sites the fresh fish is displayed in canoes or 

on sand. The fish is sold to different buyers who include consumers and fish mongers. A fishermen who owns the 

gears may sell his catch to his relatives or traders and other people. Some fishermen obliged to sell the catch to 

boat/gear owner who will then retail the catch. The price is normally reached by negotiations and experience, 

sometimes butter trade operates. 

The fish traders comprise adult men and women. The fresh fish buyers carry their fish to either rural markets or 

Kyela Central market on the head or on bicycles. The fish are put in bamboo baskets and gunny bags. 

At the rural markets fish are displayed on raised racks with or without roof. At Kyela Central market there is a wellestablished 

fish-receiving block erected since 1987. Inside the fresh fish is displayed on raised concrete tables 

Fishers 

Canoes 

Catch 

Labeo 

1% 

Rhamphochromis 

19% 

Haplochromines 

12% 

Bagrus 

4%

145 


supplied with tap water. There is also a room installed with measuring balance for recording weights of fish received 

every day. 

In rare cases smoked fish (Opsaridium sp.) and sun-dried dagaa (Engraulicypris sardella) from Ludewa (Iringa 

Region) and Mbinga (Ruvuma Region) reach Itungi Port by ship. Most of the processed fish find market in the 

villages and Kyela Central market. 

Future Strategies 

The future strategies for fisheries development in Kyela district should focus on the following areas. 

Boat construction 

The dugout canoes used in the lake are quite inefficient as fishing crafts and are risky because the lake becomes 

rough quite often. The poor development in the fishery of the Tanzania side of the lake is partly due to poor fishing 

crafts. Offshore pelagic fishing require better fishing crafts. Investors should look into the possibility of constructing 

appropriate boats in Kyela area for the northern lake zone (Tanzania side). Hard wood-“Mninga” (Pterocarpus 

angolensis) used for constructing boats is available and the demand for such boats is high which justifies investment 

in boat building. So far there is only one centre (at Liuli in Ruvuma Region) which trains few people how to build 

modern boats. That of Itungi Port should be rehabilitated. 

Exploitation of pelagic resources by use of open water seine nets -chilimira nets (ndaturu) 

Suitable gears for Kyela area are known to be chilimira nets, gillnets and longline. Chirimila nets however are 

gaining popularity especially in capturing Copadichromis sp., Diplotaxodon sp., Oreochromis sp., Rhamphochromis 

sp. and Engraulicypris sardella as mentioned earlier. 

With this gear there has been a marked shift of fishing effort from shallow inshore waters, which serve as nursery 

areas for many cichlid species towards deeper waters. Therefore investment in chilimira nets is viable, as the target 

fish species is plentiful. A proper write up on the investment however should be made before investment is 

undertaken. TAFIRI is proposing to do a study on the open seine fishery within the Tanzanian water with the 

objective of alleviating poverty through increased income and protein consumption. 

Exploitation of pelagic resources by use of drifting longline 

After recommendation of long lining in the pelagic zone by UK/SADC Lake Malawi Fisheries Project, TAFIRI 

carried out a project on the long lining in Kyela area and the results are encouraging (Mlay et al. 1994). Investment 

in longlining is rather low as the prices of hooks and twines are relatively cheap. Ordinary artisanal fishermen can 

afford them. The only expensive part of the project is the purchase of boat and engine. Investors would need to make 

feasibility studies before investing but on the whole it will need small investment. 

Fish marketing 

Most fish consumers prefer eating fresh fish. With the good tarmac road connecting Dar es Salaam and Kyela and 

the major towns in the country fresh fish from Lake Nyasa could be transported from the lake area to the towns if 

the traders invest in insulated containers. These containers are now being made locally in Dar es Salaam. The use of 

iced insulated containers will reduce post harvests loses of the fish and improve the quality of fish that reach the 

consumer. 

Ornamental fish trade 

Lake Nyasa is known for its unique fish species. The lake ranks very high in ornamental fish trade. In Kyela two 

companies are engaged in ornamental fish trade. TAFIRI is considering collaborating with any institution into the 

identification of the ornamental fishes together with the fishing grounds. Investors are invited to take up the 

ornamental fish trade, which could be done, on joint venture basis. The potential is high. For proper management 

suitable areas for ornamental fish should be mapped out and Lake Shore communities should benefit from the trade. 

References 

Mlay, M.K.L., Mwambungu, J.A. and Kihedu, K.J.1994. Longline Fishery Prospects in Lake Nyasa. TAFIRI Annual Research 

Report. Tanzania. 

Nga’ng’a, P.1993.Deltaic sedimentation in a lacustrine environment of Lake Malawi, Africa.J. Afr. Earth Sci. 16,253-264. 

Ngatunga, B.P., Mwambungu, J.A., Kihedu, K.J., Mlay, M.K.L.and Waya, R.1997.Biodiversity Conservation and Fisheries 

Management on Lake Nyasa, Tanzania. Proceedings of the Workshop of the Senga Bay,Malawi, February 1997.

146 


Hard choices for chambo management in Area A of the southeast arm of Lake 

Malawi 

Olaf L.F. Weyl 

National Aquatic Resource Management Programme (NARMAP), P.O.Box 27, Monkey Bay, Malawi. 

E-mail: narmapbay@malawi.net 

Abstract 

Management of the fishery in Area A of the southeast arm of Lake Malawi (SEA) is complicated by multi-gear utilisation and by the 

multi-species nature of the catch. One option for management is to prioritise key species for management. The suitability of 

chambo (Oreochromis) species as a key management species group for Area A is discussed. The stocks were modelled using 

multi-gear yield- and spawner biomass-per-recruit models to investigate four management target reference points (TRP’s). These 

TRP’s were Fmax, F0.1 and FSB40 and FSB50 . Fmax corresponds with the asymptote of the yield-per-recruit curve and approximates 

Maximum Sustainable Yield, F0.1 is the rate of fishing mortality at which the slope of the yield-per-recruit curve falls to 10% of its 

value at the origin, FSB40 and FSB50 are the points on the spawner-biomass-per-recruit curve corresponding to a reduction of 40% and 

50% of the pristine spawner-biomass. It was estimated that the current fishing mortality rate (F) for Area A chambo approximated 

0.34 yr -1 . This exploitation level was below the Fmax TRP but exceeded the more conservative F0.1 TRP. In addition, current 

exploitation rates exceeded both spawner biomass based TRP’s. To attain the FSB40 TRP a 10% reduction in fishing effort was 

necessary and a 23% reduction in effort was necessary to attain the FSB50 TRP. Different management options taking into account 

the effect in total landings and employment are discussed. 

Introduction 

Through the participatory fisheries management (PFM) framework of the Department of Fisheries (DoF), fishing 

communities have the right to obtain fisheries management recommendations for their specific fishing grounds 

(GOM 2000). The provision of such management recomendations is the responsibility of the Fisheries Research 

Unit (FRU) of the DoF. Since these recommendations are to cover finite areas for PFM, they have to refer to 

localised areas that are small enough to allow for community based management but large enough so that 

management interventions will have an effect on the fish stock. Furthermore, management recommendations should 

be based on the best scientific knowlege available and take into consideration the multi-gear and multi-specie nature 

of the fisheries in Malawi. 

The fishery in Area A of the southeast arm of Lake Malawi (SEA; Figure 1) has both small scale and large scale 

commercial components. The small scale fishery comprises some 420 gill nets, 119 chilimira nets, 27 kambuzi 

seines, 18 chambo seines and at least 12 nkacha nets (Weyl et al. 2000). This small scale fishery currently employs 

in excess of 1900 fishers (Weyl et al. 2000). The commercial fishery currently comprises 2 pair trawl units and one 

ring net unit. However, 4 pair trawl licences and two ring net licences are issued (Banda & Chirwa 2000). 

However, the consolidation of area A under one PFM management unit is considered feasible (Weyl 2001). 

Fisheries management in Area A is complicated by the multi species nature of the catch, and over 100 species 

contribute to the catches in Area A (Chisambo 2001, Manase 2001, Sipawe 2001, Nyasulu 2001). It has long been 

recognised, that when a fishery harvests a number of species it is impossible to manage each species at its optimum 

level (Murawski 1984). In such cases, one possible option is to base management on the least resilient species in a 

multi-species fishery (Weyl 1998). However, this requires prior knowledge of the biological and life history 

information on all target species. Such information is not available for most of the species targeted by Malawi’s 

small-scale fisheries. Until such data is available, it is recommendable to prioritise key species for management in 

specific areas. Such key species should contribute significantly towards the fishery and show declining catch trends 

that indicate that management intervention is necessary and the species should have life history traits such as limited 

distribution ranges, limited migration patterns, localised spawning and localised nursery areas, which would make 

locality specific management feasible.

Figure 1. The south east arm of Lake malawi showing Area A. 

147 


The chambo (Oreochromis ‘Nyasalapia’ spp.) stock in Area A fulfills these criteria. It has been shown that chambo 

contributes more than 30% to the total catch in Area A and therefore contributes significantly to the fishery (Figure 

2). The decline in the chambo stocks (Figure 2) in the SEA is well documented and management for this species is 

considered necessary (FAO 1993, Tweddle et al. 1994, Palsson et al. 1999, Bulirani et al. 1999). Kanyerere (2001) 

showed that the chambo stock of the SEA was concentrated in Area A and over 70% of the SEA chambo catch 

originates from this area (FRU unpublished catch data). While migration of chambo from the SEA into Lake 

Malombe have been considered, there is currently little substantiating evidence (FAO 1993) it is likely that Lowe’s 

(1952) observation that the chambo stock is localised and prone to local overfishing remains valid. Hence the 

management of the chambo population in Area A is feasible. 

Catch (tons) 

4500 

4000 

3500 

3000 

2500 

2000 

1500 

1000 

500 

0 

1984 

1985 

1986 

1987 

1988 

1989 

1990 

Figure 2. Chambo catch in the southeast arm of Lake Malawi. The contribution of Area A to the total 

chambo catch is represented by the shaded portion. 

1991 

1992 

Year 

1993 

1994 

1995 

1996 

1997 

1998 

1999

148 


In an attempt to define optimum fishing mortality from a per-recruit perspective the use of target reference points 

(TRPs) has become common practice in fisheries management (Clarke 1991, Punt 1993, Caddy & Mahon 1995, 

Booth & Buxton 1997, Griffiths 1997, Bulirani et al. 1999). The TRP approach is already advocated by the DoF 

and the 1999 Resource Report makes management recommendations for the fisheries based on TRP’s based on 

catch per unit effort in the small scale fishery (Bulirani et al. 1999). 

The yield-per-recruit (YPR) approach allows for the determination of at least two commonly used TRPs: firstly, the 

fishing mortality which corresponds to the maximum of the yield-per-recruit curve (Fmax) and secondly, the marginal 

yield or F0.1 strategy (Gulland & Boerema 1973, Deriso 1987), which is the rate of fishing mortality at which the 

slope of the yield-per-recruit curve falls to 10% of its value at the origin. However, the yield-per-recruit approach 

assumes that recruitment is constant and independent of spawner biomass. Since, there is little doubt that in most 

fish species the stock-recruitment relationship is density dependent (Shepherd 1982, Sissenwine & Shepherd 1987). 

Since, Oreochromis species are mouthbrooders with extended parental care (Trewavas 1983), a density dependent 

spawner biomass-recruitment relationship is implied. For similar reasons, scientists concerned with the management 

of marine species have tended to base their TRP recommendations on the results of spawner biomass-per-recruit 

(SBR) models (Butterworth et al. 1989, Smale & Punt 1991, Booth & Buxton 1997, Griffiths 1997). The definition 

of a spawner-biomass TRP (FSB(x)) involves setting the fishing mortality to a level at which spawner biomass-perrecruit 

is reduced to x% of its pristine level. Although there is no conventional FSB(x) TRP, spawner biomass-perrecruit 

recommendations lie between 25% and 50% of unexploited levels (Deriso 1987, Sissenwine & Shepherd 

1987, Butterworth et al. 1989, Smale & Punt 1991, Clark 1991, Punt 1993, Booth & Buxton 1997). In the absence 

of information on the spawner biomass-recruitment relationship, the FSB(x) TRPs are currently considered the most 

robust, allowing for the determination of a fishing mortality rate that will provide relatively high yields at lower 

risks (Clarke 1991, Punt 1993). 

Three Oreochromis species O. squamipinnis, O. lidole and O. karongae, comprise the chambo stock, identification 

of juveniles to species level is problematic (FAO 1993). However, the similarity in age-at-maturity and growth rate 

in the three species allows for their treatment of all three as a single management unit. This allows for the 

application of fisheries models such as the multi gear yield per recruit MGYPR (Djama & Pitcher 1997) and multi 

gear spawner-biomass per recruit MGSBR (Weyl 1998) to determine the effect of yield and spawnner biomass to 

varying levels for fishing mortality. 

The aim of this study was to investigate the effect of the management interventions suggested by the Fisheries 

Management and Conservation Act (GOM 2000) on the chambo fishery in Area A. These management interventions 

include: a closed season for seines during November and December, total bans on nkacha and kambuzi seines, a 

total ban on the light attraction chilimira/kauni fishery for chambo and minimum mesh size restrictions for gill nets 

from a per-recruit perspective. In addition, MGYPR and MGSBR approaches were applied to investigate four Target 

Reference Points (TRP’s). These were the YPR based Fmax and F0.1 TPRs and the SBR-based FSB40 and FSB50 TRP’s. 

as well as two spawner biomass dependent TRPs. 

Methods 

Per Recruit Models 

The multi-fishery per-recruit models used in this paper are based on an extension of the traditional per-recruit 

models (Djama& Pitcher 1997, Weyl 1998). Input parameters for the per-recruit simulations were derived from the 

literature available on the chambo stocks (Table 1). Yield- per recruit was calculated by the equation: 

YPR 

⎡ 

max 

= ∑ 

a= 

0 

⎢ ~ 

⎢( 

wa 

N a ( 1− 

exp( −( 

M 

⎢ 

⎣ 

+ ∑ S 

j 

aj 

F j )))) 

M 

∑ 

j 

+ 

S 

j 

aj 

∑ 

F 

S 

j 

aj 

⎤ 

⎥ 

∆a 

F ⎥ 

j ⎥ 

⎦

149 


and spawner-biomass-per-recruit for each species i (SBRi) in a multi-gear fishery was determined by: 

max 

SBR = ∑ 

a= 

0 

where wa is the weight-at-age a, described by: 

Wa= α( la ) β 

where la is the length-at-age determined by the von Bertalanffy growth equation; α and β are parameters describing 

the length-weight relationship; Saj is the selectivity for age class a by fishery j, Fj is the instantaneous rate of fishing 

mortality (yr -1 ) for all fisheries j under consideration, M is the rate of natural mortality, and max is the maximum 

recorded age. N ia 

~ is the relative number-at-age of species i and was described as: 

~ 

N 

ia 

where Sax is the selectivity of gear x for age class a and Fax is the proportional fishing mortality rate (yr -1 ) for that 

fishery. Fax was determined by the proportional contribution of each gear to the total chambo catch in all gears 

under consideration. Catch estimates for the year 2000 in the SEA were used (Table 2). All summations were 

∆ a of 0.10 of a year. 

conducted with a step size ( ) 

Input parameters 

~ 

ψ w N ∆a 

a 

a 

a 

⎧ 1 

if a = 0 

⎪ 

= ⎨ 

⎪ ~ 

⎩N 

i, 

a−1 

exp( −( 

M + ( S a1Fa1 

+ S a2 

Fa 

2 + ... S a5 

Fa5 

))) otherwise 

Growth parameters 

The average value for K (0.23) and L∞ (40.7cm) presented by Seisay et al. (1992) for Oreochromis spp. was used 

in all computations, while a to value of 0.26 was used as this gave realistic lengths at low ages (Palsson et al. 1999). 

The input parameters for the length weight equation (Equation 3) were derived Seisay et al. (1992). 

Mortality rates 

To obtain estimates for the instantaneous rate of total annual mortality (Z), length frequency data collected from 

trawl surveys undertaken by the R.V. Ndunduma in Area A in December 1995, January 1997, January 1998, June 

1999 and January 2000 were analysed by means of a linearised length-converted catch curve (Pauly 1983, 1984a, 

1984b). This method uses von Bertalanffy growth parameters 1 to plot of ln (F/dt) against t, where F is the 

frequency of individuals in each length class, t is the relative age of the fish. The value dt is the time taken for the 

fish to grow through a particular length class and allows for decreased growth with increased age. The negative of 

the slope of the resultant linear regression line through the descending data points gives a first approximation of Z. 

For natural mortality, the FAO (1993) estimates M = 0.4 yr -1 for fish younger than 2 years, M =0.3 yr -1 for fish 

between 2 and 4 years old and M = 0.2 yr -1 for fish older than 4 years were used. Having obtained Z and M, fishing 

mortality was derived by subtraction (F = Z – M). 

Age specific selectivity 

Age-specific selectivity was estimated by age-converting length-based selectivity functions for chambo. Length 

based selectivity was obtained from available literature (Sipawe 2001, Chisambo 2001, FAO 1993). Both the reparameterised 

Gamma (Punt et al., 1996) and Logistic distributions were investigated to approximate gear 

selectivity patterns. The Gamma distribution was considered most suitable for gill nets as selectivity increases to a 

maximum and then decreases. The selectivity pattern of chambo seines, kambuzi seines, chilimira/kauni and the 

commercial fisheries was described by logistic distribution as selectivity was assumed to increases to, and remain at, 

a maximum. 

1 The average value for K (0.23) and L∞ (40.7cm) presented by Seisay et al. (1992) for Oreochromis spp. was 

used in all computations, while a to value of 0.26 was used as this gave realistic lengths at low ages (Palsson et al. 

1999).

Both distributions are described as 

S 

a 

a 

= ( ) 

φ 

φ 

p 

e 

( φ −a) 

/ p 

( ) 1 

−( 

a−φ 

) / σ − 

1+ 

e 

( 4 ) / 2 

2 2 

φ + σ −φ 

150 


p = 

(Gamma distribution) 

S = 

(Logistic distribution) 

a 

where Sa is the selection of the gear on fish of age a, φ is the age at maximum selectivity (in the Gamma 

distribution) or age-at-50% selection (in the Logistic distribution) and σ the width of the selectivity function. 

Table 1. Input parameters used for the application of yield and spawner-biomass-per-recruit models for 

the chambo (Oreochromis spp.) stock in Area A of the southeast arm of Lake Malawi. 

Results 

Parameter Value Function 

Growth Parameters 

K= 0.23; L∞ = 40.5cm; to = -0.23 Von Bertalanffy 

Maximum age 

Maturity 

Length Weight 

Natural mortality 

10 yrs 

φ =2.93 yrs; σ = 0.64 yrs Logistic 

α = 0.017; β = 2.99 Equation 3 

Age 0-2 = 0.4 yr -1 ; Age 2-4 = 0.3 yr -1 ; 

Age 4+ = 0.2 yr -1 

Exponential decay 

Fishing Mortality 

F (all fisheries) 0.34 yr -1 Exponential decay 

2” gill net 0.0003 yr -1 Exponential decay 

3” gill net 0.046 yr -1 Exponential decay 

Chilimira/Kauni 0.25 yr -1 Exponential decay 

Nkacha/Kambuzi seine 0.001 yr -1 Exponential decay 

Chambo seine 0.03 yr -1 Exponential decay 

Commercial fisheries 0.01 yr -1 Exponential decay 

Selectivity 

2” gill net φ = 1.6; σ = 0.16 Gamma 

3” gill net φ = 3.68; σ =0.67 Gamma 

Chilimira/Kauni φ = 3.17; σ = 0.43 Logistic 

Nkacha/Kambuzi seine φ = 0.878; σ = 0.156 Logistic 

Chambo seine φ = 3.5; σ =0.4 Logistic 

Commercial fisheries φ = 3.19; σ = 0.48 Logistic 

Per-recruit analysis 

Input parameters used in the per-recruit analysis are summarised in Table 2. Length-converted catch curves for 

Oreochromis spp. are shown in Figure 2. Estimates of Z varied between surveys ranging from 0.50 yr -1 to 0.69 yr -1 . 

The mortality rate estimated from the January 1998 survey (0.69 yr -1 ) was considered exceptionally high and was 

therefore not used in further analysis. An average of the remaining Z estimates was used as a first approximation of 

Z in Area A. Z was estimated to be 0.54yr-1. The resultant Fishing mortality rate was therefore 0.34 on fish older 

than 4 years. Estimates for gear specific fishing mortality and selectivity are shown in Table 1.

Ln (n/dt) 

151 


(a) Dec 95 

Z = 0.51 

R2 (b) Jan 97 

y = 0.58 

R 

7 

= 0.80 

6 

5 

4 

3 

2 

1 

0 

0 2 4 6 8 10 

Age (yrs) 

2 (c) Jan 98 

= 0.91 

6 

Z = 0.69 

5 

R 

4 

3 

2 

1 

0 

0 2 4 6 8 10 

Age (yrs) 

2 6 

5 

= 0.92 

4 

3 

2 

1 

0 

0 2 4 6 8 10 

Age (yrs) 

(d) Jun 99 

Z = -0.52 

R2 (e) Jan 00 

Z = 0.55 

6 

= 0.78 

R 

5 

4 

3 

2 

1 

0 

0 2 4 6 8 10 

Age (yrs) 

2 7 

= 0.82 

6 

5 

4 

3 

2 

1 

0 

0 2 4 6 8 10 

Age (yrs) 

Ln (n/dt) 

Figure 3. Length converted catch curves for Oreochromis spp. derived from trawl net length frequency 

distributions in Area A of the southeast Arm of Lake Malawi from December 1995 to January 2000. 

Yield- and spawner biomass-per-recruit as a function of fishing mortality are presented in Figure 3. Current fishing 

mortality (Fcur) was below the Fmax TRP (0.54yr -1 ) but it was higher than the more conservative F0.1 TRP (F0.1 = 0.32 

yr -1 ). At Fcur, the chambo spawner biomass was reduced to 37% of pristene levels (Figure 1). At Fmax SBR would 

be reduced to 25% of pristene levels. The FSB40 strategy approximated the F0.1 TRP. 

YPR 

80 

70 

60 

50 

40 

30 

20 

10 

Ln (n/dt) 

Ln (n/dt) 

F SB50 

F 0.1 

F SB40 

F cur 

F max 

0 

0 

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 

F 

Figure 4. Yield-per-recruit (YPR, closed circles) and spawner biomass-per-recruit (SBR; open circles) as 

a response to fishing mortality (F) for Oreochromis ‘Nyasalapia’ spp. In Area A of the SEA of Lake 

Malawi. Common target reference points are shown. 

Ln (n/dt) 

600 

500 

400 

300 

200 

100 

SBR

152 


Using the multi-fishery per-recruit models it was possible to assess the response of chambo SBR to different 

combinations of F in the fisheries. Various management options were explored and resultant effort at the FSB40 and 

FSB50 TRP are shown in Table 2. Effort limitations to below current levels were necessary to attain the FSB40 and 

FSB50 TRP in all scenarios investigated. With all fisheries active, 10% reduction in F was necessary to obtain and 

FSB40 TRP and a 23% reduction was neccesary to obtain an FSB50 TRP. The closure of the 2 inch gill net fishery or 

the nkacha net and kambuzi seine fishery only resulted in marginal increases in the allowable effort at FSB40 and 

FSB40 TRP. In all scenarios that maintained the chilimira/kauni fishery at current effort levels an FSB50 TRP was not 

attainable and the FSB40 TRP was only attainable with large effort reductions in all other fisheries. 

Table 2. Allowable effort levels at different management scenarios maintaining FSB40 and FSB50 TRPs with 

proportional increase in effort in all fisheries.(GN2 = 2inch gill net, GN3 = 3inch + gill net, NK = Nkacha 

net, KS = Kambuzi seine, CH = Chilimira/Kauni, CS = Chambo seines, PT = Pair trawl fishery. 

Management strategy 

Current Fishery 

All fisheries active 

Ban of NK + KS 

Number of gears 

F GN2 GN3 CH PT KS+NK CS 

0.34 75 343 119 2 39 18 

FSB50 0.22 49 222 77 1 25 12 

FSB40 0.31 68 313 109 2 36 16 

FSB50 0.23 50 229 79 1 - 12 

FSB40 0.32 70 319 111 2 - 17 

Ban of GN2 

FSB50 0.22 - 225 78 1 25 12 

FSB40 0.32 - 313 109 2 36 16 

Ban on Pair Trawl FSB50 0.23 56 232 80 - 26 12 

FSB40 0.32 78 323 112 - 37 17 

Ban on all seines 

FSB50 0.24 54 245 85 1 - - 

FSB40 0.34 76 346 120 2 - - 

CH maintained at current levels 

CH + CS maintained at current levels + ban on NK + 

KS 

KS + NK banned and all other fisheries except CH 

maintained at current levels 

FSB50 

Not attainable 

FSB40 0.16 35 162 119 1 18 8 

FSB50 

Not attainable 

FSB40 0.01 1 6 119 2 - 18 

FSB50 0.20 75 343 69 2 - 18 

FSB40 0.31 75 343 108 2 - 18 

Discussion 

Current exploitation rates for chambo in Area A lie between the F0.1 and Fmax TRP levels and corresponds to an SBR 

reduction to 37% of pristene levels (Figure 2). The declining chambo catch in the SEA of Lake Malawi over the 

past 10 years (FRU unpublished catch statistics, Weyl 1999, Bulirani et al. 1999) imply that this level of SBR is too 

low to sustain the chambo stock. This may, in part, be a consequence of the reproductive behaviour of chambo. 

Oreochromis spp. are mouthbrooders which brood their eggs and young for extended periods (Trewawas 1983). For 

this reason the recruitment of chambo is likely to be highly dependent on the density of adult fish and chambo 

directed management should focuss on the maintenance of the SBR at FSB40 or even FSB50 levels. Management at 

these TRPs would require a decrease in 10% in F to attain the FSB40 and a 23% reduction in F to attain the FSB50. 

The declinning catch rates for chambo indicate that current management strategies (GOM 2000) have not adequately 

protected chambo and have failed at the maintenance of sustainable yields. For this reson it is believed that the 

closed season is an inadequate measure to sustain catches in this fishery. While the use of small meshed seines and 

nkacha nets is already banned in the SEA, it is evident that their use is continued in the SEA Arm (pers obs). 

However, due to the low contribution of these gears to the total chambo catch, the removal of these gears from the 

fishery only had a marginal effect on chambo YPR and SBR. However, it has been noted that in addition to 

catching predominantly juvenile chambo (Manase 2001, FAO 1993), nkacha nets and beach seines cause

153 


considerable habitat damage by scouring the lake bottom and the destroying weedbeds (Tweddle et al. 1994) and is 

recommended that these gears remain prohibited in the SEA. The removal of gill nets with mesh sizes of less than 2 

inches from the fishery would also only have a marginal effect on YPR and SBR of chambo. This is a result of this 

gear not actively targeting chambo (Sipawe 2001). Management therefore needs to foccuss on effort reductions in 

all chambo harvesting fisheries rather than on the removal of a few highly destructive fishing gears. 

An isopleth diagram showing the response of SBR to increasing effort in the number of chilimira/kauni nets and gill 

nets (with nkacha and kambuzi seines removed from the fishery and chambo seine and commercial effort set at 

current levels) in Area A are presented in Figure 3. Essentially this isopleth shows that the number of allowable gill 

nets is dependent on the number of chilimira nets and vice versa. If FSB50 TRP was to be maintained, then any 

combination of the two gears following the FSB50 contour would be appropriate. At FSB40, slightly higher effort 

levels would be feasible. While the removal of the the kauni fishery would allow for well over 3000 gill nets in 

Area A, a total ban of kauni may not be feasible. The chilimira net is an important gear for the harvest of species 

other than chambo and a closure of this fishery would reduce landings of non-chambo species by an estimated 1500 

tons (FRU catch statistics). In addition, consideration must be given to the current employment of over 1000 fishers 

in this fishery in Area A. It is therefore recommended that an acceptable effort balance between the chilimira/kauni 

and other fisheries be considered rather than a a total closure. 

Number of Gill nets 

5000 

4000 

3000 

2000 

1000 

50 

40 

60 

30 

40 

30 

20 

70 

80 

50 

30 

20 

0 

0 60 120 180 240 300 

Number of Chilimira nets 

Figure 5. Isopleth diagram showing chambo spawner-biomass-per-recruit as a proportion of pristine 

levels for different combinations of effort in the gill net and chilimira/kauni fisheries in Area A of the SEA of 

Lake Malawi. 

An example of different management options maintaining an FSB40 TRP, with a limit of 90 chilimira/kauni units is 

shown in Table 3. With the ban of nkacha and kambuzi seines and all other current fisheries active at 1999 levels, 

the reduction of the chilimira/kauni fishery to 90 gears would allow for 1100 gill net licences in “Area A’. At this 

level a total of 1400 people could be employed and a total yield of 2800 tons could be attained (assuming no change 

in catch rate for species other than chambo). If all commercial licences were activated, (a total of 4 pair trawlers), 

only 900 gill net licences could be sold, employment would decrease to about 1300 fishers, but production would 

increase to 3600 tons. The closure of chambo seines would allow for an increase in the number of gill nets, but only 

1200 fishers would be employed in the small scale sector and the yield would be arround 3300 tons. The closure of 

the commercial sector in area A would increase allowable gill nets to 1600, allow for 1600 fishers to be employed 

and a yield of 2600 tons. 

It is evident that the management of the chambo stocks in Area A will have costs. None of the suggested strategies 

would be able to employ all 1900 fishers currently employed in the fishery. A maximisation of employment in the 

small scale sector would decrease potential landings of species other than chambo. Alternatively, maximisation of 

20

154 


landings through commercialisation would decrease potential employment. A management strategy for chambo in 

Area A therefore needs to take into account both socio-economic as well as biological constraints. 

Table 3. Estimated number of gill net units, employment and total annual yield for all species in Area A of 

the SEA of Lake Malawi at different management scenarios maintaining an FSB40 TRP for Chambo, 

when Chilimira/Kauni effort is limited to 90 gears. 

All possible 

Closure of the 

All fisheries Active commercial licences Closure of all seines commercial fishery 

Number of gill nets < 1100 gill nets

155 


Pauly, D. 1980. On the interrelationship between natural mortality, growth parameters and mean environmental temperature in 

175 stocks of fish. J. Cons. Int. Explor. Mer, 39: 175-192. 

Pauly, D. 1983. Length-converted catch curves: a powerful tool for fisheries research in the tropics. ICLARM, Fishbyte 1(2): 9- 

13. 

Pauly, D. 1984a. Length-converted catch curves. A powerful tool for fisheries research in the tropics. ICLARM. Fishbyte 2(1): 

17-19. 

Pauly, D. 1984b. Length-converted catch curves. A powerful tool for fisheries research in the tropics (conclusion). ICLARM, 

Fishbyte 2(3): 9-10. 

Pikitch, E.K. 1987. Use of a mixed-species yield-per-recruit model to explore the consequences of various management policies 

for the Oregon flatfish fishery. Can. J. Fish. Aquat. Sci. 44: 349-359. 

Punt, A.E. 1993. The use of spawner-biomass-per-recruit in the management of linefisheries. Special Publication of the 

Oceanographic Research Institute, Durban, 2: 80-89. 

Seisay M.D.B., van Zalinge N.P & Turner G.F., 1992. Population Dynamics and stock estimates of Chambo (Oreochromis spp.) 

in the south east arm of Lake Malawi and lake Malombe-Length based approach. FI:DP/MLW/86/013. Field Document 

19. 34p. 

Shepherd, J.G. 1982. A versatile new stock-recruitment relationship for fisheries, and the construction of sustainable yield curves. 

J. Cons. Int. Explor. Mer, 40: 67-75. 

Sipawe, R.D. 2001. Gear and species selectivity of the gill net fishery in Lake Malawi. pp 133 –141, In: Weyl, O.L.F. & Weyl, 

M.V. (eds) Lake Malawi Fisheries Management Symposium – Proceedings. 4 th – 9 th June 2001. NARMAP, Department 

of Fisheries, Malawi. 248 pp. 

Sissenwine, M.P. & Shepherd, J.G. 1987. An alternative perspective on recruitment overfishing and biological reference points. 

Can. J. Fish. Aquat. Sci. 44: 913-918. 

Smale, M.J., & Punt, A.E. 1991. Age and growth of the red steenbras Petrus rupestris (Pisces Sparidae) on the southeast coast of 

South Africa. S. Afr. J. Mar. Sci. 9: 249-259. 

Trewawas, E. 1983. Tilapiine Fishes of the genera Sarotherodon, Oreochromis and Danakila. British Museum (Natural History) 

Publication No. 878. 583pp. 

Tweddle, D., Alimoso, S.B. & Sodzapanja, G. 1994a. Analysis of catch and effort data for the fisheries of the South East Arm of 

Lake Malawi 1976-1989. Government of Malawi, Fisheries Department Bulletin No. 13. 

Weyl, O.L.F. 1998, The dynamics of a sub-tropical lake fishery in central Mozambique. PhD Thesis, Rhodes University, 202p. 

Weyl, O.L.F. 2001. Area management (area management committees) as a concept for PFM implementation – Discussion Paper. 

Appendix I(g), In: National Aquatic Resource Management Programme (NARMAP), Report on the participatory 

Fisheries Management programme (PFMP) strategy review workshiop. Sun ‘n’ Sand Holiday Resort, Mangochi, 30 

January – 1 February, 2001. 

Weyl, O.L.F., Banda, M., Sodzqabanja, G., Mwenekibombwe, L.H., Namoto, W. & Mponda, O.C. 2000. Annual Frame Survey, 

September 1999. Government of Malawi, Fisheries Bulletin No. 42.

156 


The nkacha fishery in Lake Malombe and the southeast arm of Lake Malawi 

Moffat M. Manase 

Malawi Fisheries Research Unit, PO Box 27, Monkey-Bay, MALAWI.Tel. +265 587 440/ 432. Email: fru@malawi.net. 

Abstract 

The nkacha net is an open water seine with a uniform mesh size ranging from 6 to 25 mm and is rectangular in shape. It is a main 

fishing gear of the small-scale fishery in Lake Malombe but is prohibited in Lake Malawi. However, this gear is used illegally in the 

southeast arm of the lake since the mid 1990s. This study compares catch and effort data as well as the catch composition of the of 

the nkacha net fishery for Lake Malombe and the southeast arm of Lake Malawi. Higher catch per unit effort (CPUE) in the 

southeast arm as opposed to Lake Malombe might be the primary drive for this gear’s usage in Lake Malawi. Species size-classes 

caught in the southeast arm were considerably larger than of those species constituting catches from Lake Malombe. This suggests 

that catches from the southeast arm fetched higher market prices than Lake Malombe catches. Considering the impact of the 

nkacha net with regard to growth over-fishing and habitat degradation, management recommendations are therefore made for this 

fishery. 

Introduction 

Lake Malombe has a maximum length of 29 km and a maximum width of 17 km with a maximum depth of 17 m 

and a surface area of about 390 km 2 (FAO, 1993). The lake is about 16 km from Lake Malawi through the Upper 

Shire River, and this makes it easier for fishers to migrate between Lake Malombe and the southeast arm of Lake 

Malawi (Figure 1). 

In the 1980s, the more valued chambo fishery (Oreochromis spp.) in Lake Malombe was reported to be in a state of 

collapse and that the fishery was eventually replaced by a fishery for the less valued haplichromine cichlid species 

called kambuzi using open water seines called nkacha nets. Mesh size for this gear is uniform ranging from 6 mm to 

25 mm across the net panel. The nkacha net is operated by small-scale fisheries in Lake Malombe using small plank 

boats and dug out canoes. The operation of this gear does not involve use of light during night fishing. Recent 

studies on Lake Malombe indicate that this kambuzi fishery is also in a declining state (Banda & Hara 1994, Weyl 

1999, Weyl et al. 2001). 

Though nkacha gear is legal only in Lake Malombe, there has been unprecedented increase in the use of this gear in 

Lake Malawi since the mid 1990s, particularly in the southeast arm of the lake. Operation of this gear in Lake 

Malawi is prohibited on account that the gear destroys nursery grounds for some important fish stocks, including the 

chambo (Banda & Hara 1994). As such, this gear is likely to compromise the economic value of the fisheries of 

Lake Malawi. The purpose of this study was to describe and compare the nkacha net fishery in Lake Malombe with 

that of the southeast arm of Lake Malawi in reference to catch rates, catch value, species and size composition. This 

study will aid in the development of appropriate management interventions that will go a long way towards 

safeguarding the existing high valued fish stocks of Lake Malawi. 


Data from research surveys undertaken by the Fisheries Research Unit of the Fisheries Department on Lake 

Malombe and the southeast arm of Lake Malawi were used. On each sampling day, gears were described by type, 

mesh size, gear length and mode of operation. Catch samples were selected at random and every fish in each sample 

was sorted to species level. Each species group was weighed and each fish in the sample was measured for total 

length (TL) to the nearest millimetre. Catch and effort data collected during routine monitoring surveys by the 

Department of Fisheries for Lake Malombe and southeast arm of Lake Malawi for the period 1994-1999 were also 

used in this study to compare catch rates and catch value for these water bodies. Methods used in the collection of 

catch and fishing effort data from the traditional fisheries of Malawi followed the Malawi Traditional Fisheries 

(MTF) system described by FAO (1993). 


Ownership of nkacha gear is shown in Figure 2. The nkacha net fishery started booming up around 1993 in the 

southeast arm of Lake Malawi and registered a high level of some 70 gears in 1999. This showed a sharp contrast 

with the situation of Lake Malombe where ownership of nkacha nets started decreasing from a high level of over 

300 gears in 1995 to a level of 164 in 1999. This suggests some driving force that has been attracting nkacha gear 

owners to operate in Lake Malawi apart from Lake Malombe where the gear is recommended.

157 


The species composition for the nkacha fishery in Lake Malombe is not distinct from the species composition of the 

nkacha fishery of the southeast arm of Lake Malawi. Catch composition of the nkacha fishery in Lake Malombe as 

well as the southeast arm of Lake Malawi has been predominantly kambuzi species. The main species for nkacha 

fishery in Lake Malombe and the southeast arm of Lake Malawi include species belonging mainly to the genera 

Copadichromis, Lethrinops and Otopharynx (Figure 2). 

Figure 1. Map of the study area. 

No. of gears 

No. of gears 

80 

70 

60 

50 

40 

30 

20 

10 

0 

400 

300 

200 

100 

0 

93 94 95 96 97 98 99 

malombe 

93 94 95 96 97 98 99 

Figure 2. Ownership of nkacha nets in Lakes Malawi (southeast arm) and Malombe, 1993-1999. 

Year 

southeast arm 

Catch per unit of effort (CPUE) for nkacha net fishery in the southeast arm was significantly higher than that of 

Lake Malombe between 1995 and 1997 (Figure 3). The highest CPUE of 140 kg per trip was attained in southeast 

arm nkacha net fishery in 1996. Relating the CPUE levels with catch value, catches from the southeast arm have 

been registering consistently high value. In 1999, the value of the catch reached a level of MK 1,400.00 per fishing 

trip whereas catch value from Lake Malombe was MK 1,175.00 per trip in the same year. This indicates that the 

high catch rates in the southeast arm of Lake Malawi from 1995 might have attracted nkacha gear owners from Lake

158 


Malombe. These nkacha net operators have not withdrawn from southeast arm, despite high CPUE levels from 1997 

in Lake Malombe, due to low economic value of the harvested species in Lake Malombe (Table 1). 

a b 

Otopharynx 

agyrosoma 

7% 


26% 

Copadichromis 

virginalis 

11% 

Otopharynx 

tetrastigma 

16% 

Lethrinops 

'pinkhead' 

23% 

Copadichromis 

chrysonotus 

17% 

Bagrus meridionalis 

6% 

Ramphocromis spp 

6% 

Pseudotropheus 

20% 

Otopharynx 

argyrosoma 

10% 


9% 

Lethrinops 

'pinkhead' 

6% 

Figure 3. species catch composition for nkacha nets in a) Lake Malombe and b) the southeast arm of 

Lake Malawi. 

Table 1. Mean beach price ± 95% confidence interval for kambuzi at landing sites at Lake Malombe and 

the southeast arm of Lake Malawi 

Lake Malombe Lake Malawi 

Year Mean 

(MK/kg) 

std n 

Mean 

(MK/kg) 

std n 

1994 1.16 1.04 1466 1.98 0.97 958 

1995 4.45 2.48 1590 4.62 2.60 782 

1996 4.23 1.43 928 4.92 1.88 309 

1997 4.87 1.70 652 5.42 2.02 442 

1998 5.86 2.99 877 11.49 3.96 379 

1999 9.00 4.11 1537 16.33 8.78 583 

Copadichromis 

virginalis 

43% 

Other than catch rates, it appears that length frequency distributions of species could also be playing an attractive 

role for nkacha netters in Lake Malawi since the value of the harvested species is related to size of fish offered for 

sale at the markets. 

For Copadichromis virginalis (utaka), most of the fish harvested from Lake Malombe were of 65 mm TL whereas 

those from the southeast arm registered 110 mm TL (Figure 5). On the other hand, Otopharynx argyrosoma, another 

Kambuzi species, registered 60 mm TL in the southeast arm of Lake Malawi compared to 65 mm TL from Lake 

Malombe and a good number of fish sampled in southeast arm registered over 80 mm TL (Figure 6). In overall 

terms, mean lengths for Otopharynx argyrosoma were 69 mm TL and 64 mm TL in southeast arm of Lake Malawi 

and Lake Malombe respectively (Table 2). As for Lethrinops ‘pinkhead’ (Figure 7), the catches from southeast arm 

registered 90 mm TL whereas those from Lake Malombe registered 60 mm TL. Interestingly, a relatively high 

proportion of the catch from nkacha net fishery operating in the southeast arm of Lake Malawi constituted chambo 

(Table 1). Although chambo registered only 9% of this gear’s total catch, the situation is more threatening 

considering the fact that small sized individuals of chambo, locally known as kasawala, were targeted (Figure 8).

159 


Figure 4. Nkacha net (A) catch per unit effort ± 95 % confidence intervals and (B) Mean catch value in 

Malawi Kwacha in Lake Malombe and the southeast arm of Lake Malawi from 1994 to 1999. 

Frequency 

CPUE (Kg per trip) 

Catch Value (MK/trip) 

Frequency 

140 

120 

100 

300 

250 

200 

150 

100 

50 

200 

150 

100 

80 

60 

40 

20 

0 

50 

0 

0 

1500 

1250 

1000 

750 

500 

250 

0 

20 

20 

(A) 

(B) 

40 

40 

Figure 5. Length frequency distribution of Copadichromis virginalis (Utaka) 

60 

60 

80 

80 

100 

100 

Year 

120 

120 

Length mm TL 

Lake Malombe 

Southeast arm 

94 95 96 97 98 99 


140 

malombe 

140 

160 

160 

180 

180 

200 

200

Frequency 

Frequency 

Figure 6. Length frequency distributions of Otopharynx argyrosoma (Kambuzi) 

35 

30 

25 

20 

15 

10 

5 

0 


30 50 70 90 110 130 150 170 

1000 

Frequency 

Frequency 

120 

100 

80 

60 

40 

20 

0 

600 

500 

400 

300 

200 

100 

0 

800 

600 

400 

200 

0 

20 

20 

40 

40 

malombe 

30 50 70 90 110 130 150 170 190 

Length mm TL 

Figure 7. Length frequency distributions of Lethrinops ‘pinkhead’ (Kambuzi). 

60 

60 

80 

80 

100 

100 

120 

160 



140 

malombe 

120 

Length mm TL 

140 

160 

160 

180 

180 

200 

200

Frequency 

20 

15 

10 

5 

0 

20 

40 

60 

80 

161 


Figure 8: size distribution for juvenile Oreochromis sp. caught in the southeast arm 

100 

120 

Length mm TL 


According to breeding studies done on chambo in the southeast arm of Lake Malawi, length at maturity for this 

species is 200 mm TL (Palsson et al. 1999), and this was not reflected in the individuals of chambo caught by the 

nkacha net fishery. Most of the species registered 110 mm TL at capture, with a length range of 40 to 145 mm TL 

(Figure 8). This indicates that the nkacha net fishery was targeting immature fish in this area. Chambo, being one of 

the threatened fish species in Malawi, therefore deserves appropriate exploitation regimes that would seek 

sustainable harvesting of this important fishery (Alimoso, 1987). 

Table 2. Mean length TL mm, 95% confidence interval for the species caught in southeast arm of Lake 

Malawi and from Lake Malombe. 

Species Lake Malawi Lake Malombe 

mean 95%CI n mean 95%CI n 

Copadichromis virginalis 96.18 1.27 1012 70.82 0.83 1260 

Otopharynx argyrosoma 69.42 1.31 782 64.70 0.52 2971 

Lethrinops 'pinkhead' 89.84 3.53 125 67.76 0.82 2761 

Oreochromis spp. 96.06 4.66 107 NA NA NA 

Size distributions for target species of nkacha in the southeast arm of Lake Malawi were generally larger that those 

of Lake Malombe (Table 2). This implies that the market value of fish from Lake Malawi was superior to those of 

Lake Malombe. Based on this scenario, it appears market forces are among the main factors that relate to the 

continued existence of nkacha net fishery in the southeast arm of Lake Malawi although this gear is illegal in this 

area. This could be supported by the fact that small-scale fishers are increasingly becoming business-orientated in 

nature other than subsisting on their fishing operations (Turner and Mdaihli, 1992). 

It is also vital to consider that managing the nkacha fishery in Malawi has a considerable level of difficulty in the 

sense that it is legal to operate nkacha gear in Lake Malombe but illegal to operate in Lake Malawi, a nearby water 

body. Such types of regulations are complex bearing in mind the many attributes that one would expect to be 

accomplished by the fisheries enforcement section of the Department of Fisheries. 

Furthermore, the nkacha net fishery promotes loss of nesting habitat for sand nesting cichlids due to scouring effect 

of the gear (Banda and Hara, 1994). The problems of the nkacha net fishery will eventually lead to loss of 

biodiversity of Lake Malawi’s fish fauna and flora if proper management interventions are not put in place and 

followed. Against this background, management options for nkacha net fishery would include complete withdrawal 

of all nkacha netters and all small-meshed seine nets from Lake Malawi. Furthermore, a review of all existing 

fisheries regulations seems pertinent with regard to conflicting interests among stakeholders of the fisheries 

resources of Malawi. 

There is adequate evidence that migration of nkacha netters does occur between Lake Malombe and the southeast 

arm of Lake Malawi due to their close proximity to each other. This scenario can also be attributed to a lack of 

effective enforcement of fisheries regulations and therefore calls for multi-sectoral collaboration in the management 

of fisheries resources of Malawi. 

140 

160 

180 

200

162 



The author extends gratitude to the following sponsors- GTZ assisted National Aquatic Resource Management 

Programme (NARMAP), Icelandic International Development Agency (ICEIDA) and Fisheries Research Fund 

(FRF) for financing target oriented research surveys that yielded the data used in this report. Special regards also 

extend to fellow researchers at Monkey-Bay Fisheries Research Station for their valuable criticisms as well as proof 

reading of the manuscript. Special thank you goes to Dr Olaf Weyl, NARMAP Research Adviser, for fruitful 

contributions. 

References 

Alimoso SB, 1987 An Assessment of the effects of traditional fishing on Chambo (Oreochromis spp.) in Lake 

Malombe. LUSO: J.Sci. Techn. (Malawi) (1987) 8 (1&2) :1-10. 

Alimoso SB and Tweddle D, 1994 Seine net fisheries of Lake Malombe. Traditional Fisheries Assessment Project 

Working Paper TFAP/1. 

Banda MC and Hara MM, 1994 Seine nets and habitat degradation: How they have caused collapse of the Chambo 

(Oreochromis spp.) fishery of Lake Malombe and the Upper Shire. 

Bazigos GP, 1972 The Improvement of the Malawian fisheries Statistical system. A report to the integrated project. 

FAO/MLW/16. Zomba, Malawi. 

FAO, 1993. Fisheries management in the southeast arm of Lake Malawi, the Upper Shire 

River and Lake Malombe, with particular reference to the fisheries of chambo (Oreochromis spp.) CIFA 

Tech. Pap. 21, Rome FAO. 113p. 

Palsson OK, Bulirani A and Banda M 1999 A review of biology, fisheries and population dynamics of chambo 

(Oreochromis spp., CICHLIDAE) in Lakes Malawi and Malombe. Fisheries Bulletin No.38. Fisheries 

Department. Malawi 

Turner G.F. and Mdaihli M. 1992 A Bioeconomic Analysis of the Industrial and Semi-Industrial Fisheries of 

Southern Lake Malawi. FI:DP/MLW/86/013 Field Document 22, July 1992. 

Weyl, O.L.F. 1999. Lake Malombe artisanal fishery catch assessment 1994-1998. NARMAP Technical Report No. 

1. 23p. 

Weyl, O.L.F., Banda, M.C., Manase, M.M., Namoto, W. & Mwenekibombwe, L.H. 2001. Analysis of catch and 

effort data for the fisheries of Lake Malombe, 1976-1999. Fisheries Bulletin No. 45.

163 


The state of the large scale commercial fisheries on Lake Malawi 

Moses Banda 

Fisheries Research Unit, P.O. Box 27, Monkey Bay, MALAWI-AFRICA. Tel. 265-587-432, E-mail: fru@malawi.net. 

Abstract 

The large-scale commercial fisheries are mechanised and based in the southern part of the lake. This report is based on the data 

collected from monitoring surveys as well as an analysis of fisheries statistics in the past decade. The data were analysed by area 

and depth ranges: shallow < 50 m, 51-100 m and very deep >100 m. The catch was dominated by cichlids and the cpue decreased 

with increasing in depth in both areas. Biomass estimates based swept area methods and the principal precautionary approach 

indicate that the fishery is not overexploited. 

Key words: Large-scale fishery, biomass, swept area method, principal precautionary approach. 

Introduction 

The large-scale commercial fisheries in Lake Malawi are mechanised and capital intensive and use mainly trawling 

and purse seining (‘ring net’). The fishery consists of pair trawlers, (wooden boats about 8 m long with a 20-40 hp 

inboard engine), stern trawler (90-385 hp) units and ring nets (90 hp), which are confined to the southern part of the 

lake. The origins of the large-scale commercial fishery can be traced back as early as 1943 when purse seining 

started for chambo in the south east arm by the Europeans (FAO, 1976). In the mid 1960s experimental trawling led 

to the establishment of a pair trawl fishery to harvest the demersal fisheries in the shallow parts of the south east arm 

in 1968 (Tarbit, 1972). This fishery developed rapidly in the shallow waters of the southern part of the lake and in 

1976 it was established on the western side of the lake. The more powerful stern trawlers were introduced in 1972 

for bottom trawling (85 hp) and in 1976 for mid water trawling (180 hp) in the south east arm (FAO, 1976; Turner, 

1977). Four bottom trawlers have joined the fishery since then; one started fishing in the same area in 1983; the 

other two in 1997 and the fourth one in 1998 (Banda, 2000). There are five stern trawlers now operating, although a 

new 380 hp boat replaced one bottom trawler. Bottom trawling was restricted to depths ranging between 50 and 

70 m, but it has been expanded to deeper water (up to 100 m) by the two new powerful boats (380 hp). The reported 

average large-scale commercial catch is currently around 5,600 tons per annum, approximately 21% of the total 

annual fishing landings from Lake Malawi. 

The management framework of the commercial fishery is based on the recommendations from monitoring 

programmes carried out by the Fisheries Department, which include the assessment of sustainable yields and 

allowable effort through biomass surveys, and the investigation of changes in the species composition of the catches 

and size distribution. Sustainable yields were estimated from the data for the fishing areas using the Schaefer (1954) 

surplus production model with the assumption of equilibrium conditions (Pitcher and Hart, 1982). However, 

continuous changes in species composition and population structure render this assumption invalid (Turner, 1995). 

Consequently, the precautionary principle was used as stock status indicator to generate interim management 

measures in the absence of stock assessment. The monitoring surveys were started in the early 1970s and have 

continued to date with some discontinuation (FAO, 1976; Turner, 1977; FAO, 1993; Turner, 1995; Banda et al., 

1996; Banda and Tómasson, 1997; Pálsson et al. 1999). 

The fishing regulations include restriction on the number of licences to be issued in each demarcated area and the 

minimum mesh size of the trawl cod-end to be used. At present the number of the fishing fleet is 100% i.e. all 

licences have been issued, although not all are operational, especially the pair trawl units. The stern trawlers are 

licensed to fish waters greater than 50 m and the pair trawlers between 18 and 50 m water depth, or not less than one 

nautical mile from the shoreline. The recommended mesh sizes for the gears are 102 mm for the ring nets and 

38 mm for other vessels. 

In this study, data from monitoring surveys and historical trends of catch and effort from the commercial fishery 

have been analysed to assess the state of the commercial fisheries with particular reference to changes in species 

composition, mean size and catch rates. The results could be used to improve current and future management of the 

commercial fishery.


164 


Monitoring trawl survey 

The data in this study were collected during the ongoing regular stock assessment surveys carried out by the 

Fisheries Department in the south east arm (SEA) and south west arm (SWA) of the lake (Figure 1). These surveys 

were designed to monitor changes taking place in the southern part of the lake, which is heavily exploited. The 17 m 

long Fisheries Department research vessel, Ndunduma powered with a 380 hp engine which pulls a Gulloppur 

bottom trawl net with a 23 m headrope and a 38 mm mesh cod end was used. The effective mesh size in the cod end 

was 4 mm due to the net liner used to ensure that escape through the meshes was minimal. 

A total of 97 fixed trawl stations (54 in the SEA and 43 in SWA) were sampled during each survey. The stations 

were stratified into shallow 0-50 m, deep 50-100 m and very deep 100-150 m (Figure 1). Each trawl lasted for 30 

minutes. All trawling was done during the day at depths between 10– 150 m; it was not possible to sample in water 

less than 10 m deep because of restrictions imposed by the draught of the vessel. 

Latitude 

-14.4 -14.2 -14.0 -13.8 

34.60 34.75 34.90 

Longitude 

35.05 35.20 

Figure 1. Map of southern of Lake Malawi showing the sampling stations used in the monitoring 

surveys. 

The catch was sorted into four categories: Clariid catfish, Bagrus meridionalis, small cichlids and others (including 

cyprinids, mormyrids and large cichlids such as Buccochromis spp. and Oreochromis spp.). Each of them was 

measured on board the vessel to the nearest 0.5 cm (total length) and most fish were weighed to the nearest 0.01 kg 

and those that were not weighed individually were collectively weighed as a species group to the nearest 0.1 kg. 

The catch per unit effort (C/f) assumed as index of abundance was expressed as: 

C/f = catch (kg) /swept area (ha), 

SEA 

and sustainable yields were estimated to be 40% of the stock biomass (Banda et al. 1996).

165 


Commercial trawl data 

The catch and effort data from monthly-submitted returns of the commercial fishing companies from 1980 to 2000 

were used in this analysis. Four datasets were summarised according to the operational fishing method: inshore 

demersal (pair trawlers), offshore demersal (stern bottom trawlers), midwater (midwater trawlers) and ringnet 

(chambo) fisheries. Trends in effort were described using boat-days and trends in cpue using catch/day except for 

the purse seines. The effort for the purse seine were expressed as pulls and cpue and catch/pull. Effort for all the 

fisheries has been standardised according to Sparre and Venema (1992) because fishing power has changed with 

time. 

The precautionary reference points for the exploitation of the fisheries resources using standardised cpue as an 

estimator of stock biomass or size was adopted as described in Bulirani (et al. 1999). Four reference points were 

calculated: maximum biomass (Bmax ), biomass at precautionary approach (Bpa), biomass limit (Blim) and current 

biomass point estimate (Bcur). The mean cpue over a period of relatively high cpue represented maximum biomass 

(Bmax) and 45% of the Bmax was considered as Bpa, a precautionary point where management must be undertaken. 

Blim was calculated as 20% of the Bmax and was considered as a point where management must be undertaken and 

Bcur was defined as the current level of biomass (cpue) relative to Bmax. 

Results 

Monitoring surveys 

The catch composition 

The dominant species (by weight) throughout the surveys are shown in Figure 2. Copadichromis virginalis, 

Otopharynx argyrosoma, Oreochromis spp, Bagrus meridionalis and Bathyclarias species were the dominant 

species groups in the shallow waters of both arms of the lake (Figure 3). Lethrinops longipinnis and C. eucinostomus 

were also important in the south west arm of the lake. Years of high catches of C. virginalis coincided with low 

catches of C. eucinostomus and visa versa. Capadichromis virginalis, Bagrus meridionalis, Bathyclarias spp. 

Diplotaxodon limnothrisa, Lenthrinops gossei, L. oliveri, L. alba and Synodontis njassae remained generally an 

important component of the catch throughout the surveys in the deep waters of the southern part of the lake (Fig.4). 

Catch per unit of effort (cpue) 

The cpue for each year is shown in Figures 5 & 6 and where there was more than one survey conducted an average 

cpue was calculated. The mean cpue in shallow water from the SEA and SWA were 931.5 and 714.7 kg hr -1 , 

respectively, and were significantly different (t-test, p < 0.05) while that from the water depth > 50 m was 849.5 and 

532.1 kg hr -1 , respectively, which was also significantly different (t-test, p < 0.05). The cpue in the shallow water < 

50 m show a general decline since 1995 in the SEA and 1994 in the SWA (Figure 5). In contrast, the cpue in waters 

> 100 m seemed to have no proper trend in both arms (Figure 6). 

Biomass 

The biomass estimates and calculated MSY are shown in table 1. The biomass and MSY estimates of the SEA were 

generally greater than those from the SWA. Similarly, the biomass and MSY estimates are greater in waters < 50 m 

than in waters > 50 m in both arms. 

Table 1. Biomass and Maximum Sustainable Yield (MSY) estimates from RV. Ndunduma surveys in the 

southern part of Lake Malawi from 1994 to 2000. 

SEA SWA 

0-50 m > 50 m 0-50 m > 50 m 

Biomass 7230 8410 3130 6190 

MSY 2890 3360 1250 2480

others Diplotaxodon 

limnothrisa 

Otopharynx 

argyrosoma 

166 


Figure 2. Species composition of the bottom trawl catch in the southern part of Lake Malawi from 

surveys from 1994 to 2000. 

others 

Bagrus 

meridionalis 

Bathyclarias 

spp. 

Lethrinops 

gossei 

Oreochromis 

species 

Copadichromis 

virginalis 

Figure 3. Species composition of the bottom trawl catch in the shallow waters (0-50 m) of the southern 

part of Lake Malawi from surveys from 1994 to 2000. 


Letnrinopd alba 

Lethrinops 

oliveri 

Bagrus 

meridionalis 

Bathyclarias 

spp. 

Oreochromis 

species 

Copadichromis 

virginalis 

Otopharynx 

argyrosoma 

Copadichromis 

eucinostomus 

Bagrus 

meridionalis 

Bathyclarias 

spp. 

Lethrinops 

gossei 

Synodontis 

njassae 

Diplotaxodon 

limnothrissa 

Copadichromis 

virginalis 

Figure 4. Species composition of the bottom trawl catch in the shallow waters (0-50 m) of the southern 

part of Lake Malawi morning surveys from 1994 to 2000.

Catch rate (kg/hr) 

1400 

1200 

1000 

800 

600 

400 

200 

167 


Figure 5. The mean catch rate (kg/hr) made from RV. Ndunduma surveys from 1994 to 2000 in waters 

(0-50m) of the southern part of Lake Malawi. 

Catch rate (kg/hr) 

Figure 6. The mean catch rate (kg/hr) made from RV. Ndunduma surveys from 1994 to 2000 in waters (> 

50 m) of the southern part of Lake Malawi. 

Commercial fisheries catch and effort data 

0 

1200 

1000 

800 

600 

400 

200 

1994 1995 1996 

Year 

1997 1998 1999 2000 

Sea Sw a 

0 

1994 1995 1996 1997 1998 1999 2000 

Year 

Sea Sw a 

Pair trawl fishery 

The catches of the pair trawl fishery have been relatively stable between 1980 and 1988, and peaked up in 1990 to 

about 1 700 tons in the south east arm and since then there has been a general decline. The sudden drop in catch in 

1993 was due to the closure of the fishery in Area A, south of Boadzulu Island (Figure 1). The cpue and effort 

showed a similar pattern to that of the catch. In contrast, the catches in SWA increased from lowest levels of about 

150 tons in 1980 to highest levels of about 2 200 tons in 1988 and then stabilised around 1 500 tons between 1989 

and 1994. After 1994, the catches have dropped down to levels of about 600 tons and have remained relatively 

stable since then. The effort has followed the same pattern as the catch while the catch rates have remained relative 

constant around 1.2 tons a day from 1980 to 2000. The Surplus Production Models did not fit the data as a result the

168 


precautionary approach principle was applied (Table 2). The pair trawl fishery in the SEA operates close to Bpa 

while that in the SWA above Bpa and hence no need for immediate action. 

Table 2. Precautionary approach reference points for various large scale commercial fisheries in the 

southern part of Lake Malawi. PTF = pair trawl fishery, SBTF = stern bottom trawl fishery, MWTF = 

midwater trawl fishery and Ring net fishery. 

PTF PTF SBTF MWTF RNF 

Reference points SEA SWA SEA SEA SEA 

Bmax 1.5 1.5 1.5 1.3 1.5 

Bpa 0.7 0.7 0.7 0.6 0.7 

Blim 0.4 0.4 0.4 0.3 0.4 

Bcur 0.8 1.1 0.7 0.8 0.4 

(a) 



(b) 

2.0 

1.5 

1.0 

0.5 

0.0 

2.0 

1.8 

1.6 

1.4 

1.2 

1.0 

0.8 

0.6 

0.4 

0.2 

0.0 

80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 

Year 

Yie ld Relative cpue No rmo lis ed effo rt 

80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 

Year 

Figure 7. Catch, effort and CPUE in Pair trawl fishery in (a) south east arm and (b) south west arm of 

Lake Malawi from 1980 to 2000. 

1600 

1400 

1200 

1000 

800 

600 

400 

200 

0 

Yield Relative cpue Normalised effort 

2500 

2000 

1500 

1000 

500 

0 

Catch (tonnes) 

Catch (tonnes)

169 


Stern bottom trawl fishery 

The catches of the demersal fisheries based in the SEA reached the highest levels of around 2 400 tons in 1984 and 

then since it started declining reaching levels below 400 tons in 1993 (Figure 8). After 1993 the catch rose to 

2 100 tons in 1995 and since then it has stabilised to around 1 500 tons/year. The trend of effort follows that of the 

catch while the catch rates indicate a down trend since the early 1980s. The Surplus Production Models did not fit 

the data as a result the precautionary approach principle was applied (Table 2). The stern bottom trawl fishery in the 

SEA currently operates at Bpa which calls for immediate action. 

Midwater trawl fishery 

The midwater fishery catches, which is also confined in the SEA were relatively stable between 1980 and 1992 and 

ranged between 1 500 and 2 200 (Figure 9). The total catch dropped to half in 1993 and has steadily increased since 

then. The catch rates trend was similar to that of the total catches. The total effort has been stable throughout except 

in 1997 (Figure 9). The Surplus Production Models did not fit the data as a result the precautionary approach 

principle was applied (Table 2). The stern bottom trawl fishery in the SEA currently operates above Bpa and hence 

no need for immediate action. 


2.0 

1.8 

1.6 

1.4 

1.2 

1.0 

0.8 

0.6 

0.4 

0.2 

0.0 

80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 

Year 

Figure 8. Catch, effort and CPUE in stern bottom trawl fishery in the south east arm of Lake Malawi from 

1980 to 2000. 


1.8 

1.6 

1.4 

1.2 

1.0 

0.8 

0.6 

0.4 

0.2 

0.0 

Total Yield Relative cpue Normalised effort 

80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 

Year 

Total yield Rel. cpue Norm. effort 

Figure 9. Catch, effort and CPUE in midwater trawl fishery in the south east arm of Lake Malawi from 

1980 to 2000. 

2500 

2000 

1500 

1000 

500 

0 

3000 

2500 

2000 

1500 

1000 

500 

0 

Catch (tonnes) 

Catch (tonnes)

170 


Ringet fishery 

Unlike the other fisheries, the ringnet fishery targets chambo (Oreochromis spp.) south of Boadzulu Island in the 

SEA (Figure 10). The highest catches of chambo were manifested in 1983 and have declined steadily since then, 

reaching their lowest levels in 1997. From 1998 to 2000 there has been a slight increase in the catches. The total 

effort and catch rates have followed the pattern of the catch. Similarly, the Surplus Production Models did not fit the 

data as a result the precautionary approach principle was applied (Table 2). The ringnet fishery in the SEA currently 

operates at Bpa which calls for immediate management action. 


3.00 

2.50 

2.00 

1.50 

1.00 

0.50 

0.00 

80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 

Year 

2000.0 

1800.0 

1600.0 

1400.0 

1200.0 

1000.0 

800.0 

600.0 

400.0 

200.0 

To tal yie ld Rel. cpue No rm. effo rt 

Figure 10. Catch, effort and CPUE in ringnet fishery in the south east arm of Lake Malawi from 1980 to 

2000. 

Discussion 

Changes in species composition, cpue and catch contribution have been used as indicators of stock status for the 

commercial fisheries in Malawi (Turner, 1977; Banda, et al. 1996). In this case, declines of 75% of the cpue or more 

indicate overexploitation (Bulirani et al. 1999). The 75% reference point was however, determined arbitrarily. 

Species composition, cpue and catch 

No changes in species composition was noted during the survey period. The catch was dominated by small cichlid 

species such as C. virginalis, C. eucinostomus, O. argyosoma and a few large species such as B. meridionalis, 

Bathyclarias spp. and Oreochromis spp. which is consistent with the earlier studies (Banda et al. 1996;Turner et al. 

1997). All the dominant species in this survey were dominant in both small- and large-scale fisheries (Manase 2000; 

Nyasulu, 2001). This is indicative of resource use overlap between the small- and large-scale fisheries. The Fisheries 

Conservation and Management Act 1997 regulations specify that different fisheries should operate in specific areas, 

time and depth to reduce conflicts. Therefore, this overlap can mainly be attributed to non-compliance of the 

regulations by fishers resulting in all different fishers operating in the same fishing grounds. Indeed large scale 

fisheries such as pair trawlers have been observed to fish in waters < 20 m in both arms. It has also been noted 

recently that RV. Ndunduma has damaged many gillnets during its fishing experimental surveys in waters depth of 

50-60 m. The small-scale fishers set their 2 inch size gillnets during daytime to catch (Utaka) Copadichromis 

virginalis which is contrary to the law. The gillnets are supposed to be set between 6.00 p.m. and 6.00 a.m. The 

small scale fishers have also thrown stumps or nkokwe as fish aggregates in some trawlable grounds. The presence 

of the fishers waiting for their nets during daytime and stumps of trees makes trawling in these grounds difficult. 

The standardised index of abundance which is assumed to be linearly proportional to exploitable biomass, indicate 

that the relative abundance has decreased between 23% and >80% since the early 1980s in various large-scale 

fisheries. The highest decrease was evident in the cambo fishery (> 80%) and the lowest in the pair trawl fishery 

(23%) in the SWA. Since the pattern of effort corresponds with the catch, the changes in catch rates can be 

attributed to changes in fishing effort especially for the pair trawl fishery in the SWA, demersal, chambo and 

0.0 

Catch (tonnes)

171 


midwater fisheries. The same argument applies to the pair trawl fishery in 1980’s in the SEA but not between 1999 

and 2000. The catches for the last two years were around 185 tons annually as opposed to the expected catch of 448 

tonnes based on the Bcur. Therefore, the low catches were perhaps a result of other factors such as poor maintenance 

of the boats and inefficient fishing power of the engines. Currently, of the sixteen licensed pair trawl units only six 

are operational and their operational schedule is sporadic. 

Biomass 

The Surplus Production Models are fitted to catch and effort data with the assumption that the fishery is in 

equilibrium and catch rates decrease with increasing effort. There was no significant relation between cpue and 

effort in all large-scale commercial fisheries indicating that the models did not fit the data. The misfit was indicative 

that the fisheries are not in equilibrium perhaps because of frequent break down of fishing vessels and changes in 

fishing power, inaccuracy of catch and effort data submissions, fluctuations in stocks of dominant species such as C. 

virginalis and changes in species composition, mesh size, areas fished and depth. 

A comparison of the estimated MSY (Table 1) and the present catch landings (~ 180 and 440 t in the SEA and 

SWA, respectively for pair trawl fishery, 1570 t for stern trawl) show that the pair trawl and stern bottom trawl 

fisheries are operating below the MSY which means that there is no need for management recommendation. The 

fisheries resources of the SWA deep water (> 50 m) remain untapped because of the non-existence of the fishing 

activities in the area. A comparison of the estimated MSY and the actual landings by the pair trawlers however, need 

to be taken with caution because of differences in fishing powers. The RV. Ndunduma fishing power needs to be 

calibrated in relation to the pair trawlers. 

Unlike the swept area method, which can be applied only to the bottom trawl fisheries, the precautionary approach 

application is cross-cutting. The adoption of the precautionary approach indicates that there is currently no need for 

management recommendations for the pair trawl and mid water fishery but there is a need to consider reduction in 

effort of the stern bottom trawl and ring net. The recommendation of reduction in effort for the stern bottom trawl 

using the precautionary approach contradicts with that by the swept area method. This may be attributed to the fact 

that the cpue for the stern bottom trawl fishery does not represent mean Figure for the entire area because the fishery 

tends to be localised i.e. effort is not wide spread covering the entire area (Banda et al. 1996). 

In view of the above it is clear that there is competition among small and large scale fishers over various stocks such 

as Copadichromis virginalis, Oreochromis spp. Bagrus meridionalis etc. The pair trawl and midwater trawl fishery 

are not fully exploited. The pair trawl fishery needs rehabilitation for it to reach its maximum potential production 

while there is room for expansion for the midwater trawl fishery. The stern trawl fishery is currently fully exploited 

whereas the Chambo fishery is overexploited. 

Management 

The conventional management techniques for the sustainable utilisation for the commercial fisheries resources have 

been included in the Fisheries Conversation and Management Act 1997 and regulations include a limited number of 

licenses in demarcated areas, trawling in specific areas, time and depth and a minimum codend mesh size of 38 mm. 

The management of this fishery has been emphasized mainly by effort control and fishing area. However, the fishers 

have followed none of these regulations and this has been exacerbated by ineffective enforcement. All 

recommended licenses have been issued but very few are operational especially the pair trawl fishers. This is 

attributed to lack of capital for new licences and poor business management. Some people have had licenses for a 

very long time and have never fished. Equally, some fishers in the industry have failed to maintain their fishing 

fleet, which has led them to stop fishing. The factors have led to the reduction in fishing effort and decreased 

catching efficiencies of the fishing vessels. Of the eight and twelve licences allocated to the shallow waters of the 

SEA and SWA, respectively, only two are operational in each area. In contrast, the stern trawlers performance is 

better than that of the pair trawlers. All the existing stern trawlers are operational although they are fishing in the 

same areas of the SEA. Four licenses are recommended for this area but six are fishing there.). 

Despite the fact that each fishing vessel is licensed to fish in a specific area and depth range, pair trawlers usually 

fish in shallow of less than 6 and 20 m while stern trawlers start fishing in depth > 30 m. Some fishers have never 

fished in areas according to their licences and yet, the returns still indicate operation in the licenced area. One stern 

trawl is licensed for the deep waters of the SWA but fishes in the SEA. Experiments on gear selectivity indicate that 

clogging of the 38 mm mesh size happens within the first 10 minutes of trawling while that of the 50 mm mesh size

172 


in 15 minutes (Kanyerere, 2000). Consequently, the present recommended mesh size at the codend has been 

ineffective. 

In view of the above, there is a need for effective enforcement of fisheries regulations to reduce the conflict amongst 

different fishers and if effort is to be reduced. The control of the large-scale fisheries should be feasible because they 

are small and confined to a relatively small area. However, there is also an urgent need for detailed studies on stock 

identification and identification of resources overlaps between the small-scale and large- scale fisheries. 


Thanks to the staff of Malawi Fisheries Research Institute especially to Technical Assistants for commercial data 

compilation. 

References 

Banda, M.C. 2000. The biology and ecology of the catfishes of the Genera Bathyclarias and Bagrus in Lake Malawi. PhD. 

Thesis. University of Zimbabwe. 188pp. 

Banda, M.C. and T. Tomasson. 1997. Demersal fish stocks in southern Lake Malawi: Stock assessment and exploitation. 

Government of Malawi, Department of Fisheries. Fisheries Bulletin No.35. 

Banda, M.C., Tomasson, T. and Tweddle, D. 1996. Assessment of the deep water trawl fisheries of the south east arm of Lake 

Malawi using exploratory surveys and commercial data. In: Stock Assessment in Inland Fisheries. Cowx, I.G. (Ed.). 

Oxford: Fishing News Books, pp. 53-75. 

Bulirani, A., M.C. Banda, O.K. Palsson, O.L.F. Weyl, G.Z. Kanyerere, M.M. Manase and R.D. Sipawe 1999. Fish stocks and 

fisheries of Malawian waters: Resource report. Government of Malawi, Department of Fisheries, Fisheries Research 

Unit. 54 pp. 

FAO, 1993. Fisheries management in the south east arm of Lake Malawi, the Upper Shire River and Lake Malombe, with 

particular reference to the fisheries on chambo (Oreochromis spp.). FAO, Rome, CIFA Technical Paper, 21, 1-113. 

FAO, 1976. Promotion of intergrated fishery development, Malawi. Analysis of the various fisheries of Lake Malawi. FAO, 

FI:DP/MLW/77/516. Technical report 1. 

Kanyerere, G.Z. 1999. Trawl selectivity and the effect of clogging on selectivity in standard trawl surveys in Lake Malawi. . 

Government of Malawi, Department of Fisheries. Fisheries Bulletin No. 39. 

Manase, M. 2000. Traditional gear selectivity. (in press). 

Nyasulu, T.E. 2001. Resource use overlaps between pair-trawl and small-scale fisheries in the south east arm of Lake Malawi: 

preliminary results (Proceedings of this symposium). 

Palsson, O.K., M.C. Banda and A. Bulirani. 1999. Review of demersal monitoring surveys in southern Lake Malawi. 

Government of Malawi, Department of Fisheries. Fisheries Bulletin No.40. 

Pitcher, T.J. and Hart, P.J.B. 1982. Fisheries ecology. Croom Helm, Beckenham, 414 pp. 

Schaefer, M.B. 1954. Some aspects of the dynamic of populations important to the management of commercial marine fisheries. 

Bulletin of the Inter-American Tropical Tuna Commission.,1,27-56. 

Sipawe, R. 2001. Gear and species selectivity of the gillnet fishery in Lake Malawi. (Proceedings of this symposium). 

Tarbit, J. 1972. Lake Malawi trawling survey – Interim report 1969-1971. Malawi Fisheries Bulletin , (2), 16 pp. 

Turner, J. L. 1977. Some effects of demersal trawling in Lake Malawi (lake Nyasa ) from `968 to 1974. Journal of Fish Biology, 

10, 261-71. 

Turner, G.F. 1995. Management conservation and species changes of exploited fish stocks in Lake Malawi. In: The Impact of 

the Species in African Lakes. Eds. Pitcher, T.J. and Hart, P.J.B. Chapman & Hall, London. 

Turner, G.F., Tweddle , D. and Makwinja, R.D. 1995. Changes in demersal cichlid communities as a result of trawling in 

southern Lake Malawi. In: The Impact of the Species in African Lakes. Eds. Pitcher, T.J. and Hart, P.J.B. Chapman & 

Hall, London.

173 


Spatial and temporal distribution of some commercially important fish species in 

the southeast arm of Lake Malawi: A Geostatistical Analysis. 

Geoffrey Z. Kanyerere 1 & Anthony J. Booth 2 

1 Fisheries Research Unit, P.O. Box 27, Monkey Bay, Malawi. 

2 Dept. of Ichthyology & Fisheries Science, Rhodes University, Grahamstown 6140, South Africa. E.mail- t.booth@ru.ac.za 

Abstract 

The spatial and temporal distribution of the following fish species; Alticorpus mentale, Buccochromis lepturus, Copadichromis 

virginalis, Diplotaxodon elongate, and Oreochromis spp. in the southeast arm of Lake Malawi were analysed using bottom trawl 

catch per unit effort (CPUE) data collected during the annual demersal monitoring surveys of 1995 and 1999. Geostatistical 

techniques were used to model and estimate the spatial structure of abundance and ordinary kriging was used to predict local 

abundance. Small-scale intra-area variation that was detected was used to model the spatial structure. Experimental variograms 

were calculated and fitted using the spherical variogram model. Generally the structure of the variograms varied with population 

density. The results indicate that A. mentale occurs mainly in deep waters of Area C with density increasing slightly over the past 

four years from 13 kg nm -2 in 1995 to 15 kg nm -2 in 1999. However its distribution has shrunk especially in Area B over the same 

period. C. virginalis principally occurs along the inshore waters of southeast arm especially in Area C off Makanjila and in Area B off 

Masasa and Nkhudzi Bay on the western shore and off Kadango on the eastern shore. The species has declined in abundance in 

Area C. Density has decreased from 1,100 kg nm -2 in 1995 to 680 kg nm -2 in 1999. D. elongate is distributed widely occurring in 

offshore waters of southeast arm from Area A to Area C in relatively high densities. The species occurs in highest densities in Area 

B, the distribution in Area C is not as wide as it used to be. The density has declined from 20,000 kg nm -2 in 1995 to 14,000 kg nm -2 

in 1999. Oreochromis spp. occur mainly in Area A and to a smaller extent in shallow inshore waters of Area B. Over the past four 

years there has been a marked decrease in the distribution range with a corresponding decline in density from 319 kg nm -2 to 26 kg 

nm -2 in 1999. B. lepturus is evenly distributed in shallow waters of the southeast arm. The 1999 distribution pattern indicates that the 

species has slightly declined in abundance in the northeastern part of Area C and Area A. However density has increased from 37 

kg nm -2 in 1995 to 222 kg nm -2 in 1999. Although sampling was not originally designed specifically for geostatistics, these results 

indicate that geostatistics can be successfully used to detect changes in the spatial and temporal distribution of fish stocks or 

species in lake Malawi using the existing data. 

Introduction 

Lake Malawi, situated in the African rift valley between 9 o 30"S and 14 o 30"S and bordered by Malawi, Mozambique 

and Tanzania, is the third largest lake in Africa with an average depth of 292 m. The shallow Areas are the most 

productive and are found in the southeast and southwest arms of the lake (Patterson & Kachinjika, 1995). 

The fisheries on Lake Malawi are mainly distinguished by their degree of mechanisation and are classified into 

traditional and commercial components. The commercial fisheries which are relatively mechanised and capital 

intensive are dominated by stern and pair trawlers (Banda & Tomasson, 1996). Of all the fishing gears, bottom trawl 

nets are the most important. Pelagic trawl nets and pulse seine nets are seldom used. 

The commercial fishery was established in 1968 after successful experimental trials revealed the existence of large 

demersal stocks in the southern portion of the lake (Turner, 1976). All commercial fishing now occurs in the 

southern part of the lake. 

As the trawl fishery intensified, large changes in species composition were observed. Large cichlid species mostly 

belonging to the genus Lethrinops, clariid catfishes and Bagrus meridionalis have declined in abundance. This 

decline was followed by a corresponding increase in small cichlids most of which were Otopharynx spp. and 

Pseudotropheus spp. (Turner, 1976). Several follow-up projects like the “Demersal Fisheries Re-assessment Project 

(1989-1994)” were initiated in 1991 and one of the main findings of this project was that there was a 50-70% 

decline in biomass estimates in Areas A, B and C (Figure 1). Some changes in species composition were also noted, 

particularly in Area B where the relative contribution of some of the larger cichlid species to the catch had decreased 

(Turner et al, 1995). 

In view of these developments, the government of Malawi through the Department of Fisheries introduced 

standardised demersal surveys in 1994 to monitor the status of the stocks in the two arms of the lake. These surveys 

were initially conducted on quarterly basis but due to logistical problems and lack of adequate funding the frequency 

was reduced to two surveys a year. Recent information from these surveys and other sources indicate that the 

chambo (Oreochromis spp.) stocks in Lake Malawi have declined to the lowest level ever and as such it was 

proposed that all gears targeting Chambo be banned in Area A (Bulirani et al, 1999). Declining catch per unit effort

174 


(CPUE) is also observed for the stocks of bombe (Bathyclarias spp), kampango (Bagrus meridionalis) and utaka 

(Copadichromis spp.) but the fisheries for usipa (Engraulicypris sardella) and kambuzi (small non-catfish spp.) 

have remained relatively stable. The findings also indicate that despite the decline of some individual stocks, the 

overall CPUE for the deep-water stocks has remained relatively stable (Bulirani et al, 1999). 

The fact that the overall CPUE has remained relatively stable, despite the decline in CPUE of the large species 

suggests that the small species are gradually increasing in abundance at the expense of the larger ones. 

Unfortunately these small species fetch low market prices. Due to low catch rates of commercially important species 

and the increasing abundance of small species, most trawlers and other fishermen have resorted to targeting specific 

Areas in the lake where most of the species are caught in reasonable quantities. At present, the species that are 

highly sought after are chambo (Oreochromis spp.), utaka (Copadichromis spp.), ndunduma (Diplotaxodon spp.), 

ncheni (Rhamphochromis spp.), kampango (Bagrus spp.) and bombe (Bathyclarias spp). 

Concentrating fishing effort on a particular stock can have serious management implications especially in cases 

where the stock involved is localised and has a low fecundity. For instance, most mouth brooders produce relatively 

few offspring because of limited brooding space. Most of the Malawi cichlids fall into this category. In situations of 

intense fishing pressure and the use of small-meshed gears, chances of such stocks being depleted are high. At this 

point, it is worth noting that the 38 mm cod-end currently in use on the Lake Malawi trawl fishery is too small, and 

most of the demersal species being caught are immature (Kanyerere, 1999). Such factors have undoubtedly 

contributed to the depletion of stocks in the lake. 

The decline of some stocks within the southern portion of the lake suggests that the fishery is already facing serious 

problems, which might include intense and localised fishing pressure, growth over-fishing and recruitment failure. It 

is imperative that regulatory management procedures such as closed seasons and aggregate quotas, controlled Area 

fishing and restrictions on fishing gears and technology be instituted where necessary. However, identification and 

formulation of such regulations require that; 

i) Areas or localities where significant reduction in CPUE and species abundance has occurred be 

identified. 

ii) those Areas where commercially important species are found and what changes the species might 

have undergone be identified. 

iii) temporal changes of abundance for species with wide distributions and those with localised 

distributions in relation to Area, depth and time be assessed. 

iv) maps of species abundance by Area, depth and time be produced. 

The spatial and temporal distribution of commercial species needs to be assessed over a suitable time period. This 

can only be accomplished through the use of geostatistics as no traditional method is suitable for such assessments. 

Geostatistics takes the spatial autocorrelation between samples into consideration and through kriging, allows the 

analysis and modelling of the variability of a population in space (Freire 1992). Ignoring spatial patterns when using 

catch per unit effort (CPUE) data to estimate stock abundance can sometimes lead to inaccurate assessments 

(Pelletier and Parna, 1994). As geostatistics is a relatively new technique to the Malawian fisheries, this study is 

preliminary and will investigate the feasibility of using geostatistics: 

- to model the abundance, spatial and temporal distribution of Alticorpus mentale, 

Buccochromis lepturus, Copadichromis virginalis, Diplotaxodon elongate and 

Oreochromis spp. and 

- to assess if the existing sampling locations meet the requirements of 

geostatistical analysis. 


Sampling sites 

The data were collected onboard the research vessel, R.V. Ndunduma using demersal trawl gear. A total of 54 and 43 

fixed stations are sampled bi-annually in the southeast and southwest arms of Lake Malawi respectively (Figure 2). 

The southeast arm is divided into Areas A, B and C and the southwest arm into Areas D, E and F for management 

purposes (Figure1). A Global Position System (GPS) was used for fixing the position of shooting and hauling for 

each station, with each trawl lasting 30 minutes.

latitude 

Map of Southern Lake Malawi showing Area boundaries 

F 

175 


Figure 1. Area boundaries for both southeast and southwest arms of Lake Malawi. 

-14.4 -14.2 -14.0 -13.8 

latitude 

-13.60 

-13.70 

-13.80 

-13.90 

Chipoka 

-14.00 

-14.10 

-14.20 

-14.30 

-14.40 

Chilambula 

E 

D 

Malembo 

C 

Monkey Bay 

Figure 2. Map of southern Lake Malawi showing sampling stations in southeast (SEA) and southwest 

(SWA) arms. 

Data Collection 

All trawling was conducted during the day (between 6 am and 5 pm) and for each station depth (m), trawling speed, 

actual time trawled and position at start and end of each haul (latitude and longitude) were recorded. A total of 8 

surveys covering the southeast and southwest arms of lake Malawi have been conducted between 1995 and 1999. 

Data for only 1995 and 1999 was, however, used in this analysis. 

The sampling and recording of the catch closely followed guidelines outlined by Sparre et al, (1989). All catfish of 

the genera Bathyclarias, Clarias and Bagrus as well as extremely large cichlid and non-cichlid species were sorted 

out of the main catch. Thereafter the catch was sub-sampled; the fraction of which depended on the quantity of catch 

landed. The catch was then divided on deck into four categories namely; small fish (mainly cichlids but also 

including Synodontis njassae), Bagrus meridionalis, Clariid catfishes and all large cichlid and non-cichlid fishes. 

Each category was sorted into species, length measured to the nearest half centimetre and weighed. 

B 

Makaw a 

Makanjira 

A 

N 

Kadango 

-14.50 

34.40 34.50 34.60 34.70 34.80 34.90 35.00 35.10 35.20 35.30 35.40 

Chipoka 

Chilambula 

S W A 

Malembo 

longitude 

S E A 

Monkey Bay 

Makawa 

Makanjira 

34.6 34.8 35.0 35.2 

longitude 

N 

Kadango

Data analysis 

176 


Geometric mean 

From the trawl data, estimates of the percentage composition and CPUE for A. mentale, B. lepturus, C. virginalis D. 

elongate and Oreochromis spp. were obtained. The geometric mean (GM) was used to calculate mean CPUE. This 

estimation was chosen over the arithmetic mean because preliminary analysis showed that the data had an 

asymmetric right-skewed distribution. It was therefore assumed that abundance was also log-normally distributed. 

The GM was obtained by calculating the mean across stations of the log-transformed values and back transforming 

(Rosner, 1995) such that; 

n 

∑ 

( CPUES 

+ 1) 

1 log 

n 

GM = 10 s= 

1 

−1 

where CPUEs is the catch per unit effort of station s. 

Geostatistical analysis 

The abundance of a species measured at fixed locations in a lake or ocean is known as random field data and such 

data is suitable for geostatistical analysis. Geostatistical data often exhibit small-scale variation that may be 

modelled as spatial correlation and incorporated into estimation procedures. Spatial variability is modelled as a 

function of the distance between the sampling sites, where the sites closer together in space have more similar data 

values than those that are far apart. The variogram provides a measure of such correlation by describing how sample 

data are related with distance and direction (Kaluzny et al, 1998). 

Exploratory data analysis 

In two-dimensional space, a sampling location is defined by a longitude and latitude position. In order to 

successfully model an underlying random spatial process it must be assumed that: 

i) the observation and the spatial process differ only through white-noise measurement error and that 

ii) the spatial distribution of the species in question was stable throughout the survey period so that 

all CPUE observations reflect the same underlying spatial process (Pelletier and Parma, 1994). 

Although fish movement might make the second assumption difficult to satisfy, Lake Malawi fishes are typically 

resident and therefore, at a broader spatial scale, their distribution can be considered relatively stable. The spatial 

process in geostatistical data can be decomposed into a large-scale deterministic component and a small-scale 

stochastic component with the random field in this case not having a constant mean. As the existence of the 

variogram is based on a process with a constant mean and variance defined only through the magnitude of distance, 

then a variogram based on a random field with both large-scale trend and small-scale random variation will not meet 

the necessary assumption (Kaluzny et al, 1994). This implies that the first step in geostatistical data analysis is the 

detection and removal of trend from the data before using the variogram to estimate the underlying random process. 

Detecting and removing spatial trends 

Procedures used in this study for purposes of detecting and removing trend from data are those outlined in Kaluzny 

et al (1998). They include: 

i) rotation of the longitude and latitude axes to assess spatial invariance. 

ii) modelling the logged data as a smooth function of the longitude and latitude using a generalised 

additive model (GAM) and 

iii) fitting a local regression model (loess) to the whole trend surface. 

Residuals from this model are later on used to form kriging predictions.

177 


Trend in the southeast arm 

In the southeast arm the major trend was generally from south east to northwest. The angle of rotation was positive 

and therefore towards the north (90 o ) and ranged between 16 o and 45 o . 

Analysis 

The empirical variogram provides a description of how the data are correlated with distance. The semi-variogram 

function, γ(h), is defined as half the average squared difference between points separated by a distance h (Kaluzny et 

al, 1998) and is described as: 

where N(h) is the set of all pairwise Eucledian distances h = i - j, |N(h)| is the number of distinct pairs in N(h), and zi 

and zj are data values at spatial locations i and j, respectively. The letter h represents a distance measure with 

magnitude only but when direction is also considered, it becomes a vector h ~ (Kaluzny et al, 1998). 

A variogram has at least three parameters namely the nugget effect, sill and range. The nugget effect represents 

micro-scale variation or measurement error and is estimated from the empirical variogram as the value of γ(h) for h 

= 0. The sill is the variance of the random field while the range is the distance (if any) at which data are no longer 

autocorrelated. 

Additional parameters used in this study to customise the variograms were the maximum distance over which the 

variogram is calculated, minimum number of pairs used for calculating the variogram of which 30 is the minimum, 

lag and number of lags. The variograms were estimated using the robust variogram estimation method as it has the 

advantage of reducing the effect of outliers without removing specific data points from a data set (Kaluzny et al, 

1998). The robust estimation method is based on the fourth power of the square root of absolute differences as 

follows: 

1 

| zi−zi| { ∑ Nh ( ) } 

2| 

Nh ( )| 

γ ( h) 

= 

0. 457 + 0. 944/| 

Nh ( )| 

1 2 4 

Modeling the Empirical Variogram 

Anisotropy occurs when the spatial autocorrelation of a process changes with direction. A variogram from an 

anisotropic process is not purely a function of the distance h, but is a function of both the magnitude and direction of 

h ~ (Kaluzny et al, 1998). An anisotropic variogram is geometrically anisotropic if: 

2γ(h) = 2γ o (||Ah||), h ∈ℜ d 

1 

γ ( h) 

= ∑ z 

| Nh ( )| i − z 

2 

j 

N( h) 

where A is a d x d matrix and 2γ 0 is a function of a real variable. 

( ) 

In this situation, the Eucledian space is not appropriate for measuring distance between locations, but a linear 

transformation of it is (Cressie, 1993). Since variograms are only valid for an isotropic process, anisotropy has to be 

corrected before fitting a theoretical model to the empirical one. In this study geometric anisotropy (range of the 

2

178 


variogram changing in different directions, while sill remains constant) was identified by using directional 

variograms and was subsequently corrected by a linear transformation of the lag vector h ~ . 

After estimating the parameters of the empirical variogram, a theoretical model is then fitted to it. This is necessary 

because it ensures that the variance of predicted values is positive. This study makes use of the spherical model. 

Other models tested besides the spherical were exponential, gaussian, linear and power. Its choice was mainly based 

on the observation that it was the one that fitted well with the shape of the empirical variogram. Besides this, Freire 

et al (1992) state that the spherical model is the most common in the analysis of animal populations and geostatistics 

in general. According to Cressie, (1993) the spherical model has the form: 

⎧ 0, 

⎪ 

γ ( h; 

θ ) = ⎨CO 

+ C 

⎪ 

⎩CO 

+ C 

S 

S 

{( 3 

2)( 

|| h || a ) − ( 1 2)( 

|| h || a ) 

where θ = (co,cs,as)', co ≥ 0, cs ≥ 0, as ≥ 0 and C0 is the nugget effect, due to the variability between samples, the 

microstructure which remains undetected because of the size of the sample, or errors in measurement or location. Cs 

represents the sill-nugget effect, where the sill is the asymptotic value of semivariance, reached with a value of h = 

a, the range, which represents the maximum distance at which spatial effects are detected. 

The spherical model was fitted to the empirical variogram by minimising the residual sum of squares between the 

theoretical model and the empirical variogram using the weighted least-squares estimation as follows: 

K 

∑ 

j= 

1 

⎧ γ ( h( 

j)) 

⎫ 

| N( 

h( 

j)) 

| ⎨ −1⎬ 

⎩γ 

( h( 

j); 

θ ) ⎭ 

where K is the number of lags, θ = (co,c,a), γ(h(j)) = spherical variogram model and 

= empirical variogram. 

γ 

( h( 

j)) 

s 

2 

A summary of variogram parameters that were used in this analysis is displayed in Table 4. The prediction process 

also requires the values that are used for correcting anisotropy in addition to the range, sill, and nugget. Parameters 

used for correcting geometric anisotropy in this study were angle and ratio. The ratio ranged from 1.25 to 2 while 

the angle ranged from 45 o to 180 o . These were different for each species and year. 

Spatial prediction using kriging 

Kriging is a linear interpolation method that allows predictions of unknown values of a random function from 

observations at known locations. Kriging incorporates a model of the covariance of the random function when 

calculating predictions of the unknown values. There are two categories of kriging namely universal and ordinary 

(Kaluzny et al, 1998). Ordinary kriging uses a random function model of spatial correlation to calculate a weighted 

linear combination of available samples, for prediction of abundance and standard errors for unsampled locations. 

Weights for this model are chosen to ensure that the average error for the model is zero and that the modelled error 

variance is minimised. 

In this analysis, ordinary kriging was used to predict the value of the spatial process S(x) for every location x 

(latitude, longitude) in the Area, from a linear combination of the observed values {Z(xi), i = 1,…,g}. Based on the 

assumption that Z(x) and S(x) differ only through measurement error (Pelletier and Parma, 1994), it is possible to 

model the spatial covariance of Z(x) directly instead of that of S(x). Ordinary kriging requires the stationarity of the 

first differences of Z(x) as demanded by the intrinsic hypothesis of Matheron (Cressie, 1993) such that: 

s 

3 

}, 

h = 0 

0 < || h || ≤ a , 

|| h || ≥ a 

s 

s

Ε(Z(x + h ~ ) - Z(x)) = 0, 

V(Z(x + h ~ ) - Z(x)) = 2γ(h) 

179 


where h ~ is a vector of length h = |h|. The first equation means that Z(x) has the same expected value over the whole 

Area regardless of location. In ordinary kriging, this value is assumed to be unknown. In the second equation, the 

variance of the difference is a function of h ~ only. If second-order stationarity is assumed (as is usual for kriging) 

such that: 

Cov(Z(x + h), Z(x)) = C(h), 

the covariance between two points is only a function of their relative position, the semivariogram can then be 

expressed in terms of the spatial covariance as: 

γ(h) = δ 2 - C(h) 

where the sill δ 2 is the variance of Z(x) in the model (Pelletier and Parma, 1994). 

Thus the spatial correlation structure of Z(x) is characterised by the variogram, γ(h). The kriging variable was the 

residuals from the spatial loess model, while spatial locations were the rotated coordinates. The spatial correlation 

was modelled as spherical covariance based on the spherical variogram model fitted to the empirical variogram. The 

sill used was specified as the sill minus the nugget effect. 

Results 

Observed CPUE and catch composition 

Presented in Table 2 is a summary of CPUE for all five species. 

Table 2. CPUE (kg. 0.5hr -1 ) for each of the species in southeast arm of lake Malawi for 1995 and 1999. 

South East Arm (CPUE- kg 0.5 hr -1 ) 

Species 1995 1999 

A. mentale 0.93 1.00 

C. virginalis 7.79 2.18 

Oreochromis spp. 1.58 0.59 

B. lepturus 0.54 0.61 

D. elongate 2.65 4.65 

In Table 3 is a summary of the percentage composition for each species presented separately for each Area. 

Table 3. Percentage composition in the total catch for each of the species in southeast arm of lake 

Malawi. 

A B C 

Species 1995 1999 1995 1999 1995 1999 

A. mentale 0.00 0.00 0.06 0.20 1.55 2.1 

C. virginalis 0.10 15.26 30.87 14.50 28.74 0.91 

D. elongate 0.32 4.28 7.01 21.46 1.48 8.80 

Oreochromis spp. 31.22 11.78 0.51 0.29 0.02 0.00 

B. lepturus 0.10 0.41 0.36 0.48 0.24 0.53

180 


Variograms 

Table 4 presents summaries of variogram parameters for both the southeast and southwest arms of lake Malawi 

respectively. 

Table 4. Variogram parameters for species analysed from the southeast arm of lake Malawi for 1995 and 

1999. 

Southeast Arm -1995 Southeast Arm - 1999 

Species Nugget Sill Range Nugget Sill Range 

A. mentale 0.005 0.33 0.19 0.005 0.36 0.28 

C. virginalis 0.005 5.2 0.18 0.005 0.75 0.2 

Oreochromis spp. 0.005 0.21 0.2 0.005 0.04 0.22 

B. lepturus 0.005 0.13 0.36 0.005 0.19 0.3 

D. elongate 0.005 2.35 0.18 0.005 2.8 0.19 

Table 5. Changes in density for each of the species in southeast arm of lake Malawi for 1995 and 1999. 

Density (kilogrammes/nm -2 ) 

Species 1995 1999 

A. mentale 13 15 

C. virginalis 1100 680 

Oreochromis spp. 319 26 

B. lepturus 37 222 

D. elongate 20000 14000 

Results presented correspond to isotropic variograms. 

A Mentale 

A. mentale attains a maximum size of about 25 cm total length (TL) and attains maturity at a relatively large size. It 

occurs in water depths of between 60 and 128 m (Turner, 1996). It is of significance to deep water bottom trawling 

where it contributes about 2% to the total catch (Table 3). 

The 1995 and 1999 variograms for A. mentale differ (Figure 3, Table 4). The range at which the data is no longer 

spatially correlated is higher for 1999 (0.28 nm) than for 1995 (0.19 nm) suggesting that a change in the distribution 

pattern of the species has occurred. The sill value for 1995 (0.33) is lower than that of 1999 (0.36) suggesting that 

there was more variance in the random field in 1999 than in 1995. Both observations indicate that a change in the 

density and distribution pattern of the species has occurred over the past four years. 

The distribution maps in Figure 4 indicate that A. mentale occurs mainly in the deep waters of Area C and the 

distribution seems to have generally shrunk over the past four years as indicated by the 1999 distribution map. 

However density has slightly increased from 13 kg nm -2 in 1995 to 15 kg nm -2 in 1999 (Table 5). This agrees with 

the observed increase in percentage composition from 1.6% in 1995 to 2.1% (Table 3) and the increase in CPUE 

from 0.93 kg 0.5 hr -1 in 1995 to 1.00kg 0.5 hr -1 in 1999 (Table 2). 

gamma 

0.0 0.2 0.4 0.6 0.8 

Mentale 1995 

0.0 0.1 0.2 0.3 

distance 

objective = 0.64 

Figure 3. Variograms for A. mentale in the southeast arm of lake Malawi. 

0.0 0.2 0.4 0.6 0.8 1.0 1.2 

gamma 

Mentale 1999 

0.0 0.1 0.2 0.3 0.4 

distance 

objective = 1.2

-32.6 -32.4 -32.2 -32.0 -31.8 

Latitude 

181 


Figure 4. A. mentale abundance and distribution patterns during 1995 and 1999 in the southeast arm of 

Lake Malawi. 

C. virginalis 

C. virginalis is a small shoaling species that is very abundant on steep rocky coasts or submerged rock reefs of Lake 

Malawi. The species attains a maximum size of about 11-15 cm TL. It has been recorded at depths ranging from 9 m 

to 74 m. It is a very important species to a variety of mechanised and artisanal fisheries (Turner, 1996). 

gamma 

Mentale 1995 

Density (kg/sq. nm) 

0 2 4 6 8 10 

13 

0 

19.1 19.2 19.3 19.4 

Longitude 

19.5 19.6 19.7 

virginalis 1995 

objective = 53 

distance 

0.00 0.05 0.10 0.15 0.20 0.25 

Figure 5. Variograms for C. virginalis in the southeast arm of lake Malawi. 

0.0 0.5 1.0 1.5 2.0 

gamma 

36.2 36.4 36.6 

Longitude 

36.8 37.0 37.2 

The 1995 and 1999 variograms for C. virginalis differ (Figure 5, Table 4). The range at which the data is no longer 

spatially correlated is higher for 1999 (0.2 nm) than for 1995 (0.18 nm) suggesting that a change in the distribution 

pattern of the species has occurred. The sill value for 1995 (5.2) is much higher than that of 1999 (0.75) suggesting 

that there was more variance in the random field in 1995 than in 1999. Both observations indicate that a change in 

the density and distribution pattern of the species has occurred over the past four years. 

The distribution maps in Figure 6 indicate that C. virginalis occurs in Areas A, B & C. However the distribution 

seems to have generally shrunk over the past four years especially in Areas B and C as observed from the 1999 

distribution map. Density has decreased from 1100 kg nm -2 in 1995 to 680 kg nm -2 in 1999 (Table 5). This agrees 

with the observed decrease in percentage composition from 30.87% in 1995 to 14.5% in Area B and from 28.74% in 

1995 to 0.91% in 1999 in Area C (Table 3). CPUE has likewise decreased from 7.79 kg 0.5 hr -1 in 1995 to 2.18 kg 

0.5 hr -1 in 1999 (Table 2). 

8.2 8.4 8.6 8.8 

Latitude 

Mentale 1999 


Density (kg/sq.nm) 

0.0 0.1 distance 0.2 0.3 

15 

0 

objective = 2.9

182 


Figure 6: C. virginalis distribution pattern and abundance during 1995 and 1999. 


In Lake Malawi, the genus Oreochromis is composed of four species of which three are endemic to the lake. The 

endemic ones are O. squamipinnis (37 cm TL), O. karongae (38 cm TL) and O. lidole (37 cm TL) while O. shiranus 

(37 cm TL) is the only non-endemic species (Turner, 1996). They occur mainly in shallow waters and have been 

recorded at depths ranging from 2 m to 50 m although they are most abundant at depths greater than 20 m (Palsson 

et al, 1999). 

The 1995 and 1999 variograms for Oreochromis spp. differ (Figure7, Table 4). The range at which the data is no 

longer spatially correlated is slightly higher for 1999 (0.22 nm) than for 1995 (0.2 nm) suggesting that a change in 

the distribution pattern of the species has occurred. The sill value for 1995 (0.21) is much higher than that of 1999 

(0.04) suggesting that there was more variance in the random field in 1995 than in 1999. Both observations indicate 

that a change in the density and distribution pattern of the species has occurred over the past four years. 

0.0 0.2 0.4 0.6 0.8 

gamma 

- 

31 

- 

32 

m - 

ap 32 

2[, 

2] 

- 

32 

- 

32 

virginalis 

19. 19. 19. 19. 

map2[, 

19. 19. 19. 

oreochromis 1995 

Density (kg/sq. 

) 

1100 

0 


0.00 0.05 0.10 0.15 0.20 0.25 

distance 

0.00 0.05 0.10 0.15 

distance (nm) 

0.20 0.25 0.30 

Figure 7. Variograms for Oreochromis spp in the southeast arm of Lake Malawi. 

The distribution maps in Figure 8 indicate that Oreochromis spp. occurs mainly in Area A and parts of Area B. 

However the distribution has generally shrunk over the past four years especially in Areas A as observed from the 

1999 distribution map. Density has likewise decreased from 319 kg nm -2 in 1995 to 26 kg nm -2 in 1999 (Table 5). 

This agrees with the observed decrease in percentage composition from 31.22% in 1995 to 11.78% in Area A and 

from 0.51% in 1995 to 0.29% in 1999 in Area B. CPUE also decreased from 1.58 kg 0.5hr -1 in 1995 to 0.59 kg 0.5 

hr -1 in 1999. 

0.00 0.05 0.10 0.15 0.20 

gamma 

-32.6 -32.4 -32.2 -32.0 -31.8 

Latitude 


19.1 19.2 19.3 19.4 19.5 19.6 19.7 

Longitude 


Density (kg/sq. nm) 

680 

0 

objective = 0.0395

-33.0 -32.8 -32.6 -32.4 -32.2 

latitude 


183 


Figure 8. Oreochromis spp. abundance and distribution patterns during 1995 and 1999. 

B. lepturus 

B. lepturus like all other members of the genus is a large heavily-built species that attains a maximum total length of 

about 42 cm. It is a shallow water species, and has often been sampled from depths ranging from 10 to 26 m 

(Turner, 1996). It is probably one of the most delicious fishes of the lake although it is not found in large numbers. 

The 1995 and 1999 variograms for B. lepturus differ (Figure 9, Table 4). The range at which the data is no longer 

spatially correlated is slightly higher for 1999 (0.22 nm) than for 1995 (0.2 nm) suggesting that a change in the 

distribution pattern of the species has occurred. The sill value for 1995 (0.21) is much higher than that of 1999 

(0.04) suggesting that there was more variance in the random field in 1995 than in 1999. Both observations indicate 

that a change in the density and distribution pattern of the species has occurred over the past four years. 

The distribution maps in Figure 10 indicate that B. lepturus occurs mainly in the shallow waters of Areas B and C. 

However the distribution has shrunk over the past four years in both Areas as observed from the 1999 distribution 

map. 

Density has however increased from 37 kg nm -2 in 1995 to 222 kg nm -2 in 1999 (Table 5). This agrees with the 

observed increase in percentage composition from 0.36% in 1995 to 0.48% in 1999 in Area B and from 0.24% in 

1995 to 0.53% in 1999 in Area C. CPUE also increased from 0.54 kg 0.5 hr -1 in 1995 to 0.61 kg 0.5 hr -1 in 1999. 

0.00 0.05 0.10 0.15 0.20 0.25 0.30 

gamma 

Lepturus 1995 

Density(Kg sq. nm) 

319 

0 

18.6 18.7 18.8 18.9 

longitude 

19.0 19.1 19.2 


0.0 0.1 0.2 0.3 

distance 

0.4 0.5 

ga 

mm 0.3 

a 

Figure 9. Variograms for B. lepturus in the southeast arm of lake Malawi. 

-23.0 -22.8 -22.6 -22.4 -22.2 

latitude 

0.5 

0.4 

0.2 

0.1 

0.0 


Lepturus 1999 

Density (Kg sq. nm) 

26 

0 

29.8 30.0 30.2 30.4 

longitude 

0.0 0.1 0.2 0.3 0.4 

distance 

objective = 0.18

-30.8 -30.6 -30.4 -30.2 -30.0 

Latitude 

184 


Figure 10. B. lepturus abundance and distribution patterns during 1995 and 1999 in the southeast arm of 

Lake Malawi. 

D. elongate 

Members of the genus Diplotaxodon belong to the "Ndunduma" group of fishes that are principally deep-water 

piscivores and zooplanktivores with upwardly angled mouths. D. elongate and its relatives are truly pelagic 

zooplankton feeders, which occupy the feeding niche of the "Utaka" (Copadichromis) in offshore waters. The 

species attains a total length (TL) of about 19 cm. It has been recorded throughout the pelagic zone of the lake from 

the surface to a depth of at least 220 m (Turner, 1996). Commercially, it is also an important species to both the 

commercial and artisanal fisheries. 

The 1995 and 1999 variograms for D. elongate differ (Figure 11, Table 4). The range at which the data is no longer 

spatially correlated is slightly higher for 1999 (0.19 nm) than for 1995 (0.18 nm) suggesting that the distribution 

pattern of the species has not changed much. However the much higher sill value for 1999 (2.8) than that of 1995 

(2.35) suggests that there was more variance in the random field in 1999 than in 1995. The increased variance in the 

random field indicates that a change in the density of the species has occurred over the past four years. The 

distribution maps in Figure 12 indicate that D. elongate occurs throughout the pelagic waters with highest 

concentration occurring in Area B. 

0 1 2 3 

gamma 

Lepturus 1995 

elongate 1995 


37 

0 

21.9 22.0 22.1 

Longitude 

22.2 22.3 22.4 

0.0 0.1 0.2 0.3 

distance 


Figure 11. Variograms for D. elongate in the southeast arm of lake Malawi. 

-30.8 -30.6 -30.4 -30.2 -30.0 

Latitude 

0 1 2 3 4 

gamma 

Lepturus 1999 

elongate 1999 


222 

0 

21.9 22.0 22.1 

Longitude 

22.2 22.3 22.4 

objective = 19 

0.0 0.1 0.2 

distance 

0.3 0.4

-30.8 -30.6 -30.4 -30.2 -30.0 

Latitude 

elongate 1995 

Density (Kg. sq. nm) 

20000 

0 

21.9 22.0 22.1 

Longitude 

22.2 22.3 22.4 

185 


21.9 22.0 22.1 

Longitude 

22.2 22.3 22.4 

Figure 12. D. elongate abundance and distribution patterns during 1995 and 1999 in the southeast arm of 

lake Malawi. 

As observed from the 1999 distribution map the species has disappeared in shallow waters of Areas B and C but has 

instead extended into Area A and deeper parts of Area C. This agrees with the increase in percentage composition 

from 0.32% in 1995 to 4.28% in 1999 in Area A, from 7.01% in 1995 to 21.46% in 1999 in Area B and from 1.48% 

in 1995 to 8.8% in 1999 in Area C (Table 3). This is also reflected in the overall increase in CPUE from 2.65 kg 

0.5hr -1 in 1995 to 4.65 kg 0.5 hr -1 in 1999 (Table 2). Density has however decreased from 20000 kg nm -2 in 1995 to 

14000 kg nm -2 in 1999 (Table 5). 

Discussion 

Stock assessment methods that disregard spatial distribution of species may provide inaccurate information pertinent 

to management. This is so because such methods usually assume that the existing density Areas or strata are 

internally homogeneous unless the sampling design is perfectly random (Freire et al, 1992). Having the ability to 

map abundance gradients and integrating them into space, geostatistics gives a more realistic view of a species 

distribution. This information can also be used to trace the spatial and temporal allocation of fishing effort for both 

effort-controlled and open-access fisheries. 

All the five species assessed in this study show different distribution patterns. This is expected because distribution 

patterns are in most cases linked to food availability, water temperature, dissolved oxygen, bottom sediment type 

and other ecological factors such as presence or absence of aquatic vegetation and submerged rocky reefs. The 

distribution pattern can at times be highly influenced by human interference such as excessive exploitation that can 

bring about localised extinction of a species or group of species. 

Alticorpus mentale 

The results from this study indicate that A. mentale mainly occurs in deep waters of both the southwest and 

southeast arms. In southeast arm it mainly occurs in deep waters of Area C. It is however more abundant in 

southwest arm than in southeast arm. The restricted distribution range greatly endangers the continued existence of 

the species in the southeast arm because of high fishing pressure from deepwater demersal trawlers. However, the 

results from the same southeast arm indicate that the species has gradually increased in abundance over the period 

from 1995 to 1999 despite intense fishing pressure. This can only be attributed to the fact that because deepwater 

species have a poor market value, fishing pressure has over the years been redirected to high value shallow to 

medium water-depth (20-70m) species such as C. virginalis, D. elongate and Oreochromis spp. among others. The 

decline in the catch rates and distribution of species such as C. virginalis, D. elongate and Oreochromis spp. bears 

testimony to the redirection of fishing effort away from deep waters. 

Copadichromis virginalis 

This study has shown that C. virginalis occurs along the shores of the southeast arm especially in Area C off 

Makanjila, and in Area B off Masasa, Chirombo and Nkhudzi bay on the western shore and to a smaller extent on 

eastern shore just off Kadango fishing village (Figure 1). The results based on the 1999 data indicate that the species 

has greatly declined in abundance in Area C but there is instead a marked increase in abundance of the species in 

Area A. 

-30.8 -30.6 -30.4 -30.2 -30.0 

Latitude 

elongate 1999 

Density (Kg. sq. nm) 

14000 

0

186 


The decline in the abundance of this species in Areas C and B can be attributed to the uncontrolled fishing pressure 

that is exerted on the species from both the artisanal, commercial and semi-commercial fisheries. Fishers along the 

shores of the southeast arm know exactly where to fish for this species resulting into localised declines in abundance 

within its preferred grounds while the abundance has increased in Areas that are less known to fishers. An example 

here is Area A. 

Diplotaxodon elongate 

The results from this survey indicate that D. elongate is a widely distributed species occurring in offshore waters of 

the southeast arm from Area A to Area C in relatively high densities. The species occurs in highest densities in Area 

B. As observed from the distribution pattern for both 1995 and 1999 respectively, the species has over the period 

1995-1999 increased in abundance with much higher densities occurring in Area B and partly extending into the 

southern portion of Area C. However its distribution range has decreased over the period between 1995 and 1999. 

D. elongate being a pelagic species, is not always available to bottom trawlers and therefore suffers relatively less 

fishing mortality than other species that are permanently demersal. There is also relatively less fishing pressure from 

the lone midwater trawler operating in Area B which occasionally lands about 53% of D. elongate as bycatch 

(Turner, 1996). Although there are indications that the current distribution is not as wide as it used to be especially 

in Area C, the species unlike the others, is not greatly affected by fishing activities. It seems to be more or less 

governed by the existing environmental conditions. 


These species are commonly called "Chambo" and have for decades been the mainstay of both the commercial and 

artisanal fisheries in southern Lake Malawi. Results from this study indicate that Oreochromis spp. occur mostly in 

Area A and to a smaller extent in shallow inshore waters of Area B. The species have greatly declined in abundance 

and their distribution range has narrowed very much. 

The drastic decline in the abundance and distribution of the species has mostly been blamed on localised and 

excessive fishing pressure from both commercial and artisanal fisheries over the years. Much of the blame has been 

laid on the light attraction fishery commonly known as Kauni, beach seines, and undermeshed gillnets. To redress 

the situation, it was recently proposed that beach seines and Kauni fisheries be banned in Area A (Bulirani et al, 

1999). 

The situation in southeast arm requires urgent intervention from the fisheries department because if nothing is done 

to curb the current trends in fishing practices, the remaining few pockets will also disappear. 

Buccochromis lepturus 

The results from this study indicate that in the southeast arm, B. lepturus is evenly distributed in shallow waters of 

Areas B and C. Abundance has generally increased as manifested by the higher biomass in 1999 than in 1995. 

However by 1999 the species had disappeared from the eastern part of Area C and the southern most portion of the 

lake (Area A). Thus the distribution range has decreased. The general decline in abundance of this species in parts of 

Area C and Area A can be attributed to excessive fishing effort in these Areas mainly because of localised fishing 

effort in Area A for Oreochromis spp. and in Area C for C. virginalis. 

Conclusions 

This study has demonstrated that geostatistics can be successfully used to analyse the existing CPUE data in Lake 

Malawi to detect changes in the spatial and temporal distribution of fish stocks. The reasonably low nugget effect 

(0.005 max) indicates that the micro-scale variation or measurement error is small for all the species analysed. 

Biomass estimates are generally good for species found offshore or in deep water. This is because during kriging, 

this part of the water body is analysed more thoroughly than shallow waters. However, there is another method that 

can be used to improve biomass estimates for shallow water species. This method was not employed in the current 

analysis because a lot of time is required to manually define the Area to be kriged. 

In order to enhance the efficacy of this new technique, the number of stations needs to be increased in southeast arm 

especially in Area B. The seven stations in Area A need to be evenly redistributed.

187 


Generally, all the species studied show localised occurrence in various Areas of the lake, and as such they are very 

vulnerable to overfishing and are in imminent danger of facing localised extinction. To protect these stocks it is 

suggested that stock-specific management measures be constituted and implemented as soon as possible. Options to 

be considered include closed seasons and aggregate quotas, controlled-Area fishing, restrictions on fishing gear and 

technology.

188 



I would like to thank ICEIDA (Malawi) for sponsoring my studies at Rhodes University and my project supervisor, 

Dr. A. Booth, who untiringly guided me through all the data analysis and write-up. 

References 

Banda, M.C. and Tomasson, T. (1996). Surveys of Trawling Grounds and Demersal Fish Stocks in Central Lake Malawi from 

Domira Bay to Nkhata Bay in 1994 and 1995. Government of Malawi, Fisheries Department. Fisheries Bulletin No. 

33,34 pp. 

Bulirani, A.E., Banda, M.C., Palsson K.P., and Weyl, O.L.F. (1999). Fish Stocks and Fisheries of Malawian Waters. Resource 

Report 1999. Government of Malawi, Fisheries Department. 54pp. 

Cressie, N.A.C. (1993). Statistics for Spatial Data. John Wiley & Sons, New York. 900 pp. 

Freire, J., Gonzalez-Gurriaran, E. & Olaso, I.(1992). Spatial Distribution of Munida intermedia and M.sarsi 

(Crustacea:Anomura) on the Galician Continental Shelf (NW Spain): Application of Geostatistical Analysis. Estuarine, 

Coastal and Shelf Science (1992) 35, 637-647. 

Gulland, J.A. (1983). Fish Stock Assessment: A Manual of Basic methods. Volume 1. FAO, Rome. 223 pp. 

Hall, S.J. (1999). The Effects of Fishing on Marine Ecosystems and Communities. Blackwell Scientific Publ., London. 274pp. 

Kaluzny, S.P., Vega, S. C., Cardoso, T.P., and Shelly, A,A (1998). S+Spatial Stats- Users Manual for Windows and Unix. 

Springer- Verlag New York, Inc., New York. 321 pp. 

Kanyerere, G.Z. (1999). Trawl Selectivity and the Effect of Clogging on Standard Trawl Surveys on Lake Malawi. Government 

of Malawi, Fisheries Department. Fisheries Bulletin No. 39. 25pp. 

Palsson, O.K., Bulirani, A., and Banda, M (1999). A Review of Biology, Fisheries and Population dynamics of Chambo 

Oreochromis spp., Cichlidae) in Lakes Malawi and Malombe. Government of Malawi, Fisheries Department, Fisheries 

Bulletin No. 38. 35 pp. 

Patterson, G. & Kachinjika, O. (1995). Limnology and Phytoplankton Ecology. In A. Menz (ed.), The Fishery Potential and 

Productivity of the Pelagic zone of Lake Malawi/Niassa. Natural Resources Institute, Chatham, UK. pp307-349 

Pelletier, D. & Parma, A.M.(1994). Spatial Distribution of Halibut (Hippoglossus stenolepsis): An Application of Geostatistics to 

Longline Survey Data. Can. J. Fish. Aquatic(1994). Sci. 51, 1506-1517. 

Turner, G.F. (1996). Offshore Cichlids of Lake Malawi. Cichlid Press, D-31864 Lauenau, 236 pp. 

Turner, G.F.,Tweddle, D., and Makwinja, R.D. (1995). Changes in Demersal Cichlid Communities as a result of Trawling in 

Southern Lake Malawi. In T.J. Pitcher and P.J.D. Hart (eds.), The Impact of Species changes in African Lakes. 

Chapman and Hall, London, pp.397-412. 

Turner, J. (1976). Promotion of Integrated Fishery Development-Malawi. An Analysis of the various fisheries of Lake Malawi. 

Technical Report 1, Rome, FAO. 73pp. 

Rosner, B. (1995). Fundamentals of Biostatistics. Wadsworth Publ. Company, California 94002. 661 pp. 

Sparre, P., Ursin, E. & Venema, S.C. (1989). Introduction to Tropical Fish Stock Assessment. Part 1-Manual. FAO, Fisheries 

Technical Paper No. 306/1, 337 pp.

189 


Preliminary Investigations of Community Level Responses to Benthic Trawling in 

the Demersal Fish Fauna of Lake Malawi/Niassa, Africa. 

Will Darwall 

Department of Biological Sciences, University of Hull, UK, Overseas Development Group / School of Development Studies, 

University of East Anglia, UK 

Introduction 

A demersal trawl fishery was initiated in 1968 in the far south of the South East Arm (SEA) of Lake Malawi in an 

area since designated as "Area A" (Tarbit 1971) (Fig. 1). The fishery in Area A has a mean annual yield of 800 

metric ‘red’ (Bulirani et al. 1999) and targets an estimated 70-80 species, mostly from the family Cichlidae. 

Management has been largely based on the application of single species models (Tweddle & Magasa 1989) which 

are unable to account for species specific responses to fishing which result from the wide range of life histories 

expressed (see (Adams 1980) for a review of life history influence on response to fishing). Life history 

characteristics of the target species range from those maturing at 37mm (Pseudotropheus livingstonii) with a mean 

fecundity of 25, to those maturing at 160mm (Buccochromis lepturus) with an estimated mean fecundity of 450 

(Duponchelle & Ribbink 2000). Growth rate estimates for these species still remain uncertain but include values of 

K ranging from 0.49 - 1.82 (Iles 1971; Tweddle & Turner 1977; Duponchelle & Ribbink 2000). 

Kyela 

Nkhata Bay 

Chintheche 

Nkhotakota 

Domira Bay 

Senga Bay 

Mbamba Bay 

SWA SEA 

Metangula 

Maponda 

Area A 

Figure 1. Lake Malawi showing the location of the study area, Area A, in the South-East Arm (SEA). 

Previous studies in Area A have already identified a reduction in fishing yields, changes in species composition, 

local extirpation of species, decline in abundance of large species, increased abundance of small species, and an 

initial decline in species richness (FAO 1976; Turner 1977a; Turner 1977b; Turner et al. 1995; Stauffer J.R. et al. 

1997). The application of multispecies models, which may attempt to account for the varied species responses to 

fishing, is unlikely to be successful because not only are the data requirements extensive but it would be hard to 

implement management recommendations based on a whole suit of measures tailored for each individual species, or 

group of species. An alternative approach is to assess and manage the fishery at the level of the fish community or 

ecosystem. The ecosystem / community level approach to fishery management is becoming more accepted at the 

global level (Hollingworth 2000) and, given the multispecies, multigear nature of the fishery, it should be considered 

for Lake Malawi. 

In this paper community level responses in the demersal fish fauna of Area A were investigated throughout the thirty 

year period of benthic trawling with an aim to identifying a suitable "indicator" parameter for future monitoring of 

community level changes in the fishery. The study includes analyses of both species and trophic composition.

Methods 

190 


Species Compositions 

Species compositions were obtained from experimental trawl surveys conducted by the Malawi Fisheries Research Unit (FRU) since 1970. 

Analyses were restricted to catches from within the 20-40m depth range which is mainly targeted by the trawl fishery in Area A. Inconsistencies 

in taxonomic identifications and catch sampling methodologies combined with the replacement of the research vessel Ethelwyn Trewavas with 

the RV Ndunduma in 1993/4 necessitated extensive standardisation of data. 

Data standardisation 

Sampling methodologies 

There were a number of inconsistencies in the way trawl catches were sampled through the years. The large catfish 

and 'others' (large cyprinids, mormyrids and large cichlids such as Oreochromis spp. and Buccochromis spp.) had 

already been discarded from the data obtained for the 1991-92 surveys (Turner et al. 1995). Consequently, 

community changes can only be assessed for the 'small' fish subset of the community, mainly comprising cichlids 1 . 

There may still be inconsistencies between survey periods for Oreochromis and Buccochromis spp. as the size at 

which they were considered to be "large" and thus excluded from the "smalls" category has not been strictly 

determined such that different surveys may have discarded different proportions of these species. 

Taxonomic inconsistencies 

The data set for this study spans a period of 30 years during which time many of the species names and cheironyms 

used have changed. Considerable effort has been put into tracing these changes in nomenclature and in the 

production of a standardised list of species names that can be applied to all surveys. Many name changes were 

successfully traced through the literature or sorted out following discussions with Prof. George Turner who was 

responsible for many of the new cheironyms introduced during the 1991-92 survey. For all remaining cases with 

potential for taxonomic confusion taxa have been consolidated into species complexes or identification has been 

restricted to the level of genus. 

Standardisation of catches from Ethelwynn and Ndunduma trawls 

In 1993 both the RV Ethelwynn Trewavas and RV Ndunduma operated simultaneously allowing a direct 

comparison of trawl catch rates from the two vessels (Banda et al. 1996). The difference in codend mesh size, 38mm 

on Ndunduma and 25mm on Ethelwynn, was reported not to have a major impact on catch biomass and the two 

gears were considered to have "identical selectivity curves for the size at first capture". Size frequency distributions 

for catches from each vessel were very similar for fish below approximately 10cm (TL) above which the Ndunduma 

catches were significantly greater. The greater catches by the Ndunduma were said to be a product of its greater 

trawling speed enabling increased capture rates for larger fish such as catfish, cyprinids and large cichlids. The 

greater swept area, for both horizontal and vertical dimensions, would also have accounted for the greater yields. 

Unfortunately no detailed comparison of species compositions for catches from the two vessels was reported and the 

original data for these trawls appears to have been lost (Banda pers. comm.). The reactivation of the Ethelwynn 

Trewavas would be most beneficial for direct comparison of the species selectivities of the two vessels. 

In recognition of the differences in catch compositions from the two vessels all shoaling 'off-bottom' species which 

would be caught preferentially by the Ndunduma but not by the Ethelwynn Trewavas were excluded from the 

analyses. Species groups excluded under this category were all Copadichromis spp., Diplotaxodon spp, and 

Rhamphochromis spp. The larger, faster swimming species, also preferentially caught by the Ndunduma, were 

automatically excluded as they would have been sorted into the 'large fish' portion of the catch. The remaining 

portion of the catch, for which time series analyses could be conducted, comprises the group of fish known locally 

as 'Chisawsawa' which constitutes the truly demersal cichlid community. The only non-cichlid included is the small 

catfish, Synodontis njassae. As all species compositions are presented as percentages of total biomass 

standardisation of trawling times and swept areas was not required. 

Fish diets and allocation of trophic guilds 

Trophic analyses were based on the results of dietary studies. Fish dietary components were determined from 

stomach analysis of 114 species from across the whole demersal fish community. Over 3,000 fish were sampled 

with the sample sizes weighted towards those species most dominant in the catch. This sampling program provided 

1 Data on the "large fish" component has since been obtained and will be included in future analyses.

191 


dietary data for 95% of the mean catch biomass. The results of stomach analysis were confirmed in a number of 

cases by results from analysis of stable isotope ratios. 

Statistical analysis 

Dominance curves, diversity indices, and Multi-Dimensional Scaling (MDS) ordination techniques were used in the 

analyses of temporal changes in community composition (Primer Software 1994). MDS ordinations were based on 

Bray-Curtis Similarity matrices of 4 th root transformed percentage biomass for each species or trophic guild. The 

sample unit was taken as each individual trawl catch. 

Results 

The demersal fish community 

The trophic structure of the demersal fish community includes 4-5 trophic levels based predominantly on the 

consumption of detritus and diatoms (Fig. 2). The main food items were, in descending order of dominance, detritus 

and diatoms (combined in a "diatomaceous ooze"), molluscs, copepods (mainly cylcopoids), other fish, chironomid 

larvae, cladocera, other insect larvae, and oligochaetes. 

Tilapines / 

Haplochromines 

Catfish 

The Fishery 

Pisciv. Haplochromines 

Haplochromines / Synodontis 

Molluscs Caridina 

Detritus / diatoms / algae 

Figure 2. Schematic diagram of the demersal fish community food-web. 

Temporal changes in community composition 

Non-cichlids 

Herbivorous / 

detritivorous 

Carnivorous Inverts 

Species composition 

Species compositions were obtained from surveys over the 1970-75 period and from 1990 to 1998. No surveys were 

conducted during the 1980's. MDS ordination demonstrates a clear temporal reduction in diversity between 

individual trawl catches within each survey period (Fig. 3). In effect the similarity between individual trawl catches 

within a survey period has increased in more recent years (the mean similarity between trawls was 38.89 in 1973 

and 54.28 in 1998 - Bray-Curtis similarity index). K-dominance curves show that the community has become 

dominated (in terms of biomass) by fewer species in recent years (Fig. 4). There was no significant reduction in 

species richness (at the restricted taxonomic level used in the analyses) or change in species distributions over time. 

The temporal trend for increasing dominance was reversed during the last survey period (1995-98) when dominance 

was shown to have declined. This apparent reversal towards the original state of evenness may indicate a degree of 

community resilience to fishing, as originally suggested by Tweddle and Magasa (1989). 

Bacteria

192 


Figure 3. MDS ordination in which each point represents the species composition of a single trawl catch. 

The distance between points is a measure of their similarity where points closest together are the most 

similar. Trawls from each survey period are enclosed in ellipses. 

Cumilative % biomass 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

1 11 21 31 41 

Ranked species 

Figure 4. K-dominance curves for species biomass. 

1971-73 

1975 

1991-92 

1995-98 

1971-73 

1975 

1991-92 

1995-98

193 


Trophic guilds 

Species were allocated to one of ten trophic guilds defined as collections of species using similar combinations of 

food types. MDS ordination shows a clear change in guild composition over time (Fig. 5). The most significant 

change was seen between the 1971-73 and 1975 survey periods in the early stages of development of the fishery 

(ANOSIM, R statistic = 0.316, p < 0.001). Little change was observed between 1975 and 1991-92 but by the time of 

the 1997-98 survey guild composition showed no significant difference from the original 1971-73 state. The return 

to the original state following a period of change follows the pattern of change described above for species 

compositions. 

1971-73 

1971-73 

1975 

1991-92 

1995 

1997-98 

1991-92 

Figure 5. MDS ordination of trophic guilds. Each point represents the trophic guild composition of a single 

trawl catch. The distance between points is a measure of their similarity where points closest together are 

the most similar. 

Trophic biomass pyramids 

Pyramids of biomass were constructed by multiplying the percentage compositions of food items for each species by 

the species percentage composition in the community (Fig. 6). The mean trophic level dropped from an initial value 

of 3.02 in 1971-73 to a low of 2.71 in 1975. By the time of the 1995-98 survey period the trophic level had risen 

back to near its original level at 2.95. 

Trophic Level 

Trophic Level 

5 

4 

3 

2 

5 

4 

3 

2 

1971-73 

0 10 20 30 40 50 

% Com position 

1991-92 

0 10 20 30 40 50 

% Composition 

Trophic Level 

Trophic Level 

Figure 6. Pyramids of biomass. Each plot depicts the mean relative proportion of each trophic level within 

the community for that period of survey. 

5 

4 

3 

2 

1975 

1995-98 

1975 

1995 

5 

4 

3 

2 

0 10 20 30 40 50 

% Composition 

0 10 20 30 40 50 

% Com position

194 


Conclusions and Recommendations 

Species compositions, species diversity (dominance/evenness) and trophic composition have all changed 

significantly over the 30 year period of study. All of these community descriptors have followed a similar pattern of 

change. If these patterns are to be interpreted as a community response to fishing it would appear that the initiation 

of the fishery has stimulated a rapid alteration in community structure followed by a period of relative stability and, 

finally, a gradual return towards the original state by the 1995-98 period. Such a response would be typical of that 

expected from a community resilient to change. A more useful interpretation of this last observation, however, 

awaits the analysis of an extended time series (including data from 1999-2000) combined with analyses for possible 

correlation of dominance with levels of fishing effort, primary productivity and other environmental variables (work 

in progress). 

Changes in species dominance parallel the changes recorded for all other aspects of community structure 

investigated in this study and may, as such, serve as a suitable "indicator" for monitoring community level responses 

to fishing pressure. It is well documented that alterations to community structure following fishing may lead to 

significant changes in fishery productivity and stability (Jennings & Kaiser 1998). Such community level changes 

would not be detected by the current approaches to fishery assessment in Lake Malawi. As species dominance can 

easily be calculated from the data already being collected as part of the regular program of monitoring by the FRU it 

is recommended that this measure be included as a monitoring parameter for future fishery assessment on Lake 

Malawi. 


Thanks are given to Mr Davis Mandere (FRU) who was responsible for all recent species identifications and to 

Valerie Choiseul (EU project) and Tony Mhango (FRU) for their work on the dietary analysis. This work is based 

on data obtained as part of the EU funded INCO-DC project: "The trophic ecology of the demersal fish community 

of Lake Malawi/Niassa, Africa." 

References 

Adams, P. B. 1980. Life history patterns in marine fishes and their consequences for management. Fishery Bulletin, 78, 1-12. 

Banda, M., Tomasson, T. & Tweddle, D. 1996. Assessment of the deep water trawl fisheries of the South East Arm of Lake 

Malawi using exploratory surveys and commercial catch data. Stock Assessment in Inland Fisheries (ed I. G. Cowx), 

pp. 53-75. Fishing News Books, Oxford. 

Bulirani, A. E., Banda, M. C., Pálsson, O. K., Weyl, O. L. F., Kanyerere, G. Z., Manase, M. M. & Sipawe, R. D. 1999. Fish 

Stocks and Fisheries of Malawian Waters. Resource Report. 1999. Lilongwe, Malawi, Government of Malawi 

Fisheries Department. 

Duponchelle, F. & Ribbink, A. J. 2000. Fish Ecology Report: Lake Malawi/Nyassa/Niassa Biodiversity Conservation Project. 1- 

139. SADC/GEF. 

FAO. 1976. Promotion of integrated fishery development, Malawi. FI:DP/MLW/75/109, Technical Report 1. Rome. 

Hollingworth, C. E. 2000. Ecosystem Effects of Fishing. ICES Journal of Marine Science, 57, 465-792. 

Iles, T. D. 1971. Ecological aspects of growth in African cichlid fishes. J.Cons.int.Explor.Mer., 33, 363-385. 

Jennings, S. & Kaiser, M. J. 1998. The Effects of fishing on Marine Ecosystems. Advances in Marine Biology, 34, 201-352. 

Stauffer J.R., Arnegard M.E., Centron M., Sullivan J.J., Chitsulo L.A., Turner, G. F., Chiotha S. & McKaye, K. R.1997. 

Controlling vectors and hosts of parasitic diseases using fishes. Bioscience, 47, 41-49. 

Tarbit, J. 1971. Lake Malawi Trawling Survey: Interim Report 1969-1971. 2, 1-16. 1971. Zomba, Malawi, Ministry of 

Agriculture & Natural Resources. Fisheries Bulletin Ref Type: Report. 

Turner, G. F., Tweddle, D. & Makwinja, R. D. 1995. Changes in demersal cichlid communities as a result of trawling in southern 

Lake Malawi. The Impact of Species Changes in African Lakes (eds T. J. Pitcher & P. J. B. Hart), pp. 397-412. 

Chapman & Hall, London. 

Turner, J. L. 1977a. Changes in size structure of cichlid populations of Lake Malawi resulting from bottom trawling. Journal of 

the Fisheries Research Board of Canada, 34, 232-238. 

Turner, J. L. 1977b. Some effects of demersal trawling in Lake Malawi (Lake Nyassa) from 1968 to 1974. Journal of Fish 

Biology , 10, 261-272. 

Tweddle, D. & Magasa, J. H. 1989. Assessment of mulitspecies cichlid fisheries of the Southeast Arm of Lake Malawi, Africa. 

J.Cons.int.Explor.Mer., 45, 209-222. 

Tweddle, D. & Turner, J. L. 1977. Age, growth and natural mortality rates of some cichlid fishes of Lake Malawi. 

Journal of Fish Biology, 10, 385-398.

195 


Resource use overlaps between the pair trawl and small-scale fisheries in the 

southeast arm of Lake Malawi: preliminary results. 

Thomas Eddie Nyasulu 

Fisheries Research Unit, P. O. Box 27, Monkey Bay 

Abstract 

Since the development of the pair trawl fishery, there has been a conflict between the small-scale and this fishery. 

This study assesses the abundance, species composition and size distribution of fish exploited by small-scale and 

pair trawl in Lake Malawi. Preliminary results have shown that pair trawlers appear to target some shallow-water 

fish species at small sizes. Furthermore, there appears to be overlaps in the species harvested by pair trawl and 

small-scale fisheries. 

Introduction 

Lake Malawi, the southernmost of the great lakes of Africa covers an area of about 30,800km 2 which ranks it ninth 

in order of size among the world’s lakes. Experimental bottom trawling in 1965 revealed the presence of 

commercially exploitable stock of demersal fish, mostly cichlids (Turner 1975). In 1968 a commercial trawl fishery 

was initiated in the southeast arm of Lake Malawi (Figure 1) in Areas A & B by Maldeco Fisheries Ltd. The pair 

trawl fishery was developed to eliminate the expensive winching machinery and by 1972, trawling started in Area B 

(Turner 1975). In the early 1970s, the pair trawl fishery realized a mean CPUE value of approximately 2 tons per 

day and a total of 22 pair trawl units were, then, licensed in this fishery (Jambo, in press). However, only five units 

presently operate in both southeast and southwest arms of the lake and present catch-effort data for the same area 

indicates that the CPUE is approaching 1 ton per day (Banda and Tomasson 1997). This decline in the number of 

units in operation and the catch per unit effort suggests that the fishery is collapsing. 

The pair trawlers are restricted to fish in areas more than 20 metres deep and at least 1 nautical mile from the shore. 

This is similar to the area fished by the small-scale fishery and there are indications that there is competition 

between the pair-trawlers and the small-scale fisheries is over the fishing ground. Small-scale fishers operate using 

a variety of gears including open water seines, beach seines, gill nets and hook and line (Weyl et al. 2000). During 

the early days of operations, there were no restrictions over the type of fishing gears to be used neither were there 

any restrictions on any fishing activity until 1952 when a minimum of 102 mm mesh size for seine nets with the 

headline length of greater than 274 metres was put in place (Lowe 1952). Presently most of the small-scale fisheries 

involve traditional gears which operate manually while a few chilimira, kauni and gill netters use motorized boats. 

Chambo was the main species caught in the 1970s by both the commercial and the small-scale fisheries. The 

collapse of the chambo fishery in the southern arms of Lake Malawi and Lake Malombe (Banda et al 1997, Weyl et 

al 1999) and the growing demand for fish by the growing Malawi population has aggravated some conflicts between 

pair-trawlers and the small-scale fishers, each sector accusing the other over the depletion of the fish stocks. This 

paper tries to look at the causes and effects of the conflicts and provide possible preliminary recommendations that 

would be applied to both fishing sectors. This survey was designed to determine if there is any resource use 

overlaps between the small-scale and pair trawl fishery in the southeast arm of Lake Malawi and attempt to 

determine the causes for the conflicts which exist between the pair trawl and small-scale fisheries over the fisheries 

resources in the study area. It should however be noted that this study is in its initial phase and that results presented 

here are preliminary. 

Methods of data collection 

The survey was conducted in the southeast arm of Lake Malawi (Figure 1) and small scale and pair trawl catches 

were sampled at Kela beach, Mpwepwe and Namiasi.Fish samples were collected from pair-trawlers and all smallscale 

fishermen to determine the species composition and size selectivity of the catches. A representative sample 

was taken from the total catch which was later sorted to the species level. Each species component in the sample 

was weighed and all fish in the sample were measured for total length to the nearest millimetre.

Figure 1. Map of the study sites in the southeast arm (SEA) of Lake Malawi. 

196 



The in both pair-trawl and small-scale fisheries, the species composition is multi-species and target species are 

determined by the mode of operation of the gear rather than by the preference of the fisher. More than 60 species 

were recorded from the small scale fishery and the pair trawl fishery. The genera comprising the majority of the pair 

trawl catch composition and their contribution to various small scale gears is shown in Table 1. It is evident that 

there is considerable overlap in the genera harvested by the small-scale and pair-trawl fisheries (Table 1). 

Furthermore, a number of shallow water species, including Otopharynx argyrosoma, Nyassachromis argyrosoma, 

and Pseudotropheus livingstonii were recorded in pair trawl catches. This indicates that the 1-nautucal mile 

offshore rule is not abided to by the pair-trawl fisheries. 

Table 1. Percent contribution of the genus Oreochromis, Nyassachromis, Ctenopharynx, Lethrinops, 

Otopharynx, Rhamphochromis and Copadichromis to the catch in the pair trawl fishery, gill net fishery, 

chilimira/kauni, hand line and beach seine fishery. 

Pair Trawl 

GN 

1-2 inch 

GN 

2-3 inch 

GN 

3inch + 

Chilimira Kauni Hand lines 

Beach 

seines 

Oreochromis 2.3 0.2 24.0 68.0 9.9 34.0 3.5 14.5 

Nyassachromis 4.0 0.0 15.2 0.0 0.0 0.0 0.0 0.0 

Ctenopharynx 5.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 

Lethrinops 8.3 3.4 2.7 0.0 0.0 0.2 2.6 11.2 

Otopharynx 12.9 1.8 0.0 0.0 1.5 0.0 18.5 8.2 

Rhamphochromis 17.1 2.5 0.0 2.7 42.5 37.0 13.3 5.7 

Copadichromis 21.3 70.2 14.8 0.0 25.0 5.0 4.1 12.2 

Table 1 shows that the main areas of overlap between the pair trawl and the small scale fishery are in the genus 

Oreochromis, Rhamphochromis and Copadichromis. While Oreochromis was not a major component of the pair 

trawl catch (

197 


different length sizes within the range of 110 cm and 230 cm. From Figure 2 it is clear that a high proportion of 

Rhamphochromis spp. caught by the pair-trawlers were immature (Figure 2). In the small-scale fishery, on the other 

hand, a large proportion of the catch was mature. In the gill net fishery almost all the fish caught were mature. The 

chilimira kauni and the beach seine fisheries on the other hand appear to target smaller Rhamphochromis. Although, 

greater amounts of the fish caught in the chilimira/kauni fishery are mature, there is also a considerable amount of 

immature fish in the catch. This contrasts with the pair-trawl fishery in which immature individuals dominated the 

catch. It is therefore evident that there is very high competition for this resource between the two fishing sectors, 

both for the mature and the immature fish. 

Pair Trawl 

Frequency 

250 

200 

150 

100 

50 

0 

20 50 80 110 140 170 200 230 260 290 320 350 380 410 440 


Chilimira/kauni 

Frequency 

Frequency 

1500 

1000 

500 

0 

20 50 80 110 140 170 200 230 260 290 320 350 380 410 440 


Gill net 

30 

25 

20 

15 

10 

5 

0 

20 50 80 110 140 170 200 230 260 290 320 350 380 410 440 


Beach seine 

Frequency 

20 

15 

10 

5 

0 

20 50 80 110 140 170 200 230 260 290 320 350 380 410 440 


Figure 2. Length frequency of Rhamphochromis spp., in the pair trawl, chilimira/kauni, gill net and beach 

seine fisheries of the southeast arm of Lake Malawi.

198 


Copadichromis virginalis matures at lengths of 10.6 cm. This species is one of the most abundant and economically 

important fish in the lake. This fish is by most fishing gear and there appears to be a degree of user overlap in this 

resource. Copadichromis virginalis are harvested by the pair trawl fishery before maturity and there was only small 

contribution of mature individuals to the total catch (Figure 3). The situation, in general, differs from chilimira/kauni 

fishery where most of the fish caught were mature (Figure 3). Gill nets are selective for mature individuals and most 

of the catch is dominated by fish that range between 115 mm and 140 mm (Figure 3). However, not all the fish 

found in gill nets are mature since there some evidence of fish which were less than 106 mm. This shows that both 

fishing sectors harvest a wide length-range of C. virginalis there is high competition between these two fishing 

sectors for Copadichromis virginalis. 

Pair Trawl 

Frequency 

250 

200 

150 

100 

50 

0 

Chilimira/kauni 

Frequency 

20 40 60 80 100 120 140 


600 

500 

400 

300 

200 

100 

0 

20 40 60 80 100 120 140 


Gill net 

Frequency 

1500 

1000 

500 

Nkacha 

Frequency 

0 

20 40 60 80 100 120 140 


60 

50 

40 

30 

20 

10 

0 

20 40 60 80 100 120 140 


Figure 2. Length frequency of Copadichromis virginalis in the pair trawl, chilimira/kauni, gill net and 

nkacha net fisheries of the southeast arm of Lake Malawi.

199 


Conclusions and recommendations 

These preliminary results have indicated that there is some evidence of an overlap for some species between pair 

trawl and small-scale fisheries. This has been due to the fact that both fisheries fish in water less that 50 m deep. 

While pair trawlers are licensed and bound by the regulation to fish in areas that are at least 1 nautical mile offshore 

and deeper than 18 meters, the presence of shallow-water species like Otopharynx argyrosoma, Nyassachromis 

argyrosoma and Pseudotropheus spp. shows that chances are very high that the pair-trawl fishery operates in waters 

shallower than 20 meters and closer than one nautical mile from shore. 

While the catch compositions in small-scale gears such as chilimira/kauni and gill nets have shown that most of the 

fish caught in these fisheries are mature, juvenile fish predominate the catch in the pair-trawl fishery. In addition, 

juvenile cichlids, too small to be identified even down to the genus level, were also found in the catch sampled from 

the pair trawlers at Namiasi fish landing site. This composition contributed to about 2% of the total catch sampled 

from two fishing units from the Namiasi landing site. A different scenario became evident from Mpwepwe where 

the unit belongs to the Fisheries Department. All fish collected from this pair-trawl unit were comparatively large 

and approximated the size at maturity for individual species. While pair-trawlers are bound by the regulation to use 

a 38 mm mesh at the bunt, the presence of juvenile fish from these private-owned units from Namiasi indicates that 

these pair-trawlers use very small meshed bunts for their trawl nets. 

It is therefore imperative that there is a strong need for some management efforts to be put in place for the pair-trawl 

fishery before the situation gets out of hand. Some of the many management efforts required for the fishery are to 

enforce the 1-nautical mile/18m depth limitation of the gear and to enforce the 38mm minimum mesh size for pairtrawlers. 

Acknowledgements. 

This Programme was funded by the Department of Fisheries and NARMAP. Thanks to my fellow researchers, 

Moffat Manase, Richard Sipawe and Jacqueline Chisambo for jointly using part of their data. 

References 

Banda, M., and Tomasson, T., 1997. Demersal fish stocks in Southern Lake Malawi: Stock Assessment and 

Exploitation. Fisheries Department of Malawi, Bulletin No. 35, pp 39. 

Eccles, D. H., and Trewavas, E., 1989. Malawi Cichlid fishes. The classification of some Haplochromine genera. 

Herten (W. Germany), Lake Fish Movies, pp335. 

Jambo, C. M. (in press). Commercial and artisanal fisheries in Lake Malawi: Who is to blame for the declining 

catches? 

Lowe, R. H. 1952. Report on the Tilapia and other fish and fishes of Lake Nyasa1945-47, Part 11. London, HMSO, 

Colonial Office Fisheries Publ. 1(12), pp126. 

Turner, G. F. 1992. The mechanised fisheries of Lake Malawi. FI:DP/MLW/86/013 Field Document 23.FAO, Rome, 

24pp. 

Turner, J. L. 1975. Some effects of demersal trawling in Lake Malawi (Lake Nyasa) from 1968 to 1974. 

Weyl, O. L. F., Banda M. C., Manase, M., Namoto, W., Mwenekibombwe, L. H., 1999. Analysis of Catch and 

Effort Data for the Fisheries of Lake Malawi.

200 


Population biology of the catfish Bagrus meridionalis from the southern part of 

Lake Malawi. 

Moses Banda 

Fisheries Research Unit, P.O. Box 27, Monkey Bay, MALAWI-AFRICA. Tel. 265-587-432, E-mail: fru@malawi.net 

Abstract 

The study investigated the endemic catfish, Bagrus meridionalis of Lake Malawi in order to understand their ecological relations, 

distribution, abundance and life history. The catfish are the second most important group of targeted species in Lake Malawi’s 

fisheries. Surveys were done using the research vessel, Ndunduma, using a bottom and mid-water trawl from June 1994 to March 

1996. The management implications of the ecology in relation to development of the deep trawl fishery are discussed. 

Key words: Catfish, ecology, distribution, abundance, life history, Lake Malawi. 

Introduction 

The demersal catfish Bagrus meridionalis (Günther) (local name: kampango), is the only bagrid species that occurs 

in Lake Malawi. It is one of the large fish species found in the lake and is a predator endemic to the lake and feeds 

mainly on small lake demersal cichlids. B. meridionalis is one of the most common and widely distributed species in 

the lake, and inhabits a variety of habitats (Jackson et al. 1963). 

B. meridionalis is important commercially and is valued as a food fish by the rural community. It is caught mainly 

by gillnets and long lines in the small-scale commercial fisheries and by bottom trawlers in the large commercial 

fisheries (Jackson et al. 1963, Tweddle 1982, Sipawe 2001). The small-scale fisheries operate mainly in waters 

shallower than 20 m and the large-scale commercial fisheries operate in deeper water (Alimoso 1989). The mean 

total annual landings are about 1300 tonnes representing approximately 2.5 % of the mean annual total fish catch in 

Malawi for the past decade (Banda 2000). 

While several authors have described some aspects of its biology, previous studies have focused mainly on feeding 

(Jackson et al. 1963, McKaye 1986, LoVullo et al. 1992) and reproductive biology (Jackson et al. 1963, Tweddle 

1982, FAO 1976, McKaye 1985, LoVullo et al. 1992, McKaye et al. 1992). Three notable exceptions have looked at 

the age and growth (Tweddle 1975, Roberts 1990, LoVullo et al. 1992). However, no full account of the biology and 

ecology of Kampango in a single area has been published. 

This study provided information on the distribution and biology of the Kampango around the southern part of Lake 

Malawi, which is under ever-increasing pressure from both the small- and the large-scale commercial fisheries. The 

paper contributes to the knowledge of the life history of Bagrus, and may be of great importance to the assessment 

and implementation of the management plan for the sustainable use of the deep-water fishery, which is rapidly 

expanding. 


The data in this study were collected during the stock assessment and biological surveys, from June 1994 to July 

1996, carried out in the southeast arm (SEA) and southwest arm (SWA) of Lake Malawi (Figure 1) by the Fisheries 

Department. Stock assessment surveys were done on a quarterly basis and were designed to monitor changes taking 

place in the SEA and SWA, the heavily exploited areas, whilst biological surveys were done in between the stock 

assessment surveys to provide additional information on the reproduction of the fish. The 17 m long Fisheries 

Department research vessel, Ndunduma powered with a 380 hp engine which pulls a Gulloppur bottom trawl net 

with a 23 m headrope and a 38 mm mesh cod end was used. The effective mesh size in the cod end was 4 mm due to 

the net liner used to ensure that escape through the meshes was minimal. 

Eight stock assessment surveys were done and during each survey a total of 97 fixed trawl stations (54 in the SEA 

and 43 in SWA) were sampled. The stations were stratified into shallow 0-50 m, deep 50-100 m and very deep 100- 

150 m. Each trawl lasted for 30 minutes. Twelve additional surveys were done in between the quarterly surveys and 

29 stations were sampled in each survey in the same depth ranges. The duration of each pull varied between 30-120 

minutes in order to obtain a large sample. All trawling was done during the day at depths between 10 and 150m 

depth. It was not possible to sample in water less than 10 m deep because of restrictions imposed by the draught of 

the vessel.

201 


B. meridionalis were sorted out from the main catch which was always composed of many small cichlids and some 

larger fish. Each of them was measured on board the vessel to the nearest 0.5 cm (total length) and most fish were 

weighed to the nearest 0.01 kg and those that were not weighed individually were collectively weighed as a species 

group to the nearest 0.1 kg. 

Figure 1. Map of southern of Lake Malawi showing the sampling stations used in the monitoring 

surveys. 

Sex was determined after dissection and, in some cases the gonads were removed and weighed to the nearest 0.1g. 

The state of sexual maturity was assessed by the Nikolsky (1963) method, based on the external appearance of the 

gonads. The average length at first maturity was estimated by using the logistic equation: 

p= 1/(1 + e -r(L-Lm) ) 

where p = the proportion of fish with ripe gonads, r = the slope of the curve, L = the length of fish and Lm = the 

mean length at 50% maturity. 

The catch per unit effort (C/f) assumed as index of abundance was expressed as: 

Results 

-14.4 -14.2 -14.0 -13.8 

Latitude 

34.60 34.75 34.90 

Longitude 

35.05 35.20 

C/f = catch (kg) /swept area (ha). 

Distribution and relative abundance 

The distribution and abundance of the total number of individuals caught during the stock assessment surveys are 

shown in Figure 2. B. meridionalis was widespread, being found at each station, in both arms. The mean biomass of 

B. meridionalis from the SEA and SWA were 4.8 and 4.0 kg ha -1 respectively and the mean biomass from the former 

was significantly different (t-test, p < 0.05) from the latter. The mean biomass decreased with increasing depth and 

it was not significantly different in the 0-50 and 51-100 m depth strata, but the average biomass in the > 100 m 

depth strata was significantly (t-test, p < 0.05) lower in both arms (Table 1). There was no significant difference 

between biomass values in different surveys (t-test, p < 0.05) (Table 2). 

Size structure 

The length-frequency distributions of the Bagrus in different depth strata in both SEA and SWA were generally 

unimodal with few at fish < 5.0 cm (Figures 3-6). This absence was not attributed to net selectivity because the cod 

end was covered with a liner, which retained small cichlids of about 1.0 cm. B. meridionalis ranged from 4.5 cm to 

105 cm TL with single modes which varied with depth. The smaller size- classes generally dominated and showed 

asymmetrical distribution, particularly in waters deeper than 50m. In contrast, bimodal distribution patterns were 

noticeable in 0-20 m depth range, especially in the SEA indicating the presence of two distinct sizes of fish, small 

and large (Figures 7-10). The mean length decreased with increasing depth and many large individuals were caught 

SEA

202 


in shallower water (< 20 m) in both arms (Figures 3-6). Females were significantly larger (t-test, p < 0.05) than 

males in both arms and also fish (both sexes) caught in the SEA were generally larger than those from the SWA 

(Table 3). 

Sex ratios 

The sex ratio, based on all Kampango sampled, did not deviate from 1:1 ratio (Table 4). Similarly, the sexes were 

equally represented in all depth strata except in shallow water of the SWA in which there were significantly more 

males (χ 2 –test, p < 0.05) than females. Sex ratios grouped into 5-cm length classes varied within the length intervals 

and in some cases there were significant departures (χ 2 –test, p < 0.05) from the 1:1 ratio (Table 5). Males 

predominated in all the small size intervals and female in the large size intervals. 

Table 1. The mean biomass (kg ha -1 ) and density (number per unit area) with standard error (s.e.) of 

Bagrus meridionalis in the southern part of Lake Malawi, June 1994 – March 1996. 

Area Depth (m) Biomass Density 

Mean s.e. Mean s.e. 

SEA 0-50 4.90 (1.47) 5 (2.03) 

50-100 5.40 (1.50) 11 (2.74) 

>100 2.20 (1.63) 5 (5.24) 

0-150 4.80 (0.98) 8 (1.77) 

SWA 0-50 5.20 (1.84) 8 (4.13) 

50-100 4.40 (1.34) 16 (4.82) 

>100 1.20 (0.72) 5 (3.70) 

0-150 4.00 (1.01) 11 (2.96) 

Table 2. The mean biomass (kg ha -1 ) with standard error (in brackets) of Bagrus meridionalis in the 

southern part of Lake Malawi during the quarterly surveys, June 1994 – March 1996. 

SEA SWA 

Surveys Biomass s.e. Biomass s.e. 

1 22.1 (2.16) 15.5 (2.33) 

2 26.9 (2.90) 21.1 (3.00) 

3 21.2 (2.56) 25.3 (3.93) 

4 25.1 (2.70) 24.2 (2.87) 

5 31.6 (3.53) 19.1 (2.51) 

6 23.3 (2.36) 25.1 (2.92) 

7 22.2 (2.43) 15.6 (2.14) 

8 20.4 (3.09) 17.6 (2.73)

-30.8 -30.6 -30.4 -30.2 -30.0 

Latitude 

14.6 14.8 15.0 

Latitude 


CPUE (Kg/hr) 

54 

0 

21.9 22.0 22.1 22.2 22.3 22.4 

Longitude 


34.2 34.4 34.6 

Longitude 

34.8 35.0 

203 


CPUE (Kg/hr) 

Figure 2. The distribution of Bagrus meridionalis caught by trawling during monitoring surveys in the 

south east arm (above)and south west arm (below) of Lake Malawi. 

Table 3. The mean length (cm TL) with standard error of Bagrus meridionalis in the southern part of Lake 

Malawi during the quarterly surveys, June 1994 – March 1996. 

Area Sex Mean s.e Total 

SEA Male 34.2 0.13 7582 

Female 38.4 0.18 7708 

both sexes 36.4 0.11 15290 

SWA Male 30.0 0.11 9002 

Female 31.3 0.15 8753 

both sexes 30.5 0.09 17753 

148 

0

204 


Table 4. The ratio of males to females in relation to area of Bagrus in the southern part of Lake Malawi. * 

denote values that are significantly different (χ 2 –test, p < 0.05) from a 1:1 ratio. N = number of fish 

examined. 

Area Depth (m) Ratio 

SEA & SWA 0-150 1.0 

SEA 0-50 1.0 

Table 5. The ratio of males to females in relation to size (20 cm size classes) of Bagrus in the southern 

part of Lake Malawi. Asterisks denote values that are significantly different (χ 2 –test, p < 0.05) from a 1:1 

ratio. N = number of fish examined. 

Depth (m) 

SEA SWA 

Length class (m) 0-50 51-100 >100 0-50 51-100 >100 

0-20 1.1 1.1 1.0 1.5* 0.9 1.0 

21-40 1.2* 1.1* 1.0 1.1* 1.0* 0.9 

41-60 1.3* 0.9* 0.7 1.5* 0.8* 0.4 

61-80 0.1* 0.1* 0.5 0.2* 0.1* 0.4 

81-100 0.1* 

Length-weight relationships 

The length-weight relationships of Bagrus indicated that the growth is isometric i.e. the body length of an individual 

fish grows in proportion to the cubic root of its body weight. The relationship between log total length and log 

weight is: 

Log10W = – 5.13 + 3.04*Log10L. 

51-100 1.0 

>100 1.0 

SWA 0-50 1.1* 

51-100 1.0 

>100 0.9 

Reproductive patterns 

The breeding periodicity of Kampango is shown in Figure 11. Ripe individuals were found throughout the year 

indicating that continuous breeding occurs, although there was a considerable seasonal variation in the proportion of 

adult fish with ripe gonads. The breeding peak, which was between November and December, coincided with the 

rainy season. Breeding individuals were recorded at all depths, but the species appear to prefer shallow water (< 50 

m) (Table 6). Bagrus meridionalis reached sexual maturity at 52 cm TL in females and 43.8 cm TL in males in the 

southern part of Lake Malawi and the difference between sexes were significant (χ 2 -test, p < 0.05) (Figure 12). 

Table 6. The percentage of ripe Bagrus in relation to depth in Lake Malawi. N = number of fish examined. 

Area N 0-50 51-100 >100 m 

SEA 3403 77.4 22.3 0.3 

SWA 2292 78.1 21.6 0.3

Percent 

20.0 

18.0 

16.0 

14.0 

12.0 

10.0 

8.0 

Percent (a) Shallow water (0-50 m) 

6.0 

4.0 

2.0 

0.0 

10 20 30 40 50 60 70 80 90 100 

Total length (cm) 

(b) Deep water (51-100 m) 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

40.0 

35.0 

30.0 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

205 


N = 2131 

x = 39.9 cm 

s.e. = 0.25 

N = 4535 

x = 33.0 cm 

s.e. = 0.14 

10 20 30 40 50 60 70 80 90 100 


Percent (c) Very deep water (> 100 m) 

N = 916 

x = 27.0 cm 

s.e. = 0.31 

10 20 30 40 50 60 70 80 90 100 


Figure 3. The length-frequency distribution of male Bagrus meridionalis in the south east arm of Lake 

Malawi collected between June 1994 and March 1996. All mean length numbers (x) are significantly 

different (t-test, p < 0.05).

Percent 

Percent 

(a) Shallow water (0-50 m) 

12.0 

10.0 

8.0 

6.0 

4.0 

2.0 

0.0 

10 20 30 40 50 60 70 80 90 100 



18.0 

16.0 

14.0 

12.0 

10.0 

8.0 

6.0 

4.0 

2.0 

0.0 

35.0 

30.0 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

206 


N = 2147 

x = 46.5 cm 

s.e. = 0.38 

N = 4572 

x = 36.7 cm 

s.e. = 0.20 

10 20 30 40 50 60 70 80 90 100 



N = 989 

x = 28.5 cm 

s.e. = 0.34 

10 20 30 40 50 60 70 80 90 100 


Figure 4. The length-frequency distribution of female Bagrus meridionalis in the south east arm of Lake 



Percent 

20.0 

18.0 

16.0 

14.0 

12.0 

10.0 

8.0 

Percent (a) Shallow water (0-50 m) 

6.0 

4.0 

2.0 

0.0 

10 20 30 40 50 60 70 80 90 100 



25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

40.0 

35.0 

30.0 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

207 


N = 3046 

x = 35.2 cm 

s.e. = 0.21 

N = 5229 

x = 27.6 cm 

s.e. = 0.12 

10 20 30 40 50 60 70 80 90 100 



N = 725 

x = 26.0 cm 

s.e. = 0.34 

10 20 30 40 50 60 70 80 90 100 


Figure 5. The length-frequency distribution of male Bagrus meridionalis in the south west arm of Lake 



Percent 

Percent 


20.0 

18.0 

16.0 

14.0 

12.0 

10.0 

8.0 

6.0 

4.0 

2.0 

0.0 

10 20 30 40 50 60 70 80 90 100 



35.0 

30.0 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

35.0 

30.0 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

208 


N = 2630 

x = 37.9 cm 

s.e. = 0.31 

N = 5350 

x = 28.6 cm 

s.e. = 0.16 

10 20 30 40 50 60 70 80 90 100 



N = 773 

x = 27.1 cm 

s.e. = 0.41 

10 20 30 40 50 60 70 80 90 100 


Figure 6. The length-frequency distribution of female Bagrus meridionalis in the south west arm of Lake 



Percent 

Percent 


35.0 

30.0 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

10 20 30 40 50 60 70 80 90 100 


209 


N = 368 

x = 42.4 cm 

s.e. = 0.76 


20.0 

18.0 

N = 1763 

16.0 

x = 39.4 cm 

14.0 

12.0 

10.0 

8.0 

6.0 

4.0 

2.0 

0.0 

s.e. = 0.26 

10 20 30 40 50 60 70 80 90 100 


Figure 7. The length-frequency distribution of male Bagrus meridionalis in the south east arm of Lake 


different (t-test, p < 0.05). 

Percent 

Percent 


16.0 

14.0 

12.0 

10.0 

8.0 

6.0 

4.0 

2.0 

0.0 

14.0 

12.0 

10.0 

8.0 

6.0 

4.0 

2.0 

0.0 

N = 359 

x = 53.6 cm 

s.e. = 1.09 

10 20 30 40 50 60 70 80 90 100 



N = 1788 

x = 45.1 cm 

s.e. = 0.40 

10 20 30 40 50 60 70 80 90 100 


Figure 8. The length-frequency distribution of female Bagrus meridionalis in the south east arm of Lake 



Percent 

Percent 


25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

10 20 30 40 50 60 70 80 90 100 



25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

210 


N = 576 

x = 38.9 cm 

s.e. = 0.55 

N = 2470 

x = 34.3 cm 

s.e. = 0.22 

10 20 30 40 50 60 70 80 90 100 


Figure 9. The length-frequency distribution of male Bagrus meridionalis in the south west arm of Lake 


different (t-test, p < 0.05). 

Percent 

Percent 


12.0 

10.0 

8.0 

6.0 

4.0 

2.0 

0.0 

14.0 

12.0 

10.0 

8.0 

6.0 

4.0 

2.0 

0.0 

N = 359 

x = 53.6 cm 

s.e. = 1.09 

10 20 30 40 50 60 70 80 90 100 



N = 1788 

x = 45.1 cm 

s.e. = 0.40 

10 20 30 40 50 60 70 80 90 100 


Figure 10. The length-frequency distribution of female Bagrus meridionalis in the south west arm of Lake 



% mature 

30.0 

25.0 

20.0 

15.0 

10.0 

5.0 

0.0 

J94 A94 O94 D94 F95 A95 J95 A95 O95 D95 F96 A96 

Months 

211 


% F %M 

Figure 11. Breeding periodicity of male and female Bagrus meridionalis in the southern part of Lake 

Malawi. Samples were collected between June 1994 and April 1996. 

Proportion mature 

Proportion mature 

(a) 

0.80 

0.60 

0.40 

0.20 

0.00 

(b) 

0.80 

0.60 

0.40 

0.20 

0.00 

0 10 20 30 40 

Length (cm) 

50 60 70 80 

Observed Calculated 

0 20 40 60 80 100 

Length (cm) 

Observed Calculated 

Figure 12. The mean length at first maturity in (a) male and (b) female Bagrus meridionalis. (Lm = 43.8 

cm TL for males and 52.0 cm TL for females).

212 


Discussion 

Bagrus meridionalis were commonly caught in the bottom water trawl, which is consistent with previous studies 

(Turner 1982, Thompson et al. 1996). Depth is the most important factor determining its distribution. B. 

meridionalis were described as broadly eurytopic in their distribution patterns because they have a wide 

geographical distribution and inhabit a variety of habitats (Figure 2). They can live in a wide range of conditions, 

often moving from one habitat to another. 

The length-frequency distributions of B. meridionalis are ‘dome-shaped’, a phenomenon usually attributed to gearselectivity, 

in particular gillnets (Ricker 1975). In this case it is negated because the gear used in this study is almost 

non-selective (since the smallest specimen of Bagrus caught in this gear was about 4.5 cm TL). These distributions 

probably reflect the size gradation of Bagrus with depth in the lake. The absence of the smallest length classes of 

Bagrus suggests that they occur in water < 10 m deep or in areas where trawling was not possible. These small fish 

are being guarded in their nests, which are often situated adjacent to rocks (Jackson et al. 1963). The decrease in the 

numbers of large fish in shallow waters may reflect high fishing pressure by the small-scale commercial fisheries. 

The bi-modal distribution Bagrus in < 0-20 m water depth range was attributed to gillnet fishery (Figures 7-10). The 

gillnet selectivity studies (Jackson et al. 1963, Tweddle 1975, Sipawe, 2001 in prep.) indicate that the gill-net 

selectivity peak is between 35-55 cm TL if all meshes are combined. Therefore, fish over 40 cm TL were less liable 

to be caught, and fish less than 35 cm TL more exposed to higher fishing mortality. 

There was size gradation with depth indicating that some zonation or restricted movement occurs (Figures 3-6). The 

presence of adults in shallow water (< 50 m) and juveniles in deeper water (> 50 m) may be a way of avoiding 

interspecific competition and cannibalism. The difference in mean length between fish caught in the SEA and SWA 

may be explained by the following: 1) the fish stocks of Kampango from the SEA is different from that of the SWA, 

and 2) there is less competition of food resources in the SEA than SWA which perhaps promotes faster growth of B. 

meridionalis in the former. The numbers of fish per km 2 are 473 and 813 in the SEA and SWA, respectively. Fishing 

pressure is higher in the SEA than in the SWA if the number of fishing gear and fishermen are considered to be 

indicative of fishing mortality (Table 5). Similar fishing pressure could account for the reduced number of large fish 

in the 0-50 m depth stratum. The intensity of gillnet fishing pressure on the Kampango population structure is 

evident in Figures (7-10). 

B. meridionalis was most abundant in < 100 m water depth. The variation in abundance with depth reflected depth 

preferences and numbers. Size gradation indicates that their depth distribution was determined by depth preferences. 

The effect of numbers on estimates of abundance is clearly observed in this species where few large fish in the 0-50 

m depth range gave a similar estimate of biomass to that of many small fish in the 51-100 m depth range. The 

abundance of Bagrus throughout the study period was remarkably constant and there was no evidence of 

seasonality. This suggested that recruitment remained constant because its population structure was mainly 

composed of small individuals. 

The spatial distribution of the overall sexes was within the expected ratio. The sex ratio however, was different in 

shallow water (< 50 m) in the SWA and in the length classes and this discrepancy was mainly attributed to fishing 

pressure. Studies by Jackson et al., (1963) and Tweddle (1982) indicated that more females were caught in gillnets 

than males and the most vulnerable fish size was between 35-55 cm TL. This is confirmed by this study because 

61.3% (32.8% males, 28.5% females) of the 33052 fish examined were in the 35-55 cm TL range. The difference in 

sex ratio was also significant between 45-55 cm TL (Table 3). Females were fished more than the males probably 

because of their fast growth rate. The significant difference in sex ratios above 55 cm TL could be attributed to 

growth pattern. Males seemed not to grow as big as females. 

The main breeding period of Bagrus meridionalis occurred in the rainy season and the actual duration depends on 

the length of the rains, which is consistent with other studies (McKaye 1986: Jackson et al. 1963, Tweddle 1975, 

1982) and other bagrids (Dadzie and Ochieng-Okach 1989). The breeding peak coincided with the zooplankton 

production peak (Irvine 1995) that ensures that food is available for the young. Breeding occurred within the same 

depth range in which they foraged (Table 6) which suggests that there was no spawning migration (Banda, 2000). 

The inshore migration in spawning B. meridionalis (Jackson et al. 1963, Tweddle 1982) may reflect the need for 

breeding fish to find suiTable spawning sites. 

The length at first maturity differed in male and female B. meridionalis, with females maturing at a larger size, 

probably because of different growth rates between sexes. The results were similar to estimates by Eccles (1972),

213 


but not to those by Tweddle (1982) and Bernacsek et al., (1983). The difference in mean lengths could be explained 

by: 1) the difficulty in determining the gonadal stages, thereby erroneously including or excluding small fish in the 

ripe category, 2) different stocks were studied and 3) the sample size was too small in the early studies as higher 

values were obtained by Eccles (1972) with a larger sample (1409) compared to Tweddle (1982) and Bernacsek et 

al. 1983 who examined about 200 and 170 specimens, respectively. 

Conclusion 

The present study indicates that fishing is having a major impact on the population structure of Bagrus. Different 

fisheries have different impact on the population. The selective cropping of certain length classes most of which are 

immature is carried out by small-scale commercial fisheries (gillnet fishery and longline fishery) in shallow water 

while unselective fishing is undertaken by trawlers in deep waters. The result is that both fisheries will invariably 

eliminate large fish species including B. meridionalis. 

B. meridionalis has maintained a substantial relative stable population especially in the southern part of the lake 

where both fisheries operate because the fisheries have been limited to coastal areas and the decline in large fish has 

always been compensated by improved recruitment and the fact that the nursery areas are in the less exploited deep 

water. The contention that the decline in the catches of B. meridionalis was caused by high fishing pressure on the 

immature fish (FAO 1976, Fryer 1984, Alimoso 1989, Alimoso, et al. 1989) is not substantiated by these studies 

because of their life history. Therefore, this decline may reflect the absence of certain length classes of fish as the 

small-scale commercial fisheries are selective in nature. This raises a question as to whether the noted general 

decline in the catch is indicative of overfishing or gear selectivity. However, the population of Bagrus will definitely 

decline with the expansion of the expansion of the deep-water fishery and may lead to a collapse because both adults 

and juveniles will be vulnerable to exploitation. 

This size structure of Bagrus from the two arms indicates that the populations seem to be different which 

necessitates the establishment as to whether these populations are indeed distinct fish stocks. Future research 

priorities should focus on mapping the populations and distinguishing them for management purposes. These studies 

should be accompanied by studies on the biology and population dynamics of the most important commercial fish 

species. 


This work forms part of my PhD studies, which was funded by Icelandic Development Agency. 

References 

Alimoso, S.B., 1989. Management of catfish (Bagrus meridionalis Günther) in southern Lake Malawi. Fishbyte, 7(2):10-12. 

Alimoso, S.B., Magasa, J.H. and van Zalinge, N.P. 1989. Exploitation and Management of fish resources in Lake Malawi. In: 

Fisheries of the African Great Lakes. Wageningen, Netherlands, Occassional Paper No. 3, 83-95. 

Banda, M.C. 2000. The biology and ecology of the catfishes of the Genera bathyclarias and Bagrus in Lake Malawi. 

Unpublished PhD. Thesis. University of Zimbabwe. 

Bernacsek, G.M., Massingo, A. and Contreras, A. 1983. Exploratory gillnetting in Lake Niassa, Moszambique, with biological 

profiles of the main taxa caught. FAO/GCP/MOZ/oo6/SWE (phase II). Field Document, 5, 129 pp. 

Dadzie, S. and Ochieng-Okach, J.L. 1989. The reproductive biology of a suluriod catfish, Bagrus docmak 

(Forsskål)(Cypriniformes, Bagridae) in the Winam Gulf of Lake Victoria. Revue Zoology of Africa, 103, 143-154. 

Eccles, D.H. 1972. Bagrus meridionalis (Kampango). Department of Agriculture and Fisheires Annual Report 1964-65, Part 

II:6-13. 

FAO, 1976. Promotion of intergrated fishery development, Malawi. Analysis of the various fisheries of Lake Malawi. FAO, 

FI:DP/MLW/77/516. Technical report 1. 

Fryer, G. 1984. The conservation and rational exploitation of the biota of Africa’s Great Lakes. In: Conservation of Threatened 

natural Habitats. Hall A.V. (Ed). South African Natural Science Programmes Report (92), CSIR, Pretoria, pp. 135-146. 

Irvine, K. 1995. Standing biomasses, production, spatial and temporal distributions of the crustacean zooplankton. In: The 

Fishery Potential and Productivity of the pelagic Zone of Lake Malawi/Niassa. Menz, A. (Ed). Chatham, U.K: Natural 

Resources Institute, pp. 85-108. 

Jackson, P.B.N., Iles, T.D. Harding, D. and Fryer, G. 1963. Report on the survey of northern Lake Nyassa, 1954-55. Joint 

Fisheries Research Organisation. Government Printers. Zomba, Malawi. 

LoVullo, T.J., Stauffer, J.R. and McKaye K.R. 1992. Diet and growth of brood of Bagrus meridionalis Günther (Siluriformes: 

Bagridae) in Lake Malawi, Africa. Copie, 4, 1084-1088. 

McKaye, K.R. 1985. Cichlid-catfish mutualistic defence of young in Lake Malawi, Africa. Oecologia, 66, 358-363. 

McKaye, K.R. 1986. Trophic egg and parental foraging for young by the catfish Bagrus meridionalis of Lake Malwi, Africa. 

Oecologia, 69(3), 358-363.

214 


McKaye, K.R. Mughogho, D.E. and LoVullo, T.J. 1992. Formation of the shellfish school. Environmental Biology of Fisheries, 

35, 312-218. 

Nikolsky, G.V. 1963. The ecology of fishes. New York: Academic Press. 

Ricker, W.E. 1975. Computation and interpretation of Biological Statistics of Fish Populations. Bulletin of the Fisheries Research 

Board of Canada, 191, 1-382. 

Roberts, K.S. 1990. Ageing the Lake Malawi catfish Bagrus meridionalis using otoliths. Unpublished MSc. Thesis. University of 

Wales. 

Sipawe, R. 2001. Gear and species selectivity of the gillnet fishery in Lake Malawi. (Proceedings of this symposium) 

Thompson, A.B. Allison, E.H. and Ngatunga, B.P. 1996. Distribution and breeding biology of offshore pelagic cyprinids and 

catfish in Lake Malawi/Niassa. Environmental Biology of Fishes, 47, 27-42. 

Turner, J.L. 1982. Analysis of purse seine and gillnet catch data off Nkhata Bay, Lake Malawi, from March 1980 to April 1981. 

In: Biology Studies on the Pelagic Ecosystem of Lake Malawi, FI: DP/MLW/75/019, Technical Report 1, FAO, Rome, 

pp. 154-162. 

Tweddle, D.1975. Age and growth of the catfish Bagrus meridionalis Günther in the southern Lake Malawi. Journal of Fish 

Biology, 7, 667-685. 

Tweddle, D. 1982. Fish breeding migrations in the North Rukuru area of Lake Malawi with a note on gillnet colour selectivity. 

Journal of Science and Technology (Malawi), 3(2), 67-74.

215 


Feeding habit and development of feeding-related morphological characters in 

Oreochromis shiranus (Boulenger, 1896) larvae and juveniles in Malawi 

Shinsuke Morioka and Daniel Sikawa 

Department of Aquaculture and Fisheries Science, Bunda College of Agriculture, University of Malawi, P.O. Box 30321, c/o JICA 

Malawi Office, Lilongwe, Malawi. TEL: 265-277-214, FAX: 265-771-125, E-mail: morioka@malawi.net 

Abstract 

Stomach content, gut length, morphology of pharyngeal teeth and gill rakers of Oreochromis shiranus larvae and juveniles were 

investigated in order to elucidate the timing of feeding habit conversion, and the relationships between feeding habit and the 

feeding-related morphological characters. Stomach content showed that the feeding habit of fish drastically changed between 14-20 

mm TL from zooplanktovorous to herbivorous (phytoplanktovorous). Pharyngeal plate and teeth were already ossified in the 

smallest fish of this study (9.95 mm TL) and teeth number increased as fish grew. Percentage length of gut ((gut length/TL) x 100 

%) became more than 100 % when TL reached 15-20 mm, falling mostly under the size with feeding habit conversion. Gill rakers in 

4 th gill arch started appearing at 14 mm TL when the feeding habit started converting. Consequently, the conversion of feeding habit 

in Oreochromis shiranus started when the percentage intestine length became over 100 % and gill rakers in 4 th gill started 

appearing, and pharyngeal teeth did not seem to relate directly with such feeding habit conversion. 

Key words: Oreochromis shiranus, feeding habit, morphological development, zooplanktovorous, phytoplanktovorous 

Introduction 

Lake Malawi provides the main fish protein resources to the Malawian public. However, the fish catch in Malawi 

has declined since the late 1980s from over 70,000 t in 1989 to 40,000 t in 1999 (Zidana, 2000), leading to the 

decrease of fish consumption by the Malawian public. Therefore, the government of Malawi has attempted to 

promote the finfish aquaculture since late 1980s. However, there have been no remarkable developments in 

aquaculture production in Malawi, although the number of small-scale aquacultural farmers and their production has 

recently increased (Sidira, 2001). 

It is well known that Oreochromis species are phytoplanktovorus (Appler, 1985) and that the protein requirement, 

for a relatively high growth rate, is low. Therefore, Oreochromis shiranus is considered to be the most 

recommendable species for aquaculture in Malawi, and the majority of the small-scale farmers (92.9 %) grow this 

species (Sikawa, 1999). However, the feeding performance of larvae and juveniles of Oreochromis shiranus in 

Malawi has been scarcely reported so far, although the feeding and growth of fish larvae and juveniles are 

indispensable aspects for further development of seed production and nursing techniques in aquaculture. 

When considering the feeding performance of fish larvae and juveniles, the morphological development relevant to 

feeding habit and/or behavior, as well as the internal secretion, should be elucidated. In this study, therefore, the 

stomach content and the feeding-related morphological characters, such as gut length, gill rakers, pharyngeal teeth 

and stomach content of Oreochromis shiranus larvae and juveniles were investigated and the relation between the 

morphological development and feeding habit was discussed. 


This study was conducted using larvae and juveniles of Oreochrommis shiranus collected from the experimental 

earthen pond (15 x 50 m/m, 60 cm depth) of Bunda College of Agriculture, University of Malawi, during 24 May ~ 

15 June 2000. Water temperature during this period was 17.5 ~ 21.0 o C. In this pond, 333 males and 667 females 

were stocked in September 1999. The fish mostly relied on natural food (plankton) although 20 kg of maize bran 

was given as supplemental diet to the fish once a week. 

Fifty larvae and juveniles were used in this study. The size (total length) rage of fish was 9.95 ~ 50.30 mm. Fish 

were preserved in 5-10 % formalin immediately after collection. Twenty specimens were treated as the double 

stained transparent specimen, following the method described by Kawamura and Hosoya (1991). These specimens 

were used for determination of the number of pharyngeal teeth on the posterior edge of lower pharyngeal plate and 

the ossified gill rakers of 4 th gill arch. Other specimens were used to observe the stomach contents and intestine 

length (mm). The percentage length of intestine (PLI) was then calculated (PLI = [intestine length / total length] x 

100 %). The percentage of zooplankton mass and algal mass in stomach in each fish was obtained on the basis of the 

acreage of each organism in the microscopic image. The identification of zooplankton and phytoplankton in 

stomachs were attempted to higher than family level.

Results 

216 


Development of pharyngeal teeth and gill rakers 

Both lower and upper pharyngeal plates were already ossified in the smallest fish used in this study (9.95 mm TL). 

The number of pharyngeal teeth on the most posterior portion of lower pharyngeal plate increased as fish grew from 

8 at 9.95 mm TL to 36 at 43.65 mm TL (Figure 1). The relationship between TL and the number of pharyngeal teeth 

was expressed by the following two regressions intersecting at (19.91, 22.66); P = 1.29·L – 3.02 (91 = 0.95, n=9) 

and P = 0.50·L + 12.71 (r = 0.86, n = 11) where P is the number of teeth of lower pharyngeal plate and L is total 

length. The width of lower pharyngeal plate linearly increased as fish grew (Figure 2). The relationship between TL 

and the width of the lower pharyngeal plate was expressed by W = 0.097·L - 0.059 (r = 0.99, n = 20), where W is 

width of lower pharyngeal plate and L is total length. Algal cells attached on the surface of pharyngeal plate with 

mucus were frequently observed in fish larger than 20 mm TL. 

Gill rakers in the 4 th gill arch did not appear in fish smaller than 14.00 mm TL although the 4 th gill arch was already 

formed. Gill rakers in 4 th gill arch started appearing at 14.05 mm TL and increased in number as fish grew (Figure 

3). The relationship between TL and the number of gill rakers of the 4 th gill arch, after gill rakers appeared, was 

expressed by GR=1.52·L-14.46 (r=0.98, n=17) where GR is the number of gill rakers and L is total length. 

No. of phartngeal te 

40 

35 

30 

25 

20 

15 

10 

5 

0 

N = 1.29L - 3.02 

r = 0.91 

N = 0.50L + 12.71 

r = 0.86 

0 10 20 30 40 50 60 


Figure 1. Relationship between total length and the number of pharyngeal teeth in O. shiranus larvae and 

juveniles. 

Width of lower pharygeal plate 

6 

5 

4 

3 

2 

1 

0 

W = 0.097L - 0.059 

r = 0.99 

0 10 20 30 40 50 60 


Figure 2. Relationship between total length and the width of lower pharyngeal plate.

Number of gill rakers of 4th gil 

70 

60 

50 

40 

30 

20 

10 

0 

GR=1.52L-14.46 

r=0.98 

0 10 20 30 40 50 60 


217 


Figure 3. Relationship between total length and the number of gill rakers of 4 th gill arch. 

Intestine length 

Gut length was measured in 30 specimens and the intestines were elongated from 8.00 to 215.00 mm in 12.75 ~ 

41.30 mm TL of fish. Percent length of intestine (PIL = [intestine length / total length] x 100 %) smaller than 15.00 

mm TL was less than 100%. It increased as fish grew and reached more than 500% when TL was over 40 mm. The 

relationship between TL and PIL was expressed as PIL = 1.01·L 1.64 (r = 0.95, n=30), where PIL is percent intestine 

length and L is total length. 

Feeding habit 

Fish smaller than 14.00 mm TL mainly fed on zooplankton, such as Rotifers, Copepods, Brachiopods (Table 1). 

Feeding habit was changed drastically between 14.00 and 20.00 mm TL from zooplanktovorous to 

phytoplanktovorous including mono-cellular green algae, diatoms etc. (Table 1). Fibrous matter was occasionally 

observed in the stomach contents and seemed to originate from water-plants (Table 1). Fish larger than 20.00 mm 

TL dominantly fed on phytoplankton or algae even though zooplankton existed in water, as indicated by the stomach 

content of smaller fish that mainly fed on zooplankton. 

Table 1. Zooplankton and algae found in stomachs of Oreochromis shiranus larvae and juveniles. 

Fish size Main content in stomach 

< 15.00 mm TL 

>15.00 mm TL 

Rotifers 

Branchionidae, Euchlandae 

Copepods 

Diatomidae, Cyclopidae 

Branchiopods 

Daphnidae 

Diatoms 

Rhizosoleniaceae, Naviculaceae, Nitzschiaceae 

Green algae 

Characiaceae, Palmellaceae, Coelastraceae 

Zygnemataceae 

Blue-green algae 

Chroococcaceae 

Water plant (unidentified)

Percentage length of intes 

600 

500 

400 

300 

200 

100 

0 

IL = 1.01L 1.64 

r = 0.96 

0 5 10 15 20 25 30 35 40 45 


218 


Figure 4. Relationship between total length and the percentage length of intestine. 

Percentage of 

zooplankton/phytoplankton in st 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0 5 10 15 20 25 30 35 40 45 


Figure 5. Relationship between total length and the percentage contents of zooplankton/algae in 

stomach. 

Discussion 

Although Oreochromis species are known to be phytoplanktovorous or herbivorous (Appler, 1985, Hofer, 1988), the 

feeding performance at larval and early juvenile stages has scarcely been considered. In this study, the conversion of 

feeding habit from zooplanktovorous to phytoplanktovorous or herbivorous was observed to take place between 

14.00 and 20.00 mm TL on the basis of stomach content analysis (Table 1). This observation elucidated that O. 

shiranus changed the feeing habit when TL reached more than 20 mm TL, although they may be able to adapt to 

feed on zooplankton if phytoplankton is insufficient in quantity. When the conversion in feeding habit took place, 

the gill rakers of 4 th gill arch had just appeared (Figure 5) and percentage intestine length was over 100 % (Figure 4). 

In herbivorous and omnivorous fish, the function of the gill raker as a filter for phytoplanktonic organisms is well 

known. So, in O. shiranus, the development of gill rakers was also strongly related to herbivorous feeding habit 

development. Digestive ability of fish increases with the development of digestive organs and the enzyme 

production systems. Morioka (1993) reported that the increase in percent intestine length drastically increased after 

the transitional phase from larval to juvenile stages in a tropical omnivorous fish Chanos chanos. During the 

transitional phase from larvae to juveniles, the digestive systems develop morphologically and physiologically and 

they subsequently have the specific peculiarity in feeding habit (Tanaka, 1975). In O. shiranus, the transitional 

phase from larval to juvenile stages and the drastic changes in morphology and physiology are considered to take 

place at lengths of 15 – 20 mm TL.

219 


Although characters in pharyngeal teeth, i.e. the number of pharyngeal teeth (Figure 1) and the width of lower 

pharyngeal plate (Figure 2), may be related with the feeding habit conversion, the function of pharyngeal teeth for 

feeding could not be identified in this study. The decline of increasing rate of pharyngeal teeth after 20 mm TL 

(Figure 1) may demonstrate that the pharyngeal teeth are more functional for feeding on zooplankton and less 

functional when fish becomes herbivorous or phytoplanktovorous. In addition, the algae adhering to the surface of 

the pharyngeal plate by mucus was often observed in this study as well as in Tilapia rendalii (K. Utsugi, pers. 

comm.). This suggests that the pharyngeal teeth (plate) may function to transport phytoplankton to the digestive tract 

after the filtering of planktonic organisms by the gill rakers. 

In this study, the larvae and juveniles were sampled in May and June, this period was the dry season and water 

temperature was relatively low as 17~21ºC. This indicates, as additional information, that O. shiranus breeds almost 

throughout the year even though the lowest water temperature were less than 20ºC. This is in contrast to Yata and 

Miyashita (1988) report that O. niloticus did not breed under less than 23ºC. 

Consequently, the stage at the occurrence of feeding habit conversion and the development of several characters 

relevant to feeding of O. shiranus larvae and juveniles were elucidated in this study. We believe that this 

information is useful not only for the technical development in aquaculture but the resource management fields for 

further understanding the performance of the species in the wild. Since the decline in resource level has been 

recently pointed out especially in the lake chambo (O. karongae etc.) (Zidana, 2000). Researches on the early life 

history of fish are indispensable when considering the resource management aspects, and the surveys on feeding 

biology of fish larvae and juveniles have important roles for providing the information relevant to fish survival at 

early developmental phases which influences recruitment success. Taking the recent decline in the fish resource 

level into account, the research on the early biology of fish should be more emphasized in order to obtain more 

information for understanding the early life history of fish. 


We express our great thanks to DR. J.S. Likongwe, Department of Aquaculture and Fisheries Science, Bunda 

College of Agriculture, University of Malawi, for his advisory comment. We also thank Mr. O. Matambo and 

technical staff of the department for sample collection. Lastly, We are grateful to Japan International Cooperation 

Agency (JICA) to give us the opportunity to conduct this research. 

References 

Appler, H.N. 1985. Evaluation of Hydrodictyon reticulatum as protein source in feed for Oreochromis (Tilapia) niloticus and 

Tilapia zillii. J. Fish. Biol., 27: 327-334. 

Brooks A.C and Maluwa A.O., 1997, Technical Supplement, Fish Farming in Malawi, A case study of the Central and Northern 

Regions Fish Farming Project, Malawi Government, Ministry of Natural Resources. Malawi. 

Hofer, R. 1988. Morphological adaptations of the digestive tract of tropical cyprinids and cichlids to diet. J. Fish. Biol., 33: 399- 

408. 

Kawamura, K. and K. Hosoya (1991) A modified double staining technique for making a transparent fish-skeletal specimen. 

Bull. Natl. Res. Inst. Aquaculture, 20: 11-18. 

Morioka, S. 1993. Ecology of milkfish Chanos chanos larvae in the surf zone. Doctoral dessertation, Tokyo University of 

Fisheries. (in Japanese) 

Sidira, F.D. 2001. Annual report on the fish farming activities for the year 2000, Lilongwe, Mchinji, Kasungu and Dowa West. 

Department of Fisheries, Lilongwe. 

Sikawa D.C. 1999. On station and on farm investigation of optimal conditions for mass fingerling production of Tilapia, 

Oreochrmois shiranus (Boulenger 1896) in small scale hatchery systems in Malawi. Fish Farming Center, Mzuzu, 

Malawi. 

Tanaka, M. 1975. Digestive system of fish juveniles. Pages 7-23. In: Development and feeding of fish juveniles (eds.) T. Iwai 

and H. Tsukamoto, Koseisya Koseikaku, Tokyo. 

Zidana F.G. 2000, Malawi annual statistical bulletin. Ministry of Agriculture and Irrigation, Lilongwe. 56pp. 

Yata, T. and T. Miyashita. 1988. Seed production of Tilapia. In: Tilapia (ed.) T. Hidaka, Midori-shobo, Tokyo.

220 


Otolith growth increments in three cyprinid species in Lake Malawi and 

information of their early growth 

Shinsuke Morioka 1 & Emanuel Kaunda 2 

1 Department of Aquaculture and Fisheries Science, Bunda College of Agriculture, University of Malawi. c/o JICA Malawi Office, P.O. 

Box 30321, Lilongwe 3, Malawi. TEL: 265-277-214, FAX: 265-771-125, E-mail: morioka@malawi.net 

2 Department of Aquaculture and Fisheries Science, Bunda College of Agriculture, University of Malawi. c/o JICA Malawi Office, P.O. 

Box 30321, Lilongwe 3, Malawi 

Abstract 

Otoliths features and growth increments of three Malawian cyprinids, i.e. usipa Engraulicypris sardella, sanjika Opsaridium 

microcephalum and mpasa Opsaridium microlepis were investigated. The sagittae of all the species were not suitable for otolith 

increment reading due to the increment invisibility and fragile structures in rostrum portions. The asterisci have ambiguous core 

structures and fewer increment counts than the actual age (in Mpasa) or the ones in the lapilli (all species). This feature shows that 

the asteriscus is not applicable for daily increments analysis. The lapilli of them had clear increments from the core to edge, and 

were considered to be the best otolith for increment analysis. Hence, the lapilli were employed to elucidate the breeding periods and 

growth patterns in three species. The breeding period of Usipa was observed to take place almost throughout the year and the one 

of Sanjika was observed to take place at least 8 months from November to July. The growth in larvae and juveniles of both Usipa 

and Sanjika was relatively slower in the dry season and faster in the rainy season. Such difference in growth is considered to be due 

to the difference in water temperature between the dry and rainy season. 

Key words: Cyprinidae, sagitta, lapillus, asteriscus, growth, Lake Malawi 

Introduction 

Information on the early growth of fish is indispensable for both the development of resource management and 

aquacultural activities. The three cyprinid species inhabiting Lake Malawi, Usipa Engraulicypris sardella, Sanjika 

Opsaridium microcephalum and Mpasa Opsaridium microlepis, are all important for commercial fishing and sports 

fishing in the latter two. However, the current situation of fishery production in those species is not positive. In 

Usipa Engraulicypris sardella, for instance, the mechanism for the great yearly catch fluctuation (Zidana 2000) has 

not been well elucidated, and little biological research has been made on the species (Thompson et al. 1995). In 

Mpasa Opsaridium microlepis, the recent decline in catch has been pointed out (Makimoto, pers. com.), and no 

research has been made on the species recently although Tweddle (1982, 1987) conducted preliminary surveys on 

the biological aspects of the species. Much of the biology on Sanjika Opsaridium microcephalum, is still unknown, 

such as breeding season, growth, seasonal distribution etc. 

Among the biological features of fish, the information of the early developmental stages is extremely important 

because the early biology could explain lots the subsequent features of fish species, i.e. the breeding activity, 

biomass fluctuation, recruitment success and so on. To obtain the information on the early biology of fish, the 

utilization of otolith daily increment has been broadly applied since Pannella (1971) and the growth analysis in 

larvae and juveniles with daily increments was developed thereafter (Brothers and McFarland 1981,Campana and 

Neilson 1985, Campana 1990, Watanabe and Kuroki 1997). In most of the earlier studies employing otoliths, the 

sagitta was used to age fish in days. However, the sagitta is occasionally unsuitable for the determination of age in 

days due to the structural complexity (occurrence of sub-daily increments or fragility) and the lapillus has resulted to 

be better material in such cases (Hoff et al. 1997, Morioka and Machinandiarena 2001, Morioka et al. 2001). 

In this study, the detection of the applicable otoliths among the sagitta, lapillus and asteriscus were primarily 

conducted on Usipa Engraulicypris sardella, Sanjika Opsaridium microcephalum and Mpasa Opsaridium 

microlepis. Secondly, the elucidation of the breeding season and early growth pattern of the species were considered 

using the growth increments of the otolith that was observed as the applicable one in the primal achievement in this 

study. 


Fish used in this study were the larvae, juveniles and sub-adults of three cyprinid species, i.e. Usipa Engraulicypris 

sardella, Sanjika Opsaridium microcephalum and Mpasa Opsaridium microlepis. Larvae and juveniles of Usipa and 

the juveniles of Sanjika were collected from Nkhotakota and Chia beach (Figure 1) using seine nets (1 mm mesh, 1 

m height, 8 m and 30 m width) operated along the shore. Collection of these two species were made on 20 July, 17 

August 2000, January 18, February 28, 5 April and 3 May 2001 (n = 149 and 15.00 – 96.7 mm TL for Usipa, n = 

181 and 14.30 - 77.45 mm TL for Sanjika). Juveniles of Mpasa were hatched at the National Aquaculture Center, 

Domasi, Malawi by artificial fertilization (dry method) and 20 juveniles were sacrificed at 65 day-old after hatching

221 


(12.75-22.25 mm TL). Ten wild sub-adults of Mpasa were also collected from Salima on 1 May 2001 and 

Nkhotakota on 3 May 2001 (131.40 – 181.10 mm TL). 

Fish was preserved in 70% ethanol immediately after collection and the otoliths (sagitta, lapillus and asteriscus) 

were extracted. Extracted otoliths were embedded on the glass slides with epoxy resin. When otoliths were opaque, 

they were ground with sand paper (# 1500) and lapping films (12 µm, 9 µm, 6 µm mesh) for being thin proximal 

sections, following the method described by Nishimura (1993). Ground surfaces of otoliths were occasionally etched 

by 0.1 N HCL to emphasize the contrast of the continuous and discontinuous zones in otoliths. Otolith 

microstructures were observed under the optic microscopy with transmitted light (x 200 - 400). For all species, 

otolith increment counts were counted and the radii of every five increments in Usipa’s lapilli were recorded. In 

Usipa, TL was back-calculated every 5 days using the relationship between otolith radii and TL, following the 

method of Watanabe and Kuroki (1997). 

Figure 1. Collection sites of fish. 

Results 

Otolith features 

External and internal features of otoliths were more or less similar among three species of this study. The sagitta was 

arrow head shaped and had an obvious nucleus with two projections (rostrums) on anterior/posterior sides (Figure 

2). The increments in the sagittae were deposited clearly from the nucleus up to c. 30 th increment at the bases of both 

anterior- and posterior-rostrums, then, they became invisible in rostrum portions (Figure 2). Rostrum portions, in 

addition, were fragile and often destroyed by otolith extracting and grinding treatments. These features indicated that 

the sagitta was not a suitable material for the age determination in days. 

The lapillus was oval-shaped and had a clear nucleus (Figure 3). The increments in the lapilli were clearly visible 

from the nuclei to edge portion after grinding (Figure 3). The increment counts in the lapilli of laboratoryhatched/reared 

Mpasa juveniles agreed with the actual age in days (65 days after hatching). Thus, the lapilli were 

considered to be the applicable otolith for daily increment analysis in the species used in this study. 

The asteriscus was oval-shaped and had an ambiguous nucleus (Figure 4). This ambiguous nucleus made the 1 st 

increment identification impossible. In the asterisci, additionally, the increment counts were much fewer than the

222 


ones in the lapilli or, they were much less than the actual age in days in laboratory-hatched/reared Mpasa juveniles. 

These features demonstrated that the asterisci were also not useful for daily increment analysis. 

On the basis of the upper observations, the lapilli were resulted to be the best material for the increment analysis of 

otolith. Hence, the lapilli were employed for the subsequent analyses. 

A B C 

Figure 2. Sagittae of juveniles in Engraulicypris sardella (A), Opsaridium microcephalum (B) and 

Opsaridium microlepis (C). Bars indicate 100 µm. 

A B C 

Figure 3. Lapilli of juveniles in Engraulicypris sardella (A), Opsaridium microcephalum (B) and 


A B C 

Figure 4. Asterisci of juveniles in Engraulicypris sardella (A), Opsaridium microcephalum (B) and 


Breeding seasons 

Breeding season of each species was estimated by the increment counts in the lapilli. In Usipa, the hatching months 

were observed almost throughout the year except January, March, May and August although the frequencies of the 

hatching month varied in each month, being particularly higher in December and April (Figure 5A). This suggested 

that Usipa bred almost throughout the year. In Sanjika, the hatching months were observed consecutively from late 

November to early July (Figure 5B), suggesting that Sanjika bred at least during 8 months a year. The highest 

frequency was observed in May and the second highest was in January. In Mpasa, the hatching months were 

observed from September to December although the number of specimens was only 10.

Frequency (% 

Frequency (% 

30 

25 

20 

15 

10 

5 

0 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

B 

Nov Dec Jan Feb Mar Apr May Jun Jul 

Hatching month 

A 

223 


Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug 

Hatching month 

Figure 5. Hatching months of Engraulicypris sardella (A) and Opsaridium microcephalum (B) estimated 

by the lapilli increment counts. 

Growth patterns 

Growth of Usipa larvae and juveniles born in the rainy season (November to February) was faster than the one born 

in the dry season (June and July)(Figure 6). The growth rate of Usipa in the rainy season was more than 0.70 mm TL 

per day while less than 0.50 mm TL in the dry season. Sanjika juveniles also showed that the growth in fish born in 

the rainy season was higher than the one born in the dry season (Figure 7). Sanjika could be separated into the 

following 3 groups: fish born and grown in the rainy season (n = 74, referred to as the summer-born group), fish 

born in the rainy season and grown during both the rainy and dry seasons (n = 3, referred to as the intermediate 

group), and fish born and grown in the dry season (n = 104, referred to as the winter-born group). Growth rate in the 

rainy season was more than 0.70 mm TL while less than 0.60 mm TL per day in the dry season. In Mpasa, the 

largest specimen (181.60 mm TL) was aged 221 days after hatching and the smallest one (131.40 mm TL) was aged 

151 days, demonstrating that Mpasa grew at the rate of more than 0.80 mm TL per day.

Total length (m 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

born in November 

born in December 

born in January 

born in Feburuary 

born in dry season 

224 


0 10 20 30 40 50 60 70 

Number of lapilli increments (age in days) 

Figure 6. Relationship between lapilli increment counts (age in days) and back-calculated total length at 

ages (mm) in Engraulicypris sardella. Back-calculated TL was individually expressed the fish groups 

hatched in different months or season. Closed circle: fish born in November, opened circle: in December, 

opened square: in January, opened diamond: in February and closed triangle: in dry season, respectively. 


120 

100 

80 

60 

40 

20 

0 

0 20 40 60 80 100 120 140 160 

Lapilli increment counts (age in days) 

Figure 7. Relationship between lapilli increment counts (age in days) and total length (mm). Closed circle: 

fish born/grown in rainy season, opened circle: fish born/grown in dry season, opened triangle: fish born 

in rainy season and grown in rainy/dry season.

225 


Discussion 

Difficulties in use of the sagittae for the determination of age in days have been occasionally pointed out because of 

the structural complexity or the occurrence of sub-daily increments (Hoff et al. 1997, Morioka and Machinandiarena 

2001, Morioka et al. 2001). In the three cyprinid species used in this study, the sagitta was considered inappropriate 

material for the daily increments analysis due to the fragility and invisibility of increments in rostrum portions 

(Figure 2). In the asteriscus, the difficulty in 1 st increment identification due to the ambiguous nuclei was observed 

(Figure 3). Moreover, in general, the asteriscus was observed to be formed later than the sagitta and lapillus, a 

certain period after hatching such as observed in Anguilla japonica (Umezawa 1988), Delitistes luxatus and 

Chasmistes brevirostris (Hoff et al. 1997). Such information also demonstrates the disadvantage of the asterisci in 

the analysis of daily increments. Thompson et al. (1995) described that the asterisci had two projections and the 

sagittae were formed later than the lapilli and asterisci in Usipa. We suspect that they misidentified the sagittae as 

the asterisci. Additionally, they employed the sagittae and asterisci in combination for the increment analysis instead 

of using the lapilli although they found the readable increments in the lapilli and the upper-noted unsuitable 

characteristics were found in other otoliths. As shown in this study, as well as observed in other cyprinids (Hoff et 

al. 1997), the lapilli are apparently superior to others for the increment counts with the consecutive readability of 

increments. The sagittae have been mainly applied for the otolith increments analysis so far. But, with taking the 

superiority in the lapilli found in cyprinids (this study and Hoff et al. 1997) and other species (Morioka and 

Machinandiarena 2001) into account, the use of the lapilli should be emphasized more and developed. 

Breeding period of Mpasa estimated in this study was from the latter half of the dry season and the beginning of the 

rainy season (September to December). However, the sexually matured adults were captured in earlier months (July 

and August)(P.B. Kataya, pers. comm.). This indicates that the breeding season of Mpasa may extend from July to 

December. 

Breeding period of Usipa was observed to take place almost throughout the year and Sanjika’s one was at least over 

8 months (Figure 5), although the water temperature in the lake seasonally varies between 23 ºC and more than 28 

ºC (Patterson and Kachinjika 1995). These long-term breeding periods in Usipa and Sanjika strongly suggest the 

existence of the plural stocks in both species, that is, the different stocks adapting to the different optimum 

temperature for reproduction. 

The growth of Usipa larvae and juveniles and Sanjika juveniles in the rainy season was higher than the one in the 

dry season (Figures 6 and 7). The differences in growth also suggest the existence of the plural stocks in both Usipa 

and Sanjika in the lake as well as the long term breeding periods discussed above. Differences in growth may be 

able to be explained by the water temperature in the lake that is higher in the rainy season and lower in the dry 

season as referred above (Patterson and Kachinjika 1995). In addition, the larger water inlet from the rivers during 

the rainy season may provide more nutrients to promote the larger biomass of the planktonic organisms that is the 

main diet of fish larvae and juveniles. However, Irvine (1995) observed that the zooplankton biomass was more 

abundant during the dry season than the rainy season (1995). 

Conclusive note 

In this study, several fundamental biological features of Usipa, Sanjika and Mpasa, the important species for both 

the commercial and sports fishing in Lake Malawi, were preliminary elucidated. More information is necessary to 

clarify the biological aspects of the early-staged fish especially in relation to the fish resource management. 

Additionally, the biological information of fish larvae and juveniles is indispensable for the further development of 

the aquacultural activity. In both senses, the long-term research on the important fish larvae and juveniles is strongly 

recommended. 


We thank Mr. S. Matsumoto and his colleagues for assisting us in fish sample collection. We are also grateful to 

Japan International Cooperation Agency (JICA) for giving us the opportunity to conduct this study.

226 


References 

Brothers, E.B. and W.N. McFarland (1981) Correlation between otolith microstructure, growth and life history transitions in 

newly recruited French grunts (Haemulon flavolineatum). Rapp. P.-V. Reun. Cons. Int. Explor. Mer., 178, 369-374. 

Campana, S.E. (1990) How reliable are growth back-calculation based on otoliths? Can. J. Fish. Aquat. Sci., 47, 2219-2227. 

Campana, S. and J.D. Neilson (1985) Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci., 42, 1014-1032. 

Hoff, G.R., D.J. Logan and D.F. Markle (1997): Otolith morphology and increment validation in young Lost Liver and shortnose 

suckers. Transact. Amer. Fish. Soc., 126, 488-494. 

Morioka, S, L. Machinandiarena (2001) Comparison of daily increment formation pattern between sagittae and lapilli of ling 

(Genypterus blacodes) larvae and juveniles collected off Argentina. New Zealand J. Mar. Freshwater Res., 35(1): 

111-119. 

Morioka, S., L. Machinandiarena and M.F. Villarino. 2001. Preliminary information on internal structures of otoliths and growth 

of ling, Genypterus blacodes (Ophidiidae), larvae and juveniles collected off Argentine. Bull. Jpn. Soc. Fish. 

Oceanogr., 65(2): 9-16. 

Nishimura, A. (1993): Age determination of walleye pollock based on the otolith (Review). Sci. Rep. Hokkaido Fish. Exp. Stn., 

42, 37-49. (in Japanese) 

Thompson, A.B., E.H. Allison, B.P. Ngatunga and A. Bulirani. 1995. Fish growth and breeding biology. In: The fishery potential 

and productivity of the pelagic zone of Lake Malawi / Niassa (ed.) A. Menz. Natural Resources Institute, UK. 

Tweddle, D. 1982. Fish breeding migrations in the north Rukuku area of Lake Malawi with a not on gillnet colour selectivity. 

Journal of Science and Technology in Malawi 3(2); 67-74. 

Tweddle, D. 1987. An assessment of the growth rate of Mpasa, Opsaridium microlepis (Gunter, 1864) (Pisces: Cyprinidae), by 

length frequency analysis. Journal of Limnological Society of South Africa 13(2): 52-57. 

Watanabe, Y. and T. Kuroki (1997) Asymptotic growth trajectories of larval sardine (Sardinops melanosticus) in the coastal 

waters off western Japan. Mar. Biol., 127, 369-378. 

Zidana F.G. 2000, Malawi annual statistical bulletin. Ministry of Agriculture and Irrigation, Lilongwe.

227 


Development of African catfish Clarias gariepinus larvae during the transitional 

phase between endogenous and exogenous energy intake. 

Seiji Matsumoto, Shinsuke Morioka and Sigeru Kumagai 

Department of Aquaculture and Fisheries Science, Bunda College of Agriculture, University of Malawi. c/o JICA Malawi Office, P.O. 

Box 30321, Lilongwe 3, Malawi. Tel: 265-277-214, Fax: 265-771-125, E-mail: morioka@malawi.net 

Abstract 

Development of newly hatched African Catfish Clarias gariepinus larvae in an aspect of energy intake conversion was investigated. 

The anus and mouth of larvae opened by 16 and 22 h after hatching, respectively. However, the onset of feeding was observed 

from day 3 (67 h) after hatching when the jaw structure started being ossified and the nerve in barbel appeared. Yolk was 

completely absorbed on day 7 after hatching (166 h after hatching). These observations demonstrated that the larvae had the 

transitional period to utilize both endogenous and exogenous energy sources for c. 100 h. Mass-mortality in larvae reared under 

starvation after hatching took place from day 11 to day 14, all specimens being dead on day 16. These phenomena indicated that 

the food-derived larvae had survived using the energy sources in their body tissues after yolk absorption for more than 4days 

(maximum 7 days). This fact showed the larvae of this species having a stronger starvation tolerance than the marine fish larvae 

reported in earlier studies. The delayed feeding experiments showed that the PNR (point of no-return) existed on day 9-10 after 

hatching, demonstrating that larvae should start feeding by 9-10 days after hatching to survive before reaching the PNR. 

Key words: Clarias gariepinus larvae, yolk, onset of feeding, survival 

Introduction 

African Catfish Clarias gariepinus (Burchell 1815) are widely distributed over Africa. This species is known as 

having an omnivorous feeding habit although the higher protein level in diet is recommendable (Viveen et al. 1985, 

Haylor 1992). Catfish species is the one of the main species under aquaculture especially in temperate and tropical 

freshwaters because of the omnivorous feeding habit and rapid growth such as observed in C. gariepinis (Haylor 

1992), Parasilurus asotus (Tsuchiya 1976) and Ictalurus punctatus (Watanabe 1988). In such background, the 

aquacultural aspects, e.g. seed production/rearing techniques and nutritional features have been broadly surveyed so 

far in catfish species. 

For further development of aquacultural techniques, the understanding of biological view in the larval stage is 

indispensable. Especially, the information on the survival, feeding performance and morphological development in 

early larvae should be well grasped in view of this critical period (Kohno 1998). Major achievements in such 

aspects have, so far, been made mainly in marine fin fish (Wiggins et al. 1985, Yamashita and Aoyama 1985, 

Khono et el. 1990). However, such features in catfish species have been scarcely made as yet. In addition, such 

information also contributes for further understanding the fish biology in the natural water in relation to the resource 

management aspects. 

In this study, therefore, the survival, onset of feeding and the early development of morphology in newly-hatched 

African catfish larvae were investigated and the features during the energy intake conversion from endogenous to 

exogenous were discussed. 


Some 30,000 African catfish larvae used in this study were produced with artificial fertilization (dry method) at the 

Department of Aquaculture and Fisheries Science, Bunda College of Agriculture, University of Malawi, on 15 

March (batch 1) and 15 April (batch 2) 2001. Newly-hatched larvae were stocked in 1000 l or 200 l plastic tanks 

immediately after hatching. Water temperature during rearing was 22-25 ºC. Artemia sp. naulii and natural 

zooplankton (Copepoda, Branchiopoda, Ploima etc.) were given at the density of 5 – 10 ind. per ml once a day. The 

hatching time was recorded in batch 1 as when the 1 st specimen hatched. Thereafter, some 500 larvae were 

intermittently sampled with the interval of more frequently than every 12 hours. Using these specimens, the trend in 

decrease of the yolk sac volume and the morphological development were observed. Yolk sac volume can be 

calculated with the formula V = 4/6·πR 3 if the shape of yolk sac is spherical, where R is the yolk radius. But, we 

applied the formula V = π/6·LH 2 (Blaxter and Hempel 1963), where L is length (mm) and H is height (mm), because 

the shape of yolk sac was not spherical in this species. One hundred nine of them were treated as being the double 

stained transparent specimen using the method of Kawamura and Hosoya (1991) to observe the jaw structures and 

barbel development. 

Sixty-five larvae from batch 1 were stocked in 1 l beaker and reared without feeding and survival rate was recorded 

until all larvae were died by starvation. Some 400 larvae from batch 2 were separately stocked in 7 beakers and 7

228 


different timings (delayed feeding) of first diet supply were set (Table 1). The survival rates in each setting were 

recorded. 

Table 1. Experimental design of delayed feeding. 

Results 

Settings Timings of 1 st diet supply 

D-4 Diet given on 4 days after hatching 

D-6 6 days after hatching 





Starvation Without diet until dead 

Yolk sac absorption and survival in food deprived specimens 

Total length (TL) of larvae at hatching was 4.07 ± 0.11 mm (mean ± SD, n = 10). TL, then, continuously developed 

and reached 9.23 ± 0.71 mm (n = 5) at 166 h when yolk sac completely absorbed (Figure 1). Yolk sac of newlyhatched 

larvae drastically declined from hatching to 97 h after hatching as expressed by the formula V = -0.0102·T + 

1.0208 (r = 0.986). It gradually decreased thereafter as expressed by the formula V = -0.0002·T + 0.0329 (r = 0.732). 

The complete yolk sac absorption was observed at 166 h (7-day) after hatching (Figure 2). The food deprived 

(starved) larvae from batch 1 highly survived until the 10 days after hatching and the mass mortality took place 

thereafter (Figure 3). They all died on day 16 after hatching. In the larvae (from batch 2) under the different timings 

of 1 st diet giving, the ones of D-4, -6 and -8 survived with relatively higher survival rates (more than 50%) while the 

ones of D-10, -12 and -14 showed the close patterns in mortality observed in starved larvae and all died after 17 

days after hatching in D-12, -14 and starvation (Figure 4). In D-10, only one larva survived. 


12 

10 

8 

6 

4 

2 

0 

onset of 

feeding 

0 50 100 150 200 

Hours after hatching 

Figure 1. Relationship between hour after hatching and total length (mm) in Clarias gariepinus larvae. 

Dotted line indicates the onset of feeding.

Yolk volume (mm 

1.2 

1 

0.8 

0.6 

0.4 

0.2 

0 

229 


Days after hatching 

0 1 2 3 4 5 6 

V = -0.0102T + 1.0208 

r = 

V = -0.0002T + 0.0329 

r = 0.732 

0 20 40 60 80 100 120 140 160 180 

Hours after hatching 

Figure 2. Relationship between hours after hatching and yolk sac volume (mm 3 ) in Clarias gariepinus 

larvae. 

Survival rate 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

yolk completely absorbed 

0 2 4 6 8 10 12 14 16 18 

Days after hatching (days of starvation) 

Figure 3. Relationship between days after hatching and survival rate (%) in starved Clarias gariepinus 

larvae.

Survival rate 

120 

100 

80 

60 

40 

20 

0 

Fed from D-4 

Fed from D-6 

Fed from D-8 

Fed from D-10 

Fed from D-12 

Fed from D-14 

Starved 

230 


yolk completely absorbed 

0 2 4 6 8 10 12 14 16 18 

Days after hatching 

Figure 4. Relationship between days after hatching and survival rate (%) in delayed feeding experiments 

of Clarias gariepinus larvae. See Table 1 for experimental design and symbols. 

Morphological development and feeding performance 

At hatching, the bony structure and barbel did not exist. The anus was observed to open at 16 h after hatching and 

the mouth opened at 22 h. The eye pigmentation was observed from 34 h and the development of barbel started as 

well. The cartilage structure of the upper / lower jaws appeared at 39.5 h. At 58 h, the upper / lower jaw structures 

started being ossified and the taste bud cell in the barbel started developing. At 67 h, the onset of feeding was 

observed. The bony structure thereafter developed and the cleithrum appeared at 70 h. At 112 h, the pectoral fin bud 

appeared and the cartilage of pectoral fin support appeared at 118 h. Morphological development at hours after 

hatching was described in Table 2. 

Table 2. Morphological development at hour after hatching. 

Hours after 

hatching 

0 

16 

22 

34 

58 

67 

70 

112 

118 

166 

Morphological features and 

feeding performance 

No bony structures and the eyes not pigmented 

Anus opened 

Mouth opened and barbel appeared 

Eye pigmentation started and upper / lower jaw structures 

appeared 

Upper / lower jaws started being ossified and taste bud cells 

in barbel appeared 

Onset of feeding 

Cleithrum appeared 

Pectoral fin bud appeared 

Pectoral fin support appeared 

Yolk completely absorbed

231 


Discussion 

Viveen et al. (1985) reported that the yolk sac of this species was absorbed by 3-day after hatching. However, in this 

study, the yolk sac was observed to retain until 166 h after hatching (7-day after hatching) although c. 98% of yolk 

sac was absorbed by 100 h (Figure 1). In other marine fish, the same trend, such as the drastic decrease in yolk sac 

volume in initial phase while it continued retaining with gradual decrease thereafter, was observed (Khono et al. 

1990). The onset of feeding was observed at 67 h after hatching and the period to utilize both endogenous and 

exogenous energy sources was elucidated for 100 h (more than 4 days) in this species. In pelagic marine fish larvae, 

generally, the overlapping period with endogenous and exogenous energy sources was much shorter such as 

observed in Chanos chanos (Kohno et al. 1990). 

The survival rate in food-deprived larvae was high until 10 days after hatching (Figures 3 and 4). In addition, the 

larvae that were given diet 8 days after hatching showed relatively high survival (more than 50%) while the ones not 

given diet for more than 10 days after hatching died (Figure 4). This indicates that the PNR (point of no-return, 

(Blaxter and Hempel 1963)) of C. gariepinus exists at around 9~10-days after hatching at 22.0-25.0ºC water 

temperature. This demonstrates that the larvae need to start feeding by the 8 th day after hatching, although the lower 

water temperature may give more allowance in time until reaching the PNR. In species that have an oil globule, i.e. 

Psetta maxima (M. Moteki, unpubl. data), the yolk sac is absorbed at around onset of feeding and the oil globule is 

retained longer. In such species, the survival of larvae after onset of feeding relies mainly on energy supply from the 

oil globule. However, C. gariepinnus is a species that does not have an oil globule, and the survival is considered to 

be influenced by the performance decline in the yolk sac. 

Rana (1985) reported that the yolk sac in Oreochromis mossambicus larvae was mainly consumed for the 

maintenance of body, not for growth. In this study, however, the larvae showed a continuous growth in TL even 

during the phase before the onset of feeding (0~67 h, Figure 1). This shows the larvae of this species utilizing the 

yolk sac both for the maintenance and growth of body. 

Although the anus and mouth opened at 16 and 22 h after hatching, respectively, the onset of feeding was observed 

at 67 h (Table 2). At the time of the onset of feeding, the initial ossification of jaw structures was observed as well as 

the occurrence of taste buds in the barbels. This process demonstrates that the larvae do not possess the feeding 

function before the start of ossification in jaw structures and the appearance of taste buds in barbel (Table 2). In 

several marine species as reported earlier, e.g. Chanos chanos (Morioka 1993) and Psetta maxima (M. Moteki, pers. 

comm.), the pectoral fin bud appeared before or at the onset of feeding. However, in C. gariepinus, the cleithrum 

appeared at 70 h (Table 2) and the pectoral fin structures appeared thereafter (112 h) (Table 2), showing that the 

pectoral fin is functionally not related with the onset of feeding in this species. This suggests that the larval C. 

gariepinus have different feeding manners from Chanos chanos, in an aspect of the utility of pectoral fin being 

relevant to swimming manner. In addition, Taki et al. (1987) reported that the swimming manner shifts from the 

undulation to caudal propulsion when the structures in caudal fin are ossified in Chanos chanos larvae, while it was 

observed that the larval C. gariepinus was not likely to have such change in the swimming manner. Although lots of 

fish species have pelagic larval stages and feed on suspended organisms as reported in marine fin fish species 

(Tanaka 1975), C. gariepinus are rather benthic even in pre-larval stage with negative phototaxis and it does not 

have the planktonic / pelagic larval phase. Such ecological differences seem to be relevant to the above-mentioned 

differences in morphological development. 

Conclusive note 

In this study, the preliminary information of the larval biology in relation to the onset feeding was obtained. The 

elucidated features are; 1) Clarias gariepinus larvae have the strong starvation tolerance supported by the long-term 

retention of yolk sac, 2) this species has a relatively long period to utilize both endogenous and exogenous energy 

sources, 3) the PNR of this species is around 9~10-days after hatching that indicates the larvae require to start 

feeding before reaching the PNR for higher survival, and 4) the larvae of this species are able to start feeding when 

they possess the ossified jaw structures and the appearance of taste buds in barbel. However, the survival of fish 

larvae is strongly influenced by the quality of bloodstock and eggs that also affect the subsequent development of 

feeding functions. In addition, the development of feeding functions is relevant to the development in morphology, 

especially in skeletal structures as shown in Taki et al. (1987). For further consideration on the survival and feeding 

aspects relevant not only to aquaculture but the biological features in the wild, the detailed investigation on the 

maternal factors such as egg quality and skeletal development would be necessary.

232 



We thank Mr. O. Matambo and his colleagues for assisting us in the hatchery activities. We are also grateful to 

Japan International Cooperation Agency (JICA) for giving us the opportunity to conduct this study. 

References 

Blaxter, J.H.S. and G. Hempel. 1963. The influence of egg size on herring larvae (Clupea harrengus). J. Cons. Int. Explor. Mer., 

28:211-240. 

Haylor, G.S. 1992. African catfish hatchery manual. Institute of Aquaculture, University of Stirling, Stirling. 

Kawamura, K. and K. Hosoya. 1991. A modified double staining technique for making a transparent fish-skeletal specimen. Bull. 

Natl. Res. Inst. Aquaculture, 20: 11-18. 

Kohno, H. 1998. Early life history features influencing larval survival of cultivated tropical finfish. In: Tropical mariculture, (ed.) 

S.S. De Silva, Pages 71-111, Academic Press, London. 

Kohno, H., M. Duray, A. Gallego and Y. Taki. 1990. Survival of larval milkfish, Chanos chanos, during changeover from 

endogenous to exogenous energy sources. Proc. Sec. Asian Fish. Forum, 437-440. 

Rana, K.J. 1985. Influence of egg size on the growth, onset of feeding, point-of-no-return, and survival of unfed Oreochromis 

mossambicus fry. Aquaculture, 46: 119-131. 

Taki, Y., H. Kohno and S. Hara. 1987. Morphological aspects of the development of swimming and feeding functions in the 

milkfish Chanos chanos. Japan. J. Ichthyol., 34(2): 198-208. 

Tanaka, M. 1975. Digestive systems in fish juveniles. In.: Feeding and Development of Fish Juveniles, Pages 7-23, (eds.) T. Iwai 

and H. Tsukahara, Kouseisya Kouseikaku, Tokyo. (in Japanese) 

Tsuchiya, M. 1976. Aquaculture of catfish. In: Freshwater Aquaculture, Pages 429-434, (ed.) D. Inaba, Kouseisya Kouseikaku, 

Tokyo. (in Japanese) 

Viveen, W.J.A.R., C.J.J. Richter, P.G.W.J. van Oordt, J.A.L. Janssen and E.A. Huisman. 1985. Practical Manual for the culture 

of the African catfish (Clarias gariepinnus). Ministry of foreign affairs of the Netherlands, Agricultural University of 

Wageningen and University of Utrecht, Hague. 

Watanabe, T. 1988. Fish nutrition. In: Fish Nutrition and Mariculture, Pages 1-77, Japan International Cooperation Agency, 

Tokyo. 

Wiggins, T.A., T.R. Bender, Jr., V.A. Mudrak and J.A. Coil. 1985. The development, feeding, and survival of cultured American 

shad larvae through the transition from endogenous to exogenous nutrition. Prog. Fish-Cul., 47: 87-93. 

Yamashita, T. and T. Aoyama. 1985. Hatching time, yolk sac absorption, onset of feeding, and early growth of the Japanese sand 

eel Ammodytes personatus. Nippon Suisan Gakkaishi, 51: 1777-1780.

233 


Effect of temperature on oocyte development of Oreochromis karongae 

(TREWAVAS) 

L.J. Kamanga, 1 E. Kaunda, 1 J.P. Mtimuni 1 , A.O. Maluwa 2 and M.W. Mfitilodze 1 

1 Bunda College of Agriculture, P.O. Box 219, Lilongwe. 

2 National Aquaculture Centre P.O. Box 44, Domasi. 

Abstract 

Oreochromis karongae (Trewavas) is one of the indigenous Tilapias that exhibit favorable traits for aquaculture in Malawi. However, 

fingerling production has been a problem. An experiment was therefore carried out to find the effect of temperature on oocyte 

development of the fish. Female O. karongae were reared under two temperature regimes, room (20.3±� 0.8 o C) and raised (26.5 ± � 

�0.5 o C) for 90 days while changes in gonadosomatic index (GSI) and oocyte developmental stages were followed every 45 days. 

Fish samples from the pond in which experimental fish were collected were used for comparison. Results showed that raising 

temperature to 26.5 ± 0.5 o C significantly enhanced oocyte development. Higher GSI (p

234 


circulating throughout the experimental period and temperature was monitored daily in the morning and in the 

afternoon. 

The fish were fed twice per day at 3 % body weight. Apart from the two treatments, some fish samples from the 

pond in which experimental fish were collected were used for comparison. 

Sampling of oocytes 

Nine fish from each treatment were sacrificed after every 45 days. The fish were weighed and dissected. Ovaries 

were removed, weighed and gonadosomatic index (GSI) was calculated as: 

GSI = gonad mass (g) x 100/ total body mass (g) (Crim and Glebe 1990) 

Macroscopic stages of the gonads were observed and classified according to Ricker, (1971). Ovaries from 3 fish in 

each treatment were fixed in Bouin’s solution for histological purposes immediately after post mortem. After 4-24 

hrs, they were washed in 50 % ethanol and stored in 70 % ethanol. 

Water quality management 

To maintain suitable water quality, biofilters were placed in each aquarium and air was supplied continuously. Fish 

excreta and leftover feed were siphoned out daily. Water in the aquaria was changed once every week. Apart from 

temperature, dissolved oxygen, pH, and ammonia levels were monitored. 

Histological analyses 

Histological analyses were done at the College of Medicine, University of Malawi, in Blantyre. The middle, distal 

and proximal ends of each ovary were prepared according to Hinton (1990). In summary, tissues was washed in a 

series of formalin, alcohol, wax and toluene and then embedded in wax blocks. The wax blocks were cut and 

stained with hematoxylin and eosin (H & E). The samples were then mounted on slides and observed under a 

microscope. 

Observations on ovarian tissue from histological samples were classified into five stages of maturity based on the 

abundant gametogenic cell types present (Table 1) after modification of the classification by Crim and Glebe (1990). 

Oocytes were observed under the microscope at x100 magnification. Three fields were sampled in each histological 

section and the different types of oocytes observed were counted. The number of a particular type of oocyte was 

then expressed as a percentage of the total number of oocytes observed in the three fields. 

Table 1. Classification of maturity stages of oocytes in Oreochromis karongae 

Stage Description 

Immature Previtellogenic oocytes: small, spherical ovarian cells containing a central nucleus and 

increasing amounts of cytoplasm. The type of oocytes were the perinucleolar oocytes. 

Maturing Vitellogenic oocytes: oocytes that had incorporated the yolky material produced by the 

liver. This consisted of early yolk vesicle and primary yolk granule oocytes. 

Mature These were largest oocytes and were filled with tertiary yolk granules 

Post 

follicles 

ovulatory Consisted of empty follicles and corpora lutea 

Dying Oocytes that showed signs of dying. 

Data analysis 

A t-test was used to analyze for differences in the GSI of fish. Percentages of oocytes from histological sections 

were transformed to natural logarithm (ln) and subjected to the general linear model (GLM) of Statistical Analysis 

System (SAS) package to test differences in the relative frequency of oocyte stages. Where significant differences 

appeared, means were separated using Scheffe’s test at 5 % probability.

Results 

235 


Gonad development 

Raised temperature was maintained at 26.5 ± 0.5 o C. Room temperature fluctuated between 18.5 ± 0.8°C in the 

morning and 19.2 ± 0.8 o C in the afternoon during the first 45 days. In the next 45 days temperature was 21.4 ± 

0.7°C in the morning and 22.0 ± 0.8 o C in the afternoon. In ponds the temperature was 22.0 ± 1.9 o C. Some fish died 

during the acclimatization period. The mortality rate in raised and room temperatures was 4 and 37 %, respectively. 

Fish from raised temperature had higher GSI than fish at room temperature (p

236 


Table 3. Mean gonadosomatic index of Oreochromis karongae females from raised and pond 

temperature 

Day Raised temperature Pond temperature 

Mean GSI SE Mean GSI SE 

0 0.13 0.012 0.13 0.012 

45 0.817 0.657 a 0.217 0.012 b 

90 1.133 0.471 a 0.257 0.027 b 

Means with different superscripts in the same row 

are significantly different (p

Percentage (%) 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

im mg mt 

Oocyte type 

po dy 

237 


room 

raised 

pond 

Figure 3. Number of immature (im), maturing (mg), mature (mt), dying (dy) oocytes and post ovultory 

follicles (po) expressed as a percentage of the total number of oocytes in each histological section after 

45 days of the experiment 

Percentage (%) 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

im mg mt po 

Oocyte type 

room 

raised 

pond 

Figure 4. Number of immature (im), maturing (mg), mature (mt) oocytes and post ovultory follicles (po) 

expressed as a percentage of the total number of oocytes in each histological section after 90 days of the 

experiment 

Temperature influenced (p

238 


Table 5. Mean number of immature, maturing, mature, dying oocytes and post ovulatory follicles 

expressed as a percentage (transformed to natural logarithm) of the total number of oocytes in each 

histological section at different times and temperature regimes. 

Period Treatment Oocyte type/Mean percentage transformed to natural logarithm 

Immature Maturing Mature *POF Dying 

Initial 3.21 1.79 1.2 0.87 0.81 

Day 45 Room 0.90±0.45 a 4.49±0.67 a 0.45±0.38 b 1.11±0.33 a 2.78±0.55 a 

Raised 0.14±0.28 a 1.54±0.41 b 3.15±0.24 a 0.45±0.20 a 0.83±0.34 b 

Pond 1.06±0.46 a 4.82±0.68 a 0.13±0.39 b 0.57±0.34 a 1.12±0.56 b 

Day 90 Room 1.63±0.44 b 4.30±0.36 a 0.07±0.30 b 0.73±0.33 a 0 

Raised 3.47±0.38 a 4.13±0.31 a 1.33±0.26 a 0.78±0.29 a 0 

Pond 1.34±0.28 b 4.05±0.23 a 1.72±0.19 a 2.02±0.21 b 0 

a, b 

Means with different superscripts within the same column and period are significantly different (p

239 


Figure 5. Oocytes from Oreochromis karongae reared at room, raised and pond temperature. A: Oocytes 

from fish samples at the onset of the experiment (24 th July 2000); B, C, and D: Oocytes from fish samples 

from room, raised and pond temperature after 45 days, respectively. E, F and G: Oocytes from fish 

samples from room, raised and pond temperature, respectively after 90 days 

Brummet (1995) suggested that temperatures outside the range of 22-36 o C during normal reproduction might act as 

a terminating cue (a factor that may interfere with final ovulation and spawning) in Tilapia. Since room temperature 

was outside the ideal range for oocyte development, oocyte development in fish from room temperature was inferior 

to that from raised temperature. Poor oocyte development observed at room temperature may mean that the low 

temperature interfered with the release of GtH by the hypothalamo-hypophyseal complex, which in turn slowed 

down vitellogenesis. Barnabe (1994) stated that temperature directly affects the production of GtH and eventually 

affect the vitellogenesis process. Jalabert and Zohar (1982) reported that exogenous vitellogenesis seems to be 

completely inhibited by low temperatures and all yolk-bidden oocytes disappear during the cold season.

240 


Low temperatures (below 20 o C) also reduce activity and feeding of tilapias (Chervinski 1982, Piper et al. 1982). 

Reduced feeding could be the other factor that might have contributed to poor ovarian development in fish raised 

under room temperature. Generally, a relatively lower feed intake was observed in fish under room temperature. 

The reduced feeding may have resulted in the fish using fat reserves from the body to provide metabolites for 

survival. Barnabe (1994), reported that if fat reserves are totally depleted, gonads provide the metabolites which 

allow the animal to survive. This may have occurred in fish that were kept at room temperature thereby resulting in 

poor gonad development in the fish. 

At the onset of the experiment, there were more immature oocytes than during the rest of the sampling days. This 

was expected for two reasons. Firstly, the experimental fish were still growing and therefore oocytes were just 

developing at that time. Secondly, this was the coolest time of the year (July), which is outside the breeding period 

of O. karongae (Brooks and Maluwa 1997). Therefore, it was expected that the ovarian activity would be dormant, 

hence the advanced oocyte stages were not expected to dominate. The percentage of immature oocytes was lowest 

after 45 days of the experiment and increased again after 90 days. Increase of immature oocytes after 90 days may 

mean that the gonadal regression that was taking place in the ovaries stimulated development of more oocytes. 

Mature oocytes were most abundant in raised temperature. This was expected because this temperature level was 

within the range suitable for normal oocyte development in Tilapia (Chervinski 1982, Brummet 1995). The mature 

oocytes were also present throughout the experimental period because immature and maturing oocytes that were 

present were developing further. 

The high relative frequency of maturing oocytes from the pond and room temperature may reflect the unfavorable 

conditions for oocyte development. Oocyte development may have been arrested to this stage while waiting for 

suitable conditions for development. It is however interesting to note that the proportions of maturing oocytes 

increased with time and were highest after 90 days implying that development from immature to maturing oocytes 

was enhanced with time. As room and pond temperatures were not controlled, the temperature increased with time 

due to change in season. The temperature was thus getting more favorable for the oocytes to develop from 

immature to maturing types. 

The presence of almost all the maturity stages of oocytes in the ovaries are in agreement with what Msiska (1998) 

found. The results show that O. karongae has asynchronous ovaries. DeVlaming (1983) as cited by West (1990) 

reported that most fish with asynchronous ovaries have protracted spawning seasons with multiple spawnings. 

Therefore there may be a possibility that O. karongae can have protracted spawning seasons and/or multiple 

spawnings. This is evident from studies on the breeding season of O. karongae, which show that it has a prolonged 

spawning period in Lake Malombe (July-October) while in Lake Malawi it has two spawning peaks, August- 

October and December-February (Pálsson et al 1999). In the southern part of Malawi (Thyolo), farmers report that 

O. karongae spawns throughout the year except during the cold season, which is from May to July (Jamu 2001, 

personal communication). The results of the current experiment show the possibility to circumvert the seasonal 

breeding of O. karongae in hatcheries by raising temperature to 26.5 ± 0.5 o C during the cold season. 

References 

Admassu, D. 1996. The breeding season of Tilapia, Oreochromis niloticus L. in Lake Awassa (Ethiopian rift valley). 

Hydrobiologia, 337:77-83. 

Barnabè, G. 1994. Biological basis of fish culture. In Barnabè, G. (ed.) Aquaculture: Biology and ecology of cultured species. 

Ellis Horwood. New York. pp. 227-372. 

Bromage, N.R. 1995. Broodstock management and seed quality. In Bromage, N.R. and Roberts, R.J. (eds.). Broodstock 

management and egg and larval quality. Blackwell Science Ltd. Oxford UK. pp. 1-24. 

Brooks, A. C. & Maluwa, A. O. 1997. Fish farming in Malawi: A case study of the Central and Northern Regions Fish Farming 

Project. Technical Supplement: On-station trials. Government of Malawi and European Commission. 37p. 

Brummett, R. E. 1995. Environmental regulation of sexual maturation and reproduction in Tilapia. Reviews in Fisheries Science, 

3:231-248. 

Chervinski, J. 1982. Environmental physiology of Tilapias. In R.S.V. Pullin and R.H. Lowe-McConnell (eds.) The biology and 

culture of Tilapias. ICLARM Conference Proceedings 432p. International Center for Living Aquatic Resources 

Management, Manila, Philippines. p.119-128. 

Chmilevskiy, D.A. 1995a. Effect of reduced temperature on oogenesis of the mouth brooder, Oreochromis mossambicus 2. Effect 

on fish twenty-two days after hatching. Journal of Ichthyology, 35:72-80. 

Chmilevskiy, D.A. 1995b. Effect of reduced temperature on oogenesis in Tilapia, Oreochromis mossambicus. 3. Impact on fish at 

age 30 and 60 days after hatching. Journal of Ichthyology, 35:119-129. 

Chmilevskiy, D.A. and Lavrova, T.V. 1990. The influence of temperature on oogenesis in Tilapia, Oreochromis 

mossambicus. Journal of Ichthyology, 30:14-24.

241 


Crim, L.W. & Glebe, B.D. 1990. Reproduction. In Shreck, C.B. and Moyle, P.B. (ed.) 1990. Methods for fish biology. American 

Fisheries Society. Bethesda, Maryland. pp. 529-553. 

Das, P. & Ponniah, A.G. 1993. Research needs in fish breeding and genetics in aquaculture for the year 2000. In aquaculture 

research needs for 2000 A.D. AA Balkema/Rotterdam. India. pp. 181-195. 

Hinton, D.E. 1990. Histological techniques. In Shreck, C.B. and Moyle, P.B. (ed.) 1990. Methods for fish biology. American 

Fisheries Society. Bethesda, Maryland. pp. 191-211. 

Jalabert, B. & Zohar, Y. 1982. Reproductive physiology in cichlid fishes with particular reference to Tilapia and Sarotherodon. 

In R.S.V. Pullin and R.H. Lowe-McConnell (eds.) The biology and culture of Tilapias. ICLARM Conference 

Proceedings 432p. International Center for Living Aquatic Resources Management, Manila, Philippines. p.129-140. 

Msiska, O.V. 1998. Reproductive biology and growth of Oreochromis (Nyasalapia) karongae in ponds and open water in 

Malawi. Ph.D. Thesis. University of Malawi. 221p. 

Msiska, O.V. & Costa-Pierce, B.A. 1997. Factors affecting the spawning success of Oreochromis karongae (Trewavas) in ponds. 

Aquaculture Research, 28:87-99. 

Pálsson, Ó. K., Bulirani, A. & Banda, A. 1999. A review of biology, fisheries and population dynamics of chambo (Oreochromis 

spp., CICHLIDAE) in Lakes Malawi and Malombe. Government of Malawi. Fisheries Department. Fisheries Bulletin 

No.38. 35p. 

Piper, R.G., McElwain, I.B., Orme, L.E., McCaren, J.P., Fowler, L.G., & Leonard, J.R. 1982. Fish Hatchery management. US 

Department of the Interior Fish and Wildlife Services. Washington D.C. 517p. 

Ricker, W.E. 1971. Methods for assessment of fish production in fresh waters. IBP Handbook 3:348p. 

SAS. 1996. The Statistical Analysis System. SAS Institute Inc. Cary, NC 27513. USA. 

Srisakultiew, P. & Wee, K.L. 1988. Synchronous spawning of Nile Tilapia through hypophysation and temperature manipulation. 

In Pullin, R.S.V.,Bhukaswan, T., Tonguthai, K. and Maclean, J.L. (eds.). The second international symposium on 

Tilapia on aquaculture. 623p. Department of Fisheries, Bangkok, Thailand and International Center for Living Aquatic 

Resources and Management. Manila, Philippines. ICLARM conference proceedings. p.275-284. 

Subasinghe, R.P. & Somerville, C. 1992. Effects of temperature on hatchability, development and growth of eggs and yolk-sac 

fry of Oreochromis mossambicus (Peters) under artificial incubation. Aquaculture and Fisheries Management, 23:31- 

39. 

West, G. 1990. Methods of assessing ovarian development in fishes: A review. Aust. J. Mar. Freshwater Res. 41:199-222.

242 


The Effect of Feeding Density on Survival 0f African Cat Fish Clarias gariepinus 

H. K. Zidana & A.O. Maluwa 

National Aquaculture Centre, P.O. Box 44, DOMASI. Malawi. E:mail-nac@sdnp.org.mw 

Abstract 

Results on the study of the relationship between feeding density and survival of Clarias gariepinus fry are presented. This study was 

carried out at National Aquaculture Centre with the objective of finding out the optimum number of zooplankton that can improve the 

survival of fry to fingerling stage. C. gariepinus, fry 4 days, old were stocked in 50 litre tanks for 30 days at a rate of 20 fry in each 

tank. Four zooplankton densities were used 5, 10 and 30 zooplankton/ml of water, the control was 0 zooplankton /ml of water. Each 

density was replicated 3 times and every 3 days the number of fish in each tank was counted. Highest survival rates (80%) were 

obtained from tanks in which the fish were fed 30 zooplankton/ml and these were significantly different (P

Copepoda 

Figure 1. Zooplankton used in the feeding experiment. 

Table 1. The lay out of the experiment according to different feeding levels 

Tank1 

5 zooplankton/ml 

Tank2 


Tank3 


Tank4 


Tank5 


Tank6 


243 


Tank7 


Tank8 

0 zooplankton 

Tank9 


Tank10 


Tank11 

30 zooplankton 

Tank12 



After 30 days of carrying out the experiment, the results showed that fish fed with 30 zooplanktons/ml had 80% 

survival rate, fish fed with 10 zooplanktons had survival rate of 67%, fish fed with 5 zooplanktons had a survival of 

50% and fish not fed with zooplankton had 0% survival rate. 

All the fish in the 0 zooplankton/ml were dead by the 15 th day of the experiment and the experiment continued with 

the other three feeding levels as shown in Figure 2. 

SURVIVAL RATE (%) 

100 

80 

60 

Cladocera Rotifer 

40 

30 Zpl/ml 

10 Zpl/ml 

20 

5 Zpl/ml 

0 

0 Zpl/ml 

6 9 12 15 18 21 24 27 30 

CULTURE PERIOD (DYS) 

Figure 2. The trend of survival rate % of African catfish fed different density of zooplankton/ml of water 

during the period of the experiment. 

The results showed that there is a positive relationship between the zooplankton density and the survival rate of the 

fry i.e. the higher the zooplankton density the higher the survival rate of fry. However the number of zooplankton 

can only be increased up to a certain point, if the number of zooplankton increases up to 50/ml all the fry were dead 

within 3 days of culture.

244 


Apart from the feed density other factors also contributed to the death of the fry in this trial. In 30 zooplankton/ml 

feeding density the death of fry was mainly contributed by the aggressive territorial behavior of the fry. This was 

also reported by Hetch and Applebaum (1988), in their trial apart from cannibalism death of fry was due to fatal 

aggressive territorial encounters between two or more individuals which is a common behavior in young Clarias 

gariepinus. 

The lowest survival rate was in the 0 zooplankton/ml density was due to starvation. There were evidences of type-1 

cannibalism during the first 7 days of culture and after that the fry were very weak so much and they died due to 

starvation. 

The data analysis showed that there is a significant difference (P < 0.05) in feeding the fish with 30 zooplankton/ml 

against all the feeding densities and that there is no significant difference (P < 0.05) in feeding the fish with 5 and 10 

zooplankton/ml. However all the feeding densities were significantly different (P

245 


Determination of the time interval between the prime and resolving dosages of 

Cyprinus carpio hormone dose to artificially induce Opsaridium microlepis to 

breed 

P. Kataya, H. Zidana & A.O. Maluwa 

National Aquaculture Centre, P.O. Box 44, Domasi, Malawi. 

Abstract 

The Results on the study of final oocyte maturation of Opsaridium microlepis, “ Mpasa”, in response to the first (prime) dose and the 

second (resolving) dose of Cyprinus carpio pituitary gland are presented. The objective was to determine the optimum time interval 

between the prime and resolving dosages to induce ovulation in the fish. Mature broodstock of 250g, that were collected directly 

from their wild source, i.e. Linthipe River, were injected with C. carpio pituitary gland extract at the rate of 5mg/kg female body 

weight. This dose was split into two, the prime and the resolving dosages. The fish received 10% of the total dose as a prime and 

90% as the resolving dose, at three time intervals of 9, 12 and 15 hours separating the dosages. All fish ovulated when the time 

interval between the first and second injection was 12 hours. When the eggs were fertilized, hatchability was 76.6%. The fish did not 

ovulate when the resolving dosages were given at 9 and 15 hours. The results are discussed with reference to application of the 

protocols developed in this study to produce mass fingerling of this species for aquaculture. 

Introduction 

Opsaridium microlepis locally known as ‘mpasa’ is endemic to Lake Malawi and it attains a weight of at least 2.5kg 

(Msiska 1990). The fish is commonly caught in the northern and central regions of Malawi, where it ascends rivers 

like Songwe, Bua and Linthipe to breed in clear running water in the rivers. O. microlepis is an important 

commercial species of Lake Malawi. The fish fetches high prices on the market and this has led to its over 

exploitation and hence the catches from the wild have declined (Tweddle 1981 in Msiska 1990). 

Research on the suitability of indigenous cyprinids for aquaculture has not been done in detail, despite the fact that 

Malawi has cyprinids in the genera, Barbus, Labeo and Opsaridum (Maluwa et al. 1997). In other countries 

cyprinids have been farmed in aquaculture for their high value and export potential, for example the Norwegian 

Salmon and Rainbow trout. 

The National Aquaculture Centre is screening indigenous cyprinids for their suitability in aquaculture in Malawi and 

O. microlepis would be the most suited candidate for commercial aquaculture development in Malawi because of its 

large maximum body weight, high market value and demand. 

Initial studies on the Gonad Somatic Index (GSI) show that the breeding season for the species is between May and 

August. As a fish that migrates to breed in clear running waters in the river, O. microlepis require artificial induced 

breeding. Previous efforts to breed the fish did not succeed because the time interval between the prime and 

resolving dosages were unknown. The objective of this study was to determine the optimum time interval between 

the prime and resolving dosages that would induce ovulation in the fish. 


This study was conducted in the hatchery at National Aquaculture Centre (NAC). Mature females and males, of 

250g body-weight were collected from Linthipe river in the central region. The fish were collected during the 

months of January – October, as they migrated to breeding grounds. The fish were transported to NAC and after 

initial quarantine exercise in tanks they were taken into hatchery for artificial induced breeding. 

The fish were given a dose of 5mg/kg body weight of C. carpio pituitary gland extract. Of this dose, 10% was given 

as a prime dose and the rest 90% was given as a resolving dose. The gland was removed from a donor C. carpio 

obtained from the farm at NAC. The treatments were the time interval between the prime and resolving dosages. 

There were 3 time intervals that were tested, which were, 9, 12 and 15 hrs. Each treatment was replicated 4 times. 

The males were given 50% of the female dosege i.e. 2.5mg/kg body weight only once, when the females were given 

the resolving dose. 

The oocyte maturation stages were determined by sampling the oocytes from the ovary by using a plastic tube, and 

then observing oocytes under the stereoscopic microscope. In order to observe the nucleus of the oocytes, they were 

placed in petri dish with Sera’s solution (60 % ethanol, 30 % formalin and 10% acetic acid). In maturing oocytes the 

nucleus (Germinal Vesicle) migrates to the animal pole where it breaks and disappears in mature oocytes. This

246 


phenomenon known as geminal vesicle migration and break down was used as a biological indicator to determine 

the oocyte maturation stages as shown in 

Figure1: Discrimination on position of Germinal Vesicle in Oocytes with Sera’s solution 

When oocyte development processes are completed (vitellogenesis), the oocytes reach stage I of maturation, in 

which the germinal vesicle is at the center. The final oocyte maturation process follows the migration of the 

germinal vesicle from the center to the animal pole, where it breaks down at stage V and the oocyte is ovulated into 

a fertilizable egg in stage VI. When the egg reaches stage VII, it is degenerated and hence not fertilizable. This 

phenomenon is known as over ripening. The injected brood stock were put in 500 litre conical tanks at a ratio of 1:1 

(Male:female). The fish were expected to ovulate within 18 hours at the incubation temperature of 26 0 C. After 18 

hours, the fish were stripped off their eggs and artificially fertilized with sperm contained in milt that was stripped 

from the males. The fertilized eggs were incubated in zug jars in the laboratory at NAC. The incubation temperature 

was 26 0 C. The eggs hatched after 24 hours and hatchability was determined from the total number of fertilized eggs 

and that number which hatched. 

Results 

All the fish that were given the second resolving dose 9 hrs and 15 hrs after the prime dose did not ovulate. The 

oocytes in fish given the resolving dose at 9 hours passed through the stages of oocyte maturation and degenerated 

in stage VII. These eggs did not hatch when fertilized. The fish that were given a resolving dose 15 hours after the 

prime dose, their oocytes never matured beyond stage IV, i.e. the oocyte maturation processes were arrested at stage 

IV. These eggs when stripped and fertilized after 18 hours, did not hatch. Ovulation at stage VI occurred when the 

fish were given a resolving dose 12 hours after the prime dose. These eggs when stripped were fertilizable and 

hatched after 24 hours of incubation. The hatchability rate was 76.6%. The relationship between time intervals, 

number of eggs stripped, fertilization and hatchability rates is shown in Table 1. 

Table 1: The relationship between time interval separating the prime and resolving dosages and number 

of eggs produced and fertilization and hatchability rates. 

Time Interval (Hours) Eggs Produced Fertilization rate (%) Hatchability (%) 

9 

12 

15 

6,500 

6,994 

6,000 

There was no significant difference in the number of eggs stripped across the three time intervals. However, the 12 

hour interval gave significantly (P< 0.01) higher rate of fertilized eggs and hatchability (Table 1). 

0 

84 

0 

0 

76.6 

0

247 


Discussion 

Most of the work on induced breeding of cyprinids, show that the fish require two injections, i.e. the prime and 

resolving dosages. The prime dose is used to induce final oocyte maturation and the second dose to induce 

ovulation. Similar work on cyprinids with the Indian major carps, the time period between the prime and resolving 

dosages was 6 hours. The priming dose was 20% and the resolving dose was 80%. With these dosages and times, 

spawning was expected 4 to 6 hours after the resolving dose (Pullin and Jhingran 1985). The results in this trial 

show that O. microlepis has lower priming dose of 10% and higher resolving dose of 90%. In addition O. microlepis 

has a longer oocyte maturation period (18 hours). 

Having defined the right time interval for the fish to breed successfully, further work is underway to establish the 

production of mass fingerlings using the protocols developed in this study. 


This work forms part of the JICA Aquaculture Project activities at the National Aquaculture Centre. The project was 

jointly funded by the Malawian and Japanese Governments. 

References 

Maluwa, A. O., Kamakura, M & Sakai, K. 1997. Induced Maturation and Ovulation of Grass Carp (Ctenopharygodon idellus). In 

proceedings of the first Regional Workshop on Aquaculture, Bunda College of agriculture, Lilongwe, Malawi. 

1997:86-92. 

Msiska, O.V. 1990. Reproductive Strategies of two cyprinid fishes in Lake Malawi and their relevance for aquaculture 

development. In aquaculture and Fisheries Management 1990, 21, 67-75. 

Msiska, O. V. 1991. Induced spawning and Early growth of Mpasa (Opsaridium microlepis) and Ntchila (Labeo mesops) in tanks 

and earthern ponds. In Aquaculture Research in Africa, Pudoc Wagenigeni,:P 159-168. 

Pullin, R. S. V & Jhingran, V.G.1985. A hatchery manual for the common, Chinese and Indian major carps. ICLARM studies 

and Reviews 11, 191 p. Asian Development Bank, Manila, Phillipines and International Centre for Living Aquatic 

Resources Management, Manila, Phillipines. 

Tweddle, D. 1987. An Assesment of the Growth Rate of Mpasa, Opsaridium miclolepis, by length frequency analysis. In Journal 

of Limnology Society Southern Africa 13(2), 52-57.

248 


Determination of biological reference points for Lake Malawi cichlids 

Anthony J. Booth 

Department of Ichthyology and Fisheries Science, Rhodes University, P.O. Box 94 Grahamstown 6140, SOUTH AFRICA Email: 

t.booth@ru.ac.za 

Abstract 

Biological reference points (BRPs) - values that represent the state of a fishery or population - are commonly used to guide 

management decisions. BRPs are expressed as a function of fishing mortality and are derived from models that describe the 

population dynamics of a resource. These models include dynamic pool, spawner-recruit and production models. Unfortunately, in 

the absence of long-term fisheries data, such as abundance indices and catches-at-age, it is only dynamic pool models, usually 

known as yield-per-recruit and spawner biomass-per-recruit models, that are known with any degree of certainty. BRPs that have 

been proposed in the literature have been calculated for temperate demersal marine species with life-history characteristics that 

include high fecundity, pelagic spawning and little or no parental care. In contrast, cichlids have low fecundities and exhibit varying 

levels of parental care ranging from guarding to mouthbrooding. Cichlid specific BRPs, calculated with combinations of two 

spawner-recruit relationships within deterministic and stochastic frameworks, were shown to be highly dependent on the degree of 

density-dependence in the spawner-recruit relationship. Overall, cichlid specific BRPs were similar to than temperate groundfish 

species and it is suggested that fishing mortality be maintained at no more than that required to reduce spawner biomass-per-recruit 

to 40% of pristine levels if the form of the spawner-recruit relationship is unknown. If a Beverton-Holt spanwer-recruit relationship is 

assumed, a biologically plausible relationship given cichlid life-histories, then it is suggested that spawner biomass-per-recruit is not 

reduced below 50% of unfished levels. 

Introduction 

Biological reference points (BRPs) are commonly used to guide management decisions. BRPs are values that 

represent the state of a resource and whose characteristics are believed to be useful for the management of a unit 

stock through the provision of information regarding its status relative to an acceptable value or range (Caddy and 

Mahon, 1995). There is also a need by management to quantify potential long-term trade-offs between maximising 

yield, either by mass or economic gain, while limiting the risk of possible reproductive failure. Applying a constant 

fishing mortality to maximise yield i.e., FMSY, would clearly be advantageous, however, knowledge of the biomass 

required to achieve MSY is difficult to determine, principally because of the lack of information on the functional 

form of the spawner-recruit relationship, hampering understanding of the biomass-production relationship 

(Sissenwine and Shepherd, 1987; Clark, 1991). 

BRPs are most often expressed as a function of fishing mortality and are derived from models that describe the 

population dynamics of a resource. These interrelated models include dynamic pool, spawner-recruit and 

production models. For a graphical illustration see Sissenwine and Shepherd (1987). Unfortunately, in the absence 

of data, which is usually the case for most fisheries including those that are cichlid dominated, only the dynamic 

pool models, often termed as yield-per-recruit and spawner biomass-per-recruit, are known with any degree of 

certainty. Quantifying and understanding the spawner-recruit relationship and the productivity of the resource at 

various abundances is, therefore, imperative. 

BRPs that have been proposed in the literature have been calculated for temperate marine species with life-history 

characteristics that include high fecundity, pelagic spawning and little or no parental care (Clarke, 1991; 1993; Mace 

1994). In contrast, cichlids have relatively low fecundities and have various levels of parental care ranging from 

guarding to mouth brooding (Lowe, 1959; Trewavas, 1983). Could BRPs be similar for two groups of fishes with 

different life-history characteristics? This study provides the first quantitative examination into the relationships 

between yield and spawner biomass-per-recruit for a wide variety of cichlid specific life-history parameters, in an 

attempt to propose BRPs that would be beneficial for management. 


Modelling framework 

A two-phase age-structured modeling framework was developed. Firstly, a deterministic model incorporating a 

stock-recruit (S-R) relationship was used to estimate instantaneous rates of fishing mortality (F) required to achieve 

several BRPs. The procedure was based on those outlined by Sissenwine and Shepherd (1987) and Clarke (1991). 

The second phase included a stochastic extension to test the robustness of the BRPs and to assess the trade-offs 

between yield and spawner biomass, and the potential risk of reproductive failure.

249 


The stochastic model was similar to the deterministic model but incorporated recruitment variability and simulated 

an age-structured population that was fished at the desired fishing mortality necessary to achieve the desired BRP. 

2 

ε −σ 

/ 2 

2 

Age-structure within the population was estimated with stochastic recruitment R = e with ε ~ N ( 0, 

σ ) . 

2 

−σ 

/ 2 

The constant e ensured that E [ R] 

= 1. 

Age-structure was estimated by projecting the population forward by 

20 years at the required fishing mortality rate. This “burn-in” period provided the initial age structure for a 50-year 

projection at the required fishing intensity. It was assumed that a CV of 0.4 was sufficient for the recruitment 

variability such that σ ≈ CV = 0. 

4 . Error was assumed to occur in the BRP determination (for reasons such as 

sampling and ageing errors) such that F = Fˆ 

BRP e where FBRP ˆ is the predicted BRP fishing mortality and 

2 

ε ~ N( 

0, 

0. 

1 ) . A total of 100 50-year simulations were conducted. 

0. 1 

2 

2 

ε − 

Dynamic pool models 

Two dynamic pool models, yield-per-recruit and spawner biomass-per-recruit as functions of fishing mortality F, 

were calculated as 

where wa is the mass at age a, Sa is the selectivity for age class a, Ma is the rate of natural mortality at age a, ψa the 

maturity-at-age a and max is the maximum recorded age and is considered a lumped-plus group. A time step of one 

month (∆a) was used in the simulations. 

The relative proportion of fish at age a ( N a 

~ ) was expressed as recursively as 

~ 

N 

a 

Weight-at-age was described as [ ( ) ] ω −Ka 

Wa 

ψ L∞ 

− e 

selection-at-age (Sa) and maturity-at-age (ψa) was estimated by a logistic of the form ( ) 1 

−( 

a−a50 

) / δ − 

= 1 where ψ and ω describe the length-weight relationship, 

P a = 1+ 

e 

where 

Pa is the proportion selected/mature at age a, a50 the age-at-(50%) selection/maturity and δ the steepness of the 

ogive. 

Spawner-recruit models 

The Beverton-Holt (1957) and Ricker (1954) stock-recruit (S-R) curves are expressed as ( ) 1 − 

R 

YPR 

⎧ 

⎪ R = 1 

⎪ ~ 

= ⎨ N a−1e 

⎪ ~ 

N a−1e 

⎪ − 

⎩ 1− 

e 

α′ 

Se 

F 

= 

−FS 

−FS 

FS 

max 

∑ 

a= 

0 

a−1 

a−1 

max 

⎡ ~ 

⎢wa 

N 

⎣ 

−M 

−M 

−M 

a = 0 

0 < a < max 

a = max 

( 1− 

e 

−S 

F −M 

R = S α + βS 

and 

−β 

′ S 

= , respectively where α ,α′ , β and β ′ determine the shape of the curves, R is the predicted 

recruitment at a specified level of spawner biomass S. 

a 

Sa 

F 

S F + M 

a 

a 

a [ 

a 

⎤ 

) ⎥∆a 

⎦ 

~ 

w N ] ∆a 

Both S-R curves were reparameterised as a function of a “steepness” parameter h , which expresses that fraction of 

pristine recruitment (R0) when spawner biomass (S0) is reduced to θ % of pristine levels, such that hR0 = f ( θS0 

) 

where f (⋅) 

is the S-R curve of interest (Figure 1). This formulation reduces valid portion of the S-R curve between 

the origin and the replacement point (S0, R0). The “steepness” parameterisation was originally proposed by Mace and 

Doonan (1988) but most commonly ascribed to Francis (1992). 

SBR 

max 

F = ∑ 

a= 

0 

a 

a

250 


Figure 1. Relationship between recruitment and spawner biomass to the replacement line that joins the 

points (0,0) and (R0, S0) and can be used to reparameterise the spawner-recruit relationship to a single 

“steepness” parameter h. The “steepness” parameter h expresses the fraction of pristine recruits when 

spawner biomass is reduced to θ% of pristine levels. In this scenario θ = 0.2. The solid lines are the 

Ricker and the dotted lines the Beverton-Holt curves used in the analysis. 

When the S-R shape parameters are simultaneously estimated then 

θS0 

( 1− 

h) 

α = 

hR ( 1−θ 

) 

0 

α θ 

, 

h −θ 

β = 

hR0 

( 1−θ 

) 

for the Beverton-Holt curve, and 

θ ⎡ ⎛ hS ⎞ ⎤ 

0 R) 

ln( hθ 

) 

′ = exp⎢ln 

⎜ 

⎟ /( 1−θ 

) ⎥ , β ′ = for the Ricker curve. 

⎣ ⎝ θS0R 

⎠ ⎦ S0 

( 1−θ 

) 

When h = 0.2, the replacement line in Figure (1), then the equations for the Beverton-Holt curve reduce to the 

S0 ( 1− 

h) 

familiar = 

4hR0 

θ 

5h 

−1 

α , β = . For the purposes of this study, pristine recruitment was set to unity, and 

4hR0 

therefore, in the S-R model pristine biomass (S0) was replaced by unexploited spawner-biomass per-recruit (SBR0). 

As recruitment was estimated from a per-recruit perspective, equilibrium spawner biomass as a function of fishing 

SBRF 

−α 

ln( α′ 

SBRF 

) 

mortality F was calculated as S * = 

for the Beverton-Holt and S* = 

for the Ricker 

β ′ 

β ′ 

curves. Equilibrium recruitment * R was estimated by substituting S* into the relevant S-R curves (Figure 5). 

Production model 

Equilibrium yield as function of spawner biomass or spawner biomass-per-recruit was calculated as Y* = YPRF 

R* 

because spawner biomass-per-recruit was calculated as a function of fishing mortality F. Yield as a function of 

spawner biomass or spawner biomass-per-recruit was obtained by iteratively solving for a specific fishing mortality 

required to reduce the spawner biomass or spawner biomass-per-recruit to the desired level.

251 


Parameter estimates used in the simulations 

A review of the cichlid literature summarizing trends between growth, maturity and natural mortality is illustrated in 

Figure 2. Growth and mortality data were selected from studies that aged fish with hard tissues such scales and 

otoliths, with natural mortality calculated using Hoenig’s (1983) empirical formula that relates mortality as a 

function of maximum age where ln( M ) = 1.44 - 0.982ln( max) 

. Length frequency methods of ageing are 

fraught with error because cichlids are long lived and their growth zones often stack in their otoliths (Booth et al., 

1995; Booth and Merron, 1996; Weyl and Hecht, 1998) decreasing age estimates of old fish. Length frequency 

analysis, therefore, over-estimates the growth co-efficient of the von Bertalanffy growth equation and introduces 

substantial errors into the commonly used Pauly’s (1980) estimate of natural mortality. It was assumed that Hoenig’s 

formula was most suitable for cichlid species because of life-history characteristics such as longevity, low fecundity 

and parental care. The regression results of natural mortality vs the growth co-efficient was unfortunately weak, 

only explaining 16% of the variation, but illustrates a slight negative correlation between M and K. 

Based on the regression equations in Figure 2, and the average growth co-efficient from the literature cited in Figure 

2, parameter estimates used in the base-case simulations are L∞ = 200 mm, K = 0.38 .yr -1 , ψ = 0.01 mm, ω = 3 .yr - 

1 -1 

, M = 0.45 .yr , ψ 50 = S50 = 2.16 yrs, and δψ = δS = 0.25 .yr -1 . A base-case scenario assumed that both the maturity 

and selectivity schedules coincided, whereas an alternative scenario, closely resembling most cichlid harvesting 

gears, assessed selection by the fishing gear one year prior to maturity i.e., ψ 50 = 1.16 yrs. 

Selection of values that determine maximum theoretical length and the length weight parameters are essentially 

arbitrary, provided that growth in weight is (close to) isometric to length. The most influential parameters are, 

therefore, the ratio of K to M and the ratio between selectivity and maturity. A maximum age of 10 years was 

considered sufficient as a plus group and had little effect on the analyses. 

Figure 2. Relationships between length-at-maturity ( 50 

ψ ) and maximum length (L∞) (top panel), and 

natural mortality (M) based on maximum age (max) and Brody’s growth coefficient (K) (lower panel). 

Lines represent linear regression and their 95% confidence intervals. Data from Bruton and Allanson 

(1974), Bruton and Boltt (1975), Tweddle and Turner (1977), Hecht (1980), Van der Waal (1985), Booth 

et al. (1995), Booth and Merron (1996), Jambo, (1997) and Weyl and Hecht (1998).

252 


The S-R curve reparameterisation expresses relative recruitment as a function of relative spawning biomass, both 

measured relative to unfished levels, and described by the “steepness” parameter θ. The value of the “steepness” 

parameter determines the potential for density-dependent increases in specific reproductive rate – the relative 

increase in recruits per unit spawning biomass. Three curves were estimated corresponding to relative increases in 

reproductive rate of 4, 8 and 16 times. Essentially when spawner biomass is reduced to extremely low levels 

spawning is 4 - 16 times more successful. Clarke (1991) argued that this range is acceptable as it covers the 

extremes of most known S-R relationships, albeit for temperate groundfish species. The values of h, when θ = 0.2, 

were estimated as 0.749, 1.249, 2.070 for the Ricker and 0.566, 0.711, 0.823 for the Beverton-Holt S-R curves. 

These values were estimated by setting θ = 1×10 -5 and h = 4 – 16 times θ and solving for the fishing mortality that 

would result in extinction, i.e., R = 0. The resultant extinction F estimates are used to recalculate h when θ = 0.2, a 

commonly used value for this parameter used in meta-analysis of stock-recruit data for stock assessments (Myers et 

al., 1985). 

Estimation of biological reference points 

Various commonly used BRPs were calculated in this study. These were FMAX (that fishing mortality that maximise 

yield-per-recruit in the absence of a S-R relationship), F0.1 (that fishing mortality that is 10% of the slope of the 

yield-per-recruit curve at the origin in the absence of a S-R relationship) (Gulland and Boerema, 1973), FSB40 (that 

fishing mortality that reduces spawner biomass-per-recruit to 40% of unfished levels in the absence of a S-R 

relationship) and FMMY (that fishing mortality that is the maximum of all minimum yields given the predetermined S- 

R relationships) (Clarke 1991). 

During the stochastic simulations, cumulative catch over the simulation period, depletion (spawner biomass as a 

fraction of pristine levels) at the end of the 50-year simulation, the lowest deletion during the simulation and the 

number of times depletion was reduced below 20% were noted. 

Results 

Typical per-recruit effects can be noticed in Figures 3 and 4 (top left panels) with relative yield-per-recruit 

increasing quicker if fish are harvested before sexual maturity. Similarly, spawner biomass-per-recruit decreased 

more rapidly with small increases in fishing mortality in the second scenario with earlier gear selection. 

The relationship between spawner biomass-per-recruit and yield provides a proxy strategy to the spawner biomass 

approach (Figures 3 and 4 - bottom panels). If spawner biomass-per-recruit was dropped with the range of 20-60% 

of unfished levels, at least 75% of MSY was realised. In contrast to spawner biomass, relative yield as a function of 

relative spawner biomass per recruit was more strongly dependent on the choice of S-R curve. The maximum of the 

minimum yields - FMMY – suggests that at least 75% of MSY can be attained if spawner biomass-per-recruit was not 

depleted below 20 and 60% of unfished levels.

253 


Figure 3. Per-recruit and yield for a typical cichlid fish that with co-incidental gear selection and maturity 

schedules. The solid lines are the Ricker and the dotted lines the Beverton-Holt curves.

254 


Figure 4. Per-recruit and yield for a typical cichlid fish that with selection by the fishing gear occurring 1 

years prior to maturity. The solid lines are the Ricker and the dotted lines the Beverton-Holt curves.

255 


Figure 5. Mean and standard deviation of 100 simulations of four management quantities to assess the 

robustness of four biological reference points to two scenarios of life-history parameters and fishery 

characteristics. Cumulative catch is expressed as the fraction of the highest mean value for all biological 

reference points.

256 


The F0.1 and FSB40 BRPs were similar, estimated between 0.43 and 0.46.yr -1 for the base-case scenario and between 

0.28 .yr -1 and 0.32 .yr -1 when fish are exposed to fishing pressure prior to maturity. FMAX was consistently the 

highest BRP estimated, ranging between 0.53 .yr -1 and 1.16 .yr -1 . Spawner biomass-per-recruit was reduced the most 

with the FMAX BRP. FMMY had intermediate values to the other BRPs ranging between 0.59 for the base case scenario 

and 0.35 .yr -1 in the lower selection scenario. Both FMMY BRPs reduced spawner biomass-per-recruit to ~32% of 

unfished levels. As fish are selected later into the fishery fishing mortalities required to harvest the resource at levels 

approaching MSY were increased due to per-recruit effects where a “reserve” of spawner biomass-per-recruit was 

unfished, thereby reducing extinction at higher levels of fishing mortality. 

Results from the stochastic trials were not unexpected (Figure 5). Fishing the FMAX strategy resulted in slightly 

higher yields at the expense of reducing spawner biomass to the lowest levels. In contrast the FSB40 and F0.1 

strategies maintained spawner biomass at the expenses of ~10% of maximum yield. FMMY was intermediate and 

achieved 95% of MSY while maintaining spawner biomass above 35% of pristine levels. The number of simulations 

where spawner biomass dropped below 20% - effectively the risk of possible recruitment failure - was the highest 

for FMAX and lowest for the FSB40 and F0.1 BRPs. 

Discussion 

Fisheries dominated by one or more cichlid species are common in tropical and subtropical Africa, Central and 

South Americas. Cichlids also dominate Sri Lanka freshwater catches since the introduction of Oreochromis 

mossambicus and O. niloticus (De Silva, 1988). Gears utilised are varied and range from mosquito meshing to 

harvest juveniles and “fry”, hook and line and different net configurations. With the exception of size-selective 

gillnets, fish are usually harvested prior to sexual maturity as fish are usually caught for subsistence or marketed by 

number rather than mass (Weyl and Booth, unpublished data). Management is at best ad hoc, with most fisheries in 

an overfished state with catches and catch-rates dropped steadily over time (Pálsson et al., 1999; Ogutu-Owayo, 

1990). 

Lake Malawi is no exception. Chambo, a complex comprising four economically and socially important 

Oreochromis species, has dominated landings. Unfortunately these stocks are considered to heavily depleted with 

large fish rarely seen in catches and catch rates at all-time low levels (Pálsson et al., 1999). Effort has now been 

shifted onto less utilised pelagic cichlid resources. Management intervention is urgently required. 

Cichlid life-history patterns are complex and have evolved in various habitats differing in biotic and abiotic 

variability. Overall, cichlids are considered K-selected as they have a high longevity, lay relatively few but large 

eggs, exhibit parental care and mature at a high fraction of their maximum length (Bruton, 1989). Parental care has 

been shown to resolve taxonomic issues as Sarotherodon spp. are paternal mouthbrooders, Oreochromis spp. 

mouthbrood, and Tilapia spp. are guarders (Trewavas, 1983). All cichlid genera in Lake Malawi, with the exception 

of Serranochromis and Tilapia, are considered to be maternal mouthbrooders. Nests of those species residing on 

unconsolidated sediments are usually constructed within a lek (McKaye, 1984). 

Other aspects relating to the interaction between cichlid life-histories and the characteristics of their harvesting 

fisheries are that maturity and gear selectivity schedules rarely co-incide and that natural mortality is, deduced from 

other life-history traits, probably density dependent. Considering the dependence on suitable inshore habitats for 

juveniles and space required for elaborate nest building, fish are easily targeted and spawning habitat destroyed by 

gears such as large seine nets. Lake Malawi cichlids are also highly habitat specific. All of these factors have 

management implications. 

It would appear that BRPs based on groundfish life histories (Clarke 1991; 1993; Mace; Thompson) are similar to 

those of cichlids. The stochastic trials, however, suggested that the deterministic BRPs were too low and prompted 

the need for a slightly more conservative approach in selecting for suitable spawner biomass–per-recruit level to 

sustain a fishery indefinitely. Conserving more spawner biomass-per-recruit would have a marginal difference in 

yield. 

Two management strategies emerged from the analysis. If spawner biomass can be estimated and then maintained 

between 30 - 60% of unfished levels then surplus production can be safely harvested. Alternatively, spawner 

biomass-per-recruit should be maintained at levels of >40%. The broad summit between relative yield and relative 

spawner biomass suggests that a biomass based strategy would be the most suitable than one based on the regulation 

of fishing mortality. At least 75% of MSY can be attained by simply maintaining the spawner biomass in the range of

257 


30-50% of pristine levels without increasing the risk of recruitment failure beyond an unacceptable limit. 

Unfortunately, for cichlid directed fisheries, the second option is the most feasible prompting an investigation into a 

suitable S-R curve to fine-tune BRP recommendations. 

The Beverton-Holt (1957) S-R curve is probably the most biologically plausible in a cichlid context. The derivation 

of the curve assumes that juvenile competition leads to linearly dependent mortality rate relative the number of fish 

alive in the cohort at any point in time. When the cohort grows in number then individuals will disappear faster. If 

fecundity remains constant then there will be compensation, i.e., recruits per spawner decrease as a function of the 

number of spawners. The net result is a plateau in recruits per spawner. This model has also been shown to 

approximate situations with territorial interactions, disruption of nesting, gradation in habitat suitability for nesting 

and juvenile survival, and competition for food or space (Hilborn and Walters, 1992). 

Figure 6. Relative yield as a function of relative spawner biomass per recruit for a typical cichlid fish given 

that harvesting occurs prior to sexual maturity. The maximin yield is illustrated for all the spawner-recruit 

curves and only the Beverton-Holt curves. The solid lines are the Ricker and the dotted lines the 

Beverton-Holt curves. 

Consider cichlid reproduction from a simplified perspective. Spawning occurs within a lek with males constructing 

nests that are often large and elaborate (McKaye, 1984). Habitats are also species specific and could be considered a 

limited resource. Utaka are an exception as they do not construct nests but form dense reproductive shoals above 

deep rocky outcrops in upwelling areas. Female mating success is not necessarily hampered by a reduction in 

suitable habitat as they have the opportunity to choose suitable mate(s) (Kellogg et al., 1995). After spawning, 

females gather the eggs and mouthbrood them before releasing few, well-developed juveniles into suitable nursery 

habitats – inshore areas in the case of the sandy substrate spawners such as Oreochromis spp. and Lethrinops spp., 

or near the parent’s territories in the case of the rock dwelling haplochromine mbuna. It can also be assumed that 

paedophagy would linearly dependent on mouthbrooding females. Suitable nursery ground habitat would

258 


unfortunately facilitate differential mortality compounding depensation resulting in a S-R curve approximating that 

of the Beverton-Holt form. 

The Beverton-Holt S-R curve is also the most conservative from a harvesting perspective, because with the lack of 

depensation (or at least reduced compensation), recruits decrease with a reduction of spawning biomass. A 

conservative approach is not necessarily a poor management option given the largely artisanal nature of cichlid 

directed fisheries. Figure 6 illustrates an alternative management strategy when the maximin BRP is calculated for 

the Beverton-Holt curve only, and is contrasted when the BRP is calculated with all S-R curves combined. Fishing 

at FMMY decreases fishing mortality and increases spawner biomass-per-recruit by ~10% than the previous BRP, a 

substantial difference in the provision against recruitment failure. Yield is unfortunately compromised for the 

maintenance of sufficient spawner biomass but only marginally, at levels of ~80% that of MSY. The deterministic 

model as shown by the stochastic simulations is probably over-estimates fishing mortality and it is therefore 

proposed that the revised BRP be decreased to increase spawner biomass-per-recruit to 50% of unfished levels. The 

precautionary approach prevails in the cichlid situation, and if effective fishing mortality is maintained at this rate, 

could assist with sustaining the livelihood of artisanal fishing communities. 

References 

Beverton, R.J.H and Holt, S.J. 1957. On the dynamics of exploited fish populations. Fish. Invest. Ser. II (London) 

19: 1-533. 

Booth, A.J., Merron, G.S. and Buxton, C.D. 1995. The growth of Oreochromis andersonii (Pisces : Cichlidae) from 

the Okavango Delta, Botswana, and a comparison of the scale and otolith methods of ageing. Env. Biol. 

Fish. 43: 171-178. 

Booth, A.J. and Merron, G.S. 1996. The age and growth of the greenhead tilapia Oreochromis macrochir 

(Pisces:Cichlidae) from the Okavango Delta, Botswana. Hydrobiologia 321: 29-34. 

Bruton, M.N. 1989. The ecological significance of alternative life-history styles. In Alternative life-history styles. 

M.N. Bruton (ed.) Kluwer, Dordrecht. 503-553. 

Bruton, M.N. and Allanson, B.R. 1974. The growth of Tilapia mossambica Peters (Pisces:Cichlidae) is Lake Sibaya, 

South Africa. J. Fish Biol. 6: 701-705. 

Bruton, M.N. and Boltt, R.E. 1975. Aspects of the biology of Tilapia mossambica Peters (Pisces: Cichlidae) in a 

natural freshwater lake (Lake Sibaya, South Africa). J. Fish Biol. 7: 423-445. 

Caddy, J.F. and Mahon, R. 1995. Reference points for fisheries management. FAO Fish. Tech. Pap. No. 347. 

Clark, W.G. 1991. Groundfish exploitation rates based on life history parameters. Can. J. Fish. Aquat. Sci. 48: 734- 

750. 

Clark, W.G. 1993. The effect of recruitment variability on the choice of a target level of spawning biomass per 

recruit. In Proceedings of the International Symposium on Management Strategies for Exploited Fish 

Populations. Kruse, G., Eggers, D. M., Marasco, R. J., Pautzke, C. and T. J. Quinn (Eds). University of 

Alaska, Fairbanks; Alaska Sea Grant College Program Report 93-02: 233–246. 

De Silva, SS. 1988. Reservoir of Sri Lanka and their fisheries. FAO Fish. Tech. Pap. No. 298. 128pp. 

Francis, R. I. C. C. 1992 – Use of risk analysis to assess fishery management strategies: a case study using orange 

roughy (Hoplostethus atlanticus) on the Chatham Rise, New Zealand. Can. J. Fish. Aquat. Sci. 49: 922– 

930. 

Gulland, J.A. and Boerema, L.K. 1973. Scientific advice on catch levels. U.S. Fish. Bull. 71: 325-335. 

Hecht, T. 1980. A comparison of the otolith and scale methods of ageing, and the growth of Sarotherodon 

mossambicus (Pisces:Cichlidae) in a Venda mpoundment (Southern Africa). S. Afr. J. Zool. 15: 222-228. 

Hilborn, R. and Walters, C. J. 1992 – Quantitative Fisheries Stock Assessment. Choice, Dynamics and Uncertainty. 

Chapman and Hall, New York. 570 pp. 

Hoenig, J.M. 1983. Empirical use of longevity rates to calculate mortality rates. U.S. Fish. Bull. 82: 898-903. 

Jambo, C. M. 1997. Aspects of the ecology and reproductive biology of three cichlid species of the southern Lake 

Malombe (Malawi). Unpublished MSc thesis, Rhodes University. 126pp. 

Kellogg, K.A., J.A. Markert, J.R. Stauffer Jr. and T.D. Kocher. 1995. Microsatellite variation demonstrates multiple 

paternity in lekking cichlid fishes from Lake Malawi, Africa. Proc. R. Soc. Lond. B.260:79-84. 

Lowe [McConnell], R.H. 1959. Breeding behaviour patterns and ecological differences between Tilapia species and 

their significance for evolution within the genus Tilapia (Pisces: Chlidae). Pro. Zool. Soc. Lond. 132: 1-30. 

Mace, P.M. 1994. Relationships between common biological points used as thresholds and targets of fisheries 

management strategies. Can. J. Fish. Aquat. Sci.51: 110-122.

259 


Mace, P.M. and Doonan, I.J. 1988. A generalized bioeconomic simulation model for fish population dynamics. N.J. 

Fish. Assess. Doc. 88/4. 

McKaye, K.R. 1984. Behavioural aspects of cichlid reproductive strategies: patterns of territoriality and brood 

defence in Central African substratum spawners and African mouthbrooders. In G.W. Potts. & R.J. Wooton 

(Eds.) Fish reproduction: strategies and tactics. Academic Press, London. pp 245-273. 

Myers, R. A, Bridson, J. And Barrowman, N. J. 1995 – Summary of worldwide spawner and recruitment data. Can. 

tech. Rep. Fish. Aquat. Sci. 2024: iv + 327 pp. 

Ogutu-Ohwayo, R. 1990. The decline of the native fishes of Lakes Victoria and Kyoga (East Africa) and the impact 

of introduced species, especially the Nile perech Lates niloticus, and the Nile tilapia, Oreochromis 

niloticus. Env. Biol. Fish. 27: 81-96. 

Pálsson, Ó.K., Bulirani and Banda, M. 1999. A review of biology, fisheries and population dynamics of chambo 

(Oreochromis spp., Cichlidae) in Lakes Malawi and Malombe. Government of Malawi, Fisheries Bulletin. 

38. 35pp. 

Pauly, D. 1980. On the interrelationship between natural mortality, growth parameters and mean environmental 

temperature in 175 stocks of fish. J. Cons. Int. Explor. Mer. 39: 175-192. 

Sissenwine, M.P. and Shepherd, J.G. 1987. An alternative perspective on recruitment overfishing and biological 

reference points. Can. J. Fish. Aquat. Sci.44: 913-918. 

Trewavas, E. 1983. Tilapiine fishes of the genera Sarotherodon, Oreochromis and Danakilia. London. 

Tweddle, D. and J.L. Turner, 1977. Age, growth and natural mortality rates of some cichlid fishes of Lake Malawi. 

J. Fish Biol. 10:385-395. 

Van der Waal, B.C.W. 1985. Aspects of the biology of larger fish species of Lake Liambezi, Caprivi, South West 

Africa. Madoqua 14: 101-144. 

Weyl, O.L.F. and Hecht, T. 1998. The biology of Tilapia rendalli and Oreochromis mossambicus (Pisces: 

Cichlidae) in a subtropical lake in Mozambique. S. Afr. J. Zool. 33: 178-188.

260 


Appendix I: Abstracts of papers not submitted for publication 

(with comments from floor) 

Ecological aspects of the ornamental fish of Lake Malawi and their implications in 

relation to exploitation and conservation 

A. Msukwa 

Department of National Parks and Wildlife, cape MacClear. Malawi 

Abstract 

Lake Malawi has diverse ichthyofauna estimated at over 1000 species. The food fish industry (artisanal and mechanized 

commercial fisheries) exploit many species with the exception of the colourful rock dwelling cichlids locally known as mbuna. The 

mbuna are exploited by licenced exporters for the aquarium trade. The food fish exploitation industry has enjoyed monitoring and 

management attention from the Fisheries Department that the aquarium fish export industry has not. Management regulations of a 

fishery are based on the way the fish are exploited and the biological and ecological information concerning the fish. Due to this 

background some biological and ecological aspects of mbuna were studied at the three Maleri Islands in Lake Malawi National Park 

with a view to assessing the impact of their exploitation for ornamental trade. The mbuna were sampled by 1 hour gillnetting at 2, 5 

and 10 m depths around the three Maleri Islands between January and May 2000. The following aspects were assessed: the 

distribution and relative abundance, female reproductive stages and size at maturity and sex ratios. Two licenced exporters of 

mbuna were visited to assess species and numbers exploited. All sites sampled had diverse fish species ranging from 19 to 25 

mbuna species and 15 to 34 non-mbuna species per site. A majority of these species were found at all three islands with a few 

restricted to one or two of the three islands. There were significant (p

261 


Determination of species diversity in various areas of Lake Malombe 

B. Chirwa*, A. Ambali, M. Chagunda & E. Kaunda 

Molecular Biology and Ecology Research Unit (MBERU), AMBAKA Building, Chancellor, College Campus, Box 403, Zomba 

Abstract 

Lake Malombe is the third largest lake in Malawi and contributes substantially to the country's fish production. The lake's fishery 

started to decline since the early 1980s when the Oreochromis Nyasalapia species were replaced by the small cichlids collectively 

known as Kambuzi. The fishery has currently collapsed and in attempt to improve the situation, the Government of Malawi 

established a reserve area in the lake that acts as stock recovery area. The study assessed species diversity in the eastern and 

western fishing sites and sanctuary area. A total of 39 species were sampled; 33 in the eastern side and 25 in the western side. Fish 

species diversity was high in the eastern than the western sites, mean H' value 2.127 and 1.77, respectively. Fish species are more 

abundant in eastern sites than western sites, mean D' value 3.93 and 2.57, respectively and more evenly distributed in eastern sites 

than western sites, mean J'' value 0.73 and 0.63, respectively. Laboratory studies are currently underway to determine whether the 

sanctuary is contributing to recruitment in the fishing grounds and if there is migration between the eastern and western sides 

Questions and Comments 

Q: What is the difference between total alleles and effective alleles? 

A: Total alleles are all alleles present. Effective alleles are those that affect the character controlled by that 

locus and can be used as a measure of inbreeding or migration between populations. 

Q: What mesh size did you use in your sampling? Does your low sample species richness (39) as opposed to 

Kissa’s over 60 species not invalidate your findings? 

A: My results are in agreement with FAO’s and Jambo’s findings, with around 40 species for the lake. We 

used 19 mm mesh size nets since that is the mesh size used by the nkacha net fishermen. 

Q: Did you get samples from fishers or did you conduct experimental fishing as this is important in relation to 

depth? 

A: My samples were obtained from fishermen; I followed up from Jambo’s study by using a similar strategy 

for sampling. 

Comment: The low species number obtained in this study may be due to species identification problems (B. 

Ngatunga). 

A: For identification, the Fisheries Research Unit, Monkey Bay, staff were used. 

Q: How many samples did you use in this study? 

A: I used four replicates per site. 

Comment: It should be kept in mind that the number of samples in this study were much less than Kissa’s study 

who used over 100 samples, thus it is not surprising that Kissa had a higher number of species in his study 

(G. Turner).

262 


River discharge and water quality 

M. Gondwe, M. Kingdon, H. Bootsma, J. Mwita, B. Mwichande & R. Hecky 

SADC/GEF Project, P.O. Box 311, Salima, Malawi 

Abstract 

Phytoplankton in Lake Malawi support a high fish biodiversity. On the other hand phytoplankton productivity and composition in the 

productive layer of the lake water is largely determined by nutrients availability. One of the major ways nutrient loading into the 

productive layer occurs is through river discharge. The level of nutrient loading by a river depends on land use and human activities 

in the watershed of the river. Deforestation, bush fires and washing away of farmland due to poor agricultural practices have been 

observed to increase nutrient levels in runoff to lakes. Such situations have led to eutrophication of the lakes and consequent 

decline and/or extinction of biodiversity as experienced in Lake Victoria. 

The importance of rivers in respect of the foregoing was recognized by the SADC/GEF Lake Malawi Biodiversity Conservation 

Project. As such, Limnology section of the project sampled Lake Malawi rivers from 1996 through 1999 with the aim of: 

Determining the annual loading of nutrients and sediments into the lake from the rivers for inclusion in the whole lake budget. 

Assessing the current water quality of the in-flowing rivers, especially for parameters such as phosphorus, nitrogen, silica and 

suspended solids that are most likely to affect the limnology of the lake and its budget. Suggesting sensitive chemical parameters 

that are particularly important to the lake's water quality and biodiversity and which should be the focus of the future 

monitoring/research programs. 

Among the many findings are: No statistically valid relationship was observed to exist between daily river discharge and nutrient 

concentration. For all rivers the early part of the flow season exhibited a flushing effect for most nutrients. In December 1996, TSS, 

TDN and TDP were 79.03mg/l, 1.352 and 0.075 mmoles/m2 lake surface area while in May 1997 concentrations were 0.45mg/l 

0.138 and 0.005 mmoles/m2 lake surface area. Loading of all nutrients and sediments was significantly higher in the southwestern 

catchment and northeastern catchment of the lake. Northwestern catchment accounts for only 6.3% of the total for all nutrients. 

Loading of all dissolved nutrients except SO4 was similar in southwestern catchment and northeastern catchment, however loading 

of total suspended solids (TSS) and all suspended nutrients was higher in northeastern catchment than in southwestern catchment. 

Majority of the flux to the lake was in form of particulate nutrients, and the concentrations of total nitrogen (N), total phosphorus (P) 

and available silica (Si) were all correlated with suspended solids concentration. 

NOTE: Southwestern catchment has Linthipe, Bua and Dwangwa rivers. Northwestern catchment has Songwe, South and North 

Rukuru rivers. Northeastern catchment has Ruhuhu river. 


Q: Are algal blooms the only factor causing fish kills? 

A: No, but mostly it is due to algal blooms. When diatoms die they sink and do not fix nitrogen; whilst some 

green algae (some of which fix nitrogen) do not sink, thus contributing to oxygen depletion. 

Q: Is there any information on the phytoplankton for Lake Malawi or the blue-green algae composition? 

A: Yes it is available, a UK/SADC publication edited by Menz.

263 


Metals, Pesticides And Other Persistent Contaminants In Water, Sediments And 

Biota From Lake Malawi/Nyasa 

A. Bulirani*, K. Kidd, W. Lockhart, P. Wilkinson & D. Muir 

SADC/GEF Project, P.O. Box 311, Salima, Malawi 

Abstract 

Lake Malawi is a valuable resource to the populations along its shores in the riparian countries. Among the major resources are fish 

that provides about 75% of the much essential animal protein and water for irrigation, domestic use and recreation. In the event of 

the availability of micro-contaminants such as PCBs and organochlorine pesticides at toxic concentrations in the lake, these uses 

would obviously constitute health risk for the people. Previous work elsewhere on PCBs and organochlorine pesticides has shown 

that they reach toxic concentrations through bioaccumulation in food chains as a result of their very strong persistency in soil, air 

and water. One strong effect of micro-contaminants such as these PCBs and organochlorine pesticides at toxic concentrations to 

fish in a lake is fish kills. In 1999, PCBs and organochlorines were linked, in the minds of some many individuals in Malawi, to the 

famous lake-wide fish kills experienced in Lake Malawi during the windy August - November period, which was unusually long. 

One of the research interests of the SADC/GEF Lake Malawi Biodiversity Conservation Project was to determine the concentration 

of microcontaminants such as PCBs and organochlorine pesticides in fish, mud and wake water. 

The project's research findings on PCBs and organochlorines in the lake water and fish biota did not confirm the fears and 

suspicions that they were a possible cause for the fish kills. Organochlorine pesticides and PCBs were detected at low 

concentrations in surface as well as in subsurface waters of the lake. PCBs were the major organochlorines in surface waters and 

their concentrations ranged from 165-854pg/l. Deep water samples (e.g. 80m) gave lower concentrations ranging from 100-187pg/l. 

Oftenly the results were found to be below the detection limit which is approximately 100pg/l. 

Concentrations of persistent pesticides and PCBs in fish from Lake Malawi were low in all fish analysed. Examination of the 

Canadian Bureau of Chemical Safety's tolerable daily intakes (TDIs) of 1.0ug.kg body weight-1day-1 for PCBs and their 

concentrations in Lake Malawi fish revealed that a 60kg person can safely consume on a daily basis over a lifetime up to 34.6kg/day 

Mcheni, 235kg/day Chambo, 2.9kg/day Mpasa, 4.1kg/day Usipa and 17.7kg/day Kampango. This confirms that consumption of 

Lake Malawi fishes does not pose any health risk to humans especially because fish is never consumed daily and in these 

quantities. 


Q: What is the significance of using muscle tissue rather than using bony structure in the analysis? 

A: To my understanding people consume the fleshy part of the fish thus the analysis used muscle tissue. 

Q: Is there any plan to study how different fish eating habits of people reflect the levels of contamination? 

A: The contamination levels so far are too low to be of concern, therefore there are no plans to do research on 

the eating habits of people. 

Q: How do chemicals accumulate in upper layers when your colleague (M. Gondwe) indicates that most of the 

chemicals accumulate below the thermocline? How do the contaminants that have settled down below the 

thermocline affect the fish? 

A: Due to wave action and upwelling, the contaminants from the bottom reach the upper layers where the fish 

are and eventually accumulate in the food web. 

Q: Why did you not use the World Health Organization’s (WHO) or the Malawi Bureau of Standards 

standards for tolerance levels of contaminants? 

A: The Canadian standards were adopted because the contract was awarded to a water quality analysis institute 

of Winnipeg, Canada. The values were thought to be representative.

264 


The Traditional Gillnet Fisheries In Metangula, Lake Malawi/Niassa, East Africa 

J. Halafo 

Fisheries Research Institute, P.O. Box 4603, Maputo, Mozambique E-mail: halafo@iip.co.mz 

Abstract 

Metangula is the principal village on the Mozambique coast of Lake Malawi/Niassa. It has 9 human settlements 

where fishing occurs over about 40 km of shoreline. This village is one of the few places where the gillnet catches of 

Labeo mesops (nchila), and Opsaridium microcephalus (sanjika) occur, due perhaps to the low fishing pressure in 

that area. Although the gillnet fishery catches a wider variety of commercial fish than any other gear type used by 

the fishermen, nchila, and sanjika are the most important species in the catches. The seasonality of habitats appears 

to be a key factor affecting many interrelated aspects of life cycle for this species. Their movements to feed or 

spawn are tied up with seasonal and other changes in the environment. 

The traditional fisheries of Labeo mesops, and Opsaridium microcephalus, mainly from gillnets, have decreased 

seriously in the past 10 years over the lake. These species are reported as being missed from the catches in gillnet 

fisheries of Malawi. Smith (1993a) found CPUE = 0 for those species in Chembe village. Few years ago, L. mesops 

was the most important fish in the riverine fishery in Malawi. It has been depleted due to intensive gillnetting of 

gravid individuals on breeding migrations. 

This paper presents preliminary results on traditional gillnet fisheries in Metangula, focusing on 2 important 

commercial species referred above. I'm also trying to arise here the concrete need for fisheries management and 

biodiversity conservation particularly related to these species. It's not clear yet whether these species are represented 

by a single population that is widespread in the lake, or by different populations with distinctive morphometric, 

genetic and life history characteristics. Therefore, the studies in near future will focus on distribution along the 

shore, comparison of its morphometric and genetic characteristics, and life history, in order to know whether these 

species return to the same rivers for reproduction or they go to any river. The studies will also include issues of 

economic value and conservation status. 


Q: The gillnet effort per 100m; are they mounted or not? 

A: Mounted gear used for effort estimate. 

Q: Don’t you think that the sampling duration of December 1998 to March 1999 was too short to come up 

with reliable conclusions and solutions? 

A: I agree. 

Q: Do fishermen use multi or monofilament gill nets? 

A: The fishermen use both types, however monofilament gill nets are uncommon. 

Comment: There is data available for the next 8 months after this study and it should be included in the analysis 

(G.F. Turner). 

Q: The disappearance of adults (Cyprinids) in December may be due to migration and not due to fishing. 

What is known about the migratory patterns of the fish studied? 

A: There is a possibility that migration may be playing a role, however it has not been studied. It may be 

looked into next time.

265 


Genetic diversity distribution of tasselled tilapias (Oreochromis nyasalapia spp.: 

Cichlidae) in Lake Malawi using microsatellite DNA markers 

D. Kafumbata* & A. Ambali 

Molecular Biology and Ecology Research Unit (MBERU), AMBAKA Building, Chancellor College Campus, Box 403, Zomba 

Abstract 

A study was carried out to investigate the genetic diversity distribution of three species of O. Nyasalapia in Lake Malawi. Five 

polymorphic loci were scored in 12 populations. The total number of alleles ranged from 13.0(4.4 to 18.4(5.9; mean effective number 

of alleles ranged from 8.0( 4.0 to 11.09(4.0 in O. karongae, 9.09(2.3 to 10.2(3.8 in O. lidole and 8.8(4.8 to 11.9(4.7 in O. 

squamipinnis. The mean heterozygosity per population ranged from 0.45(0.17 to 0.66(0.16, which is higher than reported in 

allozyme loci. There is high genetic variation in O. squamipinnis, followed by O. karongae and O. lidole. The populations are not 

highly differentiated. These results suggest that the O. Nyasalapia flock has still maintained a reasonable level of genetic variation in 

the face of intense fishing pressures. 

Comments from floor 

Q: With what gears were the samples collected from? 

A: Nkacha nets. 

Q: What about species identification, for instance, how did you deal with the juvenile chambo? 

A: We used FRU technicians for identification and small chambo were not included in the analysis. 

Q: If two populations are not genetically different, does it mean that there is migration? 

A: Not always true. 

Q: Did you statistically test for genetic differences among the three species from the same location? 

A: Yes. There was no difference between the species. The Fst values were very small and therefore cannot be 

considered as significant. There was a lot of overlap even among species of the same area. 

Q: What does the 5 migrants per generation mean? 

A: It refers to actual individuals (i.e. 5) migrating out per generation from a population.

266 


Genetic population structure of Oreochromis mossambicus in the Shire River 

system 

M. Chimenya* & A, Ambale 


Abstract 

An investigation in the Lower Shire River fishery was carried out to assess the status of genetic variation of O. mossambicus using 

microsatellite DNA markers. Five polymorphic loci (OS08, OS64, OS75, UNH154 and UNH103) were scored in 11 populations. 

Number of alleles per locus ranged from 4 in OS64 to 40 in UNH154. Cluster analysis grouped populations into three, (1) those 

found in Chikwawa, (2) those that are found at Elephant Marsh and (3) those that are found at Ndinde Marsh. 

Genepop program was used to calculate allelic diversity of each of the 11 populations. The mean number of alleles ranged from 8.2 

to 11.0 with mean heterozygosity range of 0.48 to 0.68. All populations exhibited considerable genetic diversity. WHICHRUN 

program was used to assign individuals to their right strata/populations based multi-locus genotype data. Probability of assigning of 

a population to its right strata ranged from 52-100%. Probability of assigning of an individual to its right population ranged from 57- 

95%. Populations from the market were assigned to different sources than indicated by the fish traders and these tended to reduce 

the probability of right assignment of individuals to source. 


Q: Have you used fish from the ponds, e.g. from Domasi and those in the Lower Shire? 

A: Yes, some work has been done on O. mossambicus from the ponds in Domasi by Dr. A. Ambali. 

Q: What are the differences between the fish from the ponds and those in the Lower Shire? 

A: From Dr. Ambali’s samples from Domasi he identified three alleles, however, in this study (samples from 

the Lower Shire) four alleles were identified. This may be due to different environmental conditions 

encountered by fish in captivity. 

Q: How cost effective is your analysis process? 

A: It is a very expensive process. The cost is about MK15,000.00 per 20 samples.

267 


Zoogeographical distribution and population structure of Taeniolethrinops 

praeorbitalis exploited by artisanal fishermen in the inshores of Lake Malawi 

W. Changadeya* & A. Ambali 


Abstract 

Lethrinops species flock (Chisawasawa/mbaba) is among the major commmercially important fish species exploited in Lake Malawi. 

Their catches started declining as early as 1975 due to overexploitation especially in the southern part of the lake yet there are no 

deliberate measures put in place to conserve these vital populations. A study was carried out to determine the genetic diversity and 

population structure of Taeniolethrinops praeorbitalis populations in traditional fisheries of Southeast and Southwest arms of the 

lake, Mangochi District and Nkhota -kota lakeshore area. A total of 10 populations with 40 individuals as sample size were analyzed 

at 6 polymorphic microsatellite loci. The populations were not in Hardy-Weinberg equilibrium possibly due to Wuhlund effect. This is 

supported by inter-deme migration of more than seven individuals per generation as determined by a multilocus estimate of number 

of migrants using private alleles method. However this rate of migration has not reduced population differentiation in T. praeorbitalis, 

mean Fst of 0.152. 


Q: Are the Nkhota-kota and Mangochi populations of Taeniolethrinops praeorbitalis different? 

A: Yes. 

Q: Who identified the species used? 

A: Dr. B. Ngatunga did the identification. 

Comment by Dr. B. Ngatunga: I was involved in the initial start-up of the research by helping the student with the 

identification of the fish species, however, I wish to express that one of the pictures in the student’s 

presentation showing the fish species he analysed was not Taeniolethrinops praeorbitalis as he claimed, but 

I suspect that it was an Aulonocara species.

268 


Feeding ecology of Bathyclarias nyasensis (Siluroidei: Clariidae) in Lake Malawi 

E. Kaunda 1* &T. Hecht 2 

1 Dept of Aquaculture and Fisheries Science, University of Malawi, Bunda College, Box 219 Lilongwe, Malawi. 

2 Rhodes University, P.O. Box 6140, Grahamstown R.S.A. 

Abstract 

The ecological role of B. nyasensis in Lake Malawi was identified from studies on life history raits, ecological and functional 

morphology in addition to diet and food habits of the species conducted between 1996-1998. The maximum age for B. nyasensis 

was estimated at 14 years. Growth was best described by the four parameter Schnute model: 

for female and 

for male fish. Age-at-50% maturity for females and males were estimated at 7 years and 4 years, respectively. Typically, fish grew 

rapidly in the first year, but slower during subsequent years. Smaller fish were found inshore while larger fish were found in offshore 

regions. It was hypothesised that the rapid growth in the first year and slower growth later is a consequence of change in diet from 

high quality and abundant food source to a more dilute food and that this may be associated with a shift in habitat. 

Changes in fish growth synchronised with dietary changes and morphology particularly changes in buccal cavity volume. These 

changes occurred when fish reached 500-600 mm TL, concomitant with habitat shifts that probably imply a mechanism to open the 

inshore habitat to the next cohort, thereby maintaining a stable population. In inshore areas, B. nyasensis was primarily piscivorous 

and was zooplanktivorous in offshore regions. 

On the basis of the theoretical migratory life history cycle of B. nyasensis, and “bottom-up” and trophic cascade theories, it is 

postulated that perturbations of the B. nyasensis stock would be discernible both at the top and lower trophic levels. As a piscivore 

and therefore apex predator, effects of overfishing B. nyasensis in the inshore region could cascade to unpredictable ecological 

changes in inshore areas and, due to the ontogenetic habitat shift, in the offshore regions. It is recommended that the current 

interest in increasing fishing effort in offshore areas should proceed with caution. 


Q: In your presentation you pointed out that Bathyclarias nyassensis is a pelagic species, yet during the UK 

SADC project only 27 specimens were caught. Is it really pelagic? 

A: It is not really pelagic, although sometimes it can behave like a pelagic species. 

Q: The deepwater fishery targets the small species (chisawasawa), and the catfishes caught as bycatch and are 

not considered in management. How do you manage this fishery? 

A: There is no policy for the catfishes at present. 

Q: How did you validate your aging? 

A: I used marginal zone analysis for validation. 

lt ={42+(81 1.8 - 42 1.8 )x 1-e -0.05(t-1) } 1/1.8 

1-e -(-0.05)(11) 

lt ={41+(98 1.2 - 41 1.2 )x 1-e -0.02(t-1) } 1/1.2 

1-e -(-0.02)(13) 

Q: Concerning your observation on feeding aggregations of juveniles and adults, could you verify this 

observation? 

A: My data indicated that feeding aggregations of Bathyclarias nyassensis took place.

269 


Land use patterns in the Domasi and Likangala catchments and their effects on 

soil erosion, water quality, river flow rates, siltation rates and Barbus 

reproduction in Lake Chilwa 

D. Jamu*, J. Chimphamba & R. Brummett 

ICLARM, P.O. Box 229 Zomba 

A study to examine trends in land use practices, and to establish linkages between river flow, soil erosion rates, catchment sediment 

yield, river water quality, migratory patterns and reproductive success of Barbus species in the Likangala and Domasi rivers was 

conducted between November 2000 and November 2001. Participatory rural appraisals, to obtain information on current and past 

land use and management practices and the effects of these practices on the river ecosystem health, were conducted in four 

villages in the Likangala catchment and in Mtwiche village in the Domasi catchment. Land use cover change analysis was 

accomplished by aerial photograph interpretation of black and white aerial photographs for 1982 and 1995 covering the study area. 

Soil erosion rates were estimated using the Soil Erosion Model for Southern Africa (SLEMSA) by delineating soil erosion 

management units with homogeneous areas with similar slope, class, soil group, annual mean rainfall, and vegetation cover type 

and density. Catchment sediment yields were obtained by multiplying the estimated soil loss for each soil erosion management unit 

with a delivery ratio, which was defined as the ratio of soil eroded that had been carried downstream to that remaining within the 

field. Barbus and water quality sampling in the Likangala and Domasi Rivers was conducted biweekly from December 2000 up to 

June 2000 and thereafter monthly until November 2001. The Barbus were sampled at the river mouth using a multi-mesh gillnet to 

determine their total number and they were assessed for their position in the net to determine direction of migration. Adult females 

were assessed for their gonadosomatic index. The gonadosomatic index was used to determine the seasonal reproductive status of 

Barbus. The water quality parameters measured were dissolved oxygen, electrical conductivity, pH and total suspended solids. 

River flow rate was measured using a current flow meter and the float method. Water quality, river flow, sediment yield, 

reproductive and fish migration data were subjected to multiple regression analysis to determine the major factors affecting 

reproductive status and migration. The land use analysis results showed that land use practices in the two catchments are 

characterized by a combination of destructive forces mainly the expansion and intensification of cultivation of maize on marginal 

land using inappropriate practices. The participatory rural appraisal sessions in both catchments revealed that the number of trees in 

the upland and along the rivers, the size and number of fish caught in the two rivers had drastically declined over the past fifty years. 

The highest amount of soil loss (>100t.ha -1 .yr -1 ) was estimated in the upland areas of the Likangala River catchment and field 

walks in the area revealed wide spread evidence of very severe erosion, exposed soils and numerous deep gullies. The most 

critical factors contributing to high soil loss were rainfall high kinetic energy and poor vegetation cover. High vegetation cover and 

relatively flat plains resulted in low soil loss in the wetland. The Likangala River catchment had a higher annual sediment yield of 

374 t.km -2 .yr -1 compared to 315 t.km -2 .yr -1 for the Domasi catchment. A situation analysis for the Likangala River catchment 

showed that increasing maize yield was more efficient in reducing soil loss than contour ridging, and the highest reduction in soil 

loss was predicted when a combined 20% increase of contour ridging, maize yield and tree canopy was assumed. Multiple 

regression analysis results showed that reproductive success of females migrating upstream was negatively correlated with total 

suspended solids and positively correlated with electrical conductivity in the Domasi River. On the other hand, the upstream 

migration of Barbus juveniles was positively correlated to sediment yield in both rivers, and also to river flow rate in the Domasi 

River. The Barbus reproductive status was not correlated to any of the physico-chemical variables that were measured in the 

Likangala River. The major recommendations of the study were as follows: (i) land use practices that tackle the most critical soil loss 

factor should form an integral part of soil conservation measures in each soil erosion management unit; (ii) a combination of contour 

ridging, increases in crop yield and vegetation cover should be promoted as strategies for reducing soil loss in the Likangala River 

catchment; (iii) appropriate management actions that reduce fishing pressure on breeding Barbus females in the influent rivers 

should be formulated to ensure the success of spawning migrations of breeding females into the influent rivers, and (iv) marginal 

vegetation in the breeding grounds along the lake should be protected from burning and other forms of destruction in order to 

maintain breeding and feeding grounds for juvenile Barbus in the dry season. 


Q: How does maize reduce soil erosion? 

A: Maize grows faster and provides the best canopy to reduce soil erosion during the period when the rains are 

the heaviest. 

Q: What factors were considered to use Barbus as an ecosystem indicator? 

A: Barbus species migrate from the lake to breed in the rivers and, therefore, their reproductive success is 

related to the condition of the rivers. 

Q: What mechanism would be put in place to protect breeding individuals? 

A: The Fisheries policy needs to take into consideration no fishing during the breeding season, i.e. considering 

banning fish traps or fish weirs that cross the entire river system.

Appendix II: List of participants 

270 


Name Address Tel E-mail 

Allison E. School of Development Studies, University of East 

Anglia, Norwich NR4 7TJ, UK 

Ambali A. Molecular Biology and Ecology Research Unit 

(MBERU), AMBAKA Building, Chancellor, College 

Campus, Box 403, Zomba 

Balarin J. C/O Royal Danish Embassy, Pvt. Bag 396, Lilongwe, 

Malawi 

Banda M. Fisheries Research Unit, P.O. Box 27, Monkey Bay, 

Malawi 

Bandula D. Department of Fisheries - Headquarters, P.O. Box 

539, Lilongwe, Malawi 

Booth A. J. DIFS, Rhodes University, P.O. Box 94, 6140, 

Grahamstown, South Africa 

Bulirani A. SADC/GEF Project, P.O. Box 311, Salima, Malawi 

Changadeya W. Molecular Biology and Ecology Research Unit 


Campus, Box 403, Zomba, Malawi 

Chimenya M. Molecular Biology and Ecology Research Unit 



Chirwa W. Centre for Social Research, University Office, 

University of Malawi, P.O. Box 278, Zomba, Malawi 

Chisambo J. Fisheries Research Unit, P.O. Box 27, Monkey Bay, 

Malawi 

Darwall W. Overseas Development Group University of East 

Anglia, Norwich NR4 7TJ, UK 

44 1603 593724(t)44 

1603 451999(f) 

e.allison@uea.ac.uk 

aambali@sdnp.org.mw 

836533 DESP@malawi.net 

587 440 fru@malawi.net 

6038415 T.Booth@ru.ac.za 

263432 

524888 

44 1603 593922(t) 44 

1603 505262(f) 

Dobson T. Michigan State University 13 Natural Resources Bldg. 517-1711(t) 

E. Lansing 48824-1222 

432-1699(f) 

Ellis F. Overseas Development Group University of East 

Anglia Norwich NR4 7TJ, UK 

Gondwe M. SADC/GEF Project, P.O. Box 311, Salima, Malawi 

Halafo J. Fisheries Research Institute, P.O. Box 4603, 

Maputo, Mozambique 

Hecht T. DIFS, Rhodes University, P.O. Box 94, 

6140 Grahamstown, South Africa 

Jambo C. Department of Fisheries, Mangochi District Office, 

P.O. Box 47, Mangochi, Malawi 

Jamu D. ICLARM, P.O. Box 229, Zomba 

44 1603 504455(t) 

44 1603 464267(f) 

W.Darwall@uea.ac.uk 

dobson@msu.edu 

dobson@pilot.msu.edu 

f.ellis@uea.ac.uk 

263432 nyasa@malawi.net 

halafo@iip.co.mz 

6038415 t.hecht@ru.ac.za 

584813, 584 211 

536274(t) iclarm@sdnp.org.mw

271 



Jarchau P. NARMAP, P.O. Box 31131, Lilongwe, Malawi 

Kachinjika O. Department of Fisheries - Headquarters, P.O. Box 


Kafumabata D. Molecular Biology and Ecology Research Unit 



831 825 narmapllw@malawi.net 

788865, 788511 

Kamanga L. 

277222/277214 staff@aquadept.malawi.net 

Kanyerere G. 

Dept of Aquaculture and Fisheries Science, University 

of Malawi, Bunda College, Box 219 Lilongwe, Malawi 

Fisheries Research Unit, P.O. Box 27, Monkey Bay, 

Malawi 


Kataya P. National Aquaculture Centre, P.O. Box 44, 

Domasi, Malawi 

536215 nac@sdnp.org.mw 

Kaunda E. Dept of Aquaculture and Fisheries Science, University 

of Malawi, Bunda College, Box 219 Lilongwe, Malawi 

Kihedu J. TAFIRI, Kyela Centre, P.O. Box 98, Kyela, Mbeya 

Region, Tanzania 

Makuwila M. P.O. Box 593, Lilongwe 

277214(t)(f) staff@aquadept.malawi.net 

tafirik@africaonline.co.tz 

788511 sadcfish@malawi.net 

Maluwa A. NAC, P.O. Box 44, Domasi 536215 aomaluwa@sdnp.org.mw 

Manase M. Fisheries Research Unit, P.O. Box 27, Monkey Bay, 

Malawi 

Mapila S.A. Department of Fisheries - Headquarters, P.O. Box 


Morioka S. JICA, P.O. box 3032, Lilongwe 

Msowoyo N. Department of Fisheries - Headquarters, P.O. Box 


Msukwa A. P.O. Box 48, Moneky Bay 

Mvula P. Center for Social Research, University Office, 


Mwakiyongo K. Fisheries Research Unit, P.O. Box 27, Monkey Bay, 

Malawi 

Mwambungu J.A. TAFIRI, Kyela Center, P.O. Box 98, Kyela, Mbeya 


Ngatunga B. TAFIRI, Kyela Center, P.O. Box 98, Kyela, Mbeya 


Ngochera M. Fisheries Research Unit, P.O. Box 27, Monkey Bay, 

Malawi 

Ngombe T. Department of Fisheries - Headquarters, P.O. Box 


Nyasulu T. Fisheries Research Unit, P.O. Box 27, Monkey Bay, 

Malawi 

Nyirenda M. Department of Fisheries - Headquarters, P.O. Box 


Peschke J. MBZDP, P.O. Box 627, Mzuzu, Malawi 

Russell A. Dept of Fisheries and Wild life, 16 Natural Resources 

BLDG, E. Lansing, MI 48824, USA 


788511 

755734 morioko@malawi.net 

788511 

584657/784 





788511 


788511 

333563 peschke@malawi.net 

517 332 4931 aaron_belgium@yahoo.com

272 



Sambo E. Center for Social Research, University Office, 


Seymour T. Bryn Eryr Uchaf, Llandadwrn, Anglesey LL59 5SA UK 44 1248 712540 seymourevans@hotmail.com 

Sipawe R. Fisheries Research Unit, P.O. Box 27, Monkey Bay, 

Malawi 

Turner G.F. University of Southampton, Bassett Crescent East, 

Southampton SO16 7PX, England UK 

Tweddle D. c/o DIFS, Rhodes University, P.O. Box 94, 6140, 

Grahamstown, South Africa 

Weyl O.L.F. NARMAP , P.O. Box 27, Monkey Bay, Malawi 587 754 

935 259 

Zidana H. National Aquaculture Centre, Box 44, 

Domasi, Malawi 


44 2380 593217(t) 

44 2380 594397(f) 

536215(t) 

536203(f) 

gft@soton.ac.uk 

d.tweddle@ru.ac.za 

narmapbay@malawi.net 

nac@sdnp.org.mw 


The Lake Malawi Fisheries Management Symposium was organised and funded by the GTZ supported 

NARMAP Programme of the Department of Fisheries. However, considerable inputs and support to 

participants was received from the SADC/GEF Project, the World Bank, ICLARM, JICA, Rhodes 

University, the EU and the ODG funded Malawi/Indonesia Fluctuating Fisheries Project. In addition, this 

symposium would not have been possible without the commitment of the authors, whose presentations 

are the basis for the symposium. In addition, we would also like to thank the symposium committee for 

their committed work on this project.

proceedings of the lake malawi fisheries management symposium

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?