fish ecology report - The Cichlid Fishes of Lake Malawi, Africa

FISH ECOLOGY REPORT 

LAKE MALAWI/NYASA/NIASSA 

BIODIVERSITY CONSERVATION PROJECT 

Edited by 

Fabrice Duponchelle 

Institut de Recherche pour le Développement, Lab. GAMET, 361 rue J.F. 

Breton, BP 5095, 34033 Montpellier, France. 

Email: Fabrice.Duponchelle@mpl.ird.fr 

& 

Anthony J. Ribbink 

JLB Smith Institute of Ichthyology P Bag 1015, Grahamstown, 6140 South 

Africa. 

Email: A.Ribbink@ru.ac.za 

© 2000

Contents 

Acknowledgements 

General introduction 

F. Duponchelle & A.J. Ribbink …………….……………………………………………………… 1 

Chapter 1: Temporal trends of trawl catches in the North of the South West 

Arm, Lake Malawi 

F. Duponchelle, A.J. Ribbink, A. Msukwa, J. Mafuka & D. Mandere ………………………..………….. 5 

Chapter 2: Depth distribution and breeding patterns of the demersal species 

most commonly caught by trawling in the South West Arm of 

Lake Malawi 

F. Duponchelle, A.J. Ribbink, A. Msukwa, J. Mafuka & D. Mandere …………………..……….………15 

Chapter 3: Growth patterns of some of the most important demersal fish 

species caught by trawling in the South West Arm of Lake Malawi 

F. Duponchelle, A.J. Ribbink, A. Msukwa, J. Mafuka & D. Mandere …………………………..…..… 169 

Chapter 4: Temporal diet patterns of some Lake Malawi demersal fish 

species as revealed by stomach contents and stable isotope 

analysis 

F. Duponchelle, H. Bootsma, A.J. Ribbink, C. Davis, A. Msukwa, J. Mafuka & D. Mandere ……..………189 

Chapter 5: Morphometric, genetic and ecological comparison of two 

important demersal species along a gradient from the South West 

Arm to Nkhata Bay 

F. Duponchelle, J. Snoeks, M. Hanssens, J-F. Agnèse, A.J. Ribbink, A. Msukwa, J. Mafuka & D. Mandere ..202 

Chapter 6: The potential influence of fluvial sediments on rock-dwelling fish 

communities 

F. Duponchelle, A.J. Ribbink, A. Msukwa, J. Mafuka & D. Mandere ………………….….227 

References cited ……………………………………………………………………………….....265 

Appendixes

Acknowledgements 

(F. Duponchelle) 

This report covers the work of the Ecology team of the SADC/GEF, Lake 

Malawi/Nyasa Biodiversity Conservation Project for the period from June 1998 to the end of 

the Project in July 1999. Given the late start of this part of the ecology work and the time 

constraints before the end of the project, the success of the several research programs 

undertaken by the Ecology team is indebted to many people for their support and assistance. 

We owe a special thank to the Fisheries Department of Malawi and to Alex Bulirani in 

particular for allowing Davis Mandere to join the Ecology team, for making the research 

vessel Ndunduma available in July and August 1998 and for the assistance and support he 

provided during the course of the Project and the writing up of the report. 

Most of the nearshore ecology research was done at the Maleri Islands and Thumbi 

West in Cape Maclear. We are grateful to the Department of National Parks and Wildlife for 

granting permission to work and collect fish and algae samples within the park. 

We would like to thank the Senga Bay station staff, including the two drivers, whose 

kindness and availability was precious, the administrative staff, the boat driver (Elias 

Mnenula), the ground and boat keepers. 

The implementation of the largest research program would not have been possible 

without the great help of Captain Mark Day and the crew of the research vessel Usipa, who 

always managed to make our trips successful and cheerful. 

Arriving in the last phase of such a large Project could have been uncomfortable both 

on the personal side without the warm welcome of every family on the compound and overall 

on the professional side without the kind support of the senior Limnologist and Taxonomist, 

Dr H.A. Bootsma and Dr J. Snoeks, respectively. We acknowledge their respective teams, 

who put in many hours of work in support to our research programs, especially Mr A. 

Abdallah, Mr. M. Hanssens, Mr. J. Mwitta and Mr. B. Mwichande, Mr B. Ngatunga, Mr R. 

Sululu. We would also like to thank Dr R. Hecky for useful discussions and reviewing some 

chapters. 

People who are not much interested in pure science will acknowledge as much as us 

the artist, Mr David Voorvelt for the beautiful cover of the report and excellent fish 

illustrations he provided. Once again he added professional quality to products of the project. 

A mutually beneficial collaboration developed between the Ecology team and the local 

members of the European Union Project: "The trophic ecology of the demersal fish 

community of Lake Malawi/Niassa", Mr W. Darwall and Dr P. Buat. We really appreciated 

the several interesting and motivating discussions we shared. 

Finally, as this section is not meant to be longer than the report itself, we have pleasure 

in acknowledging some of the many people who helped in various ways: 

E. Allison, E. Andre, T. Andrew, A. Banda, D.Barber, P.Bloch, R.Brooks, G. Chilambo, J. 

Chisambo, P. Cooley, M. Genner, S. Grant, G. Hartman, S. Higgins, K. Irvine, S. Kamoto, K. 

Kidd, H. Kling, B. Kumchedwa, R. Lowe-McConnell, W. Mark, F. Mkanda, T. Nyasulu, J. 

Manuel, G. McCullough, J. Moreau, P. Ramlal, R. Robinson, S. Smith, G. Turner.

General introduction

General introduction 

Lake Malawi/Niassa/Nyasa is the southern most of the East African Rift Lakes, lying 

from 9°30'S to 14°30'S between three riparian countries: Malawi, Tanzania and Mozambique 

(Figure 1). It is one of the oldest (many million years, Lowe-McConnell et al. 1994, Konings 

1995, Stiassny & Meyer 1999) and largest lakes of the world. Its mean area (29 000 km², 

Bootsma & Hecky 1993) makes it the 9 th largest lake in the world and 3 rd largest lake of 

Africa after the lakes Victoria and Tanganyika (Lowe-McConnell 1993, Ribbink 1994, 

Konings 1995). Lake Malawi is located 472 m above the sea level, its maximum depth is 785 

m and averaged about 292 m (Bootsma & Hecky 1993). An important characteristic is that 

more than 80% of the lake is deeper than 200m (Thompson et al. 1996), depth under which it 

is permanently stratified and anoxic (Eccles 1974, Lowe-McConnell 1993). This basically 

means that the living space available for the fish and the other components of the food chains 

is only about 20% of the lake volume. About one third of Lake Malawi's shoreline is steep 

and rocky whereas two thirds are gently sloping sandy beaches or swampy river estuaries 

(Lowe-McConnell 1994, Lowe-McConnell et al. 1994). One of the main distinctive features 

of the lake is its exceptional water clarity, upon which the entire ecosystem is highly 

dependent (Bootsma & Hecky 1993, Hecky & Bootsma 1999). 

However, the most well known characteristic of the lake is its exceptional fish 

richness. It harbours greatest fish species richness than any other lake in the world (Fryer & 

Iles 1972, Ribbink 1988, Turner 1996). It is currently estimated that between 500 and 1000 

different fish species are present in the lake (Konings 1995, van Oppen et al. 1998), although 

only about a third are presently described or merely catalogued by a cheironym (Ribbink et al. 

1983). All these fishes, apart from 44 species belonging to nine other families (Ribbink et al. 

1983, Ribbink 1988), belong to a single family, the Cichlidae. With the exception of chambo 

(Oreochromis spp.), all cichlids are closely related species, possibly descended from a single 

common ancestor (Meyer 1993, Meyer et al. 1990, Moran et al. 1994, Stiassny & Meyer 

1999). This tremendous cichlid fish diversity, known as a "species flock" (or a complex of 

species flocks, Greenwood 1984), has evolved in a very short evolutionary time period, some 

of which may have been within the last 200 to 300 years for some species (Owen et al. 1990). 

More than 99% of these cichlid fish species are endemic of Lake Malawi (Ribbink 1991, 

Turner 1996), which means that they can't be found anywhere else in the world. Moreover, 

there is also a high degree of intra-lacustrine endemicity, many species belonging only to 

particular islands or stretches of shore within the lake (Ribbink & Eccles 1988, Eccles & 

Trewavas 1989, Ribbink 1991). These peculiarities of the lake fishes have led to develop a 

great interest from the scientific community, challenged by the understanding of what 

constitutes the most striking example of rapid vertebrate radiation known at this day (Turner 

1998). 

Importance of the lake and its fragility 

Lake Malawi/Niassa/Nyasa is the fourth largest freshwater body in the world and 

constitutes an inestimable resource in this semi-arid region (Hecky & Bootsma 1999). It 

1

Figure 1. Lake Malawi, its catchment and the Rift valleys.

provides water for drinking, irrigation and domestic uses for people living on the lakeshores, 

but also fish. The value of the lake fishes does not lie only in their scientific interest, but also 

in their primordial nutritional status. In the Malawian part at least, they sustain vitally 

important fisheries that provide 75% of the animal protein consumed by people and work for 

an estimated 35,000 fishermen and presumably as many as 2,000,000 people through 

associated activities (Mkoko 1992, cited by Ribbink 1994 and Turner 1994b). Although the 

fish constitute, with the water itself, the most important resource of the lake and the main 

concern at this day, they are part of a complex ecosystem which needs to be preserved as a 

whole if it is to be used in a sustainable way. As mentioned previously, the fish and the other 

components of the food chains rely heavily on the water quality of the lake. The physical 

characteristics (depth, small outflow, long flushing time) of Lake Malawi/Niassa/Nyasa and 

their implications for pollution retention and ecosystem fragility have been discussed in detail 

by Bootsma & Hecky (1993). While its great depth allows the various pollutants to go 

undetected for many years, its low flushing rate makes the elimination process very long 

(several centuries) once thresholds are reached. The water quality has been a major issue of 

the SADC/GEF Project, which has provided a sound scientific knowledge about the lake 

limnology (see Bootsma & Hecky 1999 for review). Though the lake is still in rather pristine 

condition, the first signs of changes have already been observed, for the phytoplankton 

species characteristic of eutrophic systems, which were formerly rare are now becoming 

progressively dominant (Hecky et al. 1999). 

Main threats to the fish diversity 

Malawi is a weakly industrialised country in which most of the people live directly 

upon natural resources through agriculture, fisheries and associated activities. The 

demographic context, with one of the highest population density of Africa and an annual 

increase well over 3% (Ferguson et al. 1993, Kalipeni 1996), leads to a steadily increasing 

human pressure on the limited natural resources of the country. Two mains threats to the fish 

communities can be distinguished and both are related to changes in the use of natural 

resources. 

1) - Fishing activities are the more direct human influences on fish communities. In 

absence of alternative employment, the rapidly growing human population exerts an 

increasing fish demand, which entails an increased pressure on the already overexploited 

stocks (Turner 1995). Despite their huge economical and scientific interests, very little is 

known about Lake Malawi cichlid fishes. As emphasised previously, about only one third of 

the fish are described or catalogued, and new species are discovered regularly. Paradoxically, 

the most studied fish are the colourful rock-dwelling haplochromines, which are almost not 

exploited, except for the ornamental trade (Turner 1994b, 1995). The fish exploited for food 

purposes are those that inhabit the shallow and deep sandy shores. They sustain a highly 

diversified traditional fishery and a localised commercial mechanised fishery that have greatly 

expended over the past 20 years (Tweddle & Magasa 1989) and, which are according to the 

most resent assessments, already fully or over-exploited (Turner et al. 1995, Turner 1995). It 

has been stressed that mechanised fisheries might be incompatible with the continued 

existence of the highly diverse cichlid communities and that maximising the fish yield would 

lead to a decline in the number of endemic species in the exploited area (Turner 1977b, 

Turner 1995). Fisheries scientists have already shown the critical effects of over exploitation, 

such as the reduction in population size, the modification of size structure and some local 

extinction of the larger cichlid species (Turner 1977a, 1977b, Turner 1995, Turner et al. 

1995). However, while it is believed that cichlid populations are likely to slowly recover from 

overexploitation given their life-history characteristics (Ribbink 1987), it has also been 

2

suggested that cichlid fisheries were more resilient than previously thought (Tweddle & 

Magasa 1989). As pointed out by Turner (1994b), "it is essential to distinguish between the 

resilience of a multi-species fish stock and the vulnerability of individual species". The 

fishery's resilience might be achieved through the unnoticed disappearance of several species. 

Given the importance of fish for people nutrition, there is an urgent need for an appropriate 

fisheries management regulations. However, beside the huge number of species exploited, the 

extent of the shore line, the great variety of fishing techniques in use and their poorly known 

selectivity, effective fisheries management is currently hampered by the lack of knowledge 

about the fish taxonomy and life-histories. Taxonomy and systematic, which deal with species 

determination and description, provide species inventories and geographical distribution of 

ichthyofauna that are basic information for any management and conservation purposes. On 

the other hand, fisheries management relies on mathematics models to predict the evolution of 

stocks. These models are heavily dependent upon population parameters, such as breeding 

season, age and size at maturity, fecundity, growth and mortality rates, which are currently 

missing (Lowe-McConnell et al. 1994, Worthington & Lowe-McConnell 1994, Turner 1995). 

If exploited stocks are to be managed properly, the gaps in understanding have to be filled so 

that outstanding information is gathered. 

An other interesting question is: are the species which decline or disappear from trawl catches 

actually endangered? Most target fish of trawl fisheries are sandy bottom species for which 

belonging to specific areas of the lake and degree of stenotopy are poorly known. They also 

occur in other areas and/or depth of the lake where the localised mechanised fisheries do no 

longer occur (Banda & Tómasson 1996, Tómasson & Banda 1996). Their relative 

disappearance from fisheries catches in a particular area might then not be a real threat to 

Biodiversity. However, our present state of knowledge miss some very important information 

concerning the notion of “population” for the exploited species. For example, the same 

species in two distant parts of the lake could belong to different populations, presenting life 

history and/or genetic variations. They might also present morphological differences. In such 

a case the disappearance of one of these populations would be much more critical as it would 

lead to a lost of diversity. As most of the mechanised fisheries occurs in the southern part of 

the lake, studies aiming to determine the population status of the exploited species should be 

carried out in order to assess the potentiality of re-colonisation from less exploited parts of the 

lake. 

2) – Together with fishing, agriculture is the most important human activity in Malawi. The 

steadily increasing human populations and the degradation of lands in the river catchments, 

such as deforestation, burning of vegetation, destruction of wet lands on the river banks for 

agricultural purposes and the cultivation of marginal areas, are cause of major concern. All 

these activities, by removing the vegetation cover, weaken the soil, which is carried away 

with its nutrients directly in the rivers by the rains and ultimately arrive in the lake. Another 

source of nutrients and pollution are the industrial sewage. The land clearance burning is also 

suspected to strongly participate to the atmospheric phosphorus deposition in the lake. The 

limnology team of the SADC/GEF Project have identified the increasing load of sediments 

and nutrients received by the lake from rivers and atmosphere as the main threat to the water 

quality (Bootsma & Hecky 1999). The consequences of a sediment/nutrient enrichment of the 

lake on the water quality have been experienced in the Laurentian Great Lakes or Lake 

Victoria and reviewed in Bootsma & Hecky (1993). Among the main effects of increased 

sediment and nutrient loads on aquatic communities (see Patterson & Makin 1998 for review), 

the reduction of available living space as the oxic/anoxic boundary moves up (Bootsma & 

Hecky 1993), the reduction of light penetration affecting photosynthetic rates or sexual mate 

choice (Seehausen et al. 1997), the reduction of habitat complexity and destruction of 

3

spawning grounds are of direct importance for fish (Waters 1995, Evans et al. 1996, Lévêque 

1997). For instance, over-fishing and siltation resulting from deforestation have strongly 

diminished the abundance of potadromous fish species in Lake Malawi/Niassa/Nyasa 

(Tweddle 1992). In Lake Tanganyika, species richness of fish was found much lower at sites 

with high sedimentation than at less disturbed sites (Cohen et al. 1993a). Similar observation 

were reported for Lake Victoria, where increased turbidity was recognised partly responsible 

for the decline in cichlid diversity (Seehausen et al. 1997). 

Research program undertaken 

In June 1998, a new "senior Ecologist was appointed, in replacement of the former 

one, by the SADC/GEF Lake Malawi Biodiversity Conservation Project, which closing date 

was the 31/07/1999. Taking into account the main threats to the fish communities and the fact 

that a single annual cycle was left before the end of the project, we decided to focus our 

researches on the following particular aspects: 

- Provide the fisheries managers with the maximum information about the life histories 

(breeding season, age and size at maturity, fecundity, growth and mortality rates, diet) of 

the main demersal cichlid species, and the temporal patterns of their distribution, 

abundance and diversity. These research actions are detailed in Chapters 1 to 4. 

- As emphasised previously, for the conservation of biodiversity as well as for the fisheries 

management, it is crucial to known whether a species is represented by a single 

population widespread all over the lake, or by different populations (or stocks) with 

distinctive morphometric, genetic and life history characteristics. A complementary study 

has then been undertaken in collaboration with the taxonomists of the project, to compare 

the morphometrics, the genetics (microsatellites) and the life history traits of two species 

in four different locations between the SWA and Nkhata Bay. This part is detailed in 

Chapter 5. 

- Assess the potential influence of suspended sediments on the distribution, abundance, 

diversity and some life-history characteristics of the rocky shore cichlid fishes (Chapter 

6). 

4

Chapter 1: 

Temporal trends of trawl catches in 

the North of the South West Arm, 

Lake Malawi

Chapter 1: Temporal trends of trawl catches in the North of the 

South West Arm, Lake Malawi 

F. Duponchelle, A.J. Ribbink, A. Msukwa, J. Mafuka & D. Mandere 

Introduction 

Since the closing of trawling activities between Domira Bay and Nkhotakota in 1993, 

the trawl fisheries occur only in the SE and SW Arms of the lake (Tweddle & Magasa 1989, 

Banda et al. 1996, Banda & Tómasson 1996). During the last two decades, a number of 

reports and observations have pointed out the dangers of the current overexploitation of fish 

communities by trawling that has already led to drastic changes in size structures of the 

exploited stocks and to decreasing catches in the southern part of the lake (Turner 1977a, 

1977b, Turner 1995, Turner et al. 1995, Banda et al. 1996). However, the SEA, which hold most 

of the commercial trawling, has received much more attention than the SWA, where only one 

pair-trawler operates in the shallower zone (Tómasson & Banda 1996, A. Bulirani, pers. 

com.). While numerous studies have been carried out to improve knowledge of species 

distribution and abundance for a better management of mechanised fisheries (review by 

Tweddle 1991), none had focused on the seasonal or temporal trends of catches in the SWA 

until the recent two year survey with three months sampling intervals carried out by 

Tómasson & Banda (1996). As the trawler operating in the SWA fish only in the shallow 

waters of the southern part of the arm and given that traditional fisheries are mostly confined 

to shallow and inshore areas (Banda & Tómasson 1996, Tómasson & Banda 1996), the 

offshore part of the northern SWA can therefore be considered as almost unexploited, except 

for occasional surveys by the Ndunduma (A. Bulirani, pers. com.). Therefore, the north of the 

SWA appeared to be the ideal area to conduct a program designed to assess the temporal 

trends of the distribution, diversity, abundance and the life histories of the most important fish 

species caught by trawling. The unexploited aspect of the fish stocks was particularly 

favourable for the estimation of growth and natural mortality of the major species needed for 

fisheries management (Turner 1995). The following chapter deals with the temporal patterns 

of monthly trawl catches at exactly the same sites and depths in the north of the SWA over a 

complete annual cycle. 

Material and methods 

Trawl surveys 

The project's research vessel, R/V USIPA, was used for the surveys except for the 

months of July and August 1998, when the R/V NDUNDUMA was used. The NDUNDUMA, 

which belongs to the Fisheries Department, is a 17.5 m long trawler propelled by a 380 HP 

engine. R/V USIPA is a 15 m steel catamaran powered by twin 135 HP engines. The bottom 

trawl was approximately 40 m foot rope and 35 mm stretched cod end mesh. Morgère semi 

oval doors of 135kg each spread the trawl. Actual opening of the trawl was observed using 

5

Figure C1. The southern part of the Lake Malawi/Nyasa showing the South West Arm (SWA) 

and the South East Arm (SEA). The bars represent the monthly sample sites at 10, 30, 50, 

75, 100 and 125m depths.

the Scanmar height sensor, CT 150, and displayed on Scanmar’s color graphic monitor. The 

trawl opening varied between 4.1 and 4.3 m. 

Each tow was for a duration of 20 minutes at a speed of ± 4630m/h (2.5 knots, range 2.3-2.7). 

On average the distance covered by each tow was 1543m. Swept area varied between 

175,279.49 m³ at 10m depth to 277,868.04 m³ at 125m depth (Capt. M. Day 1999). 

Each month from June 1998 to May 1999, one tow was done at 10, 30, 50, 75, 100 

and 125 m depth on approximately always the same sites along a line between Chipoka and 

Lukoloma (Figure C1). The exact positions of every tow are given in Appendix 1. Owing to 

ship availability, no sample was collected in September 1998. 

Species identification 

This is of common knowledge, species identification in Lake Malawi is a real problem 

(Lewis 1982, Tómasson and Banda 1996, Turner 1995, 1996). Despite the very useful book of 

Turner (1996), fish identification remains extremely difficult on the field for many taxa. 

Moreover, as the identification problems are size-related, the small species (Aulonocara spp., 

Nyassachromis spp., and some Placidochromis spp. for examples) are more likely to lead to 

inconsistencies. 

However, we had to work along with these problems and, as this program was aimed 

to provide the fisheries department with the basic life histories of the most commonly trawled 

species, it was decided that if mistakes were to occur, they had to be consistent with the 

Fisheries Department's mistakes. For this reason, Davis Mandere, Research Assistant and 

"field identifier" at the Malawi Fisheries Department, did all the fish identifications on board. 

During the first two cruises (June and July 1998), Mark Hanssens, support taxonomist on the 

SADC/GEF Project assisted him in species identification in order to ensure the consistency of 

names used by the Fisheries Department and the SADC/GEF Project. George Turner was 

present for the August 1998 cruise and reported some inaccuracies concerning 

Rhamphochromis spp. Diplotaxodon spp. and small species groups such Aulonocara spp. It is 

believed that inaccuracies concerning the Diplotaxodon spp. encountered in the fished area 

(limnothrissa, macrops, apogon, argenteus, greenwoodii and brevimaxillaris) were solved 

during that cruise, at least for the common species (limnothrissa, macrops, apogon, 

argenteus). 

As our study mainly focused on cichlids, the catfishes were separated into three 

groups, Bathyclarias spp., Bagrus meridionalis and Synodontis njassae. No attempt was made 

to identify the species constituting the Bathyclarias spp. flock, which were lumped together 

into one group. Clarias gariepinus, rarely caught, was grouped within the Bathyclarias spp. 

complex. Despite the growing assumption that Synodontis njassae would be constituted by 

more than one species, no formal evidence has yet been provided and Synodontis were 

considered as a single species over their full depth range. 

Owing to the difficulty of identifying them accurately, Oreochromis spp. were lumped 

into one group, as were the Rhamphochromis spp. 

For the groups of small species such as Aulonocara spp., Nyassachromis spp., which 

species were not accurately identified, only the following species were recorded individually: 

Aulonocara 'blue orange', A. 'minutus', A. 'cf. macrochir', A. 'rostratum deep', Nyassachromis 

argyrosoma. 

It was suggested (J. Snoeks, pers. com.) that what we called Nyassachromis 

argyrosoma was probably a complex of different Nyassachromis spp., as these species are 

very difficult to identify and poorly known. However, for no particular anomaly appeared 

from data analysis, we kept considering it as a single species. 

Otopharynx argyrosma was also recorded as a single species, but it became evident while 

analysing the data (length-weight or fecundity-weight relationships) that more than one 

species were included under this name. 

6

As a rule, to avoid confusion given the rhythm imposed by sorting fish on board and 

to ensure the consistency of the name attributed to a given species, Davis Mandere was asked 

to consistently allocate a particular species the name he was used to, even when we knew the 

name had changed (or was wrong). The proper name was subsequently entered in the 

database. This was the case for the following species for instance: 

- Stigmatochromis guttatus was identified as 'woodi deep' on board. 

- Sciaenochromis benthicola was recorded as 'spilostichus' on board 

What we thought was Lethrinops 'longipinnis orange head' turned out to be 

Lethrinops argenteus (Snoeks, pers. com.). Actually, the characteristic L. longipinnis whose 

breeding male has a blue head and a dark striped body (see illustration p. 58 in Turner 1996) 

was never found in our fishing area in the SWA. Some males were found sometimes with a 

darker dress, but never with a blue head. The species we identified as L. 'longipinnis orange 

head' is illustrated p. 57 (top right picture) in Turner's book (1996) as L. longipinnis Domira 

Bay. The taxonomy team of the project has found that L. longipinnis was a complex of 

different species (Snoeks, pers. com.), and that Lethrinops 'longipinnis orange head' was 

definitely Lethrinops argenteus (Ahl 1927). In our case 99% of the specimen were found at 

depth between 10 and 50m, and seldom below. This tends to confirm that 'orange head' 

differs from longipinnis, which is supposed to frequently occur at greater depths (Turner 

1996). 

The spelling of species names used was that given in Turner (1996). 

Catch analysis 

For each tow, the catfishes Bathyclarias spp. and Bagrus meridionalis were separated 

from the main catch, counted and weighed. The rest of the catch was then randomly 

distributed in 50 kg boxes and the weight recorded. The total catch weight (kg) was recorded 

as the sum of Bathyclarias spp., Bagrus meridionalis and the remaining catch. 

A 50 kg filled box was taken as a representative sample of the whole catch and 

analysed. Large and medium sized fish were sorted out of this sample with rare species and 

classified according to their taxonomic status. The weight of the remaining "small fish" (< 5-8 

cm TL) from the catch was weighed and a random sub-sample of about 3 kg was removed 

from the sample and placed in the deep freeze for later examination. When the large, medium 

and rare species were processed, the sub-sample of small fishes was processed following the 

same protocol. 

For each species, the number of specimens and their total weight was recorded to the 

nearest g. The standard length (SL) of each specimen was recorded to the nearest mm for 

analysis of length frequencies. When the number of specimens for a given species was too 

large, a sub-sample (which proportion in weight of the main sample was recorded) comprising 

at least 100 specimens was taken. This procedure was mainly used for the large males schools 

of identical size. 

Nine target species were selected according to their relative abundance, depth 

distribution and basic ecological characteristics (benthic or pelagic habits, broad trophic 

category) (Tómasson & Banda 1996, Turner 1996). These were Lethrinops gossei Burgess & 

Axelrod, Lethrinops argenteus Ahl (= L. 'longipinnis orange head'), Diplotaxodon 

limnothrissa Turner, Diplotaxodon macrops Turner & Stauffer, Copadichromis virginalis 

Iles, Mylochromis anaphyrmus Burgess & Axelrod, Alticorpus mentale Stauffer & McKaye, 

Alticorpus macrocleithrum Stauffer & McKaye and Taeniolethrinops praeorbitalis Regan. 

For these species, all the females from each haul were preserved in formalin for later 

examination. 

7

Environmental data 

After each tow, a CTD cast and a grab sample were taken in the middle of the transect. 

Both CTD and the benthic grab were lowered using the hydrographic winch of the R/V 

USIPA. The CTD casts recorded, every 2 seconds during the way down and the way up, 

measures of the following parameters: depth (m), temperature (ºC), oxygen concentration 

(mg.l -1 ), conductivity (mS.cm -1 ), water clarity (% transmission), fluorescence (arbitrary unit). 

Grab samples 

After each trawl a sample of bottom sediments was taken in the middle of the trawl 

transect by using a 24 cm benthic grab sampler lowered on the hydrographic winch. The grab 

digs about 10 cm into the sediment in such a way that the upper layers form more of the 

sample than the lower layers. It therefore gives qualitative rather than quantitative 

information. Each sediment sample was placed in a bucket. A sub-sample was taken, placed 

in 250 ml plastic bottle and deep frozen for later determination of sediment particle size. In 

March, April and May 1999, after the sub-sample was removed, the remaining part of the 

sediment sample was fixed in formalin (10%) for later extraction of benthic organisms for 

stable isotope studies. 

Determination of sediment particle size: 

The deep frozen sub-sample was mixed by hand after de-freezing and a sub-sample of 200 cc 

was placed in a 1 liter measuring cylinder toped up to 1000 cc with water. The cylinder was 

then inverted and shaken several times to suspend the sediment in the water. The sediment 

was then passed through a series of sieves (2 mm, 1mm, 500 µm, 250 µm, 125 µm, 63 µm) 

starting at the largest aperture. The volume of sediment retained in each sieve was determined 

using a measuring cylinder filed with water. Size class boundaries were as follows: > 256 mm 

= boulders, 64-256 mm = cobbles, 4-64 mm = pebbles, 2-4 mm granules, 1-2 mm = very 

coarse sand, 500 µm-1 mm = coarse sand, 63 µm-500 µm = fine sand, < 63 µm = silt and clay 

(mud). According to the proportions of the different components, the sample was then 

roughly categorized as "very coarse sand" (> 1 mm), "medium sand" (250 µm-1 mm), "very 

fine sand" (63 µm-250 µm) and "mud"(

Total catch all depth (kg.pull-1) 

1600 

1400 

1200 

1000 

800 

600 

400 

200 

0 

Jul Aug Oct Nov Dec Jan Feb Mar Apr May 

1998 1999 

Figure C2. Total catches all depths pooled over the full sampling period (July 1998-May1999). 

CPUE per depth (kg. 20 min pull -1 ) 

300 

250 

200 

150 

100 

50 

0 

10m 30m 

50m 75m 

100m 125m 


1998 

1999 

Figure C3. CPUE per depth over the full sampling period (July 1998-May1999). 

Total catch per depth category (kg.pull -1 ) 

1200 

1000 

800 

600 

400 

200 

0 

Total 10-50m 

Total 75-125m 


1998 1999 

Figure C4. Total catch per depth category over the full sampling period (July 1998-May1999).

Results 

Catches per month 

Owing to non uniformity between the record sheets of June and the other months, the 

data for June 98 were not included in the analyses. The results presented below concern the 

period from July 1998 to May 1999. 

The total catches per months all depths pooled fluctuated from about 600 kg for six 20 

min pulls, to about 1000 kg (Figure C2). The high value recorded in August 1998 was due to 

an exceptional catch of Bathyclarias spp. at 50 m: 42 specimens giving a total of 400 kg, with 

a total catch of 626 kg (Figure C3). Individual catches fluctuated between 30.5 kg at 100 m in 

October and 283 kg at 75 m in July, excluding the 626 kg recorded in August (Figure C3). 

Temporal fluctuation was observed in the catches, the lowest were recorded in October 1998 

and March 1999 and the highest in July-August 1998 and January 1999 (Figure C2). This 

temporal fluctuation was also observed for each depth (Figure C3) and when depths were 

pooled per category (Figure C4). With the exceptions of July-August 1998 and May 1999, the 

catches in the shallows and the in the deep waters were very similar (Figure C4). 

Catches per depth 

The mean CPUE per depth, all months pooled (Figure C5a) showed that the highest 

catches were recorded at 50 m and the lowest at 30 m. Catches were generally higher in the 

deep zone (50 to 125m) than in the shallows (10 to 30m). Almost the same results were 

obtained when the exceptional catch of Bathyclarias spp. in August 1998 was removed, 

except that the highest catches were recorded at 75m (Figure C5b). However, no significant 

difference of catch among depths was found in either cases, respectively with (Kruskal-Wallis 

one-way ANOVA on ranks H=8.33, 5 df, p=0.139) or without the August Bathyclarias spp. 

catch (one-way ANOVA F=1.845, 5 df, p=0.118). 

125 

125 

100 

100 

Depth (m) 

75 

50 

Depth (m) 

75 

50 

30 

30 

10 

10 

0 50 100 150 200 250 300 350 

0 50 100 150 200 250 

a 

Mean CPUE (kg. 20 min pull -1 ) 

b 

Mean CPUE (kg. 20 min pull -1 ) 

Figure C5. Mean CPUE (kg / 20 min pull) per depth (± standard deviation) over the full sampling 

period in the SWA (July-98 to May 99) (a) and with the exceptional Bathyclarias spp. catch 

removed (b), see text. 

9

Table C1. Proportion in weight of the main demersal species trawled at 10m depth in the SWA (catfishes species in italic). 

Species name Jul-98 Aug-98 Oct-98 Nov-98 Dec-98 Jan-99 Feb-99 Mar-99 Apr-99 May-99 Mean 

Aulonocara blue orange - 0,8 2,4 0,4 11,7 9,3 3,1 0,7 0,1 - 2,8 

Bagrus meridionalis 7,5 - 12,9 4,0 5,1 6,5 14,2 9,1 3,6 2,4 6,5 

Bathyclarias spp. 0,6 - 5,0 5,8 10,1 10,3 17,6 8,6 4,2 6,1 6,8 

Buccochromis lepturus 8,7 3,0 7,2 1,1 0,1 - 0,0 3,0 10,0 11,4 4,4 

Buccochromis nototaenia 2,1 0,4 1,0 0,1 0,5 0,6 0,0 0,6 1,6 - 0,7 

Chilotilapia rhoadesi 4,2 2,4 1,1 2,0 0,6 0,3 0,2 2,0 0,2 0,8 1,4 

Copadichromis quadrimaculatus 0,2 - 6,5 0,1 0,0 0,1 0,4 1,6 0,3 1,0 

Copadichromis virginalis - 1,2 - - 0,2 0,2 2,3 - - - 0,4 

Ctenopharynx nitidus - - 0,8 0,1 0,3 0,0 0,1 0,1 0,2 0,4 0,2 

Lethrinops altus - - - 0,4 0,4 0,0 0,1 - - - 0,1 

Lethrinops furcifer - - 5,9 0,2 0,1 0,5 0,1 - - - 0,7 

Lethrinops argenteus 16,2 3,5 3,5 24,3 32,5 21,3 10,5 7,5 7,0 5,8 13,2 

Lethrinops macrochir - 0,2 - 0,1 0,4 1,3 10,3 - 0,0 - 1,2 

Mylochromis anaphyrmus 3,4 1,3 5,9 6,8 0,5 2,3 2,4 5,1 3,7 7,4 3,9 

Mylochromis melanonotus - 0,6 0,5 - 0,5 - - - - 0,3 0,2 

Mylochromis spilostichus 2,4 0,1 - 0,1 - - 0,2 0,6 0,2 0,8 0,4 

Nyassachromis argyrosoma - - 38,4 32,5 20,2 5,7 11,9 53,6 38,5 30,5 23,1 

Oreochromis spp. 17,7 56,6 0,3 - 12,3 31,9 16,9 - 5,4 6,3 14,7 

Otopharynx cf productus 0,4 0,2 1,7 0,1 - 0,1 1,0 1,2 0,1 1,8 0,7 

Otopharynx decorus - 1,8 0,4 0,2 0,2 - 0,1 - 0,1 - 0,3 

Placidochromis suboccularis - - - 0,1 0,0 - 0,0 0,1 - 0,1 0,0 

Pseudotropheus livingstoni - - 2,4 0,6 0,0 - 0,0 2,7 2,0 5,9 1,4 

Synodontis njassae 0,9 1,0 - 15,8 0,7 0,6 0,1 0,5 0,2 0,3 2,0 

Taeniolethrinops furcicauda 0,6 0,2 2,0 0,2 0,1 - 0,9 1,3 3,2 7,5 1,6 

Taeniolethrinops praeorbitalis - 2,6 0,2 - - 1,5 1,4 - - - 0,6 

Trematocranus placodon 2,4 2,1 - 0,1 0,5 0,3 1,4 - - - 0,7 

Total 67,5 77,9 98,1 95,1 96,9 92,7 94,8 97,2 82,0 88,0 89,1 



Aulonocara blue orange - 19,0 - 3,4 0,2 4,4 3,1 - 0,1 10,0 4,0 

Aulonocara macrochir - - 0,1 0,1 0,3 0,2 - 0,1 - 0,0 0,1 

Bagrus meridionalis 5,3 5,5 5,8 8,1 4,2 22,6 14,2 6,1 7,1 1,4 8,0 

Bathyclarias spp. 12,4 5,5 5,8 4,3 8,3 7,7 17,6 - 4,0 2,6 6,8 

Buccochromis lepturus 2,0 1,0 - - - - 0,0 1,0 - 1,4 0,5 

Buccochromis nototaenia 3,3 2,6 1,8 1,7 0,8 2,4 0,0 1,7 1,3 1,6 1,7 

Chilotilapia rhoadesi 10,7 0,4 1,0 0,4 - 0,3 0,2 0,2 0,1 0,7 1,4 

Copadichromis quadrimaculatus 3,7 11,8 0,5 0,8 - 2,6 0,1 0,1 0,3 - 2,0 

Copadichromis virginalis 1,2 21,2 0,4 1,9 67,7 1,6 2,3 - - 0,0 9,6 

Lethrinops altus - - 0,2 2,7 0,5 1,9 0,1 17,7 1,7 0,3 2,5 

Lethrinops longimanus 0,4 - - 1,2 - 0,2 - 0,2 - - 0,2 


Lethrinops matumba - - 0,4 1,0 0,2 1,6 - 2,9 0,1 0,5 0,7 


Mylochromis spilostichus - 0,8 - 0,5 - - 0,2 0,5 0,5 1,2 0,4 

Nyassachromis argyrosoma - - 37,9 15,2 3,1 23,2 11,9 34,3 44,1 34,8 20,5 

Oreochromis spp. - 7,2 - - - 0,3 16,9 - - 0,7 2,5 

Otopharynx argyrosoma 1,6 1,3 - 0,2 3,6 4,4 - 1,8 - - 1,3 

Otopharynx speciosus 0,2 - - 0,0 0,2 0,3 - 0,2 1,4 1,2 0,3 

Placidochromis long - - - 1,5 - 0,1 - 0,6 - 0,7 0,3 

Rhamphochromis spp. 2,5 1,3 0,5 11,8 0,7 0,1 0,2 2,8 2,5 1,7 2,4 

Synodontis njassae 0,5 0,3 23,3 9,4 1,4 0,7 0,1 3,1 3,4 3,1 4,5 

Taeniolethrinops laticeps 0,1 0,2 - - 0,2 0,3 - - - 0,5 0,1 

Taeniolethrinops praeorbitalis 0,8 0,4 0,2 0,1 - 0,5 1,4 - 0,1 0,1 0,4 

Total 93,7 97,3 96,4 97,2 99,7 99,0 81,1 95,3 99,5 94,0 95,3

Proportions of cichlids and catfishes 

The proportion of cichlids and catfishes (Bagrus meridionalis, Bathyclarias spp. and 

Synodontis njassae) in the catches at each month are presented in the Figures C6a and C6b, in 

number and weight respectively. The catfishes constituted regularly between 2 and 9% of the 

catches in number from July to December 1998 and less than 0.5% between January and May 

1999 (Figure C6a). On the other hand, catfishes represented consistently 8 to 25% of the 

catches in weight during the whole sampling period (Figure C6b). 

100 

100 

% of catches in number 

90 

80 

70 

60 

Catfish 

Cichlids 

% of catches in weight 

90 

80 

70 

60 



a 

50 


1998 1999 

b 

50 


1998 1999 

Figure C6. Proportions of cichlids and catfishes in the catches over the sampling period (July-98 to 

May-99), in number (a) and weight (b). 

The proportion of catfishes per depth varied from 2% at 75 and 100 m to 5% at 125 m, 

in number (Figure C7a) and from 15.3% at 10 m to 22% at 100 m, in weight (Figure C7b). 

The proportion in weight of catfishes was not significantly different among depth (F=0.445, 

p=0.815). 

Overall mean % of catches 

(number) 

50 60 70 80 90 100 

Overall mean % of catches 

(weight) 

50 60 70 80 90 100 

10 

10 

Depth (m) 

30 

50 



Depth (m) 

30 

50 



75 

75 

100 

100 

125 

125 

a b 

Figure C7. Proportions of cichlids and catfishes in the catches per depth all months pooled, in 

number (a) and weight (b). 

10



Alticorpus mentale 0,3 - - - 0,8 0,9 0,5 0,7 0,7 0,1 0,4 

Aulonocara blue orange - 0,2 - - - 0,5 0,6 - - 0,0 0,1 

Aulonocara macrochir 0,1 0,4 1,1 1,0 1,7 1,3 0,1 1,2 2,6 0,2 1,0 


Bathyclarias spp. 11,9 41,4 4,4 8,5 - 3,1 10,9 1,9 0,0 4,2 8,6 

Copadichromis quadrimaculatus - 0,8 1,4 0,4 - 0,2 0,2 0,3 0,1 - 0,3 

Copadichromis virginalis 50,4 1,0 1,8 47,7 53,8 27,9 33,1 10,5 22,2 44,2 29,2 

Diplotaxodon argenteus 0,2 - 0,9 - - 0,5 - 1,8 0,5 1,7 0,6 

Diplotaxodon limnothrissa 1,3 0,1 0,5 - 0,3 5,1 0,1 28,1 2,1 0,9 3,8 

Docimodus johnstoni 0,1 0,2 - - - 0,1 - - - 0,3 0,1 

Hemitaeniochromis insignis - - 0,1 - - - - 0,1 0,0 0,0 0,0 

Lethrinops altus 0,3 0,8 0,7 0,5 0,5 1,9 0,6 0,6 2,3 - 0,8 

Lethrinops longimanus 1,1 4,5 0,0 3,0 0,4 0,4 6,7 0,4 0,4 0,1 1,7 


Lethrinops minutus - 1,1 6,3 - - 4,3 0,9 1,1 0,5 4,3 1,8 

Lethrinops parvidens - - - - 0,0 - 0,1 0,1 0,2 - 0,0 


Mylochromis spilostichus - 7,3 - - - - 0,6 0,1 0,4 0,9 0,9 

Otopharynx speciosus 0,3 1,3 0,2 0,2 0,5 0,6 1,1 0,7 0,1 0,5 0,5 

Placidochromis long - - 2,6 1,5 3,9 1,3 0,6 0,2 0,1 5,3 1,6 


Sciaenochromis benthicola 0,7 0,5 0,8 0,3 0,0 3,4 5,1 0,6 0,9 0,6 1,3 


Trematocranus brevirostris - - 16,7 1,6 0,0 2,3 3,5 10,1 14,5 1,4 5,0 

Total 97,5 86,1 96,3 95,9 99,9 85,2 98,8 99,0 96,7 98,1 95,4 



Alticorpus spp. 0,6 - - - 2,2 2,1 0,3 - - - 0,5 

Alticorpus geoffreyi 20,1 20,4 9,0 4,7 8,0 1,1 2,2 4,2 6,3 12,2 8,8 

Alticorpus macrocleithrum 1,1 1,3 0,1 - - - - - - 0,1 0,3 

Alticorpus mentale 3,5 4,4 4,8 11,6 18,7 2,0 16,1 4,5 3,5 4,2 7,3 

Alticorpus pectinatum 0,8 0,3 0,6 0,1 1,5 1,2 5,0 3,8 1,8 2,2 1,7 

Aulonocara minutus 0,7 0,9 0,5 - 0,9 0,2 0,3 1,4 0,3 1,8 0,7 

Aulonocara rostratum - - 2,0 - - 0,1 0,2 1,9 0,8 0,9 0,6 


Bathyclarias spp. 17,6 8,8 24,9 6,1 0,7 15,0 11,7 4,0 2,1 9,7 10,1 

Diplotaxodon apogon - 1,9 0,3 - 8,4 9,2 5,6 2,2 1,9 0,5 3,0 

Diplotaxodon argenteus 1,0 0,8 1,9 1,8 3,0 5,6 3,6 3,7 3,5 1,7 2,6 

Diplotaxodon macrops 2,3 3,6 - - 0,5 13,6 4,7 9,1 3,9 2,3 4,0 

Diplotaxodon limnothrissa 3,7 2,9 12,9 2,0 0,7 1,5 8,4 4,1 19,4 21,3 7,7 

Lethrinops deep water albus 0,2 1,2 0,1 31,3 0,1 - - - - - 3,3 

Lethrinops gossei 16,2 14,4 9,2 1,9 16,1 17,1 17,3 29,5 41,4 17,3 18,0 

Lethrinops oliveri 2,7 19,4 7,2 9,8 12,9 13,5 5,3 13,7 3,2 8,9 9,7 

Lethrinops polli 5,4 5,8 2,3 0,5 1,4 1,2 2,1 4,2 0,5 6,8 3,0 

Pallidochromis tokolosh 1,3 1,3 0,4 - 1,3 1,7 1,0 0,2 0,1 0,7 0,8 


Sciaenochromis alhi - - 0,2 0,0 0,2 0,1 0,2 - - 0,9 0,2 

Sciaenochromis benthicola 0,1 0,1 - 2,0 7,1 0,3 - 0,3 0,0 0,2 1,0 


Total 96,0 97,5 94,2 83,4 91,0 95,4 97,4 96,4 98,0 95,9 94,5

Catch composition 

Fishes representing the major part of the catches at each month are presented in Tables 

C1 to C6 for the depths of 10m, 30m, 50m, 75m, 100m and 125m respectively. Although 

cyprinids and mormyrids were sometimes caught, their occurrence was so rare and their 

contribution to the catches so weak that they were negligible. Therefore, catches were 

assumed to be constituted only of cichlids and catfishes. 

The catfish species (Bathyclarias spp., Bagrus meridionalis and Synodontis njassae) were 

consistently amongst the most important species (in weight) at each depth, averaging 15.3% 

at 10m, 19.3% at 30m, 21% at 50m, 18.9% at 75m, 21.6% at 100m and 17.6% at 125m. 

Owing to their large sizes, the Bathyclarias spp. and the Bagrus meridionalis were much less 

important in number as illustrated in Figures C8a and C8b respectively. 

Overall mean catches (%) 

0 5 10 15 


0 2 4 6 8 10 

10 

30 

Weight 

Number 

10 

30 

Depth (m) 

50 

75 

Depth (m) 

50 

75 

100 

125 

100 

125 

Weight 

Number 

a 

Figure C8. Overall mean catches (in proportion of weight and number) per depth 

for Bathyclarias spp. (a) and Bagrus meridionalis (b) from July 98 to May 99. 

B. meridionalis was proportionally more abundant in the shallow waters (10 to 50m) while 

Bathyclarias spp. was better represented in the deep waters (75 to 125m). 

b 


0 5 10 15 

Depth (m) 

10 

30 

50 

75 

Weight 

Number 

100 

125 

Figure C9. Overall mean catches (in proportion of weight and number) per depth for 

Synodontis njassae from July 98 to May 99. 

11




Alticorpus macrocleithrum 3,9 2,8 - 1,2 1,6 1,0 0,9 0,5 0,1 0,2 1,2 


Alticorpus pectinatum 2,6 0,2 0,4 5,7 1,2 0,6 2,6 0,6 1,1 1,1 1,6 

Aulonocara long - 0,1 - - 0,1 - 0,0 0,0 0,0 0,1 0,0 

Aulonocara minutus 0,8 1,0 0,2 0,8 0,2 0,0 0,8 0,1 0,4 0,2 0,5 

Aulonocara rostratum - - - - - - 0,4 0,2 0,0 0,0 0,1 


Bathyclarias spp. 6,9 9,4 - 15,0 10,1 5,5 25,6 20,0 10,1 10,0 11,3 

Diplotaxodon apogon - 2,2 3,1 17,7 4,3 1,8 0,1 0,7 0,3 0,9 3,1 

Diplotaxodon argenteus 0,7 - 14,9 5,0 3,7 0,7 0,0 0,9 1,0 1,2 2,8 

Diplotaxodon macrops 0,3 6,9 2,4 2,5 25,0 21,5 1,2 18,3 5,8 20,2 10,4 

Diplotaxodon limnothrissa 0,1 - 52,8 3,6 5,9 2,6 1,5 1,1 0,8 28,5 9,7 

Lethrinops deep water altus 5,9 5,8 1,4 6,1 - 1,2 0,3 0,1 6,2 2,9 3,0 


Lethrinops oliveri 4,9 17,4 2,8 4,4 4,6 3,5 0,3 0,4 - 1,8 4,0 

Lethrinops polli 0,1 0,9 0,2 3,2 - 1,2 0,2 0,2 0,5 0,1 0,7 

Pallidochromis tokolosh 0,1 - 0,1 0,6 0,1 0,0 - 0,0 1,4 0,4 0,3 

Placidochromis "flatjaws" 0,4 - - - 5,7 0,1 0,1 0,0 0,5 - 0,7 

Placidochromis platyrhynchos 1,8 1,2 0,1 - 1,3 0,3 0,2 0,0 0,3 0,1 0,5 


Total 95,4 94,3 86,7 94,0 93,6 99,5 95,8 99,5 99,0 99,8 95,8 



Alticorpus spp. 0,1 - - - 0,9 1,9 1,8 2,5 0,5 1,6 0,9 


Alticorpus macrocleithrum - 0,1 - 0,2 2,1 0,2 0,1 - - - 0,3 


Alticorpus pectinatum 0,1 - 0,3 4,5 4,7 - 0,2 0,9 1,8 - 1,3 

Aulonocara long - - 0,6 - - 0,0 0,1 0,3 0,1 0,3 0,1 

Aulonocara minutus - 0,2 1,4 1,4 1,9 1,1 0,6 8,2 1,0 0,7 1,7 

Aulonocara rostratum - - 0,5 - - 0,1 0,5 0,4 - 0,7 0,2 


Bathyclarias spp. 15,1 6,3 2,4 2,7 13,5 7,3 8,5 3,6 2,7 4,0 6,6 

Diplotaxodon apogon - 12,5 3,1 4,0 1,8 1,7 1,2 1,6 5,4 1,9 3,3 

Diplotaxodon argenteus 0,2 0,2 3,8 1,0 1,2 1,5 1,8 0,2 3,1 2,6 1,6 

Diplotaxodon macrops 0,7 7,5 7,8 10,9 - 15,2 17,2 6,2 24,3 27,6 11,8 

Diplotaxodon brevimaxillaris - - 0,5 0,5 0,3 0,5 0,3 - - 1,4 0,3 

Diplotaxodon limnothrissa 0,1 0,2 0,7 1,0 0,1 0,9 0,6 0,8 3,3 9,4 1,7 

Hemitaeniochromis insignis - - 0,2 - - 0,3 - 0,1 0,1 0,1 0,1 

Lethrinops deep water albus 5,1 0,3 4,1 0,2 0,1 0,8 - - - 0,1 1,1 

Lethrinops deep water altus 9,1 6,6 6,5 4,4 2,1 2,6 6,0 4,2 6,0 5,4 5,3 


Lethrinops oliveri 3,5 3,6 2,7 2,8 5,1 1,7 0,8 8,6 - 0,7 3,0 

Lethrinops polli 0,2 - 0,1 0,7 0,2 - 0,2 - - - 0,1 

Pallidochromis tokolosh 1,0 3,0 3,5 0,3 1,3 2,9 2,8 0,4 1,4 2,6 1,9 

Placidochromis "flatjaws" - - - - 0,3 0,0 0,5 3,3 0,3 - 0,4 

Placidochromis platyrhynchos 1,9 10,6 4,9 0,6 1,4 2,0 3,3 6,5 1,8 2,1 3,5 


Total 88,9 96,7 90,5 95,8 85,9 97,8 98,5 99,2 99,5 98,0 95,1

The smaller S. njassae was more evenly represented in number and weight and appeared more 

abundant in the very deep zone (100-125m, Figure C9). 

A minimum of 145 (see Appendix 2) to at least 170 different species were caught 

during the sampling year from June 1998 to May 1999 (taking into account the several 

species lumped together under their generic names, such as the Aulonocara spp., the 

Bathyclarias spp., the Copadichromis spp., the Lethrinops spp., the Mylochromis spp., the 

Nyassachromis spp., the Oreochromis spp., the Otopharynx spp., the Placidochromis spp., the 

Rhamphochromis spp., the Sciaenochromis spp.). However, despite this high number of 

sampled species, relatively few cichlid species accounted for more than 50% of the catches in 

weight at all depths, respectively 51% at 10m (Lethrinops argenteus, Nyassachromis 

argyrosoma and Oreochromis spp. Table C1), 55.2% at 30m (Copadichromis virginalis, L. 

argenteus Mylochromis anaphyrmus and N. argyrosoma Table C2), 56.9% at 50m (C. 

virginalis, Diplotaxodon limnothrissa, L. argenteus, and Trematocranus brevirostris Table 

C3), 55.5% at 75m (Alticorpus geoffreyi, Alticorpus mentale, Diplotaxodon macrops, D. 

limnothrissa, Lethrinops gossei and Lethrinops oliveri Table C4), 53.9% at 100m (A. mentale, 

D. macrops, D. limnothrissa, L. gossei Table C5) and 51.6% at 125m (A. mentale, D. 

macrops, Lethrinops "deep water altus" and L. gossei Table C6). Some of these species were 

dominant over two to three depths, such as L. argenteus, C. virginalis and N. argyrosoma in 

the shallows (10 to 50m), A. mentale, D. macrops, D. limnothrissa and L. gossei in the deeper 

waters (75 to 125m). 

Added to the proportion of catfishes at each depths, about 10 fish species only accounted for 

70 to 80% of the catches in weight over the sampling period. 

A clear change in species composition appeared after 50 m, the "shallow" water species being 

encountered down to 50m whereas the characteristic "deep" water species appeared from 75 

m downwards (Tables C1 to C6). 

The results of catch per unit effort (kg / 20 min pull) for each species according to 

depth are summarised in Appendix 3. The total number of species caught over the sampling 

period decreased with increasing depth from 80 species at 10 m to 48 at 125 m (Appendix 3). 

Again, these values are underestimated owing to the several species lumped together under 

their generic names. Unlike the three catfish species, which were consistently caught at any 

depth, very few cichlid species had depth distribution covering all the sampled depths 

(Appendix 3). Only 12 out of the 133 cichlid species or species groups listed in Appendix 3 

covered all (or at least 5 of) the sampled depths. Most of the others were restricted to three or 

four depths and some species were confined to one or two depths only. 

Discussion 

During the whole sampling period (June 1998 to May 1999), no other trawlers were 

encountered in the sampled area, roughly from Chipoka to Lukoloma (Figure C1). The 

trawlers in activity in the SWA occur in the southern part of the arm and only the Ndunduma 

can occasionally trawl in the north of the SWA (A. Bulirani, pers. com.). Therefore, it can be 

considered that our sampled area is almost not commercially exploited by trawlers. We 

recorded the highest catches at 75 and 100 m, and the catches were higher at 125 m than at 10 

and 30 m, whereas the CPUE is supposed to be higher in the shallow zone (Turner 1977a, 

Tómasson & Banda 1996). This is likely to be a consequence of the light exploitation of the 

deep zone by commercial fisheries whereas the shallow zone is heavily exploited by artisanal 

fishermen in the studied area. 

Temporal fluctuations of the total catches per month (all depths pooled) were 

observed. But the same temporal patterns were also observed at each depth and when depths 

were pooled per category, suggesting that the representativeness of our sampling was good, 

despite a potential inter-haul variability. Tweddle & Magasa (1989) also reported seasonal 

12

Temperature (°C) 

23 24 25 26 27 28 29 

0 

-10 

-20 

-30 

-40 

-50 

Depth (m) 

-60 

-70 

-80 

-90 

-100 

-110 

-120 

-130 

Jun-98 

Jul-98 

Aug-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

March-99 

May-99 

Figure C10. Seasonal progression of temperature profile according to depth off Cap Maclear, 

SWA. 

Depth range (m) 

-125 

-100 

-75 

-50 

-30 

-10 

0 5 10 15 

Mean bottom type index per depth range 

Figure C11. Modification of bottom type with depth in the SWA. Each bottom type category 

was given an arbitrary value for graphic representation: 15 for "very coarse sand", 10 for 

"medium sand", 5 for "very fine sand" and 0 for "mud". The values are the means over five 

months (June to December 1998).

trends in the catch rates in the SEA with usually a peak in August and September, which is 

supported by our results. 

The catches were dominated by cichlids both in number and weight. However, the 

catfishes, represented by only 3 genera (Bathyclarias, Bagrus and Synodontis) of which two 

have a single species (Bagrus meridionalis and Synodontis njassae), consistently constituted a 

significant part of the catches. Owing to their large size (for Bathyclarias spp. and Bagrus 

meridionalis at least), their contribution to the catches was much more important when 

referred to their biomass than to their number. They consistently represented between 10 and 

25% of the catches. Tómasson & Banda (1996) found that in the SWA B. meridionalis was 

more abundant in the deep waters (50 to 100 m) but bigger in the shallows (0 to 50 m). 

During our sampling period and in the sampled area, which was restricted to the north of the 

SWA, B. meridionalis was more abundant between 10 and 50 m, as observed by Turner 

(1977), and large specimens were evenly distributed according to depth. Bathyclarias spp. 

tended to be better represented in the deep waters from 50 m downwards whereas their 

maximum catch was observed at 40-60 m by Turner (1977). As pointed out by Tómasson & 

Banda (1996), Synodontis njassae was common at all depths and displayed an increasing 

occurrence and abundance with depth, becoming much more abundant in the very deep zone 

(100 and 125 m). Although specimens from 50 to 200 mm (standard length) were recorded, 

most individuals caught were of uniform size, between 90 and 110 mm SL, which 

corresponded to previous observations of 12 to 14 cm TL (Tómasson & Banda 1996). 

When adjusted to a 30 min pull and per depth category, the CPUE per species 

(Appendix 3) were not always consistent with those reported by Tómasson & Banda (1996). 

Details will be given in Chapter 2. 

A marked change in species composition was reported to occur around 50 m in the 

SWA (Tómasson & Banda 1996). It was hypothesised to be related to the position of the 

thermocline or the substrate type. This spectacular shift in species composition between 50 

and 75 m was also observed in our study. However, the position of the thermocline does not 

seem to be the best explanation to that pattern for it fluctuates significantly with season 

(Figure C10), whereas the species distribution pattern is stable (Tables C1 to C6). As most of 

the exploited species are demersal fish and therefore closely related to the bottom, the 

sediment quality might constitute a better explanation. The grab sample analyses revealed a 

gradient in bottom type composition from the shallows to the deep waters. The bottom types 

can roughly be categorised as "very coarse sand", "medium sand", "very fine sand" and 

"mud". Each of these categories were attributed arbitrary values, respectively 15, 10, 5 and 0 

for the sake of graphic representation. The results of grab sample analyses over several 

months are summarised in Figure C11. A clear change in bottom composition from coarse 

and medium sand to very fine sand and mud appears after 50m and is likely to influence the 

species composition pattern according to depth. 

A notable observation was that throughout the year, the bulk of the catches was 

constituted by a few common cichlid and catfish species. At any given depth, despite the large 

number of species regularly recorded, 60 to 80% of the catches was made of no more than ten 

species including the three catfishes. And about twenty species only accounted for 90 to 95% 

of the catches at each depth, with some species being dominant in two or three of the sampled 

depths. This indicates that the largest part of the species caught is relatively rare or at least 

infrequent. For the rarer ones, the occurrence in the catches might be incidental to unusual 

movements out of their habitat, which would expose them to the trawl. Another potential 

explanation might be that we did sample only a restricted amount of different habitats, though 

this hypothesis is very unlikely given the surface covered by a 20 min pull. Hence, for the 

majority of the infrequent species it probably means that they do exist in small population 

number and/or have patchy distributions either because of their high specialisation to specific 

type of habitats or because of the narrowness of their trophic niche. In any case, these species 

are likely to be the first endangered by intensive exploitation. 

13

The decreasing number of species caught with increasing depth reported by previous 

authors (Turner 1977a, Tómasson & Banda 1996) was also observed in our study (Appendix 

3). The generally accepted statement that demersal cichlids usually have restricted depth 

distributions (Eccles & Trewavas 1989, Banda & Tómasson 1996, Tómasson & Banda 1996, 

Turner 1996) was also supported by our results. Another well-known trend is the decreasing 

occurrence of large cichlid species with depth (Turner 1977a). We observed that even though 

there was a higher number of large species in the shallows (Buccochromis spp., 

Taeniolethrinops spp., Serranochromis robustus…), their occurrence was weak, except for 

the Oreochromis spp., and catches were dominated by small species such as Aulonocara spp., 

Nyassachromis spp. or Copadichromis virginalis and a few larger species such as Lethrinops 

argenteus and Mylochromis anaphyrmus (Tables C1 and C2). On the other hand, the 

dominant species of the deep zone were rather large fish such as Lethrinops gossei, the 

Alticorpus spp. and mentale particularly, the Diplotaxodon spp. (Tables C4 to C6). The 

decreased occurrence of large and medium species in the catches reported by Turner (1977b) 

and Turner et al. (1995) probably also affected the shallow waters of the SWA. However, an 

interesting proportion of large species remains in the almost unexploited deep zone. Given 

that over the year the highest catches were recorded from 50 m downwards, where the 

dominant species are relatively large, any expansion of trawl fisheries in the southern part of 

the Lake should take place in the deep zone shared by the SE and SW arms. This supports the 

position of Banda et al. (1996) against FAO's (1993) recommendation that no expansion of 

the trawl fishery should take place in the deeper zone of the SEA. 

14

Chapter 2: 

Depth distribution and breeding 

patterns of the demersal species most 

commonly caught by trawling in the 

South West Arm of Lake Malawi

Chapter 2: Depth distribution and breeding patterns of the 

demersal species most commonly caught by trawling in the South 

West Arm of Lake Malawi 


Introduction 

Given the tremendous diversity of Malawi cichlids, very few studies have been carried 

out on their breeding biology so far. Earlier studies focused on some species in the north 

(Jackson et al. 1963) and central part of the lake (zooplanctivorous Utaka: Iles 1960, 1971). A 

comprehensive work highlighted the reproductive seasonality of ten rock frequenting species 

(Marsh et al. 1986). A recent survey of the pelagic zone provided information about the 

breeding of Copadichromis quadrimaculatus, Diplotaxodon limnothrissa and 'big eye' and 

Rhamphochromis longiceps (Thompson et al. 1996). Despite the fact they hold an 

economically important commercial fishery, very few species apart from chambo (Lowe 

1953, McKaye & Stauffer 1988, Turner et al. 1991) have been studied in the south of the lake, 

where the commercial fisheries take place. Maturity and fecundity were estimated for 

Copadichromis ("Haplochromis") mloto, Lethrinops parvidens , L. longipinnis, Mylochromis 

("Haplochromis") anaphyrmus and Otopharynx ("Haplochromis") intermedius (Tweddle & 

Turner 1977), while breeding season and maturity were detailed for three Lethrinops species, 

microdon, 'species A' and gossei by Lewis & Tweddle (1990). McKaye (1983) reported 

marked seasonal variations in nest numbers for Cyrtocara eucinostomus. 

Although the information is incomplete for some species, our study describes the breeding 

biology of about 40 of the most important trawled species in the SWA. 

It is important to remember that the sampling period was June 1998 to May 1999. 

However, all the information related to catches are based on the period from July 1998 to May 

1999 for the reasons explained in the previous chapter. Owing to inter species variability of 

occurrence in the catches and therefore to sample size, the results presented in this chapter 

will be of irregular quality, the information being complete and reliable for some species and 

more indicative for the rarer ones. For the reader's convenience, information about the species 

is delivered per genera, which are ordered alphabetically. For each species, whenever 

possible, the following information is displayed: size range (SL), depth distribution, 

occurrence and abundance over the full sampling period, breeding season, age and size at 

maturity, fecundity and egg size. 

For the breeding season, priority will always be given to the females pattern. Most of the time 

in cichlids, males are sexually active for longer periods than females; a way to always be 

'ready' probably. As a consequence, determination of the breeding season is more accurate 

when based upon female data. However, when the sample sizes are not optimum for females, 

information about males may be useful. On the other hand, as most Malawi cichlids form 

breeding leks to attract females (Konings 1995, Turner 1996), priority will be given to ripe 

males distribution for estimation of spawning depth. The weight of individuals was not taken 

on board, but only in the lab on ripe females, except for the nine target species (Alticorpus 

macrocleithrum, Alticorpus mentale, Copadichromis virginalis, Diplotaxodon limnothrissa, 

Diplotaxodon macrops, Lethrinops argenteus, Lethrinops gossei, Mylochromis anaphyrmus 

and Taeniolethrinops praeorbitalis), for which all the females were weighed. Most of the 

15

length-weight relationships are then based on data for ripe females, which explain their low 

sample size sometimes. 


All the fish analysed were collected during the monthly trawl catches in the north of 

the South West Arm (see Chapter 1 for details). 

The comparisons of CPUE per depth for each species with those reported in Tómasson 

& Banda (1996) are based on the values given in Appendix 3, but pooled per depth category 

(shallow zone = 0-50 m, deep zone = 51-100 m, very deep zone = >100 m) and reported to 30 

min pulls (instead of 20 min in our case) to be comparable with Tómasson & Banda (1996) 

values. 

The maturity stage of female gonads was macroscopically determined using the 

slightly modified scale of Legendre & Ecoutin (1989) (Duponchelle 1997). 

Stage 1: immature. The gonad looks like two short transparent cylinders. No oocytes are 

visible to the naked eyes. As a comparison, immature testicle is much longer and thinner, like 

two long tinny silver filaments. 

Stage 2: beginning maturation. The ovaries are slightly larger and little whitish oocytes and 

apparent. 

Stage 3: maturing. The ovaries continue to grow in length and thickness and are full of 

yellowish oocytes in early vitellogenesis. 

Stage 4: final maturation. The ovaries occupy a large part of the abdominal cavity and are full 

of large uniform sized oocytes in late vitellogenesis. 

Stage 5: ripe. Ovulation occurred, oocytes can be expelled by a gentle pressure on the 

abdomen. This stage is ephemera. 

Stage 6: spent. The ovaries look like large bloody empty bags with remaining large sized 

atretic follicles. Small whitish oocytes are visible. 

Stage 6-2: resting. The general aspect of the gonad recall a stage 2, but the ovarian wall is 

thicker, the gonad is larger, often reddish with an aspect of empty bag. This stage is 

distinctive of resting females, which have spawned during the past breeding season. 

Stage 6-3: recovering post-spawning females. The general aspect of the gonad is like a stage 3 

but with empty rooms, remaining large-sized atretic follicles and the blood vessels are still 

well apparent. This stage is characteristic of post-spawning females initiating another cycle of 

vitellogenesis. 

Males were only recorded as being either in "breeding colour" or not. 

For each species, all the stage 4 and 5 females were preserved in 10 % formalin for 

later examination. 

Nine target species were selected according to their relative abundance, depth 

distribution and basic ecological characteristics (benthic or pellagic habits, broad trophic 

category) (Tómasson and Banda 1996, Turner 1996). These were Lethrinops gossei Burgess 

& Axelrod, Lethrinops argenteus Ahl (= L. longipinnis 'orange head'), Diplotaxodon 

limnothrissa Turner, Diplotaxodon macrops Turner & Stauffer, Copadichromis virginalis 

Iles, Mylochromis anaphyrmus Burgess & Axelrod, Alticorpus mentale Stauffer & McKaye, 

Alticorpus macrocleithrum Stauffer & McKaye and Taeniolethrinops praeorbitalis Regan. 

For these species, all the females from each haul were preserved in formalin for later 

examination. 

16

All fish preserved in formalin were measured (SL) to the nearest mm and weighed to 

the nearest 0.1 g. Their maturity stage was determined and the gonads in stage 4 were 

weighed for Gonado-Somatic Index (GSI) calculation (gonad weight/total body weight × 100) 

then preserved in 5% formalin for fecundity and mean oocyte weight calculation. 

The breeding season was determined from the monthly proportions (in %) of the 

different stages of sexual maturation (Legendre & Ecoutin 1989, Duponchelle et al. 1999). In 

order to eliminate the small immature females, which would give a biased weight to the 

immature stages (1 and 2), and to define more precisely the spawning season, only females 

which size was greater than or equal to the size at first sexual maturity were considered in 

analysis. 

The average size at first maturation (L 50 ) is defined as the standard length at which 

50% of the females are at an advanced stage of the first sexual cycle during the breeding 

season. In practice, this is the size at which 50% of the females have reached the stage 3 of the 

maturity scale (Legendre & Ecoutin 1996, Duponchelle & Panfili 1998). For the estimation of 

L 50 , only the fish sampled during the height of the breeding season were considered. 

Age at maturity was calculated from the Von Bertalanffy Growth Curve (VBGC) equation: 

L t = LJ (1-exp (-K (t-t 0 )) (2) 

Where L t is the mean length at age t, LJ is the asymptotic length K the growth coefficient and 

t 0 the size at age 0. This equation can be written: 

t = (-ln (1- (L t / LJ)) / K) + t 0 

Replacing L t by the mean size at maturity (L 50 ), age at maturity (A 50 ) is then: 

A 50 = (-ln (1- (L 50 / LJ)) / K) + t 0 

LJ and K were obtained from length frequency distribution analysis and are provided in the 

Chapter "Growth" for twenty three of the species. The size at age 0 was considered null. 

Fecundity is defined here as the number of oocytes to be released at the next spawn, 

and correspond to the absolute fecundity. It is estimated, from gonads in the final maturation 

stage (stage 4), by the number of oocytes belonging to the largest diameter modal group. This 

oocyte group is clearly separated from the rest of the oocytes to the naked eye and 

corresponds approximately to oocytes that are going to be released (Duponchelle 1997, 

Duponchelle et al. 2000). 

Oocyte weight measurements were all carried out on samples preserved in 5% 

formalin. The average oocyte weight per female, was determined by weighing 50 oocytes 

(Peters 1963) belonging to those considered for fecundity estimates. 

In order to compare mean oocyte weight and diameter among the different species, the 

measurements need to be made on oocytes in a similar vitellogenic stage, then on oocytes 

whose growth is completed. A simplified version of the method applied by Duponchelle 

(1997) was used to determine the GSI threshold above which the oocyte weight do no longer 

increase significantly. For each species, the individual oocyte weights were plotted against the 

GSI. The GSI corresponded to the beginning of the asymptotic part of the curve was visually 

determined and the fish whose GSI was inferior to the defined GSI were removed. The final 

GSI threshold was reached when no correlation subsisted between the mean oocyte weight 

and the GSI. 

17

Plate 1. Alticorpus 'geoffreyi' (by Dave Voorvelt). 

125 

Number 

Weight 

Depth (m) 

100 

75 

50 

0 2 4 6 8 10 


Figure 2-1. Mean occurrence and abundance in the catches per depth of Alticorpus 'geoffreyi' 

in the SWA between July 1998 and May 1999. 

Frequency (%) 

35 

30 

25 

20 

15 

10 

5 

0 

55 65 75 85 95 105 115 125 135 145 155 165 

Standard Length (mm) 

Figure 2-2. Size range and frequencies of Alticorpus 'geoffreyi' caught in the SWA between 

July 1998 and May 1999. 

18

Results 

Alticorpus spp. 

Alticorpus 'geoffreyi' (Plate 1) 

721 females and 845 males were analysed. A. 'geoffreyi' is a deep water species rarely 

encountered at 50m. In our sampling it was most abundant at 75m and still well represented at 

125m. (Figure 2-1). It constituted in average between 2 and 9% of the catches in weight and 

between 1 and 5% in number, depending upon depth, but was more abundant at 75 m. The 

mean CPUE per depth category was 0.3 kg for the shallow zone, 37.4 in the deep zone, and 

9.1 in the very deep zone, which differed markedly with the values reported by Tómasson & 

Banda (1996) for the shallow (9.3 kg) and deep (11.4 kg) zones, but matched in the very deep 

zone (9.0 kg). They also observed A. 'geoffreyi' from 20 m depth downwards in the SWA, 

whereas we never encountered it before 50 m. This might explain the difference in CPUE in 

the shallow zone. Specimens caught ranged between 55 and 165 mm with a mode from 110 to 

150 mm (Figure 2-2). The sex ratio observed over the full sampling period was F/M 0.5/0.5. 

The breeding season for females occurred from March to October with a maximum 

activity between May and August (Figure 2-3a). The proportion of males in breeding colour 

was much higher than the proportion of ripe females, but basically confirmed the position of 

the breeding season (Figure 2-3b). Ripe females were mostly found at 75 and 100 m whereas 

males in breeding colour were much more abundant at 75m, suggesting that breeding could 

occur around 75 m depth (Table 2-1). The size at maturity of female was about 90mm (Figure 

2-4) and was reached at 14 months old. 

Table 2-1. Percentage of ripe females (stages 4 and 5), males in breeding colour and immature 

individuals (whose size is below the size at maturity) per depth for Alticorpus geoffreyi in 

the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 

breeding colour 

Males in breeding 

colour 

Immature 

specimens 

50 m 0.5 0 0 0.5 

75 m 54.1 38.7 40.2 84.4 78.8 

100 m 13.5 42.7 16.7 3.9 15.2 

125 m 31.9 18.6 43.1 11.2 6 

The length-weight and fecundity-weight relationships are given in Figure 2-5 and 2-6, 

respectively. Fecundity ranged from 86 to 231 for females weighing between 33 and 94 g. No 

relation was found between oocyte weight and body weight. The GSI threshold above which 

the oocyte weight did no longer increase significantly was 3.7% (Figure 2-7) and the mean 

oocyte weight was 17.70 mg (± 2.48 SD, N= 26). 

19

50 

% ripe females 

a 

40 

30 

20 

10 

0 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

120 120 58 64 135 24 50 58 44 38 

% males in breeding colors 

b 

100 

80 

60 

40 

20 

0 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

131 111 106 63 115 27 48 40 65 115 

Figure 2-3. Seasonal progression of the percentage of ripe (stages 4 and 5) females (a) and 

males (b) Alticorpus 'geoffreyi' in the SWA The values below the x-axis are the effective 

(number of male or females which size was above the size at maturity) for each month. 

100 

% mature females 

50 

0 

65 75 85 95 105 115 125 135 145 155 

Standard length (mm) 

Figure 2-4. Percentage of mature females (stage 3 and above) per size class (standard length) 

for Alticorpus 'geoffreyi' in the SWA. 

20

120 

100 

Weight (g) 

80 

60 

40 

y = 6E-05x 2,8643 

20 

R 2 = 0,905 

0 

100 110 120 130 140 150 160 

Length (mm) 

Figure 2-5. Length-weight relationship for Alticorpus geoffreyi females in the SWA. (R² = 

determination coefficient). 

250 

200 

Fecundity 

150 

100 

50 

y = 1,7836x + 28,037 

R 2 = 0,5783 

0 

20 30 40 50 60 70 80 90 100 

Weight (g) 

Figure 2-6. Fecundity-weight relationship for Alticorpus geoffreyi females in the SWA. (R² = 


Oocyte weight (mg) 

30 

25 

20 

15 

10 

5 

0 

y = 0,3432x + 16,224 

R 2 = 0,0079 

2,0 3,0 4,0 5,0 6,0 7,0 

GSI (%) 

Figure 2-7. Relationship between oocyte weight and gonado-somatic index (GSI) for 

Alticorpus geoffreyi. Oocytes from females whose GSI was below (in grey) and above (in 

black with regression) 3.7 %. (R² = determination coefficient) 

21

Plate 2. Alticorpus macrocleithrum (by Dave Voorvelt). 

125 

Number 

Weight 

Depth (m) 

100 

75 

0,0 0,5 1,0 1,5 


Figure 3-1. Mean occurrence and abundance in the catches per depth of Alticorpus 

macrocleithrum in the SWA between July 1998 and May 1999. 

25 

20 

Frequency % 

15 

10 

5 

0 

62 67 72 77 82 87 92 97 102 107 112 117 122 127 132 137 142 147 


Figure 3-2. Size range and frequencies of Alticorpus macrocleithrum caught in the SWA 

between July 1998 and May 1999. 

22

Alticorpus macrocleithrum (Stauffer & McKaye) (Plate 2) 

317 females and 153 males were analysed. A. macrocleithrum is a deep water species 

encountered between 75 m and 125 m, but more abundant at 100 m (Figure 3-1). It is not an 

abundant species, always constituting less than 1.5% of the catches both in weight and 

number. The mean CPUE per depth category was 5.1 kg in the deep zone and 0.6 kg in the 

very deep zone, which is about twice the values reported in Tómasson & Banda (1996) for the 

deep zone (2.7 kg), but 8 times less in the very deep zone (5 kg). As for A. 'geoffreyi', the 

depth distribution we found was more restricted than that observed by Tómasson & Banda 

(1996), who found A. macrocleithrum from 40 m downwards. Specimens caught ranged 

between 60 and 148 mm with a mode from 105 to 130 mm (Figure 3-2). The sex ratio 

observed over the full sampling period was F/M 0.7/0.3. 

Owing to weak sample number at some months both for females and males, the 

breeding season is difficult to determine with certainty (Figure 3-3a and b). It seemed from 

female data (Figure 3-3a) that breeding season occurred between April and August. Despite 

high fluctuations from 0 to 100% due to very low sample size at some months (June, October, 

April and May), the data for males supported by correct sample size tended to confirm that 

breeding season. However, the low sample size for females in September–October and in 

March do not allow us to exclude the possibility that breeding season might be a bit 

protracted, beginning a bit earlier and finishing a bit later than observed on those graphs 

(March to September ?). All the breeding females, nearly all (97%) the males in breeding 

colour (Table 3-1) were found are 100 m, suggesting that spawning might occur at that depth. 

Maturity was reached early in their second year at 18 months old at a mean size of 100 mm 

for females (Figure 3-4). However, size at maturity has to be determined at the height of the 

breeding season to be accurate but owing to the low sample size we had to consider every data 

available. As a consequence, it is likely that L 50 was overestimated. 


individuals (whose size is below the size at maturity) per depth for Alticorpus 

macrocleithrum in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

75 m 29.7 

100 m 64.4 100 83.3 96.6 88 

125 m 5.9 16.7 3.4 12 






23

60 

50 


40 

30 

20 

10 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

a 

43 100 85 8 13 24 14 5 6 8 11 

100 

80 

60 

40 

20 

- 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

% males in breeding colour 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

b 

1 18 27 2 25 40 12 10 11 1 1 


males (b) Alticorpus macrocleithrum in the SWA The values below the x-axis are the 

effective (number of male or females which size was above the size at maturity) for each 

month. 

100 


50 

0 

65 75 85 95 105 115 125 135 



for Alticorpus macrocleithrum in the SWA. 

24

Weight (g) 

80 

70 

60 

50 

40 

30 

20 

y = 9E-05x 2,7566 

10 

R 2 = 0,8485 

0 

50 70 90 110 130 150 

Length (mm) 

Figure 3-5. Length-weight relationship for Alticorpus macrocleithrum females in the SWA. 

(R² = determination coefficient). 

Fecundity 

350 

300 

250 

200 

150 

100 

50 

y = 3,3229x + 14,744 

R 2 = 0,5662 

0 

20 30 40 50 60 70 80 

Weight (g) 

Figure 3-6. Fecundity-weight relationship for Alticorpus macrocleithrum females in the SWA. 



20 

18 

16 

14 

12 

10 

8 

6 

4 

y = 0,5497x + 9,9522 

2 

R 2 = 0,0292 

0 

2,0 3,0 4,0 5,0 6,0 7,0 8,0 

GSI (%) 


Alticorpus macrocleithrum. Oocytes from females whose GSI was below (in grey) and 

above (in black with regression) 4.5%. (R² = determination coefficient) 

25

Plate 3. Alticorpus mentale (by Dave Voorvelt). 

125 

Depth (m) 

100 

75 

50 

30 

Number 

Weight 

0 5 10 15 


Figure 4-1. Mean occurrence and abundance in the catches per depth of Alticorpus mentale in 

the SWA between July 1998 and May 1999. 

14 

12 

Frequency % 

10 

8 

6 

4 

2 

0 

55 65 75 85 95 105 115 125 135 145 155 165 175 185 195 205 215 225 235 245 


Figure 4-2. Size range and frequencies of Alticorpus mentale caught in the SWA between July 

1998 and May 1999. 

26

Alticorpus mentale (Stauffer & McKaye) (Plate 3) 

1011 females and 959 males were analysed. A. mentale is a deep water species that can 

be encountered from 30 m but becomes abundant from 75 m downwards (Figure 4-1). It is an 

abundant fish always constituting between 7 and 13% of the catches in weight and a bit less in 

number (about 3%) due to its large size. The mean CPUE per depth category was 1.1 kg in the 

shallow zone, 51.1 kg in the deep zone and 20.1 kg in the very deep zone, which is about 

twice as much as reported in Tómasson & Banda (1996) for the deep and very deep zones 

(11.4 and 12.7 kg, respectively), but 2 times less in the shallow zone (2.3 kg). The depth 

distribution we found corresponded with that observed by Tómasson & Banda (1996). 

Specimens caught ranged between 53 and 245 mm (Figure 4-2). The sex ratio observed over 

the full sampling period was F/M 0.5/0.5. 

Unlike A. 'geoffreyi' and A. macrocleithrum, A. mentale breeds throughout the year 

with a peak in November and another one in January-February and a marked decrease of 

activity in December and May (Figure 4-3a). Males in breeding colour were also found at 

each month even thought peaks of activity did not necessarily match those of females (Figure 

4-3b). It is interesting to note that the period from November to March is the period when 

Aulonocara 'minutus', frequently found as the dominant prey items in A. mentale stomach 

contents (see Chapter "Diet"), was most abundant in the catches. Ripe females were more 

abundant at 75 and 100 m whereas males in breeding colour were preferentially found at 100 

m and 125 m in a lesser extent (Table 4-1). Immature individuals were mostly found at 75 m 

but were reasonably abundant at 100 and 125 m. A. mentale seems able to spawn at any depth 

between 75 and 125 m. Maturity was reached early in their second year at 16 months old at a 

mean size of 160 mm for females (Figure 4-4). 


individuals (whose size is below the size at maturity) per depth for Alticorpus mentale in 

the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

30 m 0.2 0 0 0 0 

50 m 1 0 2.6 0.5 4.7 

75 m 39 41.8 4.3 9.4 54.3 

100 m 26.9 44.5 25 68.5 19.6 

125 m 32.8 13.6 29.1 21.6 21.5 


respectively. Fecundity ranged from 92 to 356 for females weighing between 49 and 307 g. 

No relation was found between oocyte weight and body weight. The GSI threshold above 

which the oocyte weight did no longer increase significantly was 3% (Figure 4-7) and the 

mean oocyte weight was 24.36 mg (± 4.18 SD, N= 31). 

27

50 


40 

30 

20 

10 

0 

a 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

30 48 71 66 66 57 52 83 47 34 35 


b 

- 

70 

60 

50 

40 

30 

20 

10 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 


males (b) Alticorpus mentale in the SWA The values below the x-axis are the effective 


Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

56 91 83 70 114 120 75 104 79 108 48 

100 

% of mature females 

50 

0 

55 65 75 85 95 105 115 125 135 145 155 165 175 185 195 205 215 225 235 



for Alticorpus mentale in the SWA. 

28

500 

Weight (g) 

400 

300 

200 

100 

0 

y = 2E-05x 3,1181 

R 2 = 0,9794 

0 50 100 150 200 250 300 

Length (mm) 

Figure 4-5. Length-weight relationship for Alticorpus mentale females in the SWA. (R² = 


Fecundity 

400 

350 

300 

250 

200 

150 

100 

50 

0 

y = 0,9359x + 63,115 

R 2 = 0,6394 

0 50 100 150 200 250 300 350 

Weight (g) 

Figure 4-6. Fecundity-weight relationship for Alticorpus mentale females in the SWA. (R² = 


35 

30 

Occyte weight (mg) 

25 

20 

15 

10 

y = 1,3294x + 19,431 

5 

R 2 = 0,0271 

0 

1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 

GSI (%) 


Alticorpus mentale. Oocytes from females whose GSI was below (in grey) and above (in 

black with regression) 3%. (R² = determination coefficient) 

29

125 

Depth (m) 

100 

75 

Number 

Weight 

0 1 2 3 4 5 


Figure 5-1. Mean occurrence and abundance in the catches per depth of Alticorpus pectinatum 


Frequency % 

16 

14 

12 

10 

8 

6 

4 

2 

0 

57 62 67 72 77 82 87 92 97 102 107 112 117 122 127 132 137 


Figure 5-2. Size range and frequencies of Alticorpus pectinatum caught in the SWA between 


30

Alticorpus pectinatum (Stauffer & McKaye) 

259 females and 375 males were analysed. A. pectinatum is a deep water fish evenly 

distributed between 75 m and 125 m (Figure 5-1). It is not an abundant species, always 

constituting between 1 and 2% of the catches both in weight and number. The mean CPUE 

per depth category was 8.4 kg in the deep zone and 2.6 kg in the very deep zone, which is 

about twice as much as the values reported in Tómasson & Banda (1996) for the deep zone 

(3.8 kg), but about 4 times less in the very deep zone (8.9 kg). As for A. 'geoffreyi' and A. 

macrocleithrum, the depth distribution we found was more restricted than that observed by 

Tómasson & Banda (1996), who found A. pectinatum from 45 m downwards. Specimens 

caught ranged between 55 and 140 mm with a mode from 100 to 125 mm (Figure 5-2). The 

sex ratio observed over the full sampling period was F/M 0.4/0.6. 

The breeding season occurred from November to May with an increased activity in 

January-February (Figure 5-3a). The monthly progression of males in breeding colour 

indicated the same pattern and suggested that breeding season could be protracted, occurring 

also in June, when all the males (43) were in breeding colour (Figure 5-3b). About three 

quarter of the breeding females were found at 100 m and the other quarter at 75 m, whereas 

males in breeding colour and immature individuals were more evenly distributed at 75 and 

100 m (Table 5-1). Spawning might occur mostly at 100 m although males and immature 

distribution does not confirm it firmly. Maturity was reached early in their second year at 12 

months old at a mean size of 70 mm for females (Figure 5-4). 


individuals (whose size is below the size at maturity) per depth for Alticorpus pectinatum 

in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

75 m 37.6 24.1 53.3 47 42.3 

100 m 40.2 72.4 19.2 30.5 50 

125 m 22.3 3.4 27.5 22.5 7.7 




the oocyte weight did no longer increase significantly was 4% (Figure 5-7) and the mean 


31

a 


b 


100 

80 

60 

40 

20 

0 

30 

25 

20 

15 

10 

5 

0 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

12 5 5 51 29 16 49 34 28 29 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

43 11 4 12 88 63 13 57 41 20 21 


males (b) Alticorpus pectinatum in the SWA The values below the x-axis are the effective 


100 


50 

0 

45 55 65 75 85 95 105 115 125 135 145 



for Alticorpus pectinatum in the SWA. 

32

60 

50 

Weight (g) 

40 

30 

20 

y = 3E-05x 2,9656 

10 

R 2 = 0,9656 

0 

50 70 90 110 130 

Length (mm) 

Figure 5-5. Length-weight relationship for Alticorpus pectinatum females in the SWA. (R² = 


200 

150 

Fecundity 

100 

50 

0 

y = 3,2395x + 12,796 

R 2 = 0,6576 

0 10 20 30 40 50 

Weight (g) 

Figure 5-6. Fecundity-weight relationship for Alticorpus pectinatum females in the SWA. (R² 

= determination coefficient). 


20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Alticorpus pectinatum. Oocytes from females whose GSI was below (in grey) and above 

(in black with regression) 4%. (R² = determination coefficient) 

33

Plate 4. Aulonocara 'blue orange' (by Dave Voorvelt). 

50 

Number 

Weight 

Depth (m) 

30 

10 

0 5 10 15 


Figure 6-1. Mean occurrence and abundance in the catches per depth of Aulonocara 'blue 

orange' in the SWA between July 1998 and May 1999. 

Frequency % 

35 

30 

25 

20 

15 

10 

5 

0 

42 47 52 57 62 67 72 77 


Figure 6-2. Size range and frequencies of Aulonocara 'blue orange' caught in the SWA 


34

Aulonocara spp. 

Aulonocara 'blue orange' (Plate 4) 

244 females and 485 males were analysed. A. 'blue orange' is a small fish encountered 

in shallow water between 10 and 50 m but only abundant at 10 and 30 m, where it constituted 

9.5 and 12.6% of the catches in number, respectively and 3.2 and 4.5% in weight (Figure 6-1). 

It was found only once at 75 m in November 98. The mean CPUE per depth category was 

higher (17.6 kg in the shallow zone) than that reported in Tómasson & Banda (1996). The 

depth distribution was much more restricted in our case than the 10 to 130 m depth range they 

reported. Specimens caught ranged between 40 and 80 mm with a mode from 50 to 65 mm 

(Figure 6-2). The sex ratio observed over the full sampling period was F/M 0.3/0.7. 

Owing to identification uncertainty, data from June and July 1998 were not included in 

the analyses. The breeding season seemed to occur between July-August and February (Figure 

6-3a and b). However, due to the very low sample size in March and April and the high 

percentage of males in breeding colour in April and May, it can't be excluded that A. 'blue 

orange' might breed throughout the year. More than three quarter of the ripe females were 

sampled at 10 m and the other quarter at 30 m (Table 6-1). Males in breeding colour were 

evenly distributed between 10 and 30 m and most of the immature individuals were found at 

30 m. If A. 'blue orange' does spawn at a precise depth, it does not reflect clearly in breeding 

males and immature distribution. However, the results suggest that spawning would occur 

around 10 m. Maturity was reached in their first year at 9 months old at a mean size of 48 mm 

for females (Figure 6-4). 


individuals (whose size is below the size at maturity) per depth for Aulonocara 'blue 

orange' in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

10 m 47.3 76 60.9 45.5 34.2 

30 m 47.9 24 36.6 48.5 61.8 

50 m 4.8 0 2.6 6.1 3.9 


respectively. Fecundity ranged from 9 to 41 for females weighing between 2 and 7g. No 




35


90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

a 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

14 14 50 40 39 35 2 1 30 


b 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Aug-98 

Sep-98 

Oct-98 

Nov-98 


males (b) Aulonocara 'blue orange' in the SWA The values below the x-axis are the 


month. 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

6 7 95 110 93 78 8 2 86 

May-99 

100 


50 

0 

37 42 47 52 57 62 67 



for Aulonocara 'blue orange' in the SWA. 

36

Weight (g) 

8 

7 

6 

5 

4 

3 

2 

1 

0 

y = 3E-05x 2,9596 

R 2 = 0,8159 

40 45 50 55 60 65 70 

Length (mm) 

Figure 6-5. Length-weight relationship for Aulonocara 'blue orange' females in the SWA. (R² 


Fecundity 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

y = 3,3845x + 9,6689 

R 2 = 0,2613 

1,5 2,5 3,5 4,5 5,5 6,5 7,5 

Weight (g) 

Figure 6-6. Fecundity-weight relationship for Aulonocara 'blue orange' females in the SWA. 


7 

6 


5 

4 

3 

2 

1 

y = 0,2444x + 3,6419 

R 2 = 0,0837 

0 

1,0 2,0 3,0 4,0 5,0 6,0 7,0 

GSI (%) 


Aulonocara 'blue orange'. Oocytes from females whose GSI was below (in grey) and 

above (in black with regression) 2.5%. (R² = determination coefficient). 

37

50 

Depth (m) 

30 

10 

Number 

Weight 

0,0 0,5 1,0 1,5 


Figure 7-1. Mean occurrence and abundance in the catches per depth of Aulonocara 'cf. 

macrochir' in the SWA between July 1998 and May 1999. 

25 

20 

Frequency % 

15 

10 

5 

0 

52 57 62 67 72 77 82 87 92 97 102 107 112 117 122 127 132 

Size class (mm) 

Figure 7-2. Size range and frequencies of Aulonocara 'cf. macrochir' caught in the SWA 


38

Aulonocara 'cf. macrochir' 

74 females and 90 males were analysed. A. 'cf. macrochir' was a relatively rare fish 

encountered from 10 m to 50 m but more abundant at 50 m where it constituted 1% and 0.8% 

of the catches in weight and number, respectively (Figure 7-1). The mean CPUE per depth 

category (2.7 kg in the shallow zone) matched that reported in Tómasson & Banda (1996). 

The depth distribution was much more restricted in our case than the 8 to 150 m depth range 

that they reported. Specimens caught ranged between 50 and 135 mm with a mode from 95 to 


As A. 'cf. macrochir' is a relatively rare species, the following information about life 

history traits is based on low sample size and can not be considered as very reliable but rather 

as indicative. A breeding activity was observed from August to October and from December 

to March (Figure 7-3a and b). Taking into account the sparse information available about both 

females and males, it can be hypothesised that this species might breed most of the year, with 

reduced activity in June and July. As most specimens were caught at 50 m the percentages of 

ripe females (62.5%), males in breeding colour (100%) and immature individuals (87%) at 

this depth suggested that spawning probably occurs at 50 m. Maturity was reached at about 

100 mm for females (Figure 7-4). Again, owing to low sample size, L 50 was probably 

overestimated and is likely to be closer to 90 mm. 



relationship was found between oocyte weight and body weight. The GSI threshold above 

which the oocyte weight did no longer increase significantly was impossible to assess due to 

the low sample size. However, as no relationship was found between oocyte weight and GSI 

from the data available (Figure 7-7), the mean oocyte weight was estimated from all the 

available data and was 4.82 mg (± 1.13 SD, N= 8). 

39

80 


60 

40 

20 

0 

a 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

4 16 10 13 13 8 3 6 1 


b 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

1 5 14 21 6 18 20 3 


males (b) Aulonocara 'cf. macrochir' in the SWA The values below the x-axis are the 


month. 

100 


50 

0 

55 65 75 85 95 105 115 125 135 



for Aulonocara 'cf. macrochir' in the SWA. 

40

Weight (g) 

70 

60 

50 

40 

30 

20 

10 

y = 2E-05x 3,054 

R 2 = 0,9716 

0 

70 80 90 100 110 120 130 

Length (mm) 

Figure 7-5. Length-weight relationship for Aulonocara 'cf. macrochir' females in the SWA. 


Fecundity 

160 

140 

120 

100 

80 

60 

40 

20 

0 

y = 1,6714x + 18,2 

R 2 = 0,9103 

0 20 40 60 80 

Weight (g) 

Figure 7-6. Fecundity-weight relationship for Aulonocara 'cf. macrochir' females in the SWA. 



7 

6 

5 

4 

3 

2 

1 

y = -0,1133x + 4,9968 

R 2 = 0,0029 

0 

0,0 0,5 1,0 1,5 2,0 2,5 3,0 

GSI (%) 


Aulonocara 'cf. macrochir'. (R² = determination coefficient). 

41

Plate 5. Aulonocara 'minutus' (by Dave Voorvelt). 

125 

Depth (m) 

100 

75 

Number 

Weight 

0 5 10 15 


Figure 8-1. Mean occurrence and abundance in the catches per depth of Aulonocara 'minutus' 


30 

25 

Frequency % 

20 

15 

10 

5 

0 

27 32 37 42 47 52 57 62 67 72 


Figure 8-2. Size range and frequencies of Aulonocara 'minutus' caught in the SWA between 


42

Aulonocara 'minutus' (Plate 5) 

573 females and 781 males were analysed. A. 'minutus' is a deep water fish caught 

from 75 to 125 m depth (Figure 8-1). It was much more abundant at 125 m, where it 

constituted up to 14% of the catches in number and only 1.7% in weight owing to its very 

small size. A. 'minutus' is one of the smallest demersal species of Lake Malawi. Specimens 

collected ranged between 25 and 75 mm with a mode from 40 to 60 mm (Figure 8-2). The sex 

ratio observed over the full sampling period was F/M 0.4/0.6. The mean CPUE per depth 

category was 3.5 kg in the deep zone and 2.6 kg in the very deep zone, which is about two 

times less than the values reported in Tómasson & Banda (1996) for the deep zone (7.3 kg), 

and about 4 times less in the very deep zone (11.3 kg). As for A. 'cf. macrochir' and A. 'blue 

orange', the depth distribution we found was much more restricted than that observed by 

Tómasson & Banda (1996), who found A. 'minutus' from 10 to 130 m. 

Unlike A. 'cf. macrochir' and A. 'blue orange', A. 'minutus' was found to breed 

throughout the year with an increased activity in June and February (Figure 8-3a and b). 

About 50% of non breeding females and males as well as immature individuals were found at 

125 m, the other half being evenly distributed between 75 and 100m (Table 8-1). On the other 

hand, about 50% of the ripe females and males in breeding colour were caught at 100 m, 

whereas the other half were evenly distributed at 75 and 125. This suggested that spawning 

might occur at 100 m. Maturity was reached in their first year at 7 months at a mean size of 42 

mm for females (Figure 8-4). 


individuals (whose size is below the size at maturity) per depth for Aulonocara 'minutus' in 

the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

75 m 24.2 18.7 15.7 20.9 18.1 

100 m 27.7 53.8 28.6 48.1 32.2 

125 m 48.1 27.5 55.7 31 49.7 


respectively. Fecundity ranged from 50 to 134 for females weighing between 16 and 60 g and 

was not significantly correlated to female body weight (r = 0.26). No relation was found 

between oocyte weight and body weight. The GSI threshold above which the oocyte weight 

did no longer increase significantly was 2% (Figure 8-7) and the mean oocyte weight was 

3.82 mg (± 1.13 SD, N= 29). 

43


80 

70 

60 

50 

40 

30 

20 

10 

0 

a 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

32 10 27 29 44 14 36 81 122 43 58 


100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

b 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

43 60 27 57 48 20 62 120 171 67 106 


males (b) Aulonocara 'minutus' in the SWA The values below the x-axis are the effective 


100 


50 

0 

22 27 32 37 42 47 52 57 62 67 



for Aulonocara 'minutus' in the SWA. 

44

Weight (g) 

7 

6 y = 2E-05x 3,0763 

5 R 2 = 0,8435 

4 

3 

2 

1 

0 

20 30 40 50 60 70 

Length (mm) 

Figure 8-5. Length-weight relationship for Aulonocara 'minutus' females in the SWA. (R² = 


30 

Fecundity 

25 

20 

15 

10 

5 

0 

y = 2,3366x + 10,738 

R 2 = 0,0679 

1 2 3 4 

Weight (g) 

Figure 8-6. Fecundity-weight relationship for Aulonocara 'minutus' females in the SWA. (R² 



8 

7 

6 

5 

4 

3 

2 

1 

0 

y = 0,3653x + 2,6722 

R 2 = 0,1232 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Aulonocara 'minutus'. Oocytes from females whose GSI was below (in grey) and above (in 

black with regression) 2%. (R² = determination coefficient). 

45

Depth (m) 

125 

100 

Number 

Weight 

75 

0,0 0,5 1,0 1,5 


Figure 9-1. Mean occurrence and abundance in the catches per depth of Aulonocara 

'rostratum deep' in the SWA between July 1998 and May 1999. 

Frequency % 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

62 67 72 77 82 87 92 97 102 107 112 117 122 127 132 137 


Figure 9-2. Size range and frequencies of Aulonocara 'rostratum deep' caught in the SWA 


46

Aulonocara 'rostratum deep' 

56 females and 57 males were analysed. A. 'rostratum deep' is a relatively rare fish 

found from 75 to 125 m depth but more frequent at 125 m (Figure 9-1). The mean CPUE per 

depth category was 1.7 kg in the deep zone and 0.5 kg in the very deep zone, which is about 

ten times less than the values reported in Tómasson & Banda (1996) for the deep zone (5.1 

kg), and more than twenty times less in the very deep zone (13.2 kg). As for A. 'cf. macrochir' 

and A. 'blue orange' and A. 'minutus', the depth distribution we found was much more 

restricted than that observed by Tómasson & Banda (1996), who found A. 'rostratum deep' 

from 30 to 130 m. Specimens caught ranged between 60 and 140 mm with a mode from 80 to 


As A. 'rostratum deep' is a rare species, the following information about life history 

traits is based on low sample size and can not be considered as very reliable but rather as 

indicative. A breeding activity was observed in October, March and May (Figure 9-3a and b). 

From the few data available, ripe females and males in breeding colour were evenly 

distributed among depths. Maturity was reached at about 75 mm (Figure 9-4). 


respectively. Fecundity ranged from 41 to 167 for females weighing between 14 and 69 g. 

The mean oocyte weight was impossible to assess from the few data available on this species. 

47

50 


40 

30 

20 

10 

0 

a 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

17 2 4 20 4 9 


b 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

1 24 3 11 9 7 2 


males (b) Aulonocara 'rostratum deep' in the SWA The values below the x-axis are the 


month. 

100 


50 

0 

62 67 72 77 82 87 92 97 102 107 112 117 122 127 



for Aulonocara 'rostratum deep' in the SWA. 

48

Weight (g) 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 1E-05x 3,2358 

R 2 = 0,9925 

50 70 90 110 130 150 

Length (mm) 

Figure 9-5. Length-weight relationship for Aulonocara 'rostratum deep' females in the SWA. 


Fecundity 

180 

160 

140 

120 

100 

80 

60 

40 

20 

0 

y = 2,3872x + 1,3981 

R 2 = 0,9789 

0 20 40 60 80 

Weight (g) 

Figure 9-6. Fecundity-weight relationship for Aulonocara 'rostratum deep' females in the 

SWA. (R² = determination coefficient). 

49

Plate 6. Buccochromis lepturus (by Dave Voorvelt). 

30 

Number 

Weight 

Dpth (m) 

10 

0 2 4 6 


Figure 10-1. Mean occurrence and abundance in the catches per depth of Buccochromis 

lepturus in the SWA between July 1998 and May 1999. 

Frequency % 

10 

9 

8 

7 

6 

5 

4 

3 

2 

1 

0 

65 75 85 95 105 115 125 135 145 155 165 175 185 195 205 215 225 235 245 255 265 275 285 295 305 315 325 


Figure 10-2. Size range and frequencies of Buccochromis lepturus caught in the SWA 


50

Buccochromis spp. 

Buccochromis lepturus (Regan) (Plate 6) 

74 females and 58 males were analysed. B. lepturus is a large shallow water fish 

essentially encountered at 10 m, where it constituted 4.5% of the catches in weight but only 

0.5% in number owing to its large size (Figure 10-1). It was also found sometimes at 30 m. 

The mean CPUE per depth category was 6.5 kg in the shallow zone, which approximately 

matched the values reported in Tómasson & Banda (1996) (7.4 kg). The depth distribution 

observed in our study was more restricted than that of Tómasson & Banda (1996), who found 

B. lepturus down to 50m. Specimens caught ranged between 60 and 330 mm (Figure 10-2). 

The sex ratio observed over the full sampling period was F/M 0.6/0.4. 

B. lepturus being an abundant species, the number of specimen caught at each 

sampling session was very low, which severely hampered the precise determination of 

breeding season and other life history traits. From the few data available, it seems that 

breeding season might occur between March-April and August (Figure 10-3a and b). The few 

ripe females or males in breeding colour were mostly found at 10 m, suggesting a rather 

shallow spawning. Maturity was reached at about 160 mm for females (Figure 10-4). 

The length-weight and fecundity-weight relationships are given in Figure 10-5 and 10- 

6, respectively. Fecundity ranged from 267 to 627 for females weighing between 294 and 588 

g. No relation was found between oocyte weight and body weight. The GSI threshold above 

which the oocyte weight did no longer increase significantly was difficult to assess due to low 

sample size but was tentatively fixed at about 2% (Figure 10-7). The mean oocyte weight was 

19.99 mg (± 1.88 SD, N= 4). 

51

60 


a 


b 

50 

40 

30 

20 

10 

0 

100 

80 

60 

40 

20 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

5 7 8 15 7 1 4 5 2 20 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

5 8 2 10 2 2 8 21 


males (b) Buccochromis lepturus in the SWA The values below the x-axis are the effective 


100 


50 

0 

75 85 95 105 115 125 135 145 155 165 175 185 195 205 215 225 235 



for Buccochromis lepturus in the SWA. 

52

Weight (g) 

700 

600 

500 

400 

300 

200 

y = 3E-06x 3,3761 

100 

R 2 = 0,988 

0 

150 200 250 300 

Length (mm) 

Figure 10-5. Length-weight relationship for Buccochromis lepturus females in the SWA. (R² 


Fecundity 

700 

600 

500 

400 

300 

200 

100 

y = 0,9143x + 22,481 

R 2 = 0,7403 

0 

250 300 350 400 450 500 550 600 650 

Weight (g) 

Figure 10-6. Fecundity-weight relationship for Buccochromis lepturus females in the SWA. 


25 


20 

15 

10 

5 

y = 1,4963x + 16 

R 2 = 0,108 

0 

0,5 1,0 1,5 2,0 2,5 3,0 3,5 

GSI (%) 


Buccochromis lepturus. Oocytes from females whose GSI was below (in grey) and above 

(in black with regression) 2%. (R² = determination coefficient). 

53

50 

Depth (m) 

30 

10 

Number 

Weight 

0 1 2 3 4 


Figure 11-1. Mean occurrence and abundance in the catches per depth of Buccochromis 

nototaenia in the SWA between July 1998 and May 1999. 

Frequency % 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

55 65 75 85 95 105 115 125 135 145 155 165 175 185 195 205 215 225 235 245 255 265 275 285 295 

Size class (mm) 

Figure 11-2. Size range and frequencies of Buccochromis nototaenia caught in the SWA 


54

Buccochromis nototaenia (Boulenger) 

122 females and 182 males were analysed. B. nototaenia is a large shallow water fish 

found between 10 and 50 m (Figure 11-1). It was more frequently encountered at 30 m, where 

it constituted about 2% of the catches in weight and 0.5% in number. The mean CPUE per 

depth category was 5.1 kg in the shallow zone, which was a bit more than the value reported 

in Tómasson & Banda (1996) (3.9 kg). The depth distribution observed in our study matched 

that of Tómasson & Banda (1996). Specimens caught ranged between 50 and 300 mm (Figure 

11-2). The sex ratio observed over the full sampling period was F/M 0.4/0.6. 

As for B. lepturus, the low sample size hampered the precise determination of 

breeding season and other life history traits. All that can be said from females (Figure 11-3a) 

and males (Figure 11-3b) data is that we observed a breeding activity in April, July and 

August. However, it seems a weird behaviour for a Malawi cichlid to start breeding for one 

month then stop for two months and beginning again for two others months. Although there 

was no data to support it, it can be hypothesised that breeding season might occur between 

April and August as for B. lepturus. About 70% of the ripe females and males in breeding 

colour and 84% of the immature individuals were sampled at 30 m, suggesting that spawning 

could occur at this depth (Table 11-1). The plot of the percentage of ripe females against 

standard length did not give a proper sigmoïd curve (Figure 11-4). Size at maturity has to be 

determined at the height of the breeding season to be accurate but owing to the low sample 

size we had to consider all data available. As a consequence, females caught outside the 

breeding season were included in the analyses, even large resting females, which may explain 

the shape of the upper part of the curve. Nevertheless, small sized females are more 

informative than large ones and the size range between 100 mm and 130 mm was supported 

by the higher sample size (55 females), giving a relative "power" to this part of the curve. 

From the data available it can be estimated that maturity was reached around 115 mm for 

females. Maturity appears at a smaller size than anticipated given the large maximum 

observed length for this species and compared to B. lepturus. 

Table 11-1. Percentage of ripe females (stages 4 and 5), males in breeding colour and 

immature individuals (whose size is below the size at maturity) per depth for 

Buccochromis nototaenia in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

10 m 20.9 28.6 17.6 16.7 14.6 

30 m 79.1 71.4 80 66.7 84.4 

50 m 2.4 16.7 1 


6, respectively. Fecundity ranged from 100 to 315 for females weighing between 40 and 250 


which the oocyte weight did no longer increase significantly was difficult to assess due to low 

sample size but was tentatively fixed at about 2% (Figure 11-7). The mean oocyte weight was 

11.70 mg (± 3.63 SD, N= 3). 

55


20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

a 


b 

35 

30 

25 

20 

15 

10 

5 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

10 22 12 13 6 2 16 8 8 8 17 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

8 16 5 10 7 3 15 1 6 3 17 


males (b) Buccochromis nototaenia in the SWA The values below the x-axis are the 


month. 

100 


50 

0 

55 65 75 85 95 105 115 125 135 145 155 165 175 185 195 205 215 225 



for Buccochromis nototaenia in the SWA. 

56

Weight (g) 

300 

250 

200 

150 

100 

y = 5E-06x 3,3608 

50 

R 2 = 0,98 

0 

100 120 140 160 180 200 

Length (mm) 

Figure 11-5. Length-weight relationship for Buccochromis nototaenia females in the SWA. 


Fecundity 

350 

300 

250 

200 

150 

100 

50 

0 

y = 0,8115x + 96,468 

R 2 = 0,8738 

0 50 100 150 200 250 300 

Weight (g) 

Figure 11-6. Fecundity-weight relationship for Buccochromis nototaenia females in the SWA. 


20 


15 

10 

5 

0 

1,0 1,5 2,0 2,5 3,0 

GSI (%) 


Buccochromis nototaenia. Oocytes from females whose GSI was below (in grey) and 

above (in black with regression) 2%. (R² = determination coefficient). 

57


0,0 0,5 1,0 1,5 2,0 2,5 

10 

Depth (m) 

30 

50 

75 

Weight 

Number 

Figure 12-1. Mean occurrence and abundance in the catches per depth of Copadichromis 

quadrimaculatus in the SWA between July 1998 and May 1999. 

Frequency % 

35 

30 

25 

20 

15 

10 

5 

0 

55 65 75 85 95 105 115 125 135 145 


Figure 12-2. Size range and frequencies of Copadichromis quadrimaculatus caught in the 

SWA between July 1998 and May 1999. 

58

Copadichromis spp. 

Copadichromis quadrimaculatus (Regan) 

132 females and 214 males were analysed. C. quadrimaculatus was caught from 10 to 

75 m but occurred more frequently at 30 m, where it constituted 2% and 1.5% of the catches, 

in weight and number, respectively (Figure 12-1). The mean CPUE per depth category was 

6.9 kg in the shallow zone and 0.15 kg in the deep zone, which was about three times less 

than the value reported in Tómasson & Banda (1996) for the shallow zone (3.9 kg). The depth 

distribution observed in our study was much more restricted than that of Tómasson & Banda 

(1996), who reported C. quadrimaculatus from 8 to 135 m. Specimens caught ranged between 

50 and 150 mm (Figure 12-2). The sex ratio observed over the full sampling period was F/M 

0.4/0.6. 

The low sample size hampered the precise determination of breeding season and other 

life history traits. From data available about females (Figure 13-3a) and males (Figure 12-3b), 

it can be estimated that breeding season occurs from April to October. Taking into account the 

very low sample size of females in February-March and data from males, breeding season 

might actually start in February-March. This corresponds quite well with the % of active 

females observed in open waters during the SADC/ODA Project (Thompson et al. 1995, 

1996). Most ripe females, males in breeding colour and immature individuals were found at 

30 m, suggesting that spawning could occur at this depth (Table 12-1). The percentage of 

mature females (stage 3 and above) per size class is presented in Figure 12-4. No female was 

caught in the size range where maturity occurred. The 50% of mature females observed in the 

size range 80-90 mm was based on 2 females only and is probably overestimated. However, 

the second part of the curve from 120 mm upwards was based on consistent sample size and it 

can be reasonably estimated that maturity was reached around 100 mm, which was much less 

than the 15 cm TL (about 120 mm SL) reported for the same species in the open waters 

(Thompson et al. 1995, 1996). This corresponded to a mean age at maturity of 20 months. 



Buccochromis nototaenia in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

10 m 29.3 20 2 37.6 1.4 

30 52.4 50 75.5 57 76.8 

50 m 14.6 25 22.4 5.5 21.7 

75 m 3.7 5 


6, respectively. Fecundity ranged from 15 to 62 for females weighing between 53 and 75 g. 

Using the length-weight and fecundity weight relationships, and assuming a 2 to 3 cm 

difference between standard and total length for a size range of 17 to 20 cm TL, the mean 

fecundity (50 eggs for females between 17 and 20cm TL) found by Thompson et al. (1996) 

corresponded with the fecundity we got for females of similar size. No relation was found 

between oocyte weight and body weight. The GSI threshold above which the oocyte weight 

59

60 


a 

50 

40 

30 

20 

10 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

1 52 20 5 5 1 1 13 4 


b 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

38 68 63 12 1 15 2 7 7 1 


males (b) Copadichromis quadrimaculatus in the SWA The values below the x-axis are the 


month. 

100 


50 

0 

65 75 85 95 105 115 125 135 145 



for Copadichromis quadrimaculatus in the SWA. 

60

Weight (g) 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 0,001x 2,2537 

R 2 = 0,6119 

110 115 120 125 130 135 140 145 150 

Length (mm) 

Figure 12-5. Length-weight relationship for Copadichromis quadrimaculatus females in the 


Fecundity 

70 

60 

50 

40 

30 

20 

10 

y = 0,684x - 0,8577 

R 2 = 0,2695 

0 

40 45 50 55 60 65 70 75 80 

Weight (g) 

Figure 12-6. Fecundity-weight relationship for Copadichromis quadrimaculatus females in 

the SWA. (R² = determination coefficient). 

60 


50 

40 

30 

20 

10 

y = 2,3888x + 33,841 

R 2 = 0,0048 

0 

1,0 1,5 2,0 2,5 3,0 3,5 4,0 

GSI (%) 


Copadichromis quadrimaculatus. Oocytes from females whose GSI was below (in grey) 

and above (in black with regression) 3%. (R² = determination coefficient). 

61

Plate 7. Copadichromis virginalis (by Dave Voorvelt). 


a 


b 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

- 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

129 266 198 57 696 292 417 340 3 201 298 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

18 263 86 30 633 330 231 158 246 114 828 


males (b) Copadichromis virginalis in the SWA The values below the x-axis are the 


month. 

62

did no longer increase significantly was 3% (Figure 12-7). The mean oocyte weight was 42.05 

mg (± 6.96 SD, N= 8). 

Copadichromis virginalis (Iles) (Plate 7) 

2886 females and 3052 males were analysed. C. virginalis was encountered from 10 m 

to 50 m and rarely at 100 m (Figure 13-1). 


0 10 20 30 40 

Depth (m) 

10 

30 

50 

Weight 

Number 

100 

Figure 13-1. Mean occurrence and abundance in the catches per depth of Copadichromis 

virginalis in the SWA between July 1998 and May 1999. 

It was one of the most abundant species in shallow water, where it constituted about 10% and 

30% (both in weight and number) of the catches at 30 and 50 m, respectively. The mean 

CPUE per depth category was 97.7 kg in the shallow zone and 0.15 kg in the deep zone, 

which matched the value reported in Tómasson & Banda (1996) for the shallow zone (108.8 

kg). As for C. quadrimaculatus the depth distribution observed in our study was a bit more 

restricted than that of Tómasson & Banda (1996), who reported C. virginalis from 8 to 120 m. 

Specimens caught ranged between 45 and 125 mm (Figure 13-2). 

25 

20 

Frequency % 

15 

10 

5 

0 

47 52 57 62 67 72 77 82 87 92 97 102 107 112 117 122 


Figure 13-2. Size range and frequencies of Copadichromis virginalis caught in the SWA 


63


40 

35 

30 

25 

20 

15 

10 

5 


Fluorescence 

Temperature 

9 

8 

7 

6 

5 

4 

3 

2 

1 

Fluorescence index 

Temperature (-20°C) 

0 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Figure 13-4. Monthly percentage of ripe females in relation with temperature (= temperature 

minus 20°C to fit the axis scale) and fluorescence, used as an index of chlorophyll a 

concentration. 

100 


50 

0 

45 55 65 75 85 95 105 115 



for Copadichromis virginalis in the SWA. 

64

Weight (g) 

45 

40 

35 

30 

25 

20 

15 

10 

5 

y = 4E-05x 2,9091 

R 2 = 0,8956 

0 

40 60 80 100 120 140 

Length (mm) 

Figure 13-6. Length-weight relationship for Copadichromis virginalis females in the SWA. 


Fecundity 

70 

60 

50 

40 

30 

20 

10 

y = 0,4697x + 18,163 

R 2 = 0,0749 

0 

0 5 10 15 20 25 30 35 40 

Weight (g) 

Figure 13-7. Fecundity-weight relationship for Copadichromis virginalis females in the SWA. 



45 

40 

35 

y = -1,7669x + 25,142 

R 2 = 0,023 

30 

25 

20 

15 

10 

5 

0 

1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Copadichromis virginalis. Oocytes from females whose GSI was below (in grey) and 


65


C. virginalis was found to breed throughout the year (Figure 13-3a and b) with two 

peaks of activity, a punctual one in December and another one from May to July (Figure 13- 

3a). This figure markedly differed from Iles's (1971) observations who reported a very 

restricted breeding season (March to May) in the region of Nkata Bay. The percentage of ripe 

females seemed inversely related to water temperature (Figure 13-4), being higher during the 

cold water season (windy season) from April-May to August, and directly related to the 

chlorophyll a concentration. As C. virginalis is a zooplankton feeder (Iles 1960, 1971, Turner 

1996, present study see chapter "Diet"), most of the sexual activity occurred during windy 

season when the water column was mixed up and the rich cold upwelling increased the 

phytoplankton production and consequently the zooplankton abundance. The punctual peak of 

sexual activity observed in December 1998 corresponded with a drop of temperature of about 

3°C, an associated increased concentration in chlorophyll a and then an increased zooplankton 

availability. More than 70% of the ripe females and males in breeding colour were found at 50 

m (Table 13-1) and the rest at 30 m, suggesting that spawning could occur mostly at 50 m and 

in a lesser extent at 30 m. Maturity was reached in their first year at 12 months old at a mean 

size of 75 mm for females (Figure 13-5). 



Copadichromis virginalis in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

10 m 0.3 0.5 1.7 0.3 4.8 

30 17.8 26.2 24.8 7.7 25.2 

50 m 81.9 73.1 73.3 92 69.4 

100 m 0 0.2 0.1 0 0.6 


7, respectively. Fecundity ranged from 9 to 59 for females weighing between 10 and 36 g and 

was not correlated to body weight. No relation was found between oocyte weight and body 

weight. The GSI threshold above which the oocyte weight did no longer increase significantly 

was 3% (Figure 13-8). The mean oocyte weight was 18.93 mg (± 5.90 SD, N= 76). 

66


0 2 4 6 8 

Depth (m) 

75 

100 

Weight 

Number 

125 

Figure 14-1. Mean occurrence and abundance in the catches per depth of Diplotaxodon 

apogon in the SWA between July 1998 and May 1999. 

Frequeny % 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

32 37 42 47 52 57 62 67 72 77 82 87 92 97 102 107 112 117 122 127 


Figure 14-2. Size range and frequencies of Diplotaxodon apogon caught in the SWA between 


67

Diplotaxodon and Pallidochromis spp. 

Pallidochromis tokolosh (Turner) is the only species of this genus. Despite its 

anatomical distinctness that separates it from Diplotaxodon, recent molecular evidence 

suggested that this species is a member of the Diplotaxodon clade (Turner et al. 1999). It will 

then be presented together with the Diplotaxodon spp.. 

Diplotaxodon apogon (Turner & Stauffer) 

616 females and 729 males were analysed. D. apogon was found from 75 to 125 m, 

where it constituted between 3 and 4% of the catches in weight and 4 to 6% in number 

(Figure 14-1). The mean CPUE per depth category was 15.6 kg in the deep zone and 6 kg in 

the very deep zone. There was no record for this species in Tómasson & Banda (1996). 

Specimens caught ranged between 33 and 130 mm with a mode from 75 to 105 mm (Figure 

14-2). The sex ratio observed over the full sampling period was F/M 0.5/0.5. 

Breeding activity was detected in August 98 and from November 98 to April 99 for 

females (Figure 14-3a) whereas more than 60% of males in breeding colour were found at 

every sampled month (Figure 14-3b). Despite the very high percentage of males in breeding 

colour throughout the year, female data suggested a bimodal breeding season with a major 

peak from November to March and a smaller one around August. More than 60% of the ripe 

females were found at 100m (Table 14-1). More than 80 % of the males in breeding colour 

were evenly distributed at 75 and 125 m, but aggregations of such breeding males (50 to 90 

individuals) were found at all three depths. No particular indication about spawning depth was 

drawn from these results. Maturity was reached in their second year at 21 months old at a 

mean size of 88 mm for females (Figure 14-4). 


immature individuals (whose size is below the size at maturity) per depth for Diplotaxodon 

apogon in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

75 m 27.4 14.1 5.8 40.3 13.3 

100 m 43.3 62.5 79.5 16.1 53.4 

125 m 29.3 23.4 14.7 43.6 33.3 


6, respectively. Fecundity ranged from 9 to 34 for females weighing between 16 and 39 g. No 


the oocyte weight did no longer increase significantly was 2.7% (Figure 14-7). The mean 


68


50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

a 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

33 41 105 99 74 38 39 41 10 


b 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

61 65 70 67 90 86 74 82 79 


males (b) Diplotaxodon apogon in the SWA The values below the x-axis are the effective 


100 


50 

0 

55 65 75 85 95 105 115 125 



for Diplotaxodon apogon in the SWA. 

69

50 

Weight (g) 

40 

30 

20 

y = 0,0001x 2,6373 

10 

R 2 = 0,8437 

0 

70 80 90 100 110 120 130 

Length (mm) 

Figure 14-5. Length-weight relationship for Diplotaxodon apogon females in the SWA. (R² = 


Fecundity 

40 

35 

30 

25 

20 

15 

10 

5 

0 

y = 0,6738x - 1,0371 

R 2 = 0,2297 

10 15 20 25 30 35 40 

Weight (g) 

Figure 14-6. Fecundity-weight relationship for Diplotaxodon apogon females in the SWA. (R² 


60 


50 

40 

30 

20 

y = 2,2007x + 38,224 

10 

R 2 = 0,0636 

0 

0,0 1,0 2,0 3,0 4,0 5,0 

GSI (%) 


Diplotaxodon apogon. Oocytes from females whose GSI was below (in grey) and above 

(in black with regression) 2.7%. (R² = determination coefficient). 

70

Depth (m) 

50 

75 

100 


0 1 2 3 

Weight 

Number 

125 


argenteus in the SWA between July 1998 and May 1999. 

14 

12 

Frequency % 

10 

8 

6 

4 

2 

0 

45 55 65 75 85 95 105 115 125 135 145 155 165 175 185 195 205 


Figure 15-2. Size range and frequencies of Diplotaxodon argenteus caught in the SWA 


71

Diplotaxodon argenteus (Trewavas) 

278 females and 343 males were analysed. D. argenteus was found between 50 and 

125 m but became more frequent from 75 m downwards, where it constituted between about 1 

and 2% of the catches in number and 1.5 and 3% in weight (Figure 15-1). The mean CPUE 

per depth category was 1.5 kg in the shallow zone, 11 kg in the deep zone and 3.3 kg in the 

very deep zone, which matched the values reported in Tómasson & Banda (1996) for the 

shallows (1.2 kg) but was about three times more for the deep (3 kg) and very deep zones (1.3 

kg). Specimens caught ranged between 48 and 206 mm (Figure 15-2). The sex ratio observed 

over the full sampling period was F/M 0.4/0.6. 

Owing to irregularity in the catches of this species and low sample size of individuals 

above the size at maturity, precise determination of breeding season was not possible. 

Breeding females were found in October-November 98 and from February to May 99 (Figure 

15-3a). More than 30% of males in breeding colour were found all year long excluding June 

98 were only one individual was caught (Figure 15-3b). Considering the high percentages of 

breeding males and the extremely low sample size for females from June to September, it is 

likely that D. argenteus breed most of the year with a possible cessation in December- 

January. Ripe females were relatively evenly distributed between 75 and 125 m, whereas 

more than 60% of males in breeding colour were caught at 75 m, suggesting that spawning 

could mainly occur at 75 m (Table 15-1). Maturity was reached in their second year at 20 

months old at a mean size of 140 mm for females (Figure 15-4). 



argenteus in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

50 m 35.3 11.1 2.1 0.6 1 

75 m 11.2 22.2 35.4 61.7 29.5 

100 m 38.8 33.3 29.6 16.2 33.4 

125 m 14.7 33.3 32.8 21.4 36 


6, respectively. Fecundity ranged from 25 to 53 for females weighing between 55 and 139 g. 


which the oocyte weight did no longer increase significantly was 3% (Figure 15-7). The mean 


Diplotaxodon limnothrissa (Turner) (Plate 8) 

1462 females and 2723 males were analysed. D. limnothrissa was common from 50 to 

125 m and was occasionally encountered at 10 m depth (Figure 16-1), which corresponds to 

the depth distribution reported by Thompson et al. (1996) and Tómasson & Banda (1996). It 

was a dominant species in the catches at 50, 75 and 100 m, where it constituted (in number 

and weight, respectively) 6.2 and 3.8%, 8.3 and 7.7% and 11.7 and 9.7% of the catches, 

respectively. The mean CPUE per depth category was 14.7 kg in the shallow zone, 34.1 kg in 

72


50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

a 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

1 43 7 6 8 4 19 18 19 


80 

70 

60 

50 

40 

30 

20 

10 

0 

b 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

1 11 9 60 45 41 39 42 36 32 27 


males (b) Diplotaxodon argenteus in the SWA The values below the x-axis are the 


month. 

100 


50 

0 

45 55 65 75 85 95 105 115 125 135 145 155 165 175 185 195 



for Diplotaxodon argenteus in the SWA. 

73

Weight (g) 

160 

140 

120 

100 

80 

60 

40 

20 

0 

y = 0,0015x 2,1803 

R 2 = 0,925 

100 120 140 160 180 200 

Length (mm) 

Figure 15-5. Length-weight relationship for Diplotaxodon argenteus females in the SWA. (R² 


60 

50 

Fecundity 

40 

30 

20 

10 

y = 0,3006x + 6,1836 

R 2 = 0,6176 

0 

50 70 90 110 130 150 

Weight (g) 

Figure 15-6. Fecundity-weight relationship for Diplotaxodon argenteus females in the SWA. 



90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = -0,1005x + 73,431 

R 2 = 5E-05 

1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 

GSI (%) 


Diplotaxodon argenteus. Oocytes from females whose GSI was below (in grey) and above 


74

Plate 8. Diplotaxodon limnothrissa (by Dave Voorvelt). 


0 5 10 15 

Depth (m) 

10 

50 

75 

100 

125 

Weight 

Number 


limnothrissa in the SWA between July 1998 and May 1999. 

14 

12 

10 

Frequency % 

8 

6 

4 

2 

0 

42 47 52 57 62 67 72 77 82 87 92 97 102 107 112 117 122 127 132 137 142 147 152 157 162 167 172 177 


Figure 16-2. Size range and frequencies of Diplotaxodon limnothrissa caught in the SWA 


75

the deep zone and 4.1 kg in the very deep zone, which approximately matched the values 

reported in Tómasson & Banda (1996) for the deep (36 kg) and very deep zones (7.1 kg) but 

was about twice as much for the shallows (6.7 kg). Specimen caught ranged between 40 and 

175 mm (Figure 16-2). The sex ratio that was observed over the full sampling period was F/M 

0.3/0.7. 

Breeding season for females in the SWA occurred from March to August with a peak 

between April and June (Figure 16-3a) whereas males in breeding colour were caught 

throughout the year except in June (Figure 16-3b). These results being based upon pretty 

consistent sample size, the reason for the lack of fitting between females and males results is 

too be found elsewhere. Ripe females of the same species were found almost all year with a 

peak in March-April from offshore fishing locations and with a peak in May-June from 

inshore fishing location in the SEA (Thompson et al. 1996). Ripe females and males were also 

recorded from all over the Lake at any time of the year (Turner 1994a, Robinson R. L. pers. 

com.). Recent molecular analyses have shown that D. limnothrissa was constituted of a single 

wide spread population all over the Lake (Turner et al. 1999). This implies no breeding 

isolation from any part of the Lake and then large scale migrations of individuals. It is 

therefore likely that the breeding season we observed in the SWA between June 98 and May 

99 was only a fixed and reductive picture of what happens at the Lake scale. Taking into 

account all the available information, it is probable that D. limnothrissa breeds all year long, 

but with seasonal geographical peak of activity, which would explain our observed pattern. 

Ripe females and males in breeding colour were found from 50 to 125 m with a higher 

frequency at 125 m for females and at 75 m for males (Table 16-1). It was previously thought 

that D. limnothrissa was not forming demersal spawning arenas because aggregation of 

breeding males had never been observed (Turner 1994a, Thompson et al. 1996). This was 

used to emphasised that some of the pelagic cichlid species might be able to spawn 

independently of the bottom of the lake (Thompson et al. 1996), as already observed for 

Copadichromis ("Haplochromis") chrysonotus (Eccles & Lewis 1981). However, large 

aggregations of males in breeding colours (more than 300 specimens) were found at 75 and 

100 m in the SWA in April and May 99, suggesting that spawning probably occur close to the 

bottom at these depths. As already reported by Turner (1994a, 1996), D. limnothrissa females 

were observed to moothbrood young to large sizes, up to 23 mm SL. Maturity was reached 

early in their second year at 16 months old at a mean size of 105 mm for females (Figure 16- 

4), a size a bit smaller than the 14 cm TL (about 113 mm SL) reported by Thompson et al. 

(1996). 



limnothrissa in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

10 m 0.2 0 0.4 0 0.7 

50 m 14.4 25.7 18.5 4.1 34.7 

75 m 39.6 11 31.5 53.4 34.4 

100 m 35.4 17.8 41 31.2 27 

125 m 10.4 45.5 8.5 11.3 3.3 

76


50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

a 

238 128 108 316 53 41 60 65 109 232 113 


80 

70 

60 

50 

40 

30 

20 

10 

0 

b 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 


males (b) Diplotaxodon limnothrissa in the SWA The values below the x-axis are the 


month. 

Nov-98 

227 72 115 328 70 84 116 111 347 514 739 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

100 


50 

0 

45 55 65 75 85 95 105 115 125 135 145 



for Diplotaxodon limnothrissa in the SWA. 

77

Weight (g) 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 5E-05x 2,7974 

R 2 = 0,9456 

20 40 60 80 100 120 140 160 

Length (mm) 

Figure 16-5. Length-weight relationship for Diplotaxodon limnothrissa females in the SWA. 


Fecundity 

35 

30 

25 

20 

15 

10 

5 

y = 0,259x + 7,2365 

R 2 = 0,1148 

0 

20 30 40 50 60 

Weight (g) 

Figure 16-6. Fecundity-weight relationship for Diplotaxodon limnothrissa females in the 



90 

80 

70 

60 

50 

40 

30 

20 

y = 5,5821x + 41,918 

10 

R 2 = 0,1606 

0 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Diplotaxodon limnothrissa. Oocytes from females whose GSI was below (in grey) and 


78

Plate 9. Diplotaxodon macrops (by Dave Voorvelt). 


0 5 10 15 

Depth (m) 

75 

100 

Weight 

Number 

125 


macrops in the SWA between July 1998 and May 1999. 

30 

25 

Frequency % 

20 

15 

10 

5 

0 

42 47 52 57 62 67 72 77 82 87 92 97 102 107 112 117 122 127 132 137 142 147 


Figure 17-2. Size range and frequencies of Diplotaxodon macrops caught in the SWA 


79


6, respectively. Fecundity ranged from 10 to 30 for females weighing between 22 and 55 g 

and was not correlated to body weight. The fecundity range we found corresponded with the 

average fecundity of 15 eggs observed by Thompson et al. (1996) for females of 14-18 cm 

TL. No relation was found between oocyte weight and body weight. The GSI threshold above 



Diplotaxodon macrops (Turner & Stauffer) (Plate 9) 

1664 females and 2919 males were analysed. D. macrops was caught from 75 to 125 

m (Figure 17-1). It was a dominant species in the catches at 75 100 and 125 m, where it 

constituted (in number and weight, respectively) 5.3 and 4%, 11.6 and 10.4% and 10 and 

11.8% of the catches, respectively. The mean CPUE per depth category was 39.3 kg in the 

deep zone and 24 kg in the very deep zone. There was no record for this species in Tómasson 

& Banda (1996). Specimens caught ranged between 40 and 135 mm (Figure 17-2). A single 

individual measuring 150 mm was caught, which was probably a specimen of the resembling 

D. 'offshore' that grows larger (Robinson R. L., pers. com.). The sex ratio observed over the 

full sampling period was F/M 0.4/0.6. 

Ripe females (Figure 17-3a) and males (Figure 17-3b) were found throughout the year 

with a decline between October and December and a peak of activity from February to April. 

Most ripe females and males in breeding colour were found at 100 and 125 m (Table 17-1). 

All the large aggregations of breeding males (between 100 and 400 specimens) observed in 

August 98, January, February, March , April and May 99, were caught at 100 and 125 m, 

suggesting that spawning probably occurs at these depths. The mean size at maturity for 

females was about 98 mm (Figure 17-4), which corresponded to a mean age at maturity of 20 

months. 



macrops in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

75 m 11.9 5.1 24.9 11.5 33.2 

100 m 52.1 49.1 47.8 30.4 46.5 

125 m 36.1 45.8 27.3 58 20.3 


6, respectively. Fecundity ranged from 10 to 37 for females weighing between 20.6 and 51 g 

and was not correlated to body weight. No relation was found between oocyte weight and 

body weight. The GSI threshold above which the oocyte weight did no longer increase 

significantly was 3% (Figure 17-7). The mean oocyte weight was 56.04 mg (± 9.4 SD, N= 

46). 

80

60 


a 


b 

50 

40 

30 

20 

10 

0 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

135 50 137 264 58 85 222 89 233 7 155 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

29 24 273 132 140 111 538 335 315 366 656 


males (b) Diplotaxodon macrops in the SWA The values below the x-axis are the effective 


100 


50 

0 

45 55 65 75 85 95 105 115 125 135 



for Diplotaxodon macrops in the SWA. 

81

Weight (g) 

70 

60 

50 

y = 7E-05x 2,7825 

R 2 = 0,9358 

40 

30 

20 

10 

0 

20 40 60 80 100 120 140 

Length (mm) 

Figure 17-5. Length-weight relationship for Diplotaxodon macrops females in the SWA. (R² 


Fecundity 

40 

35 

30 

25 

20 

15 

10 

y = 0,245x + 8,8603 

5 

R 2 = 0,085 

0 

20 25 30 35 40 45 50 55 

Weight (g) 

Figure 17-6. Fecundity-weight relationship for Diplotaxodon macrops females in the SWA. 



90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 1,5093x + 50,438 

R 2 = 0,0116 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Diplotaxodon macrops. Oocytes from females whose GSI was below (in grey) and above 


82


0,0 0,5 1,0 1,5 2,0 2,5 

Depth (m) 

50 

75 

100 

Weight 

Number 

125 

Figure 18-1. Mean occurrence and abundance in the catches per depth of Pallidochromis 

tokolosh in the SWA between July 1998 and May 1999. 

25 

20 

Frequency % 

15 

10 

5 

0 

65 75 85 95 105 115 125 135 145 155 165 175 185 195 205 215 


Figure 18-2. Size range and frequencies of Pallidochromis tokolosh caught in the SWA 


83

Pallidochromis tokolosh (Turner) 

210 females and 134 males were analysed. P. tokolosh was common from 75 to 125 m 

and was occasionally caught at 50 m (Figure 18-1). It was more frequent at 125 m, where it 

constituted 1.1 and 1.9% of the catches in number and weight, respectively. The mean CPUE 

per depth category was 0.15 kg in the shallows, 3.6 kg in the deep zone and 4.4 kg in the very 

deep zone, which matched the values reported by Tómasson & Banda (1996) for the shallow 

and deep zone but was more than twice as much for the very deep zone (1.8 kg). Specimens 

caught ranged between 62 and 213 mm with a mode from 100 to 160 mm (Figure 18-2). The 


P. tokolosh is not an abundant fish and low sample size at some months hampered the 

correct determination of breeding season. Next, we never observed any particular breeding 

dress for males and therefore the percentage of males in breeding colour was impossible to 

assess. From the data available, it seemed that the breeding season would occur between 

October and February (Figure 18-3). The depth distribution of ripe females (Table 18-1) 

reflected the relative abundance per depth with 67% at 125 m and 30% at 75 m. No particular 

indication about spawning depth was drawn from these results. Maturity was reached early in 

their second year at 16 months old at a mean size of 135 mm for females (Figure 18-4). 



Pallidochromis tokolosh in the SWA. Sample size between bracket. 

Depth 

Non ripe females 

(70) 

Ripe females 

(27) 

Immature 

Specimens 

(185) 

50 m 0 3.7 0 

75 m 22.9 29.6 44.3 

100 m 10 0 18.4 

125 m 67.1 66.7 37.3 

The length-weight and fecundity-weight relationships are presented in Figure 18-5 and 

18-6, respectively. Fecundity ranged from 13 to 87 for females weighing between 36 and 143 

g. No relation was found between oocyte weight and body weight. The largest and heaviest 

oocytes of all cichlid species studied were produced by P. tokolosh, the record being 85 mg. 

The GSI threshold above which the oocyte weight did no longer increase significantly was not 

determined owing to low sample size of high GSI (Figure 18-7). Nevertheless, assuming a 3% 

threshold as for the other species of the Diplotaxodon clade, the mean oocyte weight was 

70.49 mg (± 12.53 SD, N= 3). 

84


80 

70 

60 

50 

40 

30 

20 

10 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

3 8 10 8 8 21 18 4 9 8 27 

Figure 18-3. Seasonal progression of the percentage of ripe (stages 4 and 5) females 

Pallidochromis tokolosh in the SWA The values below the x-axis are the effective (number 

of females which size was above the size at maturity) for each month. 

100 


50 

0 

65 75 85 95 105 115 125 135 145 155 165 175 185 195 205 



for Pallidochromis tokolosh in the SWA. 

85

Weight (g) 

160 

140 

120 

100 

80 

60 

40 

y = 5E-05x 2,7716 

20 

R 2 = 0,9287 

0 

100 120 140 160 180 200 220 

Length (mm) 

Figure 18-5. Length-weight relationship for Pallidochromis tokolosh females in the SWA. (R² 


100 

80 

Fecundity 

60 

40 

20 

0 

y = 0,5496x - 4,5233 

R 2 = 0,7257 

0 50 100 150 

Weight (g) 

Figure 18-6. Fecundity-weight relationship for Pallidochromis tokolosh females in the SWA. 



90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0,0 1,0 2,0 3,0 4,0 5,0 

GSI (%) 


Pallidochromis tokolosh. Oocytes from females whose GSI was below (in grey) and above 

(in black) 3%. 

86

Plate 10. Lethrinops argenteus (by Dave Voorvelt). 


0 5 10 15 20 

10 

Depth (m) 

30 

50 

75 

100 

125 

Weight 

Number 

Figure 19-1. Mean occurrence and abundance in the catches per depth of Lethrinops 

argenteus in the SWA between July 1998 and May 1999. 

Frequency % 

9 

8 

7 

6 

5 

4 

3 

2 

1 

0 

52 57 62 67 72 77 82 87 92 97 102 107 112 117 122 127 132 137 142 147 152 157 162 


Figure 19-2. Size range and frequencies of Lethrinops argenteus caught in the SWA between 


87

Lethrinops spp. 

Lethrinops argenteus (Ahl) (Plate 10) 

3176 females and 2424 males were analysed. L. argenteus was caught from 10 to 125 

m (Figure 19-1). It was abundant only between 10 and 50 m where it was a dominant species 

in the catches, constituting between 9 and 11% of the catches in number and between 13 and 

19% in weight. The mean CPUE per depth category was 127.4 kg in the shallows, 0.15 and 

0.3 kg in the deep and very deep zones, respectively. No reference was made to this species in 

Tómasson & Banda (1996), who probably included this species in the L. longipinnis group. 



Ripe females (Figure 19-3a) and males (Figure 19-3b) of L. argenteus were found 

throughout the year with steady decline from March to June and in October and peaks of 

activity in August 98 and between December and February 99. 70% of males in breeding 

colour and 43% of ripe females were caught at 30 m (Table 19-1). Aggregations of breeding 

males, although small (between 30 and 80 individuals) compared to those observed for 

Diplotaxodon spp., were always found at 30 m suggesting that spawning could occur at this 

depth. The mean size at maturity for females was about 108 mm (Figure 19-4), which 

corresponded to a mean age at maturity of 12 months. 


immature individuals (whose size is below the size at maturity) per depth for Lethrinops 

argenteus in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

10 m 17.4 26.2 31.2 22.6 23.1 

30 m 35.8 41.1 43.1 70.2 27.3 

50 m 46.5 32.3 25.5 6.7 49.5 

75 m 0.1 0 0.2 0.2 0.2 

100 m 0 0.4 0 0 0 

125 m 0.2 0 0.1 0.3 0 


19-6, respectively. Fecundity ranged from 43 to 218 for females weighing between 12.5 and 

103 g. No relation was found between oocyte weight and body weight. The GSI threshold 

above which the oocyte weight did no longer increase significantly was 4% (Figure 19-7). 

The mean oocyte weight was 20.09 mg (± 4.17 SD, N= 28). 

88


a 


b 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

87 219 129 191 199 179 164 180 206 362 190 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

130 207 155 184 287 202 245 260 237 285 232 


males (b) Lethrinops argenteus in the SWA The values below the x-axis are the effective 


100 


50 

0 

55 65 75 85 95 105 115 125 135 145 



for Lethrinops argenteus in the SWA. 

89

Weight (g) 

120 

100 

80 

60 

40 

20 

y = 3E-05x 3,0087 

R 2 = 0,9596 

0 

40 60 80 100 120 140 160 

Length (mm) 

Figure 19-5. Length-weight relationship for Lethrinops argenteus females in the SWA. (R² = 


250 

200 

Fecundity 

150 

100 

50 

0 

y = 1,7255x + 21,72 

R 2 = 0,6123 

0 20 40 60 80 100 120 

Weight (g) 

Figure 19-6. Fecundity-weight relationship for Lethrinops argenteus females in the SWA. (R² 


35 

30 


25 

20 

15 

10 

5 

y = 0,1084x + 19,545 

R 2 = 0,001 

0 

0,0 2,0 4,0 6,0 8,0 10,0 

GSI (%) 


Lethrinops argenteus Oocytes from females whose GSI was below (in grey) and above (in 


90

Plate 11. Lethrinops 'deep water albus' (by Dave Voorvelt). 


0,0 1,0 2,0 3,0 4,0 

75 

Depth (m) 

100 

Weight 

Number 

125 

Figure 20-1. Mean occurrence and abundance in the catches per depth of Lethrinops 'deep 

water albus' in the SWA between July 1998 and May 1999. 

Frequency % 

35 

30 

25 

20 

15 

10 

5 

0 

45 55 65 75 85 95 105 115 125 135 145 155 


Figure 20-2. Size range and frequencies of Lethrinops 'deep water albus' caught in the SWA 


91

Lethrinops 'deep water albus' (Plate 11) 

303 females and 171 males were analysed. L. 'deep water albus' was caught from 75 to 

125 m, though rarely at 100 m (Figure 20-1). It was not an abundant fish, always constituting 

less than 1% of the catches and often completely absent from the sampled part of the SWA. 

The 3.3 and 3.7% of the catches respectively in weight and number, observed at 75 m were 

due to an exceptional catch in November 98, during which L. 'deep water albus' made up to 

31 and 33% of the catch in weight and number. However, L. 'deep water albus' was very 

abundant off Domira Bay and off Leopard Bay, being the dominant species in catches of 400 

to 600 kg between 75 and 125 m. This species seemed to be dominant in the deep water 

catches when Lethrinops gossei was absent or rare. L. 'deep water albus' was often very 

abundant in the catches in these areas when we were targeting for L. gossei, and one or the 

other was dominant but never both of them at the same time. As a matter of fact, during the 

exceptional catch of L. 'deep water albus' at 75 m in the SWA in November 98, only 38 

specimens of L. gossei were caught. As L. gossei was a consistently dominant species in the 

deep zone of the SWA (see further), it might explain why L. 'deep water albus' was rare. 



Owing to irregularity in the catches of this species and low sample size for most of the 

months, precise determination of breeding season was not possible and. Ripe females were 

found from June to August 98 and in November and January 98 (Figure 20-3a) whereas males 

in breeding colour were found at each sampling date, including in May (Figure 20-3b). Size at 

maturity was estimated at about 82 mm for females and was probably slightly overestimated 

as this estimation was not done during the height of the breeding season but with all the data 

available for females (Figure 20-4). 




which the oocyte weight did no longer increase significantly was not determine precisely 

owing to low sample size, but was estimated at 4% (Figure 20-7). The mean oocyte weight 

was 10.27 mg (± 0.89 SD, N= 3). 

92


a 


b 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

19 29 15 22 216 1 1 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

15 33 7 38 64 3 9 2 


males (b) Lethrinops 'deep water albus' in the SWA The values below the x-axis are the 


month. 

100 


50 

0 

45 55 65 75 85 95 105 115 125 135 145 



for Lethrinops 'deep water albus' in the SWA. 

93

60 

50 

Weight (g) 

40 

30 

20 

y = 3E-05x 3,0396 

10 

R 2 = 0,8992 

0 

70 80 90 100 110 120 

Length (mm) 

Figure 20-5. Length-weight relationship for Lethrinops 'deep water albus' females in the 


Fecundity 

160 

140 

120 

100 

80 

60 

40 

y = 2,8411x + 22,771 

20 

R 2 = 0,7249 

0 

10 15 20 25 30 35 40 45 

Weight (g) 

Figure 20-6. Fecundity-weight relationship for Lethrinops 'deep water albus' females in the 


12 


10 

8 

6 

4 

2 

0 

1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Lethrinops 'deep water albus' Oocytes from females whose GSI was below (in grey) and 


94

Plate 12. Lethrinops 'deep water altus' (by Dave Voorvelt). 


0 5 10 15 

Depth (m) 

75 

100 

Weight 

Number 

125 

Figure 21-1. Mean occurrence and abundance in the catches per depth of Lethrinops 'deep 

water altus' in the SWA between July 1998 and May 1999. 

Frequency % 

35 

30 

25 

20 

15 

10 

5 

0 

35 45 55 65 75 85 95 105 115 125 135 145 


Figure 21-2. Size range and frequencies of Lethrinops 'deep water altus' caught in the SWA 


95

Lethrinops 'deep water altus' (Plate 12) 

625 females and 885 males were analysed. L. 'deep water altus' was an abundant 

species of the deep and very deep zones, increasing in occurrence and biomass with depth 

from 75 to 125 m to reach 5.3 and 11.5% of the catches in weight and number, respectively 

(Figure 21-1). The mean CPUE per depth category was 8.6 and 3.3 kg in the deep and very 

deep zones, respectively. No reference was made to this species in Tómasson & Banda 

(1996). Specimens caught ranged between 35 and 130 mm with a mode from 60 to 100 mm 


Breeding season occurred between December and August-September with a peak of 

activity from February to June and a cessation in October-November confirmed by the lower 

percentage of active males during this period (Figure 21-3a and b). Ripe females and 

immature individuals were evenly distributed between 75 and 125 m, whereas about 60% of 

the males in breeding colour were caught at 125 m, suggesting that spawning could occur 

mostly at this depth (Table 21-1). The mean size at maturity for females was about 60 mm 

(Figure 21-4), which corresponded to a mean age at maturity of 11 months. 



'deep water altus' in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

75 m 14.2 29.5 12.6 11.5 23.5 

100 m 46.2 34.7 37.1 31.3 37.3 

125 m 39.7 35.8 50.3 57.2 39.2 


21-6, respectively. Fecundity ranged from 10 to 84 for females weighing between 5 and 19 g 

and was not correlated with body weight. No relation was found between oocyte weight and 


significantly was 3% (Figure 21-7). The mean oocyte weight was 8.08 mg (± 2.32 SD, N= 

34). 

96


a 


b 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 


males (b) Lethrinops 'deep water altus' in the SWA The values below the x-axis are the 


month. 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

84 40 42 38 17 61 25 23 116 108 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

85 70 91 83 40 61 73 51 150 181 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

100 


50 

0 

37 42 47 52 57 62 67 72 77 82 87 92 



for Lethrinops 'deep water altus' in the SWA. 

97

Weight (g) 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

y = 4E-05x 2,9176 

R 2 = 0,7929 

50 60 70 80 90 

Length (mm) 

Figure 21-5. Length-weight relationship for Lethrinops 'deep water altus' females in the 


Fecundity 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 1,6315x + 21,615 

R 2 = 0,1206 

5 10 15 20 

Weight (g) 

Figure 21-6. Fecundity-weight relationship for Lethrinops 'deep water altus' females in the 


14 

12 


10 

8 

6 

4 

2 

y = 0,5544x + 5,7836 

R 2 = 0,0587 

0 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 

GSI (%) 


Lethrinops 'deep water altus' Oocytes from females whose GSI was below (in grey) and 


98

Plate 13. Lethrinops gossei (by Dave Voorvelt). 


0 10 20 30 

Depth (m) 

50 

75 

100 

Weight 

Number 

125 

Figure 22-1. Mean occurrence and abundance in the catches per depth of Lethrinops gossei in 


25 

20 

Frequency (%) 

15 

10 

5 

0 

40 50 60 70 80 90 100 110 120 130 140 150 160 170 


Figure 22-2. Size range and frequencies of Lethrinops gossei caught in the SWA between July 

1998 and May 1999. 

99

Lethrinops gossei (Burgess & Axelrod) (Plate 13) 

3513 females and 3425 males were analysed. L. gossei was caught from 50 to 125 m, 

becoming very abundant form 75 m downwards (Figure 22-1). It was a dominant species of 

the deep and very deep zones, where it constituted between 13 and 16% of the catches in 

number and between 18 and 25% in weight. The mean CPUE per depth category was 0.3, 108 

and 51.6 kg for the shallow, deep and very deep zones, respectively. Except for the very deep 

zone, this was very different from the values reported in Tómasson & Banda (1996), who 

caught much more L. gossei per time unit in the shallows (21.6 kg) and about two times less 

in the deep zone (61.9 kg). Specimens caught ranged between 35 and 170 mm with a mode 

from 95 to 135 mm (Figure 22-2). The sex ratio observed over the full sampling period was 

F/M 0.5/0.5. 

Breeding season occurred between November and August with a peak of activity from 

January to March-April and a stop in September-October (Figure 22-3a). It was one of the 

rare species for which the profile of the percentage of ripe males followed exactly the female's 

one (Figure 22-3b). The breeding season we observed corresponded relatively well to that 

found by Lewis & Tweddle (1990), who reported a decline in October-November and a peak 

in March for the period 1983-85. Ripe females and males were evenly distributed between 75 

and 125 m (Table 22-1), as were the aggregations of males in breeding colour (between 100 

and 200 specimens), suggesting that spawning probably takes place at all three depths. The 

mean size at maturity for females was about 92 mm (Figure 22-4), which is much less than 

the 147 mm TL estimated by Lewis & Tweddle (1990). This corresponded to a mean age at 

maturity of 11 months. 



gossei in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

50 m 0 0 0.1 0.1 0.2 

75 m 33.5 42.1 28 27.8 44 

100 m 26.9 36.8 39.3 37.8 36.4 

125 m 39.6 21.1 32.6 34.3 19.4 


22-6, respectively. Fecundity ranged from 23 to 234 for females weighing between 10 and 


above which the oocyte weight did no longer increase significantly was 4% (Figure 22-7). 

The mean oocyte weight was 21.34 mg (± 4.34 SD, N= 190). A GSI of 12.4% was recorded 

for a L. gossei female (Figure 22-7), which was the highest GSI calculated on any cichlid 

species during the course of this study. 

100


a 


b 

35 

30 

25 

20 

15 

10 

5 

0 

90 

80 

70 

60 

50 

40 

30 

20 

10 

- 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

242 313 347 237 251 127 276 263 356 362 251 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

250 312 254 183 226 236 449 298 333 450 346 


males (b) Lethrinops gossei in the SWA The values below the x-axis are the effective 


100 


50 

0 

35 45 55 65 75 85 95 105 115 125 135 145 155 



for Lethrinops gossei in the SWA. 

101

Weight (g) 

120 

100 

80 

60 

40 

20 

0 

y = 3E-05x 2,9934 

R 2 = 0,9639 

0 20 40 60 80 100 120 140 160 

Length (mm) 

Figure 22-5. Length-weight relationship for Lethrinops gossei females in the SWA. (R² = 


250 

200 

Fecundity 

150 

100 

50 

0 

y = 1,6832x + 19,458 

R 2 = 0,5994 

0 20 40 60 80 100 120 

Weight (g) 

Figure 22-6. Fecundity-weight relationship for Lethrinops gossei females in the SWA. (R² = 


40 

35 


30 

25 

20 

15 

10 

5 

y = 1,0771x + 15,785 

R 2 = 0,0657 

0 

0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 

GSI (%) 


Lethrinops gossei. Oocytes from females whose GSI was below (in grey) and above (in 


102


0,0 0,5 1,0 1,5 2,0 

30 

Depth (m) 

50 

75 

100 

125 

Weight 

Number 

Figure 23-1. Mean occurrence and abundance in the catches per depth of Lethrinops 

longimanus in the SWA between July 1998 and May 1999. 

Frequency % 

35 

30 

25 

20 

15 

10 

5 

0 

55 65 75 85 95 105 115 125 135 145 


Figure 23-2. Size range and frequencies of Lethrinops longimanus caught in the SWA 


103

Lethrinops longimanus (Trewavas) 

210 females and 154 males were analysed. L. longimanus was caught from 30 to 125 

m (Figure 23-1). It was a relatively rare fish averaging less than 1% of the catches at all 

depths except at 50 m where it made up to 1.6 and 1.7% in number and weight, respectively. 

The mean CPUE per depth category 11.7, 2 and 0.15 kg for the shallow, deep and very deep 

zones, respectively, were much lower than those reported by Tómasson & Banda (1996) 

(18.9, 23.1 and 3.7 kg, respectively). Specimens caught ranged between 56 and 147 mm with 

a mode from 100 to 120 mm (Figure 23-2). The sex ratio observed over the full sampling 

period was F/M 0.6/0.4. 

Owing to low sample size at some months and low number of ripe females caught, 

determination of precise breeding season was impossible. What can be said from the few data 

available is that breeding probably did not take place during the period from November to 

February when sample size were correct (Figure 23-3a). The only significant proportion of 

ripe females was found in August 98, and taking into account the data for males (Figure 23- 

3b), it might be hypothesised that breeding season occur from May-June to August- 

September. All the ripe females and 70% of the ripe males were found at 50m, suggesting that 

spawning might occur at this depth. However, most specimens of L. longimanus were caught 

at 50 m and this trend might reflect nothing more than the depth of occurrence. The mean size 

at maturity for females was about 107 mm, which was probably overestimated given that this 

estimation was done with all the data available for females, including females caught outside 

the breeding season (Figure 23-4). This corresponded to a mean age at maturity of 18 months, 

hence also probably overestimated. 

Owing to the very narrow size range of females measured, it was impossible to assess 

the length-weight relationship. The fecundity-weight relationship is presented in Figure 23-5. 

Fecundity ranged from 57 to 99 for females weighing between 35 and 47 g. No relation was 

found between oocyte weight and body weight. The GSI threshold above which the oocyte 

weight did no longer increase significantly was estimated at about 4% (Figure 23-6). The 


104


a 


b 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

60 

50 

40 

30 

20 

10 

0 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

8 27 1 80 25 13 46 6 2 2 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

7 32 35 25 5 44 2 4 


males (b) Lethrinops longimanus in the SWA The values below the x-axis are the effective 


100 


50 

0 

55 65 75 85 95 105 115 125 135 



for Lethrinops longimanus in the SWA. 

105

Fecundity 

120 

100 

80 

60 

40 

20 

y = 2,9578x - 43,716 

R 2 = 0,8193 

0 

30 35 40 45 50 

Weight (g) 

Figure 23-5. Fecundity-weight relationship for Lethrinops longimanus females in the SWA. 



18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

y = -0,4321x + 16,853 

R 2 = 0,0121 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Lethrinops longimanus. Oocytes from females whose GSI was below (in grey) and above 


106

Lethrinops macrochir (Regan) 

L. macrochir is a rare species caught only at 10 m. It was usually absent from the 

catches, except for November to February and once in August. 42 females and 89 males were 

analysed, giving a sex ratio of F/M 0.3/0.7. Specimens caught ranged from 59 to 150 mm 

(Figure 24-1). 

Frequency % 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

57 62 67 72 77 82 87 92 97 102 107 112 117 122 


Figure 24-1. Size range and frequencies of Lethrinops macrochir caught in the SWA between 


Among the rare species, L. macrochir was one for which we caught some ripe females, 

allowing an estimation of the fecundity-weight relationship (Figure 24-2) and oocyte weight 

(for females whose GSI was above 4%, which is the maximum encountered for the genus): 

11.87 mg (± 0.97 SD, N= 2). 

Fecundity 

180 

160 

140 

120 

100 

80 

60 

40 

20 

0 

y = 2,7129x + 12,68 

R 2 = 0,9968 

20 30 40 50 60 

Weight (g) 

Figure 24-2. Fecundity-weight relationship for Lethrinops macrochir females in the SWA. (R² 


107

Plate 14. Lethrinops 'oliveri' (by Dave Voorvelt). 


0 10 20 30 

75 

Depth (m) 

100 

125 

Weight 

Number 

Figure 25-1. Mean occurrence and abundance in the catches per depth of Lethrinops 'oliveri' 


30 

25 

Frequency % 

20 

15 

10 

5 

0 

32 37 42 47 52 57 62 67 72 77 82 87 92 97 


Figure 25-2. Size range and frequencies of Lethrinops 'oliveri' caught in the SWA between 


108

Lethrinops 'oliveri' (Plate 14) 

875 females and 1017 males were analysed. L. 'oliveri' was caught between 75 and 

125 m, with a decreasing occurrence and biomass with depth (Figure 25-1). It was a dominant 

species of the deep and very deep zones, constituting between 3 and 10% of the catches in 

number and between 8 and 22% in weight. The mean CPUE per depth category, 39.8 and 5.4 

kg for the deep and very deep zones, respectively, matched the value reported by Tómasson & 

Banda (1996) for the deep zone (36.7 kg), but was about three times less for the very deep 

zone (16.7 kg). The depth distribution observed in our study was more restricted than that of 

Tómasson & Banda (1996), who reported L. 'oliveri' from 20 to 150 m. Specimens caught 

ranged between 33 and 98 mm (Figure 25-2). The sex ratio observed over the full sampling 

period was F/M 0.5/0.5. 

Breeding occurred throughout the year with a trough from October to December and a 

peak of activity from February to April (Figure 25-3a). As for L. gossei, the profile of the 

percentage of ripe males followed exactly the female's one (Figure 25-3b). Ripe females and 

males were most abundant at 75 m, suggesting that spawning could occur mainly at this depth 

(Table 25-1). The mean size at maturity for females was about 60 mm (Figure 25-4), which 




'oliveri' in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

75 m 45.3 46.5 37.2 67.7 43.5 

100 m 38.3 31.6 38.1 27.3 36.6 

125 m 16.3 21.9 24.7 5.1 19.9 


25-6, respectively. Fecundity ranged from 19 to 81 for females weighing between 4 and 16 g. 




109


50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

a 


b 

70 

60 

50 

40 

30 

20 

10 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

5 56 114 40 53 61 120 32 129 31 91 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

10 68 148 69 97 88 148 45 169 11 164 


males (b) Lethrinops 'oliveri' in the SWA The values below the x-axis are the effective 


100 


50 

0 

32 37 42 47 52 57 62 67 72 77 82 87 92 97 



for Lethrinops 'oliveri' in the SWA. 

110

Weight (g) 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

y = 0,0005x 2,314 

R 2 = 0,7327 

40 50 60 70 80 90 100 

Length (mm) 

Figure 25-5. Length-weight relationship for Lethrinops 'oliveri' females in the SWA. (R² = 


Fecundity 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 4,5704x + 3,0726 

R 2 = 0,4043 

0 5 10 15 20 

Weight (g) 

Figure 25-6. Fecundity-weight relationship for Lethrinops 'oliveri' females in the SWA. (R² = 


12 


10 

8 

6 

4 

2 

y = 0,4753x + 5,1611 

R 2 = 0,0634 

0 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Lethrinops 'oliveri'. Oocytes from females whose GSI was below (in grey) and above (in 


111


0 2 4 6 

Depth (m) 

75 

100 

125 

Weight 

Number 

Figure 26-1. Mean occurrence and abundance in the catches per depth of Lethrinops polli in 


30 

25 

Frequency % 

20 

15 

10 

5 

0 

32 37 42 47 52 57 62 67 72 77 82 87 92 97 102 107 112 117 


Figure 26-2. Size range and frequencies of Lethrinops polli caught in the SWA between July 

1998 and May 1999. 

112

Lethrinops polli (Burgess & Axelrod) 

229 females and 198 males were analysed. L. polli was caught between 75 and 125 m, 

with a decreasing occurrence and biomass with depth (Figure 26-1). It was caught regularly in 

the deep and very deep zones, constituting between 0.1 and 3% of the catches in weight and 

between 0.3 and 5.6% in number. The mean CPUE per depth category, 12.3 and 0.3 kg for the 

deep and very deep zones, respectively, was about four times as much as the value reported 

by Tómasson & Banda (1996) for the deep zone (3 kg), and about seven times less for the 

very deep zone (2.2 kg). As for L. 'oliveri', the depth distribution observed in our study was 

more restricted than that of Tómasson & Banda (1996), who reported L. polli from 10 to 140 

m. Specimens caught ranged between 30 and 120 mm (Figure 26-2). The sex ratio observed 

over the full sampling period was F/M 0.5/0.5. 

Breeding season occurred from May to August-September, with an isolated activity in 

December, in the middle of the resting period (Figure 26-3a). As for L. gossei and L. 'oliveri', 

the profile of the percentage of ripe males followed exactly the female's one (Figure 26-3b). 

Ripe females and males were clearly most abundant at 75 m, suggesting that spawning could 

occur mainly at this depth (Table 26-1). The mean size at maturity for females was about 65 

mm (Figure 26-4), which corresponded to a mean age at maturity of 10 months. 



polli in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

75 m 68.9 76.3 50.4 85.5 83.3 

100 m 28.9 21.1 41.1 10.1 5.6 

125 m 2.1 2.6 8.5 4.3 11.1 






113


80 

70 

60 

50 

40 

30 

20 

10 

0 

a 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

1 27 19 34 26 5 28 12 26 8 42 


90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

b 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

7 27 15 20 22 8 23 20 15 10 31 


males (b) Lethrinops polli in the SWA The values below the x-axis are the effective 


100 


50 

0 

52 57 62 67 72 77 82 87 92 



for Lethrinops polli in the SWA. 

114

Weight (g) 

40 

35 

30 

25 

20 

15 

10 

y = 3E-05x 2,9546 

5 

R 2 = 0,8425 

0 

50 60 70 80 90 100 110 120 

Length (mm) 

Figure 26-5. Length-weight relationship for Lethrinops polli females in the SWA. (R² = 


Fecundity 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 2,5654x + 2,6796 

R 2 = 0,5713 

0 10 20 30 40 

Weight (g) 

Figure 26-6. Fecundity-weight relationship for Lethrinops polli females in the SWA. (R² = 



18 

16 

14 

12 

10 

8 

6 

4 

y = 0,9759x + 8,4406 

2 

R 2 = 0,0776 

0 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Lethrinops polli. Oocytes from females whose GSI was below (in grey) and above (in 


115

Plate 15. Mylochromis anaphyrmus (by Dave Voorvelt). 


0 2 4 6 8 

10 

Depth (m) 

30 

50 

75 

Weight 

Number 

Figure 27-1. Mean occurrence and abundance in the catches per depth of Mylochromis 

anaphyrmus in the SWA between July 1998 and May 1999. 

Frequency % 

9 

8 

7 

6 

5 

4 

3 

2 

1 

0 

42 47 52 57 62 67 72 77 82 87 92 97 102 107 112 117 122 127 132 137 142 147 152 157 162 167 


Figure 27-2. Size range and frequencies of Mylochromis anaphyrmus caught in the SWA 


116

Mylochromis spp. 

Mylochromis anaphyrmus (Burgess & Axelrod) (Plate 15) 

1375 females and 1058 males were analysed. M. anaphyrmus was caught from 10 to 

75 m, but mostly abundant at 10 and 30 m (Figure 27-1). Actually, a single specimen was 

caught at 75 m. It was a very common species in the shallow zone, constituting between 4 and 

7% of the catches in weight and between 2 and 6% in number. The mean CPUE per depth 

category, 23.3 kg for the shallows, was more than twice as much as the value found by 

Tómasson & Banda (1996) (9.8 kg). The depth distribution they reported was very similar to 

the one we observed. Specimens caught ranged between 40 and 164 mm (Figure 27-2). The 


The breeding season was from January to October, with a peak between March and 

June, a steady decline in November, and ceasing in December (Figure 27-3a), even though 

occurrence of males in breeding colour was more erratic (Figure 27-3b). Ripe males and 

immature individuals were evenly distributed between 10 and 30 m (Table 27-1). However, as 

more than three quarters of the ripe females were found at 30 m, spawning probably occurs at 

30 m. Maturity was reached early in their second year at 17 months old, at a mean size of 105 

mm for females (Figure 27-4), which was less than the 160 mm TL (about 130 mm SL) and 3 

years old reported by Tweddle & Turner (1977). 


immature individuals (whose size is below the size at maturity) per depth for Mylochromis 

anaphyrmus in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

10 m 25.8 11.4 35.6 48.9 39.4 

30 m 64.2 76.2 60.9 46.7 56.6 

50 m 10 12.4 3.5 4.4 4.1 






117


a 


50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

- 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

19 68 74 40 104 51 77 68 45 49 49 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

b 

100 98 105 109 63 41 165 79 42 64 114 


males (b) Mylochromis anaphyrmus in the SWA The values below the x-axis are the 


month. 

100 


50 

0 

45 55 65 75 85 95 105 115 125 135 145 155 



for Mylochromis anaphyrmus in the SWA. 

118

Weight (g) 

160 

140 

120 

100 

80 

60 

40 

20 

0 

y = 2E-05x 3,1127 

R 2 = 0,9759 

20 40 60 80 100 120 140 160 180 

Length (mm) 

Figure 27-5. Length-weight relationship for Mylochromis anaphyrmus females in the SWA. 


250 

200 

Fecundity 

150 

100 

50 

0 

y = 1,9783x + 25,864 

R 2 = 0,7413 

0 20 40 60 80 100 120 

Weight (g) 

Figure 27-6. Fecundity-weight relationship for Mylochromis anaphyrmus females in the 


16 

14 


12 

10 

8 

6 

4 

2 

y = 1,5242x + 6,0293 

R 2 = 0,0812 

0 

1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 

GSI (%) 


Mylochromis anaphyrmus. Oocytes from females whose GSI was below (in grey) and 


119

Plate 16. Nyassachromis 'argyrosoma' (by Dave Voorvelt). 


50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

a 


b 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

59 147 46 133 222 262 324 215 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

116 303 170 105 296 185 167 221 


males (b) Nyassachromis 'argyrosoma' in the SWA The values below the x-axis are the 


month. 

120

Nyassachromis spp. 

Nyassachromis 'argyrosoma' (Plate 16) 

1440 females and 1563 males were analysed. N. 'argyrosoma' is a small species caught 

from 10 to 50 m, but was most abundant at 10 and 30 m (Figure 28-1). 


0 10 20 30 40 50 

10 

Depth (m) 

30 

50 

Weight 

Number 

Figure 28-1. Mean occurrence and abundance in the catches per depth of Nyassachromis 

'argyrosoma' in the SWA between July 1998 and May 1999. 

With Aulonocara 'blue orange' and Copadichromis virginalis, N. 'argyrosoma' was a 

numerically dominant species in the shallow zone, constituting about 23% of the catches in 

weight, and 41 and 43% in number at 10 and 30 m, respectively. The mean CPUE per depth 

category was 91.8 kg for the shallows. No reference to this species (under this name at least) 

was made in Tómasson & Banda (1996). Specimens caught ranged between 38 and 97 mm 


35 

30 

Frequency % 

25 

20 

15 

10 

5 

0 

37 42 47 52 57 62 67 72 77 82 87 92 97 


Figure 28-2. Size range and frequencies of Nyassachromis 'argyrosoma' caught in the SWA 



121

100 


50 

0 

42 47 52 57 62 67 72 77 82 87 92 



for Nyassachromis 'argyrosoma' in the SWA. 

Weight (g) 

14 

12 

10 

8 

6 

4 

2 

0 

y = 0,0003x 2,3822 

R 2 = 0,7256 

50 55 60 65 70 75 80 85 90 

Length (mm) 

Figure 28-5. Length-weight relationship for Nyassachromis 'argyrosoma' females in the 


60 

50 

Fecundity 

40 

30 

20 

10 

0 

y = 3,2835x + 10,481 

R 2 = 0,449 

2 4 6 8 10 12 14 

Weight (g) 

Figure 28-6. Fecundity-weight relationship for Nyassachromis 'argyrosoma' females in the 


122


14 

12 

10 

8 

6 

4 

2 

0 

y = 0,77x - 1,0704 

R 2 = 0,3293 

0 2 4 6 8 10 12 14 

Body weight (g) 

Figure 28-7. Relationship between oocyte weight and body weight for Nyassachromis 

'argyrosoma' females in the SWA. (R² = determination coefficient). 

14 

12 


10 

8 

6 

4 

2 

y = 0,0983x + 7,2789 

R 2 = 0,0007 

0 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Nyassachromis 'argyrosoma'. Oocytes from females whose GSI was below (in grey) and 


123

Ripe females (Figure 28-3a) and males (Figure 28-3b) were found at all the sampled 

months, with peaks in December and March and decreased activity in November and 

February. Owing to inconsistencies in species identification during the first three months of 

sampling, data were removed from analyses. However, it is likely that breading occurred all 

year long. Three quarters of the ripe females and more than half the ripe males were found at 

10 m, the rest were at 30 m (Table 28-1), suggesting that spawning probably occurs between 

10 and 30 m and mostly at 10 m. As most immature individuals were caught at 30 and 50 m, 

it appears that after being released, juveniles migrate into deeper waters. The mean size at 

maturity was about 57 mm for females (Figure 28-4), which corresponded to a mean age at 

maturity of 10 months. The upper part of the sigmoïd curve does reach 100% after various 

steps, which is unexpected for an abundant species. Even though we had to take into account 

data from months outside the peaks of breeding activity to increase the sample size in each 

size class, an alternative explanation can not be excluded for this species. The taxonomy of 

the Nyassachromis spp. complex is one of the most difficult and despite the particular 

attention given to this species on board, specimen of an other larger species might have been 

included, which would explain the significant occurrence of immature specimens above the 

mean size at maturity. 



Nyassachromis 'argyrosoma' in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

10 m 48.9 74.1 47.3 57.6 16.5 

30 m 51.1 25.9 45.8 30 31.7 

50 m 0 0 6.9 12.4 51.8 



A positive correlation was found between oocyte weight and body weight (Figure 28-7). This 

is the only species for which a relationship between oocyte weight and body weight was 

found and it might be due to identification inaccuracies. The GSI threshold above which the 

oocyte weight did no longer increase significantly was 3.2% (Figure 28-8). The mean oocyte 

weight was 7.66 mg (± 2.36 SD, N= 16). The standard deviation was very high for such a low 

mean oocyte weight, which was probably due to the fact that oocyte weight was correlated 

with body weight. 

124

20 

Frequency % 

15 

10 

5 

0 

77 82 87 92 97 102 107 112 117 122 127 132 


Figure 29-1. Size range and frequencies of Otopharynx 'productus' caught in the SWA 


25 

20 

Weight (g) 

15 

10 

5 

y = 0,0016x 2,0084 

R 2 = 0,838 

0 

70 80 90 100 110 

Length (mm) 

Figure 29-2. Length-weight relationship for Otopharynx 'productus' females in the SWA. (R² 


Fecundity 

70 

60 

50 

40 

30 

20 

10 

0 

y = 2,8372x - 7,8469 

R 2 = 0,8262 

5 10 15 20 25 

Weight (g) 

Figure 29-3. Fecundity-weight relationship for Otopharynx 'productus' females in the SWA. 


125

Frequency % 

16 

14 

12 

10 

8 

6 

4 

2 

0 

55 65 75 85 95 105 115 125 135 145 155 165 175 185 195 205 215 225 235 245 255 


Figure 30-1. Size range and frequencies of Otopharynx speciosus caught in the SWA between 


Weight (g) 

400 

350 

300 

250 

200 

150 

100 

50 

0 

y = 1E-05x 3,1606 

R 2 = 0,9954 

120 140 160 180 200 220 240 

Length (mm) 

Figure 30-2. Length-weight relationship for Otopharynx speciosus females in the SWA. (R² = 


Fecundity 

350 

300 

250 

200 

150 

100 

50 

0 

y = 1,0452x - 25,644 

R 2 = 0,9918 

0 100 200 300 400 

Weight (g) 

Figure 30-3. Fecundity-weight relationship for Otopharynx speciosus females in the SWA. 


126

Otopharynx spp. 

Otopharynx 'productus' 

O. 'productus' is a rare species encountered at 10 and 30 m. 27 females and 20 males 

were analysed, giving a sex ratio of F/M 0.6/0.4. Specimens caught ranged between 75 and 

134 mm (Figure 29-1). Among the rare species, O. 'productus' was one for which we caught 

some ripe females, allowing an estimation of the length-weight (Figure 29-2) and fecundityweight 

relationships (Figure 29-3). Fecundity ranged between 22 and 61 for females weighing 

between 9 and 22 g. The mean oocyte weight was impossible to assess from the few data 

available, but the largest oocytes weighed averaged 14.2 mg for a GSI of 2.8%. 

Otopharynx speciosus (Trewavas) 

Like O. 'productus', O. speciosus is not an abundant species and was essentially caught 

between 30 and 75 m. 61 females and 38 males were analysed, giving a sex ratio of F/M 

0.6/0.4. Specimens caught ranged between 55 and 260 mm (Figure 30-1). Among the rare 

species, O. 'productus' was also one for which we caught some ripe females, allowing an 

estimation of the length-weight (Figure 30-2) and fecundity-weight relationships (Figure 30- 

3). Fecundity ranged between 51 and 322 for females weighing between 85 and 337 g. The 

mean oocyte weight was impossible to assess from the few data available, but the largest 

oocytes weighed averaged 25 mg for a GSI of 2.9%. 

127

Placidochromis spp. 

Placidochromis 'long' 

111 females and 180 males were analysed. P. 'long' is a small species caught from 10 

to 50 m, with an increasing occurrence and biomass with depth (Figure 31-1). 


0 1 2 3 4 5 

Depth (m) 

10 

30 

Weight 

Number 

50 

Figure 31-1. Mean occurrence and abundance in the catches per depth of Placidochromis 

'long' in the SWA between July 1998 and May 1999. 

It was most common at 50 m, where it constituted 1.6 and 4.3% of the catches in weight and 

number, respectively. The mean CPUE per depth category was 3.2 kg for the shallows. No 

reference to this species (under this name at least) was made in Tómasson & Banda (1996). 



Owing to low sample size or absence of data for some months, the precise 

determination of breeding season and size at maturity was not possible. Ripe females (Figure 

31-3a) and males (Figure 31-3b) were found only in October and in April-May and breeding 

season might occur from April-May to October. 


31-5, respectively. Fecundity ranged from 17 to 38 for females weighing between 4.7 and 6.7 

g. No relationship was found between oocyte weight and body weight. The GSI threshold 

above which the oocyte weight no longer increased significantly was estimated at 2% (Figure 

31-6). The mean oocyte weight was 4.36 mg (± 1.30 SD, N= 10). 

128

Frequency % 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 


Figure 31-2. Size range and frequencies of Placidochromis 'long' caught in the SWA between 


30 


a 

25 

20 

15 

10 

5 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

10 13 24 8 3 8 45 


100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Oct-98 

Nov-98 

Dec-98 

b 14 25 36 11 7 4 1 82 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 


males (b) Placidochromis 'long' in the SWA The values below the x-axis are the effective 


129

Weight (g) 

8 

7 

6 

5 

4 

3 

2 

1 

0 

y = 0,0019x 1,9262 

R 2 = 0,6995 

56 58 60 62 64 66 68 

Length (mm) 

Figure 31-4. Length-weight relationship for Placidochromis 'long' females in the SWA. (R² = 


Fecundity 

40 

35 

30 

25 

20 

15 

10 

5 

0 

y = 6,8445x - 12,583 

R 2 = 0,3545 

4 4,5 5 5,5 6 6,5 7 

Weight (g) 

Figure 31-5. Fecundity-weight relationship for Placidochromis 'long' females in the SWA. 


7 

6 


5 

4 

3 

2 

1 

y = 0,9679x + 2,0288 

R 2 = 0,0943 

0 

1,5 2,0 2,5 3,0 3,5 

GSI (%) 


Placidochromis 'long'. Oocytes from females whose GSI was below (in grey) and above 


130


0 2 4 6 

Depth (m) 

75 

100 

Weight 

Number 

125 

Figure 32-1. Mean occurrence and abundance in the catches per depth of Placidochromis 

'platyrhynchos' in the SWA between July 1998 and May 1999. 

30 

25 

Frequency % 

20 

15 

10 

5 

0 

47 52 57 62 67 72 77 82 87 92 97 102 107 112 


Figure 32-2. Size range and frequencies of Placidochromis 'platyrhynchos' caught in the 

SWA between July 1998 and May 1999. 

131

Placidochromis 'platyrhynchos' 

263 females and 225 males were analysed. P. 'platyrhynchos' is a relatively small 

species caught from 75 to 125 m, with an increasing occurrence and biomass with depth 

(Figure 32-1). It was a very common species at 125 m, where it constituted 3.5 and 4.7% of 

the catches in weight and number, respectively. The mean CPUE per depth category was 1.7 

kg for the deep zone and 6.2 kg for the very deep zone, which matched with the value 

reported by Tómasson & Banda (1996) for the deep zone (1.4 kg) and was more than twice as 

much for the very deep zone (2.8 kg). Specimens caught ranged between 46 and 115 mm 


Ripe females were found from July to August and from December to May (Figure 32- 

3a). Considering the very low sample size for June, it can be estimated than breeding season 

occurred from December to October, with a peak from January to May and a cessation in 

November. This pattern was confirmed by males data, excluding the months with very low 

sample size such as August and November (Figure 32-3b). Nearly all the ripe females and 

males and the immature individuals were caught at 125 m, suggesting that spawning could 

take place at this depth (Table 32-1). The mean size at maturity was about 80 mm (Figure 32- 

4), which was probably overestimated given that this estimation was with all the data 

available to increase the sample size, including data outside the breeding season. This 




Placidochromis 'platyrhynchos' in the SWA. 

Depth 

Non ripe 

females 

Ripe females 

Males not in 



colour 

Immature 

specimens 

75 m 0 0 1.2 0 0 

100 m 23.4 7.5 28.2 1.4 16.7 

125 m 76.6 92.5 70.6 98.6 83.3 




which the oocyte weight did no longer increase significantly was 3.5% (Figure 32-7). The 


132


a 


b 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Jun-98 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 


males (b) Placidochromis 'platyrhynchos' in the SWA The values below the x-axis are the 


month. 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

4 39 16 30 13 22 12 27 23 18 30 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

26 1 49 1 17 21 17 31 15 47 

100 


50 

0 

52 57 62 67 72 77 82 87 92 97 102 107 112 



for Placidochromis 'platyrhynchos' in the SWA. 

133

Weight (g) 

35 

30 

25 

20 

15 

10 

5 

0 

y = 4E-05x 2,8938 

R 2 = 0,8149 

60 70 80 90 100 110 120 

Length (mm) 

Figure 32-5. Length-weight relationship for Placidochromis 'platyrhynchos' females in the 


Fecundity 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 1,8798x + 14,85 

R 2 = 0,2775 

5 10 15 20 25 30 35 

Weight (g) 

Figure 32-6. Fecundity-weight relationship for Placidochromis 'platyrhynchos' females in the 


25 


20 

15 

10 

5 

y = 0,8026x + 10,028 

R 2 = 0,1582 

0 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 

GSI (%) 


Placidochromis 'platyrhynchos'. Oocytes from females whose GSI was below (in grey) and 


134


0 1 2 3 4 5 

Depth (m) 

10 

30 

Weight 

Number 

Figure 33-1. Mean occurrence and abundance in the catches per depth of Pseudotropheus 

livingstonii in the SWA between July 1998 and May 1999. 

25 

20 

Frequency % 

15 

10 

5 

0 

37 39 41 43 45 47 49 51 53 55 57 59 61 63 


Figure 33-2. Size range and frequencies of Pseudotropheus livingstonii caught in the SWA 


135

Pseudotropheus spp. 

Pseudotropheus livingstonii (Boulenger) 

68 females and 124 males were analysed. Ps. livingstonii is a small species principally 

caught at 10 and sometimes at 30 m (Figure 33-1). It constituted 1.4 and 4% of the catches in 

weight and number, respectively. The mean CPUE per depth category was 1.8 kg for the 

shallow zone, which was about five times less than the value reported by Tómasson & Banda 

(1996) (9.8 kg). Specimens caught ranged between 37 and 63 mm (Figure 33-2). The sex ratio 

observed over the full sampling period was F/M 0.4/0.6. 

As Ps. Livingstonii was sometimes relatively abundant and sometimes absent from the 

catches for a few months, precise determination of the breeding season was impossible. Each 

time it was present in the catches (October, November, March, April, May), ripe females were 

found (Figure 33-3a), which was not always the case for males in breeding colour (Figure 33- 

3b). Maturity was reached at about 37 mm for females (Figure 33-4). 




which the oocyte weight did no longer increase significantly was estimated at about 2% 

(Figure 33-7). The mean oocyte weight was 5.45 mg (± 0.59 SD, N= 7). 

136

60 


50 

40 

30 

20 

10 

a 


b 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0 

Oct-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 


males (b) Pseudotropheus livingstonii in the SWA The values below the x-axis are the 


month. 

Feb-99 

Mar-99 

Apr-99 

May-99 

11 4 20 9 24 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

22 7 1 1 22 17 54 

100 


50 

0 

32 37 42 47 52 57 62 



for Pseudotropheus livingstonii in the SWA. 

137

Weight (g) 

6 

5 

4 

3 

2 

1 

0 

y = 7E-05x 2,8076 

R 2 = 0,4355 

45 46 47 48 49 50 51 52 53 

Length (mm) 

Figure 33-5. Length-weight relationship for Pseudotropheus livingstonii females in the SWA. 


Fecundity 

35 

30 

25 

20 

15 

10 

5 

0 

y = 6,4275x + 1,2343 

R 2 = 0,3916 

3 3,5 4 4,5 5 

Weight (g) 

Figure 33-6. Fecundity-weight relationship for Pseudotropheus livingstonii females in the 


7 

6 


5 

4 

3 

2 

y = 0,0035x + 5,4322 

1 

R 2 = 6E-05 

0 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 

GSI (%) 


Pseudotropheus livingstonii. Oocytes from females whose GSI was below (in grey) and 


138

Weight (g) 

35 

30 

25 

20 

15 

10 

y = 0,0064x 1,782 

5 

R 2 = 0,6384 

0 

80 85 90 95 100 105 110 115 

Length (mm) 

Figure 34-3. Length-weight relationship for Sciaenochromis ahli females in the SWA. (R² = 


Fecundity 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 4,4172x - 55,843 

R 2 = 0,6611 

15 20 25 30 35 

Weight (g) 

Figure 34-4. Fecundity-weight relationship for Sciaenochromis ahli females in the SWA. (R² 


25 

Oocytes weight (mg) 

20 

15 

10 

5 

0 

0,0 1,0 2,0 3,0 4,0 5,0 

GSI (%) 


Sciaenochromis ahli. Oocytes from females whose GSI was below (in grey) and above (in 

black with regression) 3.5%. (R² = determination coefficient). 

139

Sciaenochromis spp. 

Sciaenochromis ahli (Trewavas) 

38 females and 37 males were analysed. S. ahli is a relatively small species 

encountered at every depth from 10 to 125 m but never numerous in any of them. The mean 

CPUE per depth category was 0.9, 0.8 and 0.3 kg for the shallow, deep and very deep zones, 

respectively. No reference to this species (under this name at least) was made in Tómasson & 

Banda (1996). Specimens caught ranged between 62 and 124 mm (Figure 34-1). 

30 

25 

Frequency % 

20 

15 

10 

5 

0 

62 67 72 77 82 87 92 97 102 107 112 117 122 


Figure 34-1. Size range and frequencies of Sciaenochromis ahli caught in the SWA between 




determination of breeding season and size at maturity was not possible. Ripe females were 

found at each sampling date except when sample size were almost null (March and April) 

(Figure 34-2). Only three males in breeding colour were found, in November and May, all at 

75 m. Breeding season could then occur at least from October to May. 


100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

11 10 5 4 1 1 1 5 


Sciaenochromis ahli in the SWA The values below the x-axis are the sample size for each 

month. 

140


0,0 0,5 1,0 1,5 

10 

Depth (m) 

30 

50 

75 

100 

125 

Weight 

Number 

Figure 35-1. Mean occurrence and abundance in the catches per depth of Sciaenochromis 

benthicola in the SWA between July 1998 and May 1999. 

30 

25 

Frequency % 

20 

15 

10 

5 

0 

45 55 65 75 85 95 105 115 125 135 145 155 165 


Figure 35-2. Size range and frequencies of Sciaenochromis benthicola caught in the SWA 


141




which the oocyte weight did no longer increase significantly was impossible to assess because 

of low sample size. However, despite the fact that oocyte weight apparently still increased 

with GSI above 3.5% (Figure 34-5), the mean oocyte weight was estimated from the three 

females whose GSI was above 3.5%, and was 19.11 mg (± 1.25 SD, N= 3). 

Sciaenochromis benthicola (Konings) 

188 females and 182 males were analysed. S. benthicola was encountered at every 

depth from 10 to 125 m but was more frequent at 50 and 75 m, where it constituted 1 to 1.3% 

of the catches in weight and 0.4 to 1.4% in number, respectively (Figure 35-1). The mean 

CPUE per depth category was 5.6, 2.4 kg and almost nothing for the shallow, deep and very 

deep zones, respectively. No reference to this species (under this name at least) was made in 

Tómasson & Banda (1996). Specimens caught ranged between 48 and 168 mm (Figure 35-2). 



determination of breeding season was not possible. Ripe females were found in June, August 

and from November to February, with peaks in August and December (Figure 35-3a). Data 

for males gave approximately the same pattern except for the period from June to August, 

when no ripe males were found (Figure 35-3b). 100% of the ripe females and 91% of the 

males in breeding colour were found at 50 and 75 m, suggesting that spawning could occur at 

these depths. The mean size at maturity was about 100 mm for females (Figure 35-4). Size at 

maturity has to be determined at the height of the breeding season to be accurate, but owing to 

the low sample size we had to consider every data available. As a consequence, females 

caught outside the peak of breeding season were included in the analyses, even large resting 

females, which explains the shape of the upper part of the curve. 




which the oocyte weight no longer increased significantly was 3% (Figure 35-7). The mean 


142


a 


b 

60 

50 

40 

30 

20 

10 

0 

30 

25 

20 

15 

10 

5 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

17 6 17 22 26 18 25 49 9 6 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

11 14 5 28 14 24 28 30 5 8 15 


males (b) Sciaenochromis benthicola in the SWA The values below the x-axis are the 


month. 

100 


50 

0 

82 87 92 97 102 107 112 117 122 127 132 137 142 147 152 157 



for Sciaenochromis benthicola in the SWA. 

143

Weight (g) 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 2E-05x 3,0263 

R 2 = 0,9874 

0 50 100 150 200 

Length (mm) 

Figure 35-5. Length-weight relationship for Sciaenochromis benthicola females in the SWA. 


120 

100 

Fecundity 

80 

60 

40 

20 

y = 1,264x - 0,3265 

R 2 = 0,5579 

0 

20 30 40 50 60 70 80 

Weight (g) 

Figure 35-6. Fecundity-weight relationship for Sciaenochromis benthicola females in the 



35 

30 

25 

20 

15 

10 

5 

0 

y = 1,8563x + 19,776 

R 2 = 0,1418 

0,0 1,0 2,0 3,0 4,0 5,0 6,0 

GSI (%) 


Sciaenochromis benthicola. Oocytes from females whose GSI was below (in grey) and 


144

Stigmatochromis spp. 

Stigmatochromis 'guttatus' 

62 females and 22 males were analysed. S. 'guttatus' was encountered at every depth 

from 10 to 125 m but was never abundant in any of them although it occurred more frequently 

at 50 and 75 m. Specimens caught ranged between 64 and 147 mm (Figure 36-1). The sex 

ratio observed over the full sampling period was F/M 0.7/0.3. 


determination of breeding season was not possible. Ripe females were found from October to 

December, in February and in April-May (Figure 36-2). The six males in breeding colour we 

found were caught at 50 m in March. The mean size at maturity was about 100 mm for 

females (Figure 36-3), which was probably overestimated as this estimation was not done 

during the height of the breeding season but with all the data available for females, to increase 

the sample size. 






145

Frequency % 

30 

25 

20 

15 

10 

5 

0 

67 72 77 82 87 92 97 102 107 112 117 122 127 132 


Figure 36-1. Size range and frequencies of Stigmatochromis 'guttatus' caught in the SWA 



100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 


Stigmatochromis 'guttatus' in the SWA The values below the x-axis are the effective 


Feb-99 

Mar-99 

Apr-99 

12 5 5 2 15 14 8 1 

May-99 

100 


50 

0 

67 72 77 82 87 92 97 102 107 112 117 122 127 132 



for Stigmatochromis 'guttatus' in the SWA. 

146

Weight (g) 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

y = 0,0001x 2,6693 

R 2 = 0,8511 

90 100 110 120 130 

Length (mm) 

Figure 36-4. Length-weight relationship for Stigmatochromis 'guttatus' females in the SWA. 


70 

60 

50 

Fecundity 

40 

30 

20 

10 

y = 1,4769x - 8,0045 

R 2 = 0,8872 

0 

10 20 30 40 50 

Weight (g) 

Figure 36-5. Fecundity-weight relationship for Stigmatochromis 'guttatus' females in the 


35 

30 


25 

20 

15 

10 

5 

y = 2,6415x + 18,855 

R 2 = 0,1457 

0 

0,0 1,0 2,0 3,0 4,0 5,0 

GSI (%) 


Stigmatochromis 'guttatus'. Oocytes from females whose GSI was below (in grey) and 


147

Taeniolethrinops spp. 

Taeniolethrinops furcicauda (Trewavas) (Plate 17) 

135 females and 88 males were analysed. T. furcicauda was mostly encountered at 10 

m where it made up to 1.2 and 1.6% of the catches in number and weight, respectively. It was 

also found sometimes at 30 m. Specimens caught ranged between 65 and 178 mm (Figure 37- 

1). 

50 

Frequency % 

40 

30 

20 

10 

0 

65 75 85 95 105 115 125 135 145 155 165 175 


Figure 37-1. Size range and frequencies of Taeniolethrinops furcicauda caught in the SWA 




determination of breeding season was not possible. Ripe females were only found in July 

(Figure 37-2), as was the single male in breeding colour caught. The mean size at maturity 

was about 130 mm for females (Figure 37-3), which was probably overestimated as this 

estimation was not done during the height of the breeding season but with all the data 

available for females, to increase the sample size. 

The length-weight relationship is not given because of the too narrow size range of 

females measured. The fecundity-weight relationship is presented in Figure 37-4. Fecundity 

ranged from 138 to 219 for females weighing between 88 and 109 g and was not correlated to 

body weight. No relation was found between oocyte weight and body weight. The GSI 

threshold above which the oocyte weight did no longer increase significantly was difficult to 

assess owing to low sample size. However, as the oocyte weight did no longer increase above 

a GSI of 1.5% (Figure 37-5), the mean oocyte weight was estimated (probably 

underestimated) at 10.13 mg (± 0.83 SD, N= 3). 

148

Plate 17. Taeniolethrinops furcicauda (by Dave Voorvelt). 


90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

7 9 1 2 14 11 27 64 


Taeniolethrinops furcicauda in the SWA The values below the x-axis are the effective for 

each month. 

149

100 


50 

0 

70 80 90 100 110 120 130 140 150 160 170 



for Taeniolethrinops furcicauda in the SWA. 

250 

200 

Fecundity 

150 

100 

50 

y = 0,5165x + 135,84 

R 2 = 0,0209 

0 

80 85 90 95 100 105 110 115 

Weight (g) 

Figure 37-4. Fecundity-weight relationship for Taeniolethrinops furcicauda females in the 


12 


10 

8 

6 

4 

2 

y = -0,0077x + 10,148 

R 2 = 8E-06 

0 

0,0 0,5 1,0 1,5 2,0 2,5 

GSI (%) 


Taeniolethrinops furcicauda. Oocytes from females whose GSI was below (in grey) and 


150

Taeniolethrinops praeorbitalis (Regan) (Plate 18) 

85 females and 78 males were analysed. T. praeorbitalis was encountered at 10 m and 

30 m, but was never frequent at any depth. Specimens caught ranged between 71 and 199 mm 


25 

Frequency % 

20 

15 

10 

5 

0 

75 85 95 105 115 125 135 145 155 165 175 185 195 


Figure 38-1. Size range and frequencies of Taeniolethrinops praeorbitalis caught in the SWA 




determination of breeding season was not possible. Ripe females were only found in August, 

and no male in breeding colour was ever caught. The mean size at maturity was about 155 

mm for females (Figure 38-2), which was probably overestimated as this estimation was not 

done during the height of the breeding season but with all the data available for females, to 

increase the sample size. The first inflexion of the lower part of the curve came from the 

August data and it is likely that the mean size at maturity would rather be around 130 mm. 


38-4, respectively. Fecundity ranged from 193 to 250 for females weighing between 135 and 


above which the oocyte weight did no longer increase significantly was estimated from the 

few data available at about 3% (Figure 38-5) and the mean oocyte weight was 25.73 mg (± 

2.02 SD, N= 3). 

151

Plate 18. Taeniolethrinops praeorbitalis (by Dave Voorvelt). 

100 


50 

0 

85 95 105 115 125 135 145 155 165 175 185 195 



for Taeniolethrinops praeorbitalis in the SWA. 

152

250 

200 

Weight (g) 

150 

100 

50 

y = 2E-05x 3,0304 

R 2 = 0,9549 

0 

60 80 100 120 140 160 180 200 

Length (mm) 

Figure 38-3. Length-weight relationship for Taeniolethrinops praeorbitalis females in the 


Fecundity 

280 

260 

240 

220 

200 

180 

160 

140 

120 

100 

y = 0,9672x + 74,587 

R 2 = 0,8658 

130 140 150 160 170 180 190 

Weight (g) 

Figure 38-4. Fecundity-weight relationship for Taeniolethrinops praeorbitalis females in the 


30 


25 

20 

15 

10 

y = 0,5292x + 23,623 

5 

R 2 = 0,0304 

0 

2,0 2,5 3,0 3,5 4,0 4,5 5,0 

GSI (%) 


Taeniolethrinops praeorbitalis. Oocytes from females whose GSI was below (in grey) and 


153


0 5 10 15 

Depth (m) 

30 

50 

Weight 

Number 

Figure 39-1. Mean occurrence and abundance in the catches per depth of Trematocranus 

brevirostris in the SWA between July 1998 and May 1999. 

Frequency % 

35 

30 

25 

20 

15 

10 

5 

0 

42 47 52 57 62 67 72 77 82 


Figure 39-2. Size range and frequencies of Trematocranus brevirostris caught in the SWA 


154

Trematocranus spp. 

Trematocranus brevirostris (Trewavas) 

162 females and 249 males were analysed. T. brevirostris is a small species 

encountered at 30 and 50 m but mostly present at 50 m where it made up to 6.3 and 11.1% of 

the catches in weight and number, respectively (Figure 39-1). The mean CPUE per depth 

category was 9.9 kg for the shallows. No reference to this species (under this name at least) 

was made in Tómasson & Banda (1996). Specimens caught ranged between 40 and 85 mm 



determination of breeding season was not possible. The monthly progression of ripe females 

(Figure 39-3a) and males (Figure 39-3b) were very similar, indicating a breeding activity 

from January to May, and in October, with a peak in March-April. According to the 

percentages of ripe females and males in May and October, it is likely that some breeding also 

activity occur between May and October. As almost all the specimens of T. brevirostris were 

caught at 50 m, spawning probably takes place at this depth. Maturity was reached in their 

first year at 11 months old at a mean size of 50 mm for females (Figure 39-4). Size at maturity 

has to be determined at the height of the breeding season to be accurate, but owing to the low 

sample size we had to consider every data available. As a consequence, females caught 

outside the peak of breeding season were included in the analyse, even large resting females, 

which explain the shape of the upper part of the curve. 


39-6, respectively. Fecundity ranged from 13 to 47 for females weighing between 4 and 13 g 

and was not correlated to body weight. No relation was found between oocyte weight and 


significantly was estimated from the few data available at about 3% (Figure 39-7) and the 


155


50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

a 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

25 8 13 16 38 53 9 


b 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

58 22 6 8 22 52 65 16 


males (b) Trematocranus brevirostris in the SWA The values below the x-axis are the 


month. 

100 


50 

0 

42 47 52 57 62 67 72 77 



for Trematocranus brevirostris in the SWA. 

156

Weight (g) 

14 

12 

10 

8 

6 

4 

2 

0 

y = 7E-05x 2,7752 

R 2 = 0,881 

50 55 60 65 70 75 80 85 

Length (mm) 

Figure 39-5. Length-weight relationship for Trematocranus brevirostris females in the SWA. 


Fecundity 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

y = 1,9389x + 17,924 

R 2 = 0,1812 

2 4 6 8 10 12 14 

Weight (g) 

Figure 39-6. Fecundity-weight relationship for Trematocranus brevirostris females in the 



10 

9 

8 

7 

6 

5 

4 

3 

2 

1 

0 

y = 0,5913x + 4,362 

R 2 = 0,1534 

0,0 1,0 2,0 3,0 4,0 5,0 

GSI (%) 


Trematocranus brevirostris. Oocytes from females whose GSI was below (in grey) and 


157

Trematocranus placodon (Regan) (Plate 19) 

34 females and 17 males were analysed. T. placodon is a common but not abundant 

species only encountered at 10 m. The mean CPUE per depth category was about 2 kg for the 

shallows, which was about three times less than the value reported (5.8 kg) by Tómasson & 

Banda (1996). Specimens caught ranged between 90 and 160 mm (Figure 40-1). 

Frequency % 

35 

30 

25 

20 

15 

10 

5 

0 

95 105 115 125 135 145 155 


Figure 40-1. Size range and frequencies of Trematocranus placodon caught in the SWA 




determination of breeding season was not possible. Ripe females were found from June too 

August and in January-February (Figure 40-2). Only three males in breeding colour were 

caught, in July 98. The mean size at maturity was about 105 mm for females (Figure 40-3), 

which was probably overestimated as this estimation was not done during the height of the 

breeding season but with all the data available for females, to increase the sample size. 




which the oocyte weight did no longer increase significantly was estimated from the few data 

available at about 2.5% (Figure 40-6) and the mean oocyte weight was 12.83 mg (± 1.15 SD, 

N= 5). 

158

Plate 19. Trematocranus placodon (by Dave Voorvelt). 


100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Jun-98 

Jul-98 

Aug-98 

Sep-98 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

4 6 4 4 2 14 


Trematocranus placodon in the SWA The values below the x-axis are the effective for 

each month. 

100 


50 

0 

85 95 105 115 125 135 145 



for Trematocranus placodon in the SWA. 

159

Weight (g) 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

y = 5E-05x 2,906 

R 2 = 0,8417 

100 110 120 130 140 150 

Length (mm) 

Figure 40-4. Length-weight relationship for Trematocranus placodon females in the SWA. 


Fecundity 

200 

180 

160 

140 

120 

100 

80 

60 

40 

20 

0 

y = 1,2303x + 50,512 

R 2 = 0,5123 

40 50 60 70 80 90 100 

Weight (g) 

Figure 40-5. Fecundity-weight relationship for Trematocranus placodon females in the SWA. 



16 

14 

12 

10 

8 

6 

4 

2 

y = 0,4948x + 11,361 

R 2 = 0,048 

0 

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 

GSI (%) 


Trematocranus placodon. Oocytes from females whose GSI was below (in grey) and 


160

Table 41. Peak of breeding activity of species per depth category. X = data supported by good 

sample size, x = data supported by low sample size. "Intermediate" means common in both 

deep and shallow zone. 

1998 1999 

Depth 

category Species Jun Jul Aug Oct Nov Dec Jan Feb Mar Apr May 

Deep 

water 

(75- 

125m) 

A. 'geoffreyi' X X X X 

A. macrocleithrum x X X x 

A. mentale X X X 

A. pectinatum X X 

Au. minutus X X 

D. apogon X X X X 

D. macrops X X X 

P. tokolosh X X X 

L. 'deep water altus' X X X X X 

L. gossei X X X X 

L. 'oliveri' X X X 

L. polli X 

Pl. 'platyrhynchos' x X X X X X 

Intermed 

-iate 

D. limnothrissa X X X 

C. quadrimaculatus x 

C. virginalis X X X 

L. longimanus x 

Sc. ahli X X 

Sc. Benthicola X X 

Shallow 

water 

(10-50m) 

Au. 'blue orange' X X X 

L. argenteus X X X X 

M. anaphyrmus X X X X 

N. 'argyrosoma' X X 

Pl. 'long' 

x 

T. furcicauda x 

T. praeorbitalis x 

Tr. brevirostris X X 

161

Discussion 

Depth distribution and mean CPUE for the major species were compared with those 

found by Tómasson & Banda (1996) in the SWA. Several of the species we regularly caught 

were absent of Tómasson & Banda (1996) survey of the SWA, and the other way round. This 

is likely due to identification differences between the two studies, some species being either 

lumped together under their generic names or accurately identified. Also, the smaller area 

covered might explain the absence of some species in our survey. However, global trends in 

CPUE were similar for both studies and the observed differences may lies in the fact that 

Tómasson & Banda covered the entire SWA whereas our work focused on the northern part 

of it. 

Breeding species were found throughout the year at any depth. Some species were 

capable of breeding throughout the year (as already reported by Marsh et al. 1986, Lewis & 

Tweddle 1990, Thompson et al. 1996) with marked seasonal variations in the proportions of 

ripe individuals, whereas others appeared to have distinct breeding season (as the 

Oreochromis spp., Lowe-McConnell 1987 or members of the Utaka group, Iles 1960, 1971, 

Jackson et al. 1963), generally over 6 months long. Again, these patterns were independent 

from the depth distribution of the species. An interesting hypothesis linking the length of the 

breeding season to the fecundity of the species was proposed by Marsh et al. (1986): the lower 

the fecundity the longer the breeding season. If this hypothesis seemed to fit for the mbuna, it 

does however not hold for the demersal cichlids. Indeed, species such as Lethrinops 

argenteus, or L. oliveri, which had continuous breeding season also presented amongst the 

highest relative fecundity (2162 eggs.kg -1 , 4931 eggs.kg -1 , respectively) whereas species such 

as Diplotaxodon apogon with definite breeding season had amongst the lowest relative 

fecundity (632 eggs.kg -1 ). Apart from a single species, Copadichromis virginalis, whose 

peaks of breeding activity were clearly related to plankton abundance, the breeding cycle of 

the other species was not directly related to any of the monitored environmental factors 

(temperature, oxygen concentration, photoperiod, algal abundance, conductivity). No 

particular breeding distribution pattern was found among the several species studied, when 

they were grouped per trophic category, per genera or per depth category. Even when 

focusing on the peaks of breeding activity per depth category (Table 41), no marked structure 

appeared. The peaks of breeding activity were evenly distributed over the year for the shallow 

water species. The only slight remarkable trend was that the highest concentration of breeding 

peaks occurred from January to May for the deep water species whereas no peak was recorded 

from January to March for the species with "intermediate" depth distribution. The period from 

December to April is the time of the year when the lake is most strongly stratified, the 

thermocline being around 40-50 m (Eccles 1974, see Figure 11 previous chapter). It also 

corresponds to the period of lowest phytoplankton and zooplankton abundance (Irvine 1995). 

However, most of the studied species are supposed to be benthic invertebrates feeders and the 

observed breeding patterns might be related to the relative abundance of their main food 

sources, as seems to be the rule for mbuna (Marsh et al. 1986). The seasonal progression of 

benthic invertebrate abundance and biomass, which is currently under study in the European 

Union Project: "The trophic ecology of the demersal fish community of Lake Malawi/Niasa", 

might shed a new light on the determinism of the observed breeding patterns. On the other 

hand, the reason for this apparent time sharing could lie in Fryer & Iles (1972) hypothesis of 

self-controlled population densities (population homeostasis) achieved through sharing of 

spawning territories. Lowe-McConnell (1979) also suggested that for tropical fishes 

asynchronous breeding could be a good way of maximising resource use for species with 

similar requirements. However, given the good correspondence between food availability and 

162

Table 42. Ratio of length at maturity (L 50 ) to maximum observed length (MOL) for several 

cichlids species caught by trawling in the South West Arm of Lake Malawi between June 

1998 and May 1999. The lengths are standard lengths. In italic are values for which 

determination of L 50 was uncertain owing to low effective. 

Species MOL (mm) L 50 (mm) Ratio 

A. 'geoffreyi' 165 90 0.55 

A. macrocleithrum 148 100 0.68 

A. mentale 245 160 0.65 

A. pectinatum 140 70 0.5 

Au. 'blue orange' 78 48 0.62 

Au. 'cf macrochir' 134 100 0.75 

Au. 'minutus' 72 42 0.58 

Au. 'rostratum deep' 140 75 0.54 

B. lepturus 326 160 0.49 

B. nototaenia 300 115 0.38 

C. quadrimaculatus 150 100 0.67 

C. virginalis 123 75 0.61 

D. apogon 130 88 0.68 

D. argenteus 206 140 0.68 

D. limnothrissa 175 105 0.6 

D. macrops 135 98 0.73 

P. tokolosh 213 135 0.63 

L. argenteus 165 108 0.65 

L. 'deep water albus' 160 82 0.51 

L. 'deep water altus' 130 60 0.46 

L. gossei 170 92 0.54 

L. longimanus 147 107 0.73 

L. 'oliveri' 98 60 0.61 

L. polli 120 65 0.54 

M. anaphyrmus 164 105 0.64 

N. 'argyrosoma' 97 57 0.59 

Pl. 'platyrhynchos' 115 80 0.69 

Ps. livingstonii 63 37 0.59 

Sc. benthicola 168 100 0.60 

St. 'guttatus' 147 100 0.68 

T. furcicauda 178 130 0.73 

T. praeorbitalis 199 130 0.65 

Tr. brevirostris 85 50 0.59 

Tr. placodon 160 105 0.66 

163

eeding patterns for rock-dwelling species and C. virginalis, this is likely to be also the case 

for demersal species. For every species, more males than females in breeding condition were 

caught at any time of the year. This was also observed by Lewis & Tweddle (1990) on the 

three studied Lethrinops spp. Although not demonstrated for every species, the general trend 

of Malawi cichlids males to form breeding aggregations is likely to account for this 

difference, breeding males being more vulnerable to trawling when they form leks (Fryer 

1972, 1984). 

Among the species for which we determined the size at maturity, a few had already 

been studied: Diplotaxodon limnothrissa (Thompson et al. 1996), Lethrinops gossei (Lewis & 

Tweddle (1990), Mylochromis anaphyrmus (Tweddle & Turner 1977). The size at maturity 

we got for L. gossei and M. anaphyrmus were much smaller than those reported by Lewis & 

Tweddle and Tweddle & Turner. Size at maturity may vary among successive years for a 

same population under environmental variation (Duponchelle & Panfili 1998), or over longer 

time periods under fishing pressure (Stewart 1988, Lowe-McConnell 1982, Trewavas 1983). 

The observed differences in size at maturity could thus lie in the distant time period and 

geographical area between the studies. However, the method they both used overestimated the 

L 50 . The average size at first maturation (L 50 ) is defined as the length at which 50% of the 

females are mature during the breeding season. Tweddle & Turner and Lewis & Tweddle 

considered the "length at the percentage point which is one-half of the maximum percentage 

of mature fish found in any length group", that is the highest point on the sigmoid curve. But 

as they considered in fact ripe fish as being mature, and that there is always only a fraction of 

the population in a "ripe" state, they consistently overestimated the L 50 . That's the reason why 

their sigmoid never reached 100%. Iles (1971) stated that the ratio of length at maturity to 

asymptotic length for cichlids was characteristically a little above 0.7. The ratios found by 

Tweddle & Turner (1977) tended to confirm that mean value. However, as they overestimated 

the L 50 , the ratios they found were also overestimated. In our study, the asymptotic length was 

not calculated for all the species (see Chapter "Growth"), thus the maximum observed length, 

considered to be close to the asymptotic length, was used to calculate the ratio (Table 42). 

Ratios ranged between 0.38 and 0.75 with an averaged value of 0.61. As the L 50 for some 

species were probably overestimated due to low sample size (in italic in Table 42) and 

considering that the maximum observed length is usually smaller than the theoretic 

asymptotic length, the average ratio is likely to be a little below 0.6. In the case of Lethrinops 

'deep water altus', the ratio value of 0.46 might have been underestimated. Indeed, this low 

value might lie in misidentification of the largest specimens (up to 130 mm), as Turner (1996) 

reported a maximum size of 10 cm TL for this species. It is interesting to note that some 

genera present ratio values that tend to be close to 0.7 (e.g. Diplotaxodon spp., Copadichromis 

spp.) whereas others show values closer to 0.5 such as the Buccochromis and Lethrinops spp.. 

Despite the importance of this population parameter in fisheries management, age at 

maturity had only been estimated on a few species so far: Lethrinops longipinnis, L. 

parvidens, Mylochromis ("Haplochromis") anaphyrmus, Copadichromis ("Haplochromis") 

mloto (Tweddle & Turner 1977). These species were believed to reach maturity at the end of 

their third year except C. mloto, which matured at the end of its second year. M. anaphyrmus 

was the only common species between Tweddle & Turner study and ours. We found M. 

anaphyrmus was reaching maturity early in its second year at 18 months old. The growth 

parameters determined by Tweddle & Turner (1977), K=0.671 and LJ=196 mm TL, being 

very similar to ours: K=0.62 and LJ=180 mm SL (see Chapter "Growth"), the two fold 

difference in age at maturity observed between the two studies (which took place in the same 

area) lies mainly in the overestimation of size at maturity by the method used in Tweddle & 

Turner (1977). Assuming the same mean size at maturity of 160 mm TL (about 130 mm SL in 

our case), they would mature early in their third year (25 months) as well. It is very likely that 

164

this overestimation of age and size at maturity hold as well for the others species they studied. 

Growth and age at maturity were also estimated for C. ("Haplochromis") virginalis and C. 

("Haplochromis") quadrimaculatus by Iles (1971). They were both reaching maturity at three 

years old whereas we found they matured in their first and second year at 12 and 20 months 

for C. virginalis and C. quadrimaculatus, respectively. Despite slight differences in growth 

parameters, the differences between the two sets of estimations lies in the fact that 

determination of the size at maturity was based on ripe females in Iles's study (Iles 1971, 

Jackson et al. 1963), which leads to an overestimation of L 50 . Indeed, the size at maturity they 

determined represented 88 and 84 % of the maximum theoretical length (asymptotic length) 

for C. virginalis and C. quadrimaculatus, respectively. It is assumed that in unexploited 

stocks, which is certainly not the case for members of the Utaka group, the oldest fish reach 

only about 95% of their asymptotic length (review in Pauly 1980). This means that in Iles's 

case, they would breed for the first time close to their maximum size, which only few fish 

reach. Also, assuming the mean maturity length ratio Iles (1971) considered as the rule for 

cichlids (0.7), age at maturity would have been 1.6 and 2.0 years old for its populations of C. 

virginalis and C. quadrimaculatus, respectively, which would place them in the range of age 

at maturity we observed. The 22 species for which we determined A 50 reached maturity before 

the end of their second year or during their first year for the smallest species (Au. 'Blue 

orange', Au. 'minutus', N. 'argyrosoma'…). It is more conceivable that highly exploited 

species with such low fecundity as the Lake Malawi cichlids, mature early in their life 

(around one year old) rather than in the last part of their lives (more than three years old). 

Indeed, very few species grew older than four years (see Chapter "Growth", see also the 

abacus of longevity versus mean maximum length drawn out of 111 species of African 

freshwater fishes by de Merona 1983). Next, as only a very small fraction of a fish population 

reach the asymptotic length, which is usually slightly higher than the maximum observed 

length, this means that very few fish indeed would reach maturity and would therefore 

reproduce to ensure the replenishment of populations or the species survival. This clearly 

appears when looking at the proportion of fish reaching the size at maturity on the length 

frequency distributions presented in Iles (1971). Species with such a life history strategy 

would have unlikely survive about thirty years of intense exploitation by fisheries, which 

systematically remove the large specimens of fish populations (Turner 1977b, Turner 1995, 

Turner et al. 1995). 

It is a well known characteristic of Lake Malawi cichlids, they usually produce few but 

large eggs (see for review Fryer & Iles 1972, Konings 1995, Turner 1996). The forty species 

studied perfectly fitted into this trend. However, as already mentioned above, distinct 

reproductive strategies are distinguished among species and genera: species of the 

Diplotaxodon clade (sensu Turner et al. 1999) and Rhamphochromis spp. (pers. obs.) clearly 

presented the lowest relative fecundities and the largest eggs of all species. Although in a 

lesser extent, C. quadrimaculatus also presented low fecundity and large eggs. On the other 

hand, for a same body weight Lethrinops, Alticorpus or Mylochromis species produced much 

smaller but more numerous eggs. It appears that the species having a more pelagic life style 

such as the Diplotaxodon spp., Rhamphochromis spp. (Allison et al. 1996, Thompson et al. 

1995, 1996, Turner et al. 1999), or C. quadrimaculatus (Allison et al. 1996, Thompson et al. 

1995, 1996) are characterised by reduced fecundity, very large eggs and delayed maturity 

(Diplotaxodon and Copadichromis spp. had the highest Lmax to L 50 ratios). This reproductive 

strategy, which enable females to moothbrood young to very large sizes (observed for D. 

limnothrissa Turner 1994a, 1996, this study), is likely to be an adaptation to the pelagic 

environment. Egg size is an essential parameter of fish reproductive strategy for it conditions 

the ability to use a large array of food sources, to avoid predators and to survive unfavourable 

165

A C 

Plate 20. Gonad of a ripe Lethrinops longimanus female with at least three batches of oocytes. 

A = oocytes in late vitellogenesis that will be laid at the next spawn. B = second batch of 

developing vitellogenic oocytes (characteristic of a stage 3). C = third batch of oocytes in 

early development (characteristic of a stage 2). 

B 

166

environmental conditions (Bagenal 1969, 1978, Ware 1977, Mann & Mills 1979, Reznick, 

1982, Marsh 1986, Sargent et al. 1987, Wootton 1990, Cambray & Bruton 1994, Wootton 

1994). An inverse relationship links the egg size and the fecundity in most fish species (Mann 

& Mills 1979, Albaret 1982, Stearns 1983, Duarte & Alcaraz 1989, Elgar 1990) and in 

cichlids particularly (Peters 1963, De Silva 1986, Legendre & Ecoutin 1989, 1996, 

Duponchelle et al. 2000, this study). According to the energetic cost of gamete production, 

food is probably one of the most important environmental factors involved in the regulation of 

fecundity (Wooton & Evans 1976, Wooton 1979). Therefore, food availability being 

relatively poor in the pelagic zone, a logical strategy to improve both parents and offspring's 

fitness is to produce fewer eggs but of larger size and to moothbrood young to large size to 

enhance their survival. Indeed, the larger the fry when released by the female the more 

effective they will be in utilising the scarce food resources of the open waters. If this 

hypothesis is true, the cichlid reproductive strategy to adapt to pelagic environment would 

differ markedly from the strategy of most other pelagic fish families, which generally produce 

large numbers of small eggs (reviews by Duarte & Alcaraz 1989, Elgar 1990), including the 

freshwater pelagic cyprinids Engraulicypris sardella in Lake Malawi (Jackson et al. 1963) 

and Rastrineobola argentea in Lake Victoria (review in Witte & van Densen 1995), or the 

clupeids Stolothrissa tanganicae and Limnothrissa miodon in Lake Tanganyika (reviews by 

Coulter 1991, Marshall 1993, and Patterson & Makin 1998). 

As pointed out by Fryer & Iles (1972), to know how many eggs each female produces at each 

brood is important but how many times a year that number is produced is even more essential 

for fisheries management. Determining the number of times a female can spawn during a 

breeding season is a difficult task in natural environments, particularly for the demersal 

Malawi cichlids. Cichlids are known world-wide for their capacity to have multiple spawning 

events during the same breeding season when environmental conditions are favourable (Lowe 

1955, Trewavas 1983 for review, James & Bruton 1992). However, based on the observation 

of only one batch of eggs at a time in ovaries of ripe female cichlids in Lake Malawi, it was 

assumed that a same female was able to spawn only once per breeding season (Lowe 1955, 

Iles 1971, Tweddle and Turner 1977). Having analysed tens of thousands demersal cichlids 

pertaining to at least 170 species (Appendix 1) during the course of this study, three sets of 

evidence have emerged to suggest that annual multiple spawning is likely. The evidence is 

presented below: 

(1) - For most of the studied species it was observed but not recorded that more than one 

batch of eggs were in the ovaries of ripe females. An example is given on Plate 20, which 

presents the ovaries of a ripe Lethrinops longimanus female. A minimum of three batches of 

developing oocytes are clearly visible: the batch that will be released soon (A), a second batch 

that already undertook vitellogenesis (B), and a third one of smaller oocytes in early 

development (C). To verify the observation, and taking the opportunity of a fishing cruise 

with the RV Ndunduma in the SWA in July 1999, a special survey was organised in order to 

check the existence of more than one batch of oocyte in cichlids ripe gonads. All the females 

found with ripe gonads had two and sometimes three batches of oocytes. The sample included 

the following species: Alticorpus 'geoffreyi', A. macrocleithrum, A. mentale, A. pectinatum, 

Aulonocara 'blue orange', Copadichromis virginalis, Diplotaxodon greenwoodi, 

Hemitaeniochromis insignis, Lethrinops longimanus, L. gossei, L. 'oliveri', L. polli, 

Mylochromis anaphyrmus, Otopharynx speciosus, O. brooksi and Sciaenochromis 'guttatus'. 

(2) - The gonad maturity stage 6-3 (see Material and methods) was frequently observed for all 

the studied species. This stage is characteristic of post-spawning females initiating another 

cycle of vitellogenesis. 

167

(3) - Moothbrooding females of many species were often found with developing ovaries 

(stages 6-3 or 3). 

These three sets of evidence strongly suggest that each female probably spawns more than 

once during the breeding season, for haplochromine cichlids at least. Indeed, it would be very 

poor energetic strategy for a female to initiate another vitellogenesis cycle if the developed 

oocytes were not to be laid. In such a stable environment as Lake Malawi, it is very unlikely 

that environmental conditions become so unfavourable that a female would resorb her eggs 

during the few weeks needed to complete a vetillogenesis cycle (four to five weeks in 

cichlids, Fishelson 1966, Gauthier et al. 1996, Tacon et al. 1996). Furthermore, although 

extensive breeding season does not mean that individual fish breed continuously, most of the 

studied species have breeding seasons lasting more than six months, and several showed 

continuous breeding seasons. It appears very reasonable in these conditions that individual 

fish, known to breed repeatedly in aquarium, may then reproduce more than once, as do 

cichlids everywhere else (reviews by Fryer & Iles 1972, Trewavas 1983, Lowe-McConnell 

1987). 

Although this study presents the most comprehensive work on Lake Malawi cichlids 

life histories, it is based on a one year survey only. As inter annual variability of reproductive 

characteristics can be important in African cichlids (Duponchelle & Panfili 1998, 

Duponchelle et al. 1999, 2000), the described breeding patterns might not represent a 

permanent situation but rather a situation representative of the prevalent environmental 

conditions. However, Lake Malawi is a stable environment where important environmental 

perturbations are unlikely. The likeliest unexpected fluctuations are in resource availability, 

which might modify the intensity of breeding activity rather than its periodicity (Duponchelle 

et al. 1999). Furthermore, the breeding season observed for C. ("Haplochromis") virginalis 

and C. ("Haplochromis") quadrimaculatus were the same over a period of five years (Iles 

1971) and Lewis & Tweddle (1990) reported very similar trends in breeding seasonality 

among the three years of their study, which suggested little or no inter annual variability of 

breeding patterns. 

168

Chapter 3: 

Growth patterns of some of the most 

important demersal fish species 

caught by trawling in the South West 

Arm of Lake Malawi

Chapter 3: Growth patterns of some of the most important 

demersal fish species caught by trawling in the South West Arm 

of Lake Malawi 


Introduction 

Several methods exist to assess fish growth parameters, such as analysis of periodic 

marks on opercular bones, scales, vertebrae or otoliths, individual tagging or analysis of 

length frequency distributions (review by Casselman 1987, de Merona et al. 1988, Wootton 

1990, King 1995). Among these methods, analysis of the modal progression of length 

frequency distributions (Ricker 1975) has been commonly used for African freshwater fishes 

despite its potential biases in tropical conditions, where lack of seasonality, long spawning 

periods, non-year events giving rise to variations in growth and survival rates (hence to age 

and size modes) may lead to erroneously interpret size modes as differing in age by units of 

year. Indeed, tropical species often have extended breeding seasons during which multiple 

broods are produced and several cohorts (a cohort is a group of fish born at the same time) are 

likely to be encountered. In these conditions following year classes is often difficult and 

hamper precise interpretation of length progression series (Fryer & Iles 1972, Casselman 

1987, Lowe-McConnell 1987, de Merona et al. 1988, King 1995). As just seen in the previous 

chapter, most of the studied species have extended breeding seasons. However, when 

reasonably accurate information on the species biology is available, such as the breeding 

season and the maximum length, it is still possible to obtain correct estimates of growth using 

modal length progression analysis. For most of the species studied below, more than one 

cohort per year was identifiable and different sets of K and L J provided reasonable fit of the 

length frequency distributions. In every case, we retained the set of parameters that best 

corresponded with the breeding patterns observed for the species (ie. which estimated date of 

birth best corresponded with breeding peaks) and that best described the distributions (ie. 

which went through the largest number of large modes). Also, as an extensive sampling was 

done monthly over a complete annual cycle, always on the same sites, it was assumed that for 

most species the maximum observed length was close to the asymptotic length (L J ), which 

participated in selecting the best set of parameters. This was particularly true for the small 

abundant species, for which we fixed LJ within a few millimetres from the maximum 

observed length. This process also permitted to diminish the tendency of ELEFAN method to 

underestimate K and overestimate LJ (Moreau et al. 1995). 

Despite the 35 mm cod end mesh size, the length frequency distributions were 

influenced either by the trawl selectivity or the absence of the juveniles from trawled areas. 

Indeed; juveniles of species were seldom caught before 50 mm and were usually accessible to 

the trawl gear between 60 and 90 mm depending on species and shapes. Consequently, adults 

were better represented in the catches than juveniles and length frequency distributions below 

the size of full selection were not adequate for mortality estimates. Whenever one or more 

suitable distributions were available, mortality estimates were based only on them. But in 

most of the cases, they were based on all the distributions. This must be kept in mind even 

169

though mortality estimates often appeared reasonable. FiSAT allows correction of the bias 

due to fishing gear's selectivity, which often leads to a better estimation of the growth 

parameters (Moreau et al. 1995, Gayanilo et al. 1996). This correction was tried for every 

species and never lead to better estimations (according to the "goodness of fit" index of the 

ELEFAN routine, see Gayanilo & Pauly 1997). The weight given to small length 

systematically flattened out the rest of the distribution and the resulting fits were bad. All the 

estimations were therefore based on non corrected data. However, for every species except the 

small ones for which it was not necessary, the smoothing of the data by calculating the 

running averages over three length bins (5 mm for most of the species and 10 mm for large 

and not abundant species) helped to track the progression of modes. 

All the lengths are standard lengths. 


All the fish analysed were collected during the monthly trawl catches in the north of 

the South West Arm (see Chapter 1 for details). 

The following equation was used to convert total lengths (TL) from literature into standard 

length (SL): 

SL = 0.785 TL + 3.477 (1) 

This equation was calculated from Table G1. 

Table G1. Growth estimates (mm in TL) of Lake Malawi cichlids species given by Iles 

(1971) and Tweddle & Turner (1977) and the same growth estimates expressed in SL by de 

Merona et al. (1988). 

Species LJ (TL) K References LJ (SL) References 

H. anaphyrmus 196 0,671 Tweddle & Turner 

(1977) 

157 De Merona et al. 

(1988) 

H. intermedius 229 0,571 Tweddle & Turner 

(1977) 


(1988) 

H. virginalis 121 0,778 Iles (1971) 99 De Merona et al. 

(1988) 

H. quadrimaculatus 190 0,65 Iles (1971) 153 De Merona et al. 

(1988) 

H. pleurostigmoides 144 0,764 Iles (1971) 116 De Merona et al. 

(1988) 

L. parvidens 208 0,487 Tweddle & Turner 166 De Merona et al. 

(1977) 

L. longipinnis 202 0,571 Tweddle & Turner 

(1977) 

(1988) 


(1988) 

Growth was estimated from the modal progressions of length frequency distributions 

of species at every sampled month. The growth parameters were calculated by the Von 

Bertalanffy Growth Curve (VBGC) equation (equation 2) fitted by the electronic length 

frequency analysis (ELEFAN) method (Pauly 1987) using the FAO-ICLARM Package 

FiSAT (Gayanilo et al. 1996, Gayanilo & Pauly 1997). 

170

L t = L J (1-exp (-K (t-t 0 )) (2) 

Where L t is the mean length at age t, L J is the asymptotic length, K the growth coefficient and 

t 0 the size at age 0. 

Among the several growth models available (VBGC, Richards, Gompertz, logistic, quadratic, 

exponential, ect. see Schnute 1981 for review), the VBGC model was retained for the 

following reasons: 1)- for African fish species the VBGC usually provides a good fit of the 

data (de Merona et al. 1988, Moreau et al. 1995). 2)- it proves useful for comparative 

purposes as it has been largely utilised for African freshwater species and cichlids 

particularly, for which synthesis are available (de Merona et al. 1983, 1988, Moreau et al. 

1995). The ELEFAN method, used by Moreau et al. (1995) on 57 stocks of African 

freshwater species was also utilised in this study for comparative purposes. 

Mortality estimates were obtained as described in Moreau et al. (1991) and Moreau & 

Nyakageni (1992) for Lake Tanganyika fishes. Total mortality (Z) was estimated by the 

method of the length-converted catch curves (LCC) (Pauly 1983). This method consists in 

pooling all the distributions while keeping their relative importance to obtain a single 

frequency distribution. This decreases part of the sampling biases. Total mortality is then 

calculated on the descending part of this single global distribution. But Z is determined in a 

given age/size range and the estimation makes sense only within this range. Natural mortality 

(M) was evaluated using Pauly's equation (Pauly 1980) based on LJ, K and the mean annual 

environmental temperature of the species concerned. For each species, the mean annual 

temperature used for natural mortality estimates corresponds to the mean annual temperature 

at the depth to which the species was more abundant (see Chapter "Breeding and depth 

distribution"). Fishing mortality (F) was calculated as F = Z-M. All these methods were 

provided by the FiSAT package, including the estimation of the probability of capture. 

The mean size at first capture by the trawl (Lc = length at which 50% of the fish entering the 

trawl are retained by the gear) was calculated for each species from the length-converted 

catch curves using FiSAT (Gayanilo & Pauly 1997). 

171

Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May 

1998 1999 

FigureG1-1.LengthfrequencyplotsforA.'geoffreyi' 

intheSWAofLakeMalawi. 

Relativeage(years-t0) 


FigureG2-2.Lengthconvertedcatchcurvefor A.macrocleithrum. 


1998 1999 

FigureG2-1.Lengthfrequencyplotsfor A.macrocleithrum 

intheSWAofLakeMalawi.

Results 


Alticorpus 'geoffreyi' 

The length frequency distributions, based on 1721 fish are presented in Figure G1-1. 

Three year classes were identified by the model with L J =181 mm and K=0.6. The calculated 

date of birth (June) corresponded to the middle of the observed breeding peak. 

Assuming a mean environmental temperature of 24°C and taking into account all the 

distributions, the mortality estimates were: Z=3.29, M=1.37 and F=1.92 for the age range 

showed in Figure G1-2. Given that the length frequency distributions were not adequate, it is 

likely that mortality was overestimated, particularly for a deep water species subjected to little 

if any exploitation in this part of the lake. The selectivity of the 35 mm cod end trawl net for 

this species was 120 mm. 

Alticorpus macrocleithrum 

A. macrocleithrum was not an abundant fish and only 437 specimens were measured 

over the sampling period. The length frequency distributions are presented in Figure G2-1. 

Two sets of growth estimates, correctly fitting the distributions and giving birth dates 

consistent with breeding observation were obtained: L J = 166, K = 0.92 (dashed line) and L J 

= 166, K = 0.6 (solid line). The calculated date of birth, May-June for the first set and July- 

August for the second corresponded to the middle of the observed breeding peak. Given the 

low number of individuals of the length frequency distributions, particularly in the small sizes 

(little peaks between 60 and 80 mm were all based on one or two fish only), it was difficult to 

decide for one or the other set of parameters. 

Assuming a mean environmental temperature of 24°C corresponding to the depth distribution 

of A. macrocleithrum (75 to 125 m) and taking into account all the distributions, the mortality 

estimates were: Z=6.04, M=1.86 and F=4.18 with L J =166, K=0.92 and Z=3.94, M=1.40 and 

F=2.54 with L J =166, K=0.6. The mean size at first capture by the 35 mm cod end trawl net 

for this species was 106 mm with both sets of parameters. In both cases, mortality estimates 

were high partly because the shapes of distributions were not adapted to calculate mortality 

parameters. However, the second set of mortality estimates, though overestimated, appeared 

more realistic, particularly for a deep water species subjected to little if any exploitation. 

Furthermore, with K=0.92, fish would reach their maximum size and die at the end of their 

second year, which appears too fast for a fish of this size compared to others related species. 

For these reasons, the set of parameters L J = 166 and K = 0.6 was chosen for the 

calculation of age at maturity and mortality estimates (Z=3.94, M=1.40 and F=2.54 for the 

age range showed in Figure G2-2). 

172


1998 1999 

FigureG3-1.Lengthfrequencyplotsfor A.mentale 

intheSWAofLakeMalawi. 



Figure G4-2.Lengthconvertedcatchcurvefor A.pectinatum. 


1998 1999 

FigureG4-1.LengthfrequencyplotsforA.pectinatum 


Alticorpus mentale 


Two year classes were identified by the software, with L J =266 mm and K=0.7 (solid line). 

The calculated date of birth, April corresponded with a period of intense breeding activity. 

Assuming a mean environmental temperature of 24°C corresponding to the depth 

distribution of A. mentale (75 to 125 m) and considering only the June distribution, the 

mortality estimates were: Z=1.63, M=1.36 and F=0.27 for the age range showed in Figure G3- 

2. The mean size at first capture by the 35 mm cod end trawl net for this species was 91 mm. 

Another set of parameters provided a good fit of the distributions: LJ=256 mm and K=0.68 

(dashed line). The birth date, October-November also matched a period of intense breeding 

activity. However, the corresponding fit was going through a smaller number of large peaks 

than the first set of parameters, which was therefore considered better. 

Alticorpus pectinatum 


Three year classes were identified by the software with L J =160 mm and K=0.58. The 

calculated date of birth, December corresponded with a period of intense breeding activity. 


of A. pectinatum (75 to 125 m) and considering only the October distribution, the mortality 

estimates were: Z=1.90, M=1.39 and F=0.51 for the age range showed in Figure G4-2. The 

mean size at first capture by the 35 mm cod end trawl net for this species was 84 mm. 


Aulonocara 'blue orange' 


Au. 'blue orange' is a small species and juveniles were not caught by the net. Consequently a 

single year class was identified by the software with L J =80 mm and K=1.21. The calculated 

date of birth, December-January corresponded with a peak of breeding activity. 


of Au. 'blue orange' (10 to 30 m) and considering all the distributions, the mortality estimates 

were: Z=4.67, M=2.83 and F=1.84 for the age range showed in Figure G5-2. The mean size at 

first capture by the 35 mm cod end trawl net for this species was 51 mm. Although Au. 'blue 

orange' is a short lived fast growing species with a high natural mortality (M), the estimated 

total mortality was probably overestimated owing to the shape of length distributions. 

173


1998 1999 

Figure G5-1.Lengthfrequencyplotsfor Au.'blueorange' intheSWAofLakeMalawi. 



FigureG6-2.Lengthconvertedcatchcurvefor Au.'minutus' . 


1998 1999 

Figure G6-1.Lengthfrequencyplotsfor Au.'minutus' intheSWAofLakeMalawi.

Aulonocara 'minutus' 


Au. 'minutus' is also a small species and juveniles were not caught by the net. Consequently a 

single year class was identified by the software with L J =75 mm and K=1.44. The calculated 

date of birth, March corresponded with a period of intense breeding activity. 

Assuming a mean environmental temperature of 23.5°C corresponding to the depth 

distribution of Au. 'minutus' (75 to 125 m) and considering all the distributions, the mortality 



Although Au. 'minutus' is a short lived fast growing species with a high natural mortality rate 

(M), mortality estimates were probably overestimated owing to the shape of length 

distributions. Indeed, fishing mortality is likely to be close to zero in this deep part of the lake 

where almost no trawling activity takes place. 


Copadichromis quadrimaculatus 

C. quadrimaculatus was a relatively rare fish and only 631 specimens were measured 

over the sampling period. The length frequency distributions are presented in Figure G7-1. 

Despite the low sample size, a reasonably good growth estimation was obtained with L J =160 

mm and K=0.58. The calculated date of birth, September corresponded with a period of 

breeding activity. 


of C. quadrimaculatus (10 to 75 m) and considering only the November and January 


showed in Figure G7-2. The mean size at first capture by the 35 mm cod end trawl net for this 

species was 59 mm. 

Copadichromis virginalis 


Two year classes were identified by the software with L J =130 mm and K=0.84. The 

calculated date of birth, May corresponded to the middle of the main peak of breeding 

activity. 


distribution of C. virginalis (10 to 50 m) and considering only the April distribution, the 

mortality estimates were: Z=3.55, M=1.93 and F=1.62 for the age range showed in Figure G8- 

2. The mean size at first capture by the 35 mm cod end trawl net for this species was 72 mm. 

174


1998 1999 

FigureG7-1.Lengthfrequencyplotsfor 

C.quadrimaculatus intheSWAofLakeMalawi. 



FigureG8-2.Lengthconvertedcatchcurvefor C.virginalis . 


1998 1999 

FigureG8-1.Lengthfrequencyplotsfor C.virginalis intheSWAofLakeMalawi.


1998 1999 

FigureG9-1.Lengthfrequencyplotsfor D.apogon intheSWAofLakeMalawi. 



FigureG10-2.Lengthconvertedcatchcurvefor D.argenteus. 


1998 1999 

FigureG10-1.LengthfrequencyplotsforD.argenteus 



1998 1999 

FigureG11-1.Lengthfrequencyplotsfor D.limnothrissa intheSWAofLakeMalawi. 

A 


B 



1998 1999 

FigureG12-1.Lengthfrequencyplotsfor D.macrops 


Diplotaxodon and Pallidochromis spp. 

Diplotaxodon apogon 


Three year classes were identified by the software with L J =140 mm and K=0.56. The 

calculated date of birth, April corresponded to the end of the main peak of breeding activity. 


distribution of D. apogon (75 to 125 m) and considering all the distributions, the mortality 



Diplotaxodon argenteus 


Two sets of parameters gave good fit of the distribution, each of them fitting three year 

classes. With the first set, L J =219 mm and K=0.78 (solid line), the date of birth was 

December. Assuming a mean environmental temperature of 24°C corresponding to the depth 

distribution of D. argenteus (50 to 125 m) and considering the November and December 

distributions only, the mortality estimates were: Z=2.19, M=1.54 and F=0.65 for the age range 

0.8 to 3 years. With the second set of parameters, L J =220 mm and K=0.62 (dashed line), the 

date of birth was April and the mortality estimates were: Z=1.74, M=1.33 and F=0.41 for the 

age range 0.8 to 4 years. 

The mean size at first capture by the 35 mm cod end trawl net for this species was 57 mm 

with both sets of parameters. 

Both fits were good, however, the dashed line went better through the middle of the major 

peaks whereas the solid line went often through the end of the peaks (ex. October or 

December). Next, a birth in April, during the major breeding season appeared more 

reasonable than a birth in December after a little isolated breeding peak based on low sample 

size. Furthermore, mortality estimates and life span (Figure G10-2) obtained with the second 

set of parameters (L J =220 mm, K=0.62, dashed line), also appeared more reasonable for a 

deep water species subjected to little if any exploitation. 

Consequently, the following parameters were considered to better represent the data and were 

used for the calculation of age at maturity: L J =220 mm, K=0.62, Z=1.74, M=1.33 and 

F=0.41. 

Diplotaxodon limnothrissa 


Three year classes were identified by the software (solid line), with L J =192 mm, K=0.64. The 

calculated date of birth, July-August corresponded to the last two months of the breeding 

season. 


of D. limnothrissa (50 to 125 m) and considering only the August, November and March 


showed in Figure G11-2A. Given that the length frequency distributions were not adequate, it 

is likely that mortality estimates were overestimated, particularly for a deep water species 

175


FigureG12-2.Lengthconvertedcatchcurvefor D.macrops. 


1998 1999 

FigureG13-1.Lengthfrequencyplotsfor P.tokolosh intheSWAofLakeMalawi. 


FigureG13-2.Lengthconvertedcatchcurvefor P.tokolosh.

subjected to little if any exploitation in this part of the lake. The mean size at first capture by 

the 35 mm cod end trawl net for this species was 77 mm. 

More than one cohort per year class were present. The major peak in March appeared to be 

from the same year class that the lower peaks fitted by the solid lined model in February and 

April. The best estimation fitting that cohort (dashed line) was obtained with L J =188 mm, 

K=0.62. The calculated date of birth, May corresponded to the first two months of the 

breeding season. Mortality estimates were: Z=3.22, M=1.39 and F=1.83 for the age range 

showed in Figure G11-2B. The selectivity of the 10 mm cod end trawl net with this set of 

parameters was 74 mm. 

These two estimations, one taking into account the cohort born in the beginning of the 

breeding season, the other considering the cohort born at the end of the breeding season were 

very similar and gave very close growth and mortality estimates. The cohort born at the 

beginning of the breeding season (dashed line) had a slightly slower growth leading to an age 

at maturity of 16 months old whereas the cohort born at the end of the breeding season (solid 

line) reached maturity at 15 months old. 

Diplotaxodon macrops 


Three year classes were identified by the software with L J =143 mm, K=0.7. The calculated 

date of birth, September-October corresponded to the last months of the second peak of 

breeding activity. No correct fit gave a birth in the major breeding peak (January-May). 


distribution of D. macrops (75 to 125 m) and considering all the distributions, the mortality 

estimates were: Z=3.12, M=1.6 and F=1.52 for the age range showed in Figure G12-2. Given 

that the length frequency distributions were not adequate, it is likely that mortality estimates 

were overestimated, particularly for a deep water species subjected to little if any exploitation 

in this part of the lake. The mean size at first capture by the 35 mm cod end trawl net for this 

species was 106 mm, which also appeared overestimated. 

Pallidochromis tokolosh 

P. tokolosh was not an abundant fish and only 375 specimens were measured over the 

sampling period. The length frequency distributions are presented in Figure G13-1. Despite 

the low sample size, a reasonably good growth estimation was obtained with L J =244 mm, 

K=0.62. The calculated date of birth, January corresponded to the middle of the observed 

breeding season. 


distribution of P. tokolosh (75 to 125 m) and considering the June, July, October, January and 

February distributions, the mortality estimates were: Z=2.06, M=1.28 and F=0.78 for the age 

range showed in Figure G13-2. The mean size at first capture by the 35 mm cod end trawl net 

for this species was 106 mm, which appeared overestimated though it translated the length 

distributions. 

176


1998 1999 

Figure G14-1.Lengthfrequencyplotsfor L.argenteus intheSWAofLakeMalawi. 



FigureG15-2.Lengthconvertedcatchcurvefor L.'dwaltus' . 


1998 1999 

FigureG15-1.Lengthfrequencyplotsfor L.'dwaltus' 



Lethrinops argenteus 


Two year classes were identified by the software with L J =182 mm, K=0.94. The calculated 

date of birth, August corresponded to a major peak of breeding activity. 


of L. argenteus (10 to 50 m) and considering only the December and January distributions, the 

mortality estimates were: Z=4.38, M=1.87 and F=2.51 for the age range showed in Figure 

G14-2. The mean size at first capture by the 35 mm cod end trawl net for this species was 55 

mm. 

Lethrinops 'deep water altus' 


Despite the relative homogeneity of length distribution among successive months, it was 

possible to fit a VBGC, which gave reasonable estimates with L J =142 mm, K=0.62. The 

calculated date of birth, March corresponded to the main peak of sexual activity. 


distribution of L. 'deep water altus' (75 to 125 m) and considering only the December and 

January distributions, the mortality estimates were: Z=2.94, M=1.48 and F=1.46 for the age 

range showed in Figure G15-2. As the deep zone in this part of the lake is almost not 

exploited, it is very likely that mortality were overestimated owing to the under-representation 

of juveniles in the length distributions. The mean size at first capture by the 35 mm cod end 

trawl net for this species was 58 mm. 

Lethrinops gossei 


The best combination was obtained with L J =185 mm, K=0.78. The calculated date of birth, 

March corresponded to the peak of breeding activity. 


distribution of L. gossei (75 to 125 m) and considering all the distributions, the mortality 


mean size at first capture by the 35 mm cod end trawl net for this species was 99 mm. Again, 

as the deep zone in this part of the lake is hardly exploited, it is very likely that mortality was 

overestimated owing to the inadequate length distributions. 

177


1998 1999 

FigureG16-1.Lengthfrequencyplotsfor L.gossei intheSWAofLakeMalawi. 



FigureG17-2.Lengthconvertedcatchcurvefor L.longimanus. 


1998 1999 

FigureG17-1.Lengthfrequencyplotsfor L.longimanus 


Lethrinops longimanus 

L. longimanus was not an abundant fish and only 553 specimens were measured over 

the sampling period. The length frequency distributions are presented in Figure G17-1. 

Despite the low sample size and the lack of clear progression in length modes, a reasonable 

growth estimation was obtained with L J =160 mm, K=0.75. As it was impossible to determine 

the precise breeding season for this species, it was difficult to assess the quality of the 

calculated date of birth, August. However, it occurred at the period when most breeding 

activity was observed. 


of L. longimanus (30 to 125 m) and considering all the distributions, the mortality estimates 

were: Z=3.00, M=1.67 and F=1.33 for the age range showed in Figure G17-2. The mean size 

at first capture by the 35 mm cod end trawl net for this species was 82 mm. 

Despite the poor appropriateness of the data set to this kind of study, the estimates were in 

agreement with the values obtained for other Lethrinops species of comparable size. 

Lethrinops 'oliveri' 


As for the other small species a relative homogeneity of length distribution among successive 

months was observed. However, it was possible to fit a VBGC, which gave reasonable 

estimates with L J =110 mm, K=0.88. The calculated date of birth, June corresponded to the 

middle of the main breeding season. 


distribution of L. 'oliveri' (75 to 125 m) and considering all the distributions, the mortality 


mean size at first capture by the 35 mm cod end trawl net for this species was 60 mm. Again, 

as the deep zone in this part of the lake is lightly exploited, it is very likely that mortality was 

overestimated owing to the non adequate length distributions. 

Lethrinops polli 


The best estimation was obtained with L J =134 mm, K=0.78. The calculated date of birth, 

August-September corresponded to the end of the main observed breeding season. 


of L. polli (75 to 125 m) and considering only the July, August and May distributions, the 

mortality estimates were: Z=3.43, M=1.77 and F=1.66 for the age range showed in Figure 


mm. Again, as the deep zone in this part of the lake is almost not exploited, it is very likely 

that mortality was overestimated owing to the non adequate length distributions. 

178


1998 1999 

FigureG18-1.Lengthfrequencyplotsfor L.'oliveri' intheSWAofLakeMalawi. 



Figure G19-2.Lengthconvertedcatchcurvefor L.polli. 


1998 1999 

FigureG19-1.Lengthfrequencyplotsfor L.polli 



1998 1999 


M.anaphyrmus intheSWAofLakeMalawi. 



FigureG21-2.Lengthconvertedcatchcurvefor 

N.'argyrosoma'. 


1998 1999 

FigureG21-1.Lengthfrequencyplotsfor N.'argyrosoma' 



Mylochromis anaphyrmus 


Three year classes were identified by the software (solid line), with L J =180 mm, K=0.62. The 

calculated date of birth, May corresponded to the main peak of breeding activity. 

More than one cohort per year class was present. The peaks of small sizes in March, April and 

May appeared to be from the same year class that the upper peaks fitted by the solid lined 

model in April and May. The best estimation fitting that cohort (dashed line) was obtained 

with L J =179 mm, K=0.52. The calculated date of birth, September corresponded to the last 

two months of the breeding season. 

However, the estimation taking into account the cohort born in the middle of the breeding 

season (solid line) fitted the distributions best and was selected for the calculations of 

mortality. Assuming a mean environmental temperature of 26°C corresponding to the depth 

distribution of M. anaphyrmus (10 to 50 m) and considering all the distributions, the mortality 




Nyassachromis 'argyrosoma' 




estimates with L J =100 mm, K=1. The calculated date of birth, December corresponded to the 

major peak of breeding activity. 


of N. 'argyrosoma' (10 to 30 m) and considering all the distributions, the mortality estimates 


at first capture by the 35 mm cod end trawl net for this species was 59 mm. Again, it is very 

likely that mortality was overestimated owing to the inadequate length distributions. 

179


1998 1999 

FigureG22-1.Lengthfrequencyplotsfor Pl.'platyrhynchos' intheSWAofLakeMalawi. 



FigureG23-2.Lengthconvertedcatchcurvefor 

T.brevirostris. 


1998 1999 


T.brevirostris intheSWAofLakeMalawi.




The best estimation, fitting two year classes, was obtained with L J =132 mm, K=0.9. The 

calculated date of birth, April corresponded to the middle of the main observed breeding 

season. 


distribution of P. 'platyrhynchos' (75 to 125 m) and considering only the March distribution, 

the mortality estimates were: Z=2.43, M=1.93 and F=0.5 for the age range showed in Figure 


mm. 

Trematocranus spp. 

Trematocranus brevirostris 




estimates with L J =100 mm, K=0.79. The calculated date of birth, September seemed to 

correspond with a period of breeding activity. 


of T. brevirostris (around 50 m) and considering all the distributions, the mortality estimates 


at first capture by the 35 mm cod end trawl net for this species was 55 mm. 

180

Table G2. Growth parameters for 23 cichlid species from the SWA of Lake Malawi. L J and K are the parameters 

of the Von Bertalanffy growth curve equation (VBGC). The TMM (mean maximum observed length) is the 

average length of the ten largest fish caught (de Merona 1983, Moreau & Nyakageni 1992). K' and L J ' are rapid 

growth estimates calculated as follows: L J ' = 1,248 TMM, K' = 153 / L J ' (de Merona 1983). ∆G: growth 

difference at 2 years old obtained by fitting the VBCG with both sets of estimates (L J ', K' and L J , K). Values 

indicate the magnitude of growth overestimation by the rapid growth estimation model. 

L J (mm) K (year -1 ) TMM (mm) L J ' (mm) K' (year -1 ) D G (mm) 

A. 'geoffreyi' 181 0,6 158,2 197 0,77 29 

A. macrocleithrum 166 0,6 138,1 172 0,89 27 

A. mentale 266 0,7 229 286 0,54 -13 

A. pectinatum 160 0,58 134,5 168 0,91 31 

Au. 'blue orange' 80 1,21 72,6 91 1,69 15 

Au. 'minutus' 75 1,44 67,4 84 1,82 11 

C. quadrimaculatus 160 0,58 147,5 184 0,83 39 

C. virginalis 130 0,84 120,4 150 1,02 25 

D. apogon 140 0,56 123,1 154 1,00 38 

D. argenteus 220 0,62 197 246 0,62 19 

D. limnothrissa 188 0,62 167,9 210 0,73 27 

D. macrops 143 0,7 131,2 164 0,93 31 

P. tokolosh 244 0,62 205,6 257 0,60 5 

L. argenteus 182 0,94 156,5 195 0,78 0 

L. 'deep water altus' 142 0,62 115 144 1,07 26 

L. gossei 185 0,78 164,4 205 0,75 13 

L. longimanus 160 0,75 136,5 170 0,90 18 

L. 'oliveri' 110 0,88 96,3 120 1,27 20 

L. polli 134 0,78 110,5 138 1,11 17 

M. anaphyrmus 180 0,62 159 198 0,77 28 

N. 'argyrosoma' 100 1 93,7 117 1,31 22 

Pl. 'platyrhynchos' 132 0,9 114 142 1,08 16 

Tr. Brevirostris 100 0,79 80,9 101 1,52 17 

50 

40 

y = -0,0038x 2 + 1,0062x - 40,253 

R 2 = 0,5414 

30 

G (mm) 

20 

10 

0 

-10 

-20 

50 100 150 200 250 

TMM (mm) 

Figure G27. Distribution of G in relation with mean maximum observed length (TMM). 

181

Discussion 

As expected from the extended breeding seasons displayed by the studied species, 

more than one cohort per year class was usually present in the length frequency distributions. 

Knowledge about the species biology significantly helped us choosing among the various sets 

of growth parameters that correctly fitted the distributions. In particular, for a similar "quality 

of distribution's fitting", we always selected the set of parameters that gave a birth date 

consistent with the breeding season observed for the species (see previous Chapter). For this, 

we considered that the potential bias associated with the assumption that growth curve 

parameters applied right to zero, and not merely to the smallest sampled sizes (between 45 

and 60 mm depending on species) was negligible. Although the breeding seasonality of 

species was studied over a single annual cycle, previous investigations over more than one 

annual cycle have suggested little or no inter annual variability of breeding patterns (Iles 

1971, Tweddle & Turner 1977). Also, we decided to keep the asymptotic length (LJ) within a 

controlled ranged. The trawl, with its 35 mm cod end mesh size is a non species-selective 

gear, catching fish from about 50 mm (see the estimated length at first capture for the studied 

species) to more than 300 mm (ex. Buccochromis spp.). Large and fast predatory species such 

as Rhamphochromis spp. were also caught, sometimes to sizes up to 500 mm. As a 

consequence and given the large numbers of specimens caught for most of the species, we 

considered that the maximum observed lengths were likely to be close to the asymptotic 

lengths, for the sampled area. Therefore, LJ was intentionally kept within a range of 1 to 2 

cm above the maximum observed length for the medium and large species, and within a few 

millimetres above for the smallest species (Au. 'bue orange', Au. 'minutus', N. 'argyrosoma', 

T. brevirostris). 

Growth factor (K) values ranged from 0.56 to 1.44, averaging 0.77 (Table G2). As 

expected given the inverse relationship between LJ and K (de Merona et al. 1988), the 

smallest species (L J < 100 mm) had the highest K, ranging from 0.79 to 1.44 with an average 

value of 1.11. However, the largest species (A. mentale, P. tokolosh and D. argenteus) did not 

have the lowest K. Medium sized species did have them. For comparative purposes, growth 

estimates of individual species were fitted by the VBGC equation and grouped per genera or 

size classes (Figure G24). The VBGC have been fitted up to the maximum observed length 

(MOL) of species. Within a single genus, species with comparable lengths had slight growth 

differences, though usually not higher than 1 cm at 2 years old as illustrated by A. 

macrocleithrum and A. pectinatum (Figure G24a), D. apogon and D. macrops (Figure G24b), 

L. argenteus and L. gossei (Figure G24c) or Au. 'blue orange' and Au. 'minutus' (Figure 

G24d). Growth performances of species within genera, as expressed by length at age, were 

proportional to their maximum length for Alticorpus, Diplotaxodon and Lethrinops spp., with 

the exception of L. 'deep water altus' having a slower growth than the smaller L. polli. 

Between genera comparison for species of similar lengths showed that Lethrinops spp. had 

better growth than Alticorpus and Diplotaxodon spp. (see L. argenteus and gossei versus A. 

'geoffreyi' and D. limnothrissa). Lethrinops versus Copadichromis spp. comparisons were less 

clear as the similarly sized C. virginalis and L. polli had equivalent growths whereas C. 

quadrimaculatus had a slower growth than L. longimanus. A. pectinatum and C. 

quadrimaculatus had the same growth estimates and logically presented exactly the same 

growth curves, as did A. 'geoffreyi' and M. anaphyrmus (Figure G24a). Apart from L. 'deep 

water altus', which growth was intermediate to those of D. macrops and D. apogon, 

Lethrinops species tended to have better growths than species of others genera with similar 

sizes. 

182

a 

300 

250 

A. 'geoffreyi' 

Length at age (mm) 

200 

150 

100 

A. macrocleithrum 

A. mentale 

A. pectinatum 

C. quadrimaculatus 

50 

C. virginalis 

M. anaphyrmus 

0 

0 1 2 3 4 5 

Age (year) 

b 

250 

200 

D. apogon 


150 

100 

D. argenteus 

D. limnothrissa 

50 

D. macrops 

P. tokolosh 

0 

0 1 2 3 4 5 

Age (year) 

Figure G24. Von Bertalanffy growth curves for cichlid species caught by trawling in the SWA 

of Lake Malawi between June 1998 and May 1999. a: Alticorpus, Copadichromis & 

Mylochromis spp., b: Diplotaxodon & Pallidochromis spp.. 

183

c 

180 

160 

140 

L. argenteus 


120 

100 

80 

60 

L. 'dw altus' 

L. gossei 

L. longimanus 

40 

L. 'oliveri' 

20 

L. polli 

0 

0 1 2 3 4 5 

Age (year) 

d 

140 

120 

Au. 'blue orange' 


100 

80 

60 

Au. 'minutus' 

N. 'argyrosoma' 

40 

Pl. 'platyrhynchos' 

20 

Tr. brevirostris 

0 

0 1 2 3 4 5 

Age (year) 

Figure G24. Von Bertalanffy growth curves for cichlid species caught by trawling in the SWA 

of Lake Malawi between June 1998 and May 1999. c: Lethrinops spp., d: miscellaneous 

small species. 

184

For a same population, values of LJ and K can significantly vary from one cohort to 

another (Craig 1978). Examples of multiple cohorts within a same year class were given by A. 

mentale, D. limnothrissa and M. anaphyrmus, though they appeared for most of the medium 

and large species. For these three species, growth estimates of the two cohorts resulted in 

different growth performances (Figure G25). 

300 

250 


200 

150 

100 

A. mentale A. mentale 2 

50 

D. limnothrissa D. limnothrissa 2 

M. anaphyrmus M. anaphyrmus 2 

0 

0 1 2 3 4 5 

Age (years) 

Figure G25. Von Bertalanffy growth curves illustrating differences among cohorts for three 

species caught by trawling in the SWA of Lake Malawi between June 1998 and May 1999. 

Alticorpus mentale = cohort born in April (L J = 266, K = 0.7), A. mentale 2 = cohort born 

in October-November (L J = 256, K = 0.68). Diplotaxodon limnothrissa = cohort born in 

May (L J = 188, K = 0.62), D. limnothrissa 2 = cohort born in July-August (L J = 192, K = 

0.64). Mylochromis anaphyrmus = cohort born in May (L J = 180, K = 0.62), M. 

anaphyrmus 2 = cohort born in September (L J = 179, K = 0.52). 

For A. mentale, the cohort born in April had a length of 200 mm at 2 years old against 190 

mm for the cohort born in October-November (5.5 % difference). For D. limnothrissa the 

respective lengths at 2 years old were 134 mm for the cohort born in May and 139 mm for the 

cohort born in July-August (4% difference). The largest length difference at 2 years old 

(10.4%) was observed for M. anaphyrmus: 128 mm and 116 mm for the cohorts born in May 

and September, respectively. These growth differences appeared after only a few months and 

were already marked at one year old. This means that for fish of a same population, growth 

depends upon the period of birth, thus of the prevailing environmental conditions. The most 

important environmental parameters influencing growth are temperature, oxygen and food 

availability (Pauly 1980, Caulton 1982, Pitcher & Hart 1982, Wootton 1990). For the deep 

water species (A. mentale, D. limnothrissa) temperature variations over the year (less than two 

degrees, Figure G26a) were unlikely to influence growth. In the depth distribution of M. 

anaphyrmus, temperature variations were more important (between 4 and 5°C). However, 

during the period separating the two cohorts (May with best growth and September) 

temperature were at their minimum and increased from August until the next cold season 

(May to August). Temperature is therefore unlikely to account for the observed growth 

185

differences. The opposite pattern is observed for oxygen (Figure G26b) with higher seasonal 

fluctuations in the deep waters (about 5 mg.l -1 ) than in the shallows (less than 2 mg.l -1 ). In the 

deep zone, oxygen concentration increased from February to August and then decreased from 

September to January. For A. mentale (which presented the highest growth difference in the 

deep zone), the cohort with the slowest growth (born in October-November) faced a two fold 

decrease in oxygen concentration during its first three months. In December and January D.O. 

got down to 1.5 mg.l -1 , which represented about 17% saturation at 23°C and 100 m depth. It 

has been shown that growth of tilapia, which are well known to tolerate very low D.O., is 

reduced below 25% saturation (review by Chervinski 1982). Exposition to such low D.O. 

during at least two months might partly account for the observed growth difference between 

the cohorts. However both cohorts encountered periods of low oxygen concentration during 

their first six months and growth differences between cohorts were also observed for species 

which did not face low D.O. (M. anaphyrmus). Variations in food availability appears a more 

plausible explanation though little is known about seasonal variations of food availability for 

these three species with marked different feeding regimes: piscivorous, zooplanktivorous and 

malacophageous for A. mentale, D. limnothrissa and M. anaphyrmus, respectively (see 

chapter "Diet"). Nevertheless, whatever caused these growth differences among cohorts, it is 

striking that differences remained over time. Indeed, compensatory growth is well 

documented in fish and cichlids (review in Weatherley & Gill 1987, Melard et al. 1997). In 

Malawi cichlids for which fasting periods imposed by mouthbrooding are frequent, genuine 

capacities to buffer these periods are expected and probably exist (see Chapter 6). 

Temperature (°C) 

D.O. (mg/l) 

22 23 24 25 26 27 28 29 

0 1 2 3 4 5 6 7 8 9 10 

-10 

-10 

-30 

-30 

Depth (m) 

-50 

-70 

June-98 

July-98 

August-98 

October-98 

Depth (m) 

-50 

-70 

November-98 

December-98 

-90 

January-99 

March-99 

-90 

April-99 

-110 

May-99 

-110 

a 

b 

Figure G26. Seasonal variations of thermocline (a) and oxycline (b) at our 100 m transect in 

the SWA, Lake Malawi. 

186

Equations allowing rapid estimation of Von Bertalanffy growth parameters were 

proposed for African freshwater fishes (de Merona 1983). These equations, based on 111 

species, were: 

L J ' = 1.248 TMM and K' = 153 / LJ' 

TMM being the mean maximum length, usually calculated as the mean length of the ten 

largest specimen caught (Moreau & Nyakageni 1992). K' and L J ' were calculated for our 

species and compared with K and LJ obtained from length progression analysis (Table G2). 

Rapid estimates of asymptotic length (L J ') were always higher than the corresponding LJ, 

except for L. 'deep water altus', L. polli and Tr. brevirostris for which values were close. The 

same trend was observed with rapid estimates of the growth factor (K'), generally much 

higher than the K obtained from length frequencies. However, for the five largest and fastest 

growing species (A. mentale, D. argenteus, P. tokolosh, L. argenteus and L. gossei), rapid 

estimates were equal or lower than the observed K. The lack of fitting between the rapid 

growth estimates model and the estimates resulting from length progression analysis may lie 

in the fact that only half of the 111 species on which the model is based were cichlids, and 

most of them were tilapiine cichlids (de Merona 1983, de Merona et al. 1988). Only 9 species 

out of 111 were haplochromine cichlids. Next, the aim to this rapid growth estimates model 

was to provide a quick and reasonably reliable way to assess growth in absence of suitable 

data for other methods, not to replace them. The rapid growth estimation model does not seem 

well adapted to haplochromine cichlids, for which it tends to overestimate growth by an 

average value of 20 mm (∆G) at 2 years old. However, the overestimation tended to be lower 

for small and large species than for medium sized ones (Figure G27). 

1,6 

1,4 

1,2 

K (year -1 ) 

1 

0,8 

0,6 

0,4 

0,2 

0 

y = 57,362x -0,8845 

R 2 = 0,5756 

0 50 100 150 200 250 300 350 400 450 

Asymptotic length (mm) 

Figure G28. Relationship between K and L J for 87 species of African cichlids. Grey spots = 

data from de Merona 1983, de Merona et al. (1988). Black spots = data from Iles (1971) 

and Tweddle & Turner (1977) for Malawi haplochromine cichlids, adjusted in SL by de 

Merona et al. (1988). Black triangles = data from Iles (1971) and Tweddle & Turner 

(1977) for Malawi haplochromine cichlids recalculated using ELEFAN by Moreau et al. 

(1995) and adjusted in SL using equation (1) (see Material and Methods). White square = 

Diplotaxodon limnothrissa from Thompson et al. (1995) adjusted in SL using equation (1). 

White triangles = data from this study. 

187

Growth data available for 86 African cichlid species from Iles (1971), Tweddle & 

Turner (1977), de Merona (1983), de Merona et al. (1988), Moreau et al. (1995), and this 

study, were used to produce a general relationship between K and LJ and to check how our 

estimates were fitting in it (Figure G28). Our estimates were within the range of reported 

values for other African cichlids and Malawian haplochromines and set new references for 

small species with asymptotic length below 100 mm. Out of the 23 species we studied, four 

had already been studied for growth parameters: Copadichromis quadrimaculatus, C. 

virginalis (Iles 1971, recalculated by Moreau et al. 1995), Diplotaxodon limnothrissa 

(Thompson et al. 1995) and Mylochromis anaphyrmus (Tweddle & Turner 1997, recalculated 

by Moreau et al. 1995). As growth estimates can vary from one cohort to another within a 

same population (Craig 1978, de Merona 1983 for review, this study), comparing our results 

with growth estimates obtained 20 to 30 years ago (Iles 1971, Tweddle & Turner 1977) in 

different geographic areas may appear useless. However, despite the time and distance 

separating the estimations, growth differences for a same species at 2 years old (calculated 

using equation 2) were within the range of the differences we found among cohorts of a same 

species, though growth performances were always better with our estimates. More striking 

were the differences of longevity we found compared to Iles's (1971) and Tweddle & Turner's 

(1977), who reported 5 to 6 years life span. As illustrated on Figure G24, the species were not 

growing older than 4.5 to 5 years, and most of them no older than 4 years. Similar life span 

were found by Moreau et al. (1995) from Iles's (1971) and Tweddle &Turner's (1977) data 

reanalysed using FiSAT package. Haplochromines and non tilapiine cichlids from other lakes 

also lived between 2 and 4.5 years (Moreau et al. 1995), which tends to confirm shorter life 

span than previously thought for non tilapiine cichlids. Attempts to determine accurately 

mortality rates were often not successful. Indeed, though within the range of mortality values 

reported for Malawi cichlids (Tweddle & Turner 1977, Moreau et al. 1995), mortality rates 

often appeared overestimated. Given the low fishing exploitation in the sampled area, 

reasonable estimates of instantaneous mortality rate (Z), particularly for the deep water 

species, should have been close to natural mortality estimates (M), which usually seemed 

reasonable. This was seldom the case. The reason mainly lie in the fact that length 

distributions were biased by lack or under-representation of smaller individuals. However, 

this is a common problem in Lake Malawi and even though juveniles were caught by fishing 

gears, it would be of little help considering the identification problems that would result. 

Consequently, though mortality rates were probably overestimated, they remain a useful basis 

for fisheries and trophic modelling. 

188

Chapter 4: 

Temporal diet patterns of some Lake 

Malawi demersal fish species as 

revealed by stomach contents and 

stable isotope analysis

Chapter 4: Temporal diet patterns of some Lake Malawi demersal 

fish species as revealed by stomach contents and stable isotope 

analysis 

F. Duponchelle, H. Bootsma, A.J. Ribbink, C. Davis, A. Msukwa, J. Mafuka & D. Mandere 

Introduction 

Cichlids have evolved an astonishing diversity of feeding adaptations and behaviours 

that enable them to utilise virtually any kind of food, from phytoplankton, epilithic and 

epiphytic algae, plants, detritus, zooplankton, molluscs, insects, benthic invertebrates, fish 

eggs, larvae, eyes, and scales, to whole fish (reviews in Fryer & Iles 1972, Ribbink 1990, 

Yamaoka 1991). The role of the feeding apparatus and trophic specialisations in the adaptive 

radiation of African cichlids has often been discussed (Fryer & Iles 1972, Liem 1980, 1991, 

McKaye & Marsh 1983, Ribbink et al. 1983, Reinthal 1990, Ribbink 1990, Yamaoka 1991). 

The understanding of how such rich and diverse fish communities with apparently similar 

food requirements can coexist still challenges ecologists. In Lake Malawi, since Fryer's 

(1959) suggestion that the mbuna community was violating the Gaussian principle of 

competitive exclusion, several studies have provided evidences that food partitioning may 

reduce interspecific competition and allow coexistence among rock-dwelling cichlids fishes 

(McKaye & Marsh 1983, Marsh & Ribbink 1985, Reinthal 1990, Bootsma et al. 1996, Genner 

et al. 1999a, 1999b, 1999c). Similar conclusions have resulted from research in Lake 

Tanganyika (Sturmbauer et al. 1992) and Victoria (Bouton et al. 1997). However, while diet, 

feeding behaviour and trophic specialisations have been, and still are, intensely studied in 

Malawian rock-dwelling species, very little is known of the divers offshore (sensu Turner 

1996) cichlid communities. Apart from the zooplanktivorous utaka group (Copadichromis 

spp.) (Fryer & Iles 1972), the chambo (Oreochromis spp., Turner et al. 1991b) and the pelagic 

species, whose feeding ecology was recently thoroughly studied (Allison et al. 1996, 

Ngatunga & Allison 1996), the only information available on the diet of offshore fishes comes 

from Eccles & Trewavas (1989) and Turner (1996). These studies resulted in useful now 

insights into fish feeding habits, but they were limited by the relatively small numbers of 

observations and the limited time span over which fish stomach contents were monitored. The 

currently running European Union Project: "The trophic ecology of the demersal fish 

community of lake Malawi/Niassa", partly aimed at filling this gap, will improve our 

knowledge. However, the seasonal variability of fish diet is not a priority of the EU Project. 

The temporal aspect of demersal fish diets was investigated in the context of a general 

program designed to assess the seasonal progression of distribution, abundance and diversity 

of the fish species exploited by demersal trawling and to determine their main life history 

characteristics. Such a study had been considered when the new "Ecology program" had 

started in June 1998, but had been cancelled because the research actions already undergone 

were too time and people-consuming to allow the addition of another program. However, 

when a new staff member detached from the World University Service of Canada (WUSC) 

joined the ecology team of the project in October 1998, the study was reconsidered and 

initiated in November 1998. However, since the program ended in June 1999, an entire annual 

189

cycle could not be studied. We decided to determine the diet and its potential seasonal 

variability for the nine target species retained for the life history study (Lethrinops gossei, 

Lethrinops argenteus = L. longipinnis 'orange head', Diplotaxodon limnothrissa, 

Diplotaxodon macrops, Copadichromis virginalis, Mylochromis anaphyrmus, Alticorpus 

mentale, Alticorpus macrocleithrum and Taeniolethrinops praeorbitalis) with both monthly 

stomach content and stable isotope analysis. Whereas stomach content analysis provides 

insight into the ingested food items over a short time period, stable isotope signatures 

represent a spatio-temporal integration of the assimilated food over long time periods varying 

from months to years depending on fish growth rates (Peterson & Fry 1987, Hesslein et al. 

1991, 1993, Bootsma et al. 1996, Gannes et al. 1997, Gorokhova & Hansson 1999, Fry et al. 

1999). Stable isotope analysis is particularly useful for deep-water species that often have 

inverted stomachs when retrieved from trawls. 


Stomach content analysis 

Seasonal variability of diet was estimated over 8 months for the nine target species. 

Fish were collected from the monthly trawl survey in the north of the South West Arm 

(SWA). Every month from November 1998 to May 1999, 20 specimens of each species were 

sampled from the main catch as soon as the total catch weight was estimated. 15% formalin 

was injected in the abdominal cavity of each fish to ensure the preservation of food items and 

the fish were fixed in 10% formalin for later examination. A frequently encountered problem 

when trawling below 50m depth is that the stomachs are often burst out of the fish mouths 

during hauling. For Alticorpus mentale, whose stomachs were almost systematically empty, 

every specimen from the whole catch was checked for intact stomachs. Even apparently intact 

stomachs often contained only very little amounts of remaining food items. 

When enough intact stomachs were available, 5 specimens of each species were 

analysed each month for diet composition. For the sake of data compatibility, the method used 

was the modified version of the "point method" (Hynes 1950) selected by the ongoing EU 

Project: "The trophic ecology of the demersal fish community of lake Malawi/Niassa" 

(Darwall, 1999). The weight of the stomach plus content and the stomach minus content were 

determined to the nearest 0.001g. Total weight of stomach contents was calculated as the 

difference between the two weights. The stomach content was then examined under binocular 

microscope (10X to 40X magnification). All identifiable items were grouped into separate 

piles and allocated one of the following value: 16, 8, 4, 2, 1, or P (if present but in negligible 

amounts). The most abundant items were allocated a 16 and the others items were allocated a 

16, 8, 4, 2, 1 or P depending upon their abundance relative to the most abundant item. All the 

small items unidentifiable under the binocular were pooled under a pile of "Small 

Unidentified" and allocated a value according to their relative proportion of the total stomach 

content. The pile of small unidentified items was then mixed with a small quantity of water 

(approximately 4× volume of the pile) and agitated thoroughly to break up any compacted 

lump. The sample was then left to settle out and the excess water removed. The remaining 

solution was homogeneously mixed and a small quantity (approx. 0.1 ml) was poured onto a 

slide under a slip cover. The slide was marked into quarters to define 4 sub-samples. The 

whole slide was then analysed under 40X to 400X magnification to identify the full range of 

food items. Each sub-sample was analysed under 40X magnification and each item was 

allocated a number of points as described above. When the four sub-samples were quite 

variable, the process was repeated with a second slide to get a total of eight estimates. The 

190

composition of the small unidentified pile was then converted in percentage composition of 

each of its constituents and expressed as a percentage of the total stomach content. The 

percentage composition was calculated for each item as its own points value divided by the 

combined total of points for all items combined, multiplied by 100. In order to avoid giving 

too much weight to stomachs with only very little content, for every month each diet item 

value was weighted by the total weight of the stomachs analysed for that given month (W. 

Darwall, pers. com.). 

Items which made up less than 2% of the diet each were lumped together and referred to as 

"others". Non identifiable materials were recorded as "No ID". 

Stable isotope analysis 

Fish samples for stable isotope analysis were collected by trawling during the January 

1999 cruise. About six specimens of each of the nine target species, three small and three 

large whenever possible, were collected. Only dorsal muscle tissue was analysed. The 

potential food sources were collected during the April 1999 cruise. Benthic invertebrates and 

gastropods were sorted out from grab samples at 10, 30, 50, 75, 100 and 125 m depth. 

Sediment samples were taken from the upper layer of grab samples at every depth. 

Zooplankton and mayfly larvae were collected by 125 m vertical tows with a 50 µm mesh 

zooplankton net. 

Stable isotope analyses were carried out at the Environmental Isotope Laboratory at 

the University of Waterloo, Canada. Samples were run for Nitrogen and Carbon analysis on 

an Isochrome Continuous Flow Stable Isotope Mass Spectrometer (Micromass) coupled to a 

Carlo Erba Elemental Analyzer (CHNS-O EA1108). Results were corrected to Nitrogen 

standards IAEA-N1 and IAEA-N2 (both Ammonium Sulphate) and Carbon standards IAEA- 

CH6 (sugar), EIL-72 (cellulose) and EIL-32 (graphite). EIL-70b, a lipid extracted/ball-milled 

fish material, is often used as a monitoring standard – (‘EIL’ denotes Internal Standards with 

values calculated using International Standards). The error for clean ball-milled standard 

material is +/- 0.2‰ for Carbon and +/- 0.3‰ for Nitrogen. This error can be expected to 

increase depending on the homogeneity, type and amount of sample used in analysis. A truer 

representation of sample reproducibility can be gained through sample repeats. Standards are 

placed throughout each run at a range of weights to allow for an additional linearity 

correction, when necessary, due to machine fluctuations or samples of varying signal peak 

areas. Nitrogen and Carbon concentrations are calculated based on Carlo Erba Elemental 

Standards B2005, B2035 and B2036 with an error of +/- 1%. 

191

N=22 

Overall 

Chironomid larvae 

Chaoborus larvae 

Insect larvae 

Sand 

No ID 

Detritus 

Diatoms 

Others 

November-1998 

January-1999 

N=5 

N=2 

February-1999 

March-1999 

N=1 

N=3 

April-1999 

May-1999 

N=7 

N=4 

Figure D1. Overall and monthly diet composition (% wet weight) of Alticorpus macrocleithrum . 

See text for details on "Others" items.

Results and Discussion 

Stomach content analysis 


A. macrocleithrum is a deep water species found between 75 and 125 m. As for most 

of the deep water species, stomachs were very often inverted during trawl hauling. Only 22 

specimens with remaining items in their stomachs were caught between November 1998 and 

May 1999. Weight of stomach contents averaged 31.2 mg and ranged from 6.4 to 96.8 mg for 

fishes of 103 to 137 mm SL (31-67 g). Diet composition (as percentage of the wet weight) at 

each sampled month and all months pooled are presented in Figure D1. 

As inferred from its anatomy by Stauffer & McKaye (1985), A. macrocleithrum appears to be 

a benthic invertebrate feeder. About 51% of the diet was not identifiable though, recorded 

either as detritus or "no id". The other 49% were constituted of chironomid larvae, Chaoborus 

larvae, insect larvae, diatoms, sand and other items (fish scales, adult insects, cladocerans, 

copepods, gastropods, oligochetes, macrophytes and other algae). The most important food 

items were chironomid larvae, present at every sampled date and lake fly (Chaoborus) larvae, 

though they were found only in March and May. Owing to its short gut and deep water 

existence, Stauffer & McKaye suggested A. macrocleithrum was not feeding on algae or 

phytoplankton. However, at every month except in February (0.3%), diatoms constituted 

between 1.4 and 10.5% of the diet. Diatoms might be ingested incidentally with sediment and 

sand while digging to catch the invertebrates. However, diatoms accounted for 10.5 of the diet 

in November, when only a small quantity of sand had been ingested (compared to February 

and April) and for 4.1% in March where almost no sand was ingested. Next the proportion of 

diatoms ingested appeared too high to be incidental. 

Although the high proportion of detritus and unidentified materials at every month tended to 

screen the potential seasonal patterns, some food items occurred in the diet only at some 

months (ex. Chaoborus and insect larvae) suggesting A. macrocleithrum feeds 

opportunistically upon these items when available, either because they were the most 

abundant items at the moment or because they are preferred items. 


A. mentale is also a deep water species mostly abundant between 75 and 125 m. As 

stomachs were almost always empty, every specimen from the whole catch was checked for 

intact stomach. Despite this effort, only 14 specimens had remaining items in their stomachs 

during the period from November 1998 to May 1999. Weight of stomachs contents averaged 

1374.8 mg and ranged from 10.2 to 6549.2 mg for fishes of 110 to 245 mm SL (25-279 g). 

Diet composition (as percentage of the wet weight) at each sampled month and all months 

pooled are presented in Figure D2. 

As expected from its morphology, A. mentale is a piscivore. At any sampled date except 

December when 47% of the stomach content was unidentifiable, more than 75% of its diet 

consisted of adult cichlid fishes, often Aulonocara minutus. Other items were cichlid fry, 

scales, eggs, chironomid larvae, Chaoborus larvae, insect pupae, crustacean zooplankton, 

nematodes, sand, macrophytes and other algae. A significant amount of diatoms was recorded 

only once from a single large specimen (245 mm SL) in December, in which they made up to 

19.7% of the stomach contents. No seasonal pattern was observed during the sampling period. 

An interesting observation, not apparent on the figures, was that of five relatively small 

192

Overall 

Cichlid adults 

N=14 

No ID 

Diatoms 

Others 

November-1998 

December-1998 

N=5 

N=1 

February-1999 

March-1999 

N=5 

N=2 

April-1999 

N=1 

Figure D2. Overall and monthly diet composition (% wet weight) of Alticorpus mentale . 


individuals measuring between 110 and 117 mm, four fed principally (61 to 84% of the diet) 

on zooplankton (mostly copepods), and the fifth one fed on fish (94%). This pattern was not 

represented in the figures because the stomach content of these four small specimens were too 

light to account for a significant part of the weighted monthly mean diet. 


C. virginalis mainly occurs at depths between 30 and 50 m in the north of the SWA. 

Inverted stomachs were not a problem and 30 specimens were analysed between November 

1998 and May 1999. Weight of stomachs contents averaged 68.7 mg and ranged from 18 to 

190 mg for fishes of 70 to 115 mm SL (8-37 g). Diet composition (as percentage of the wet 

weight) at each sampled month and all months pooled are presented in Figure D3. 

C. virginalis is known as a member of the zooplanktivorous utaka group (Iles 1971, Fryer & 

Iles 1972, Turner 1996). Indeed, more than 95% of the diet was made of zooplankton, mainly 

copepods. It was only in this species that significant amount of cladocerans were found. No 

seasonal pattern was observed, crustacean zooplankton constituting more than 95% of the diet 

at any sampled date except December 1998, when the "others" components of the diet 

accounted for 17%. Other items were fish scales, chironomid larvae, Chaoborus adults and 

larvae, insects adults and larvae, nematodes, macrophytes, other algae and detritus. No 

particular trend relative to size was observed. 


D. limnothrissa was found at depths from 50 to 125 m, but was mostly abundant 

between 75 and 100 m. Despite their deep water existence, stomachs were not always inverted 

after hauling and 31 specimens were analysed. Weight of stomachs contents averaged 66 mg 

and ranged from 8 to 272mg for fishes of 101 to 145 mm SL (20-48 g). Diet composition (as 

percentage of the wet weight) at each sampled month and all months pooled are presented in 

Figure D4. 

Turner (1994) described D. limnothrissa as a zooplankton feeder, with specimens above 12 

mm SL feeding mainly on copepods and small specimens of 3 mm feeding on chaoborid 

larvae and copepods. Allison et al. (1996) reported a mixed diet composed of crustacean 

zooplankton, Chaoborus larvae, Engraulicypris sardella (usipa) larvae and occasionally 

phytoplankton. Our observations support the statement of a mixed diet: 71% of the diet was 

made of copepods, Chaoborus larvae, adult insects and usipa larvae. The remaining part of 

the diet was composed of unidentified material, detritus and "others" items (cichlid fry, scales, 

fish eggs, chironomid larvae, cladocerans, bivalves, sand, diatoms and other algae). Unlike for 

A. mentale and C. virginalis, diet composition strongly varied among months, being either 

dominated by copepods in November 1998, April and May 1999, by usipa larvae in January 

1999 or Chaoborus larvae in March 1999. As for A. macrocleithrum, Chaoborus larvae were 

almost exclusively present in the diet in March and May 1999 (also some in February). D. 

limnothrissa appears to feed opportunistically on a few preferred food items depending upon 

their availability. No particular trend relative to size was observed in the narrow size range 

studied. 

193

Overall 

Copepods 

N=30 

Cladocerans 

Others 

November-1998 

December-1998 

N=4 

N=5 

January-1999 

February-1999 

N=5 

N=5 

Avril-1999 

May-1999 

N=6 

N=5 

Figure D3. Overall and monthly diet composition (% wet weight) of Copadichromis virginalis . 


Overall 

Copepod 


No ID 

N=31 

Usipa larvae 

Insect adults 

Detritus 

Others 

November-1998 

January-1999 

N=4 

N=4 

February-1999 

March-1999 

N=3 

N=5 

April-1999 

May-1999 

N=5 

N=10 

Figure D4. Overall and monthly diet composition (% wet weight) of Diplotaxodon limnothrissa . 


Overall 

Copepods 


Sand 

N=20 


No ID 

Detritus 

Others 

November-1998 

N=1 

December-1998 

N=4 

January-1999 

N=2 

February-1999 

N=1 

March-1999 

N=1 

April-1999 

N=6 

May-1999 

N=5 

Figure D5. Overall and monthly diet composition (% wet weight) of Diplotaxodon macrops . 



D. macrops is a deep water species found from 75 to 125 m. Stomachs with remaining 

items were not as frequent as for D. limnothrissa and only 20 specimens were analysed 

between November 1998 and May 1999. Weight of stomachs contents averaged 61.2 mg and 

ranged from 5.4 to 157.4 mg for fishes of 85 to 118 mm SL (17-42 g). Diet composition (as 

percentage of the wet weight) at each sampled month and all months pooled are presented in 

Figure D5. 

Turner (1996) stated D. macrops was zooplanktivorous. Our results tend to support this 

statement as 55% of the diet was made of copepods and Chaoborus larvae. The remaining 

part of the diet was constituted of chironomid larvae, sand, detritus non identified material 

and other items (scales, adult insects, cladocerans, macrophytes, diatoms and other algae). It 

is important to notice that at months when more than only one or two specimens were 

examined, zooplankton accounted for 84 to 99% of the diet (December 1998, April and May 

1999). Chironomid larvae constituted a significant part of the diet only once in November 

1998, with a single fish examined. Chaoborus larvae were dominant items in February and 

May 1999. Like for D. limnothrissa, diet composition strongly varied among months for D. 

macrops, which seemed to switch opportunistically on some preferred items according to their 

relative availability. No particular trend relative to size was observed over the size range 

examined. 

Diet composition of D. macrops was globally similar to that of D. limnothrissa, copepods and 

Chaoborus larvae accounting for most of their diet. However, their feeding strategy appeared 

slightly different as benthic invertebrates and important amount of sand were regularly found 

in D. macrops diet, suggesting a digging activity not observed in D. limnothrissa. 


L. argenteus (='longipinnis orange head') mainly occurs at depth between 10 and 30 

m. The stomachs of 34 specimens were examined during the period from November 1998 to 

May 1999. Weight of stomachs contents averaged 54.2 mg and ranged from 6 to 197 mg for 

fishes of 92 to 142 mm SL (37-87 g). Diet composition (as percentage of the wet weight) at 

each sampled month and all months pooled are presented in Figure D6. 

Diet composition of L. argenteus was almost identical from one month to another, being 

essentially constituted of chironomid larvae (49 to 82%), sand, detritus, non identified 

material and other items (scales, fish eggs, insects adults, larvae and pupae, crustacean 

zooplankton, nematods, gastropods, bivalvs, macrophytes, diatoms and other algae). Given 

the nature of the main food item and the presence of large amounts of sand at each month, this 

species seems to be a benthic invertebrate feeder specialised on chironomid larvae. 

194

Overall 


Sand 

N=34 

No ID 

Detritus 

Others 

November-1998 

N=4 

December-1998 

N=4 

January-1999 

N=5 

February-1999 

N=5 

March-1999 

N=5 

April-1999 

N=6 

May-1999 

N=5 

Figure D6. Overall and monthly diet composition (% wet weight) of Lethrinops argenteus . 



L. gossei is a deep water species mainly caught from 75 to 125 m depth. Despite is 

deep water existence, stomachs not inverted were found though they were seldom full. The 

stomachs of 21 specimens were examined between November 1998 and May 1999. Weight of 

stomachs contents averaged 72.7 mg and ranged from 5 to 210 mg for fishes of 101 to 155 

mm SL (33-118 g). Diet composition (as percentage of the wet weight) at each sampled 

month and all months pooled are presented in Figure D7. 

Turner (1996) stated this species fed on benthic arthropods. Our results rather indicated a 

mixed diet constituted of benthic invertebrates and zooplankton. Over the sampling period, 

dominant food items were Chaoborus larvae, chironomid larvae, diatoms and copepods, 

making up to 76% all together. The remaining part of the diet was made of unidentified 

material, detritus and other items (scales, insects adults, larvae and pupae, nematodes, 

macrophytes and other algae). Diet composition was highly variable from one month to 

another, being dominated by diatoms in December 1998 and January 1999, by chironomid 

larvae in April 1999 and by Chaoborus larvae in March and May 1999. Diatoms might appear 

to be ingested incidentally, but as only a very small amount of sand was found in the 

stomachs and given the large amounts of diatoms found, the hypothesis of accidental 

ingestion is unlikely. No particular trend relative to size was observed. L. gossei appeared to 

switch opportunistically on a few preferred food items according to their availability. 


M. anaphyrmus frequents the shallow waters between 10 and 50 m depth. The 


stomachs contents averaged 45.6 mg and ranged from 8 to 290 mg for fishes of 84 to 151 mm 

SL (16-125 g). Diet composition (as percentage of the wet weight) at each sampled month and 

all months pooled are presented in Figure D8. 

This species is known as a gastropod feeder (McKaye et al. 1986, Eccles & Trewavas 1989, 

Konings 1995, Turner 1996, Msukwa & Ribbink 1997), although copepods, chironomids, 

algal remains and arthropod material are sometimes found (Turner 1996). Most of the 

specimens we examined had lots of snail remains in their guts. However, what is presented 

here is only the stomach content analysis. Gastropods made up to an averaged 16% (4-44%) 

of the stomach content only. Chironomid larvae accounted for 41%, adult insects for 10% and 

crustacean zooplankton for 5%. The remaining part of stomach content was made of sand, 

non identified material and other items (scales, chironomid pupae, Chaoborus larvae, insect 

pupae and larvae, nematodes, bivalves, macrophytes, diatoms and other algae). Dominant 

components of stomach contents were the same from one month to another, but the relative 

proportion of these items slightly varied among months. No particular trend relative to size 

was observed except that individuals below 100 mm SL tended to have higher proportions of 

chironomid larvae in their stomachs; more than 99% for the two smallest specimens (80 and 

84 mm). 

195

Overall 



Copepods 

N=21 

Non ID 

Detritus 

Diatoms 

Others 

December-1998 

January-1999 

N=5 

N=5 

March-1999 

April-1999 

N=5 

N=2 

May-1999 

N=4 

Figure D7. Overall and monthly diet composition (% wet weight) of Lethrinops gossei . 


Overall 


Insect adults 

Zoopk 

N=36 

Gastropods 

November-1998 

N=5 

December-1998 

N=5 

Sand 

No ID 

Others 

January-1999 

N=5 

February-1999 

N=5 

March-1999 

N=5 

April-1999 

N=5 

May-1999 

N=5 

Figure D8. Overall and monthly diet composition (% wet weight) of Mylochromis anaphyrmus . 


Overall 


Copepod 

N=13 

Sand 

Detritus 

Non ID 

Others 

January-1999 

February-1999 

N=5 

N=5 

April-1999 

May-1999 

N=2 

N=1 

Figure D9. Overall and monthly diet composition (% wet weight) of Taeniolethrinops praeorbitalis 


Taeniolethrinops praeorbitalis 

T. praeorbitalis is a shallow water species encountered from 10 to 30 m depth. The 


stomachs contents averaged 49.3 mg and ranged from 5 to 142 mg for fishes of 97 to 193 mm 

SL (24-166 g). Diet composition (as percentage of the wet weight) at each sampled month and 

all months pooled are presented in Figure D9. 

This species is known to feed primarily on chironomid larvae (Fryer 1959, Eccles & 

Trewavas 1989, Turner 1996). However, specimens examined by Fryer (1959) were 

sometimes full of nematodes. Jackson (cited by Turner 1996) reported T. praeorbitalis fed 

mostly on Chaoborus larvae. Konings (1995) stated its main food is insect larvae. Detritus, 

diatoms and sand were also reported to occur in its diet (Turner 1996). The 13 specimens we 

examined fed largely on chironomid larvae, which averaged 46% of the diet. Copepods 

accounted for 13%, though they had been abundant in one specimen only in May 1999. The 

remaining components of the diet were large amounts of sand, detritus, non identified 

materials and other items (scales, insect adults and pupae, nematodes, macrophytes, 

cladocerans, diatoms and other algae). Apart from the one specimen with lots of copepods in 

May, no temporal variability in diet composition was noticed, the only variation being the 

relative proportions of chironomids and sand between months. The small specimens (90-130 

mm SL) analysed tended to have a smaller proportion of chironomid larvae in their stomachs 

and larger proportions of copepods and plant material than large individuals (165-195 mm). 

With the exception of a few species, which actually feed upon fish (A. mentale, D. 

limnothrissa…), the presence of fish scales in almost every stomach analysed, regardless of 

species, probably represents an artefact caused by the piling up of fishes during the trawl 

hauling. Indeed, at every haul, medium and large fish, including non piscivorous species, had 

small fish stuck in their mouths. 

Stable isotopes analysis 

As already emphasised by Bootsma et al. (1996), the use of both δ 13 C and δ 15 N 

signatures is very useful in separating fish with identical δ 13 C signals but different feeding 

habits. Without the δ 15 N signatures, it would have been impossible to distinguish between 

species such as C. virginalis, small D. limnothrissa, L. gossei and large A. mentale for 

example, which all have very different feeding regimes: zooplankton feeders, benthic 

invertebrate feeder and piscivore. 

Except for M. anaphyrmus, there was significant intra-specific variability of both δ 13 C and 

δ 15 N for all species (Figure D10). This was particularly striking for A. mentale, which covered 

a δ 13 C range of 4°/ oo and about 2.5°/ oo in δ 15 N. For most of the species, this variability was 

mainly explained by size differences among individuals, as illustrated by the circles and 

written sizes on Figure D10. In general, smaller specimens had a lower δ 15 N signature than 

larger ones, indicating they were on a lower trophic level. Intra-specific variations of δ 13 C 

also showed small specimens fed on different items than large ones. As a consequence, small 

and large specimens of the species A. mentale, C. virginalis, D. limnothrissa and T. 

praeorbitalis, were subsequently separated in the analysis. 

The average isotopic composition of the nine target species and their potential food 

sources are presented on Figure D11. Owing to low sample size for most of the food sources, 

samples were analysed for δ 13 C only and few δ 15 N signatures are available. Adult mayfly 

196

-17,00 

5,00 6,00 7,00 8,00 9,00 10,00 

-18,00 

90-132 mm 

-19,00 

162-188 mm 

Amen 

-20,00 

Cvir 

Delta 13 C 

-21,00 

-22,00 

-23,00 

-24,00 

-25,00 

70-80 mm 

73-75 mm 

100-128 mm 

113mm 

94-103 mm 

140 mm 

190-200 mm 

Dmac 

Larg 

Tpra 

Amac 

Mana 

Dlim 

Lgos 

-26,00 

-27,00 

142 mm 

Delta 15 N 

Figure D10. Individual isotopic composition of some demersal cichlid fish species in South 

West Arm of Lake Malawi. Amac = Alticorpus macrocleithrum, Amen = Alticorpus 

mentale, Cvir = Copadichromis virginalis, Dlim = Diplotaxodon limnothrissa, Dmac = 

Diplotaxodon macrops, Larg = Lethrinops argenteus, Lgos = Lethrinops gossei, Mana = 

Mylochromis anaphyrmus, Tpra = Taeniolethrinops praeorbitalis. Numbers represent the 

standard length or standard length ranges of specimens. 

197

isotopic composition is not displayed on this figure because they can not be a significant food 

sources for these fish owing to their high mean δ 15 N signatures: 7.82. The nine fish species 

displayed a δ 15 N range of just over 3°/ oo , which corresponds approximately to one trophic 

level (3 to 5°/ oo Peterson & Fry 1987, Hesslein et al. 1991, Bootsma et al. 1996). For most of 

the nine fish species, stable isotope results were consistent with stomach content analysis. 

A. macrocleithrum had the second highest δ 15 N signature (8.49), just below that of the 

piscivorous A. mentale. It was found to have a mixed diet composed mainly of benthic 

invertebrates, chaoborid larvae and unidentified material. Zooplankton and oligochaetes were 

also regularly found in its stomachs. The isotopic composition of this fish was consistent with 

these observations as it was intermediate between the lightest δ 13 C signatures (zooplankton 

Diaphanosoma excisum, Tropodiaptomus cunningtoni and Oligochaetes at 100 m) and the 

heavier Chaoborus larvae, average zooplankton and sediment between 75 and 125 m (Figure 

D11). 

The large specimens (190-200 mm SL) of A. mentale had the highest δ 15 N signature, 

as expected from its almost strictly piscivorous habits (Figure D11). The smaller specimens 

(113-142 mm) had a lighter δ 15 N and lighter δ 13 C composition. Stomach contents of all the 

small specimens analysed contained over 60% zooplankton, except for one individual who 

had fed on fish (95%). Carbon isotopic composition of small A. mentale matched these 

observations, being intermediate between the different zooplankton species. 

C. virginalis had amongst the lowest δ 15 N in muscle. Stomach content analysis 

revealed this species feeds almost exclusively on zooplankton, which was supported by its 

isotopic composition, right in the range of the different zooplankton species (Figure D11). 

The δ 15 N difference observed between small and large specimen is likely due to a selective 

predation upon different zooplankton species. Large specimens of C. virginalis probably feed 

more upon larger predatory zooplankton species than small ones. 

Large D. limnothrissa specimens (100-128 mm SL) had a higher average δ 15 N 

signature (7.71) than the small specimens (70-80 mm) (6.56), indicating they relied on food 

sources of a slightly higher trophic level (Figure D11). Temporal trends of stomach content 

analysis revealed a mixed diet composed of zooplankton, Chaoborus larvae, usipa and adult 

insects. This is supported by its δ 13 C signature, which is slightly above those of zooplankton 

species, Chaoborid larvae, mixed adult insects and below the signature of some insect species 

such as Eatonica shoutedini (Ephemeroptera). It must be stressed that the Coleoptera can not 

account for an important part of D. limnothrissa diet because their mean δ 15 N signature is 

only about 1°/ oo lighter. 

D. macrops mean δ 15 N signature (8.49) places this species on top of the represented 

trophic level with A. mentale and A. macrocleithrum (Figure D11). As for D. limnothrissa, 

stomach content analysis revealed a mixed diet in which crustacean zooplankton and 

Chaoborus larvae accounted for more than 50%. Its mean δ 13 C isotopic signature supported 

these results. It was about 2°/ oo above the main zooplankton and chaoborid larvae signatures, 

indicating that this species relies also on heavier 13 C sources. Adults insects such as 

Coleoptera, which indeed have heavier 13 C signals, were often found in stomachs. 

L. argenteus and T. praeorbitalis had almost exactly the same isotopic composition (Figure 

D11) and the heaviest δ 13 C signature after the small specimens of T. praeorbitalis. 

Monthly stomach analysis for both species showed very similar diet composition mainly 

made of chironomid larvae, sand, detritus and other items. However, L. argenteus and T. 

praeorbitalis δ 13 C signatures were heavier than expected if they were relying only upon 

chironomid larvae at 10, 30 and 50m (which is the depth distribution of these species in the 

sampled area, except for T. praeorbitalis only found at 10 and 30 m). As these species are 

198

-17 

5,50 6,50 7,50 8,50 9,50 

POTENTIAL FOOD SOURCES 

-18 

Chironomid adult and pupae 100m 

Tpra S 

-19 

Delta 13 C 

-20 

-21 

-22 

-23 

-24 

Larg 

Mana Tpra L 

Dlim L 

Dlim S 

Cvir L 

Cvir S 

Amen S 

Lgos 

Amac 

Dmac 

Amen L 

Eatonica shoutedini (Ephemeroptera) 

Coleoptera 100m, Sediment 10-30m 

Sediment 50m 

Chironomid adult 100m 

Gastropod 30m 

Chironomid larvae 30m 

Adult insects 

Nematode 30m, Chironomid larvae 30m, 

Oligochaete 30m, sediment 75-125m 

Chironomid larvae 50m, Phytoplankton 

Zooplankton, Chironomid larvae 

Chaoborus larvae, Chironomid adult 10m 

Oligochaete 100m 

-25 

Delta 15 N 

Diaphanosoma excisum, 

Tropodiaptomus cunningtoni 

Figure D11. Mean isotopic composition (± standard deviation) of some demersal cichlid fish 

species and their potential food sources in South West Arm of Lake Malawi. Amac = 

Alticorpus macrocleithrum, Amen = Alticorpus mentale, Cvir = Copadichromis virginalis, 

Dlim = Diplotaxodon limnothrissa, Dmac = Diplotaxodon macrops, Larg = Lethrinops 

argenteus, Lgos = Lethrinops gossei, Mana = Mylochromis anaphyrmus, Tpra = 

Taeniolethrinops praeorbitalis. S and L refer to small and large specimens of a given 

species, respectively. The numbers correspond to the depth at which the samples were 

collected. Adult insects refer to the averaged δ 13 C values of Hemiptera, Ephemeroptera and 

Corixidae. Zooplankton refers to the averaged δ 13 C values of various crustacean copepod 

species (Diaptomus vraepelini, D. dimixtus, Mesocyclops neglectus, M. leukarti). 

199

ottom feeders ingesting large amounts of sand, they probably also have other food sources 

with heavier carbon signals like periphyton (Bootsma et al. 1996) or other plant material, 

which were indeed regularly found in stomachs. 

As observed in stomach content analysis, small specimens (90-130 mm SL) of T. 

praeorbitalis tended to have a different diet composition than large individuals (160-195 

mm), which was reflected in their isotopic composition. Small individuals had heavier δ 13 C 

and lighter δ 15 N signals. Stomach analysis revealed that apart from chironomid larvae, 

zooplankton and plant material were important components of their diet. However, as the 

zooplankton has a lighter 13 C signal than chironomid larvae found in shallow water, it can not 

account for the heavier 13 C signature of small T. praeorbitalis compared to large ones. 

Oligochaetes, which are difficult to identify in stomachs when digested, have been found to 

account for more than 50% of T. praeorbitalis diet (W. Darwall, pers. com.) and might have 

constituted an important part of the unidentified material in our analysis. Unfortunately, we 

did not get enough oligochaetes at 10 m, where all the T. praeorbitalis analysed for isotopic 

composition were caught, to analyse them for isotopic composition. Nevertheless, given the 

large difference in δ 13 C signal between oligochaetes at 100 m and 30 m (Figure D11), 

oligochaetes are likely to have a heavier 13 C signature in the shallow waters, but probably not 

heavy enough to actually account for a large part of T. praeorbitalis isotopic composition. 

The sand-digging habits of this species suggests that it may rely on vegetal materials 

(periphyton debris) with heavier 13 C signals found in sediment. The higher δ 13 C and lower 

δ 15 N suggest that small T. praeorbitalis may rely even more than adults on benthic algae, and 

may occupy a shallower habitat. 

The deep water L. gossei had an average δ 15 N signature among the highest (8.12) 

(Figure D11). Its diet composition as revealed by stomach contents varied much from one 

month to another, the main food items being Chaoborus larvae, chironomid larvae, diatoms, 

crustacean zooplankton and other less important items. Its δ 13 C signature, about 1°/ oo above 

that of Chaoborus larvae and zooplankton, might indicate that this species mainly relies on 

these food sources. However, as we do not have carbon signals of chironomid larvae in deep 

waters nor of diatoms, their potential importance in L. gossei diet can not be excluded. 

M. anaphyrmus is commonly referred to as a gastropod-eater and most of the 

specimens examined had lots of snail remains in their guts. However, gastropods represented 

only a small fraction (16%) of the stomach contents in which chironomid larvae accounted for 

41%, adult insect 10% and zooplankton 5%. Despite the weak occurrence of gastropods in 

stomach contents, M. anaphyrmus 13 C average signature (–20.17) was consistent with a diet 

mostly based on gastropods, for which the average signature at 30 m (most common depth of 

the species in the sampled area) was –22.02 (Figure D11). Usually a food source is on average 

1°/ oo heavier in 13 C than its consumer. The 2% found here might be explained by a preference 

for particular gastropod species with slightly lighter signatures or by feeding partly at 

shallower depths, where gastropods can be expected to have heavier signals (Bootsma et al. 

1996). 

Conclusions 

Despite the high variability of stomach fullness encountered during this study, 

particularly for the deep water species, a good correspondence between the results of stomach 

content and stable isotope analysis was observed for most of the nine fish species. Stable 

isotope results proved very useful in clarifying the observed patterns. Species with the 

narrower feeding regime such as A. mentale or C. virginalis had an isotopic composition 

exactly matching the stomach content observations. For M. anaphyrmus, despite the 

200

dominance of other food items in the stomachs at every sampled months, the isotopic 

composition confirmed the previously reported snail diet of the species (McKaye et al. 1986, 

Eccles & Trewavas 1989, Konings 1995, Turner 1996, Msukwa & Ribbink 1997). Stomach 

content analysis indicated regular temporal trends of diet composition for both L. argenteus 

and T. praeorbitalis, mainly dominated by chironomid larvae. However, stable isotope 

analysis revealed that the apparently secondary vegetal food items were playing an important 

role in the diet of these species, particularly for the small specimens of T. praeorbitalis. 

Isotopic composition clearly illustrated the diet separation among these demersal fish: on one 

hand there is the pelagic phytoplankton food chain, centered on –23°/ oo with at least two 

trophic levels and intermediate feeding levels and on the other hand there is an ascending line 

towards heavier carbon source, which is most likely a periphyton-based food web represented 

by T. praeorbitalis. The ascending line to heavier carbon and lighter nitrogen likely represents 

mixed feeding on the periphyton-based source and the phytoplankton-based source, though 

there was no piscivore specialising on T. praeorbitalis in our collection. This is interesting in 

that it does indicate that benthic algal production is contributing to energy flow of the 

demersal fishes even well away from shore. 

Species with a complex feeding regime such as A. macrocleithrum, D. limnothrissa, 

D. macrops and L. gossei showed important temporal variations in diet composition. These 

variations, which at first look appeared to be of opportunistic nature, are likely to be 

influenced by seasonal trends in food availability. Indeed, these four species fed upon 

Chaoborus larvae when available. For each of these species, Chaoborus larvae were dominant 

items in their stomachs at exactly the same months, March and May, and only at these 

months. This strongly suggests that the observed opportunistic feeding regimes were related 

to the seasonal and/or temporal fluctuations of their preferred food sources availability. 

These results emphasise the complementary nature of stomach content and stable 

isotope approaches in the study of feeding habits and trophic patterns of complex fish 

communities. There appears to be a heavy reliance by most demersal fish on benthic 

organisms as a food source and little complete dietary overlap, supporting the belief that these 

fish must partition their resources in order to coexist (Bootsma et al. 1996, Turner 1996). If 

the habitat was to become more homogeneous (as a result of increased sedimentation, reduced 

water clarity, etc.), it can be expected that benthic organisms will be affected and that the 

potential for competitive exclusion will increase. 

201

Chapter 5: 

Morphometric, genetic and ecological 

comparison of two important 

demersal species along a gradient 

from the South West Arm to Nkhata 

Bay

Chapter 5: Morphometric, genetic and ecological comparison of 

two important demersal species along a gradient from the South 

West Arm to Nkhata Bay 

F. Duponchelle, J. Snoeks, M. Hanssens, J-F. Agnèse, A.J. Ribbink, A. Msukwa, J. Mafuka & 

D. Mandere 

Introduction 

The state of the demersal trawling fisheries in Lake Malawi has been monitored since 

the development of this industry in 1968 (Tweddle & Magasa 1989). Various authors have 

reported the consequences of trawling activities on fish communities, such as the profound 

size structure modifications, the decreased occurrence of some species and the disappearance 

of some others in the SEA, and drew the attention to the associated dangers for biodiversity 

conservation (Turner 1977a, 1977b, Turner 1995, Turner et al. 1995). However, demersal 

trawling only occurs in a restricted area in the southern part of the lake, representing less than 

five percent of its total surface. A recent report from the Fisheries Department revealed that 

most of the exploited species in the SE and SW Arms also occur in the non exploited trawling 

grounds between Dormira Bay and Nkhata Bay (Banda & Tómasson 1996, Tómasson & 

Banda 1996). When considering both fisheries management and biodiversity conservation 

issues, a logical question arises: are the species that have disappeared from the catches in the 

SEA really endangered if they also occur in other areas of the Lake? Although some of these 

species are supposed to occur only in restricted parts of the lake (Turner 1996), their degree of 

stenotopy is unknown and a decreasing stock might be repopulated from other geographical 

areas. However, a single species present in geographically distant parts of the lake might be 

composed of a single widespread population, or of different populations (or "stocks" in 

fisheries language). For the conservation of biodiversity as well as for the fisheries 

management, it appears crucial to know whether a species is represented by a single 

population distributed all over the lake, or by different populations with distinctive 

morphometric, genetic and life-history characteristics. If the threatened species of the 

southern arms were to be composed of a single widespread population, their disappearance 

from the exploited trawling areas would not represent an irreversible threat for the 

biodiversity. On the other hand, if they were distinct populations with different morphometric, 

genetic and life-history characteristics, their disappearance would lead to an irreversible loss 

of biodiversity. The starting hypothesis for this study is that shallow-water species have more 

chances to encounter physical barriers to their movements and therefore are more likely to be 

structured in distinct populations than deep-water species. A study was undertaken in 

collaboration with the taxonomists of the project, to compare the morphometrics, the genetics 

and some life-history traits of two species (a shallow-water and a deep-water species) from 

four different locations between the SWA and Nkhata Bay (Figure P1). This would allow us 

to assess whether they are part of a single widespread population or of distinct populations 

and consequently whether eventual differences are related to geographical distance and depth. 

202

Tanzania 

Chintheche 

Malawi 

Mozambique 

Nkhotakota 

Kirambi Point 

Domira Bay 

Leopard Bay 

SWA 

SEA 

Figure P1. Map of Lake Malawi showing the sampling sites for Mylochromis anaphyrmus 

(white squares) and Lethrinops gossei (black squares) populations comparison. 

203

Material and Methods 

The shallow-water species, Mylochromis anaphyrmus, was sampled in the north of the 

SWA (where the monthly survey was done), Dormira Bay, Nkhotakota and Chinteche in 

February 1999. M. anaphyrmus was chosen for this study as a representative of the shallow 

water demersal community. It was one of the target species for the life-history studies within 

the project. Reasons for this choice are the fact that this species is relatively easily to identify, 

and common within its distribution range. Eccles & Trewavas (1989) reported that M. 

anaphyrmus is very common over sandy substrates, in waters of 15 to 35 meters depth in the 

southern part of the lake. They added that it was also reported from Nkhotakota. Konings 

(1995) stated that M. anaphyrmus is endemic to the southern and western parts of the lake. 

Turner (1996) reported that this species was present in 42 out of 57 experimental trawl 

catches between 18 and 72 m depth in the South-East Arm where it is often one of the most 

abundant species. Banda & Tómasson (1996) reported it in the SW and SE Arms as well as 

from Domira Bay to Nkhata Bay. 

The deep-water species, Lethrinops gossei was chosen for the same reasons. It is an abundant 

species in deep water and was one of the target species for life-history studies within the 

project.. Eccles & Trewavas (1989) reported that it dominates the benthic community at 

depths of 92-130 m in the SEA. Turner (1996) stated it was one of the dominant species at 

depths of 90 m or more in the SEA. He also reported what appeared to be a female L. gossei 

caught off Karonga, in the far north of the lake. Banda & Tómasson (1996) reported it in the 

SW and SE Arms as well as from Domira Bay to Nkhata Bay. It was sampled during the same 

February 1999 cruise in the north of the SWA, where the monthly survey was done, off 

Leopard Bay, off Kiramby Point, in Nkhotakota and Chinteche (Figure P1). 

It was planned that at each location, 30 specimens of each species were to be collected for 

morphometric analyses, 75 to 100 specimens for genetics and 100 to 200 specimens for life 

history traits analysis. L. gossei proved to be rare in Chinteche area between Bandawe Point 

and Sanga Point so that only 35 specimens were caught as a whole. 

Morphometric analysis 

At the time we were writing this report, the morphometric analysis of L. gossei was 

not yet finished so that only the analysis for M. anaphyrmus will be presented here. 

Ideally, for each locality 15 specimens of each sex, all of similar size, should have been 

preserved. For some localities this could not be done. Hence, the results and analyses are 

based on a lower number of specimens. No female specimens were preserved from 

Nkhotakota. On all specimens 23 measurements and 17 counts were taken following Snoeks 

(1994). Two techniques were used to explore and analyse the metric data: Principal 

Component Analysis (PCA) on the log-transformed measurements and Mann-Whitney U- 

Tests on the relative measurements (percentages). 

List of abbreviations used: 

LacD : lachrymal depth ; SnL : snout length ; LJL : lower jaw length ; ChD : cheek depth ; 

EyeD : eye diameter ; IOW : interorbital width ; HW : head width ; HL : head length ; SL : 

standard length ; BD : body depth ; DFB : dorsal fin base ; AFB : anal fin base ; PrD : 

predorsal distance ; PrP : prepectoral distance ; PrV : preventral distance ; PrA : preanal 

distance ; CPL caudal peduncle length ; CPD caudal peduncle depth ; PhJL : pharyngeal jaw 

length ; PhJW : pharyngeal jaw width ; DeArL ; dentigerous area length (pharyngeal jaw) ; 

DeArW : dentigerous area width ; UJT : number of outer teeth in the upper oral jaw ; LJT : 

number of outer teeth in the lower oral jaw. 

204

Genetic analysis 

Microsatellites variability for the two species has been investigated using primers already 

defined by Rico et al. (1993,1996) and Zardoya et al. (1996). These primers have been 

already tested across a panel of very diverse fish species and successfully amplified in nonsource 

species due to a high level of conservation of the flanking regions of these 

microsatellites. Eight different loci have been tried (Table P1). Numerous amplification 

conditions were used for each pair of primers with variations in the concentration of MgCl2 

and of the annealing temperature. 

For locus GMO 2 we always observed two bands of respectively 250 and 290 bp in both 

species. Locus GMO 132 gave also two bands of 50 and 300 bp. At loci GMO 145 and CIER 

51, we did not obtain any band. A 310 bp band was observed for M. anaphyrmus at locus 

CIER 62 and a 250 bp band for both species at locus TMO M25. 

Only locus TMO M5 and TMO M 27 gave, for both species, bands varying in size 

from 300 to 400 bp for TMO M5 and from 230 to 250 bp for TMO M27. 

PCR conditions were as follows: DNA was amplified (94°C 60 s, 48°C 60s, 72°C 60s) for 35 

cycles in 20 µl volumes (1 x polymerase buffer, 1.5 mM MgCl2, 0.4 mM of each dNTP, 75 

ng of each primer, and 1 unit of Taq Promega). 

The genetic population structure has been statistically described using Wright (1969) “F 

statistics” indices with Weir & Cokerham (1984) formulas. Fis measure the deficit of 

heterozygous due to non-random mating in a (sub)population while Fst measure the loss of 

heterozygosity due to the subdivision of the sample in two or more populations. 

Isolation by distance was tested using Mantel's test (Mantel 1967). Mantel's test consists of a 

comparison of two matrices (here Fst versus geographical distances). This test determines if 

there is a correlation between the two matrices. The Mantel coefficient Z is calculated from 

the real data and then the data are permuted to obtain pseudo matrices and the corresponding 

pseudo Z’ values. The various Z’ values obtained are compared to the Z values. If Z is 

statistically different of all Z’, then the two matrices are correlated. 

All coefficient and statistical analysis have been done using GENETIX programme (Belkir et 

al. 1996) 

Life history traits analysis 

For both species, length-weight relationships, percentage of ripe females, and fecundity 

were compared between populations. As it needs to be estimated during the peak of the 

breeding season, which may vary at each site, size at maturity was not compared between 

populations. Determination of life history traits was done as described in Chapter 2. 

Comparison of the percentage of ripe females between populations was carried out using a 

Kruskal-Wallis one way ANOVA on ranks (Sherrer 1984). 

As for most of the fish species, length-weight relationship among populations of the two 

species were characterised by the following equation: W = a.L b . Direct comparison of 

populations using the maximum likelihood method (Tomassone et al. 1993) was not possible 

because regression residuals increased with length. Length and weights were then 

logarithmically transformed (ln), which lead to a linear relationship between length and 

weight. Therefore, estimation of differences between populations was investigated by 

comparing regression lines between length and weight. The regressions were compared by an 

analysis of covariance (Scherrer 1984) followed by a 2 × 2 comparison method. First, slopes 

were compared, and populations whose slopes were not significantly different were then 

205

compared for intercepts. As the type I error increases when more than two populations are 

compared pairwise (Scherrer 1984), a probability α' was calculated so that the overall α (α = 

0.05 in our case) was maintained over the k(k-1)/2 comparisons. The new α' was calculated 

by the following formula: α' = 1-(1-α) 2/(k(k-1)) . 

Rather than comparing regressions between fecundity and body weight over a narrow weight 

range, we compared relative fecundity (fecundity per kg of body weight). Over the weight 

ranges studied, no correlation was found between relative fecundity and body weight, as it 

may happen in cichlids (Legendre 1992). Relative fecundity was compared using one way 

ANOVA followed by Tukey's all pairwise multiple comparison test (Scherrer 1984). 

Table P1. Microsatellites loci and primers tested. 

Locus Primers references 

GMO 2 F ccctcagattcaaatgaagga Rico et al., 1993 

R gtgtgagatgactgtgtcg Rico et al., 1996 

GMO 132 

F ggaacccattggattcaggc 

R cgaaaggacgagccaataac 

GMO 145 

F gcattgtaggaacaacaattaac 

R gtgcatgtgctcattatagc 

CIER 51 

F gccaaaacactgacgaggtga 

R tttgcgcaagcttcaggatga 

CIER 62 

F ggtgctgtcacttttggccac 

R aactctgctggtcgccactcc 

TMO M5 F gctcaatattctcagctgacgca Zardoya et al., 1996 

R aga aca gcg ctg gct atg aaa agg t 

TMO M25 

F ctgcagtggcacatcaagaatgagcagcggt 

R caagaacctttcaagtcattttg 

TMO M27 

F aggcaggcaattaccttgatgtt 

R tactaactctgaaagaacctgtgat 

206

Results and Discussion 

Morphometric analysis 

Principal Component Analysis. 

PCA log measurements 

2 

1 

PC 3 : EyeD, ChD, BD 

0 

-1 

-2 

-2,5 -1,5 -0,5 0,5 1,5 2,5 

PC 2 : CPD, DFB, DeArL, PhJL, AFB, PhJW 

m SWA 

f SWA 

m CHIN 

f CHIN 

m DOMB 

f DOMB 

m KOTA 

Figure P2. PCA of the log-transformed measurements, all specimens included (n=60). 

A first PCA (Figure P2) on all specimens showed that morphological differences 

between males and females are generally larger than the differences observed between the 

populations. Male and female specimens are partly separated on PC2; female specimens are 

shifted towards the positive, male specimens towards the negative side of PC2. Polygons are 

drawn for the males only. All further analyses were therefore based on either males or females 

alone. 

In this plot SWA and CHIN males are relatively well separated on PC3 from the KOTA 

males. The DOMB males overlap with all other populations on PC3. This result is surprising 

since the South West Arm and Chinteche are the most distant localities sampled and we 

would have expected differences to be most prominent between these two populations. The 

third component is mainly defined by eye diameter and cheek depth (both measurements are 

obviously strongly correlated). This surprising result is probably due to the fact that all 

specimens were included in this analysis. The larger morphological differences that are found 

between sexes are mixed and analysed together with the inter population morphological 

differences and may therefore ‘blur’ the latter ones. Therefore another PCA was done, 

207

including only male specimens (Figure P3). The plot of the specimens on the second and third 

principal component shows a large overlap between almost all populations. Some distinction 

appears between the SWA and KOTA males on PC2. The SWA males are mainly on the 

positive part of PC2, the KOTA males are shifted towards the negative part of PC2. 

2 

PCA log measurements ; males 

1 

PC 3 : CPL, CPD 

0 

-1 

-2 

-3 

-2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 2,5 

SWA 

CHIN 

DOMB 

KOTA 

PC 2 : AFB, PrA, ChD 

Figure P3. PCA of the log-transformed measurements, male specimens (n=37). 

The factor loadings from this analysis indicated that the third component was almost entirely 

defined by a single character, the caudal peduncle length. After verification of the 

measurements, it appeared that one of the males collected in Chinteche was aberrant for this 

character. However, the measurement was checked and found to be correct. Consequently, the 

third principal component is unreliable to distinguish the populations (Figure P3). 

We therefore added another plot using the second and fourth principal component 

(Figure P4). On PC4, the CHIN males are relatively well separated from the males of the 

other populations. One CHIN specimen was aberrant and scored high on PC4 (Figure P4: 

specimen marked with arrow). This was again the same aberrant male with the high CPL 

value. 

208

2 

PCA log measurements ; males 

1 

PC 4 : ChD, CPD, DeArL 

0 

-1 

-2 

-3 

-2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 2,5 

PC 2 : AFB, PrA, ChD 

SWA 

CHIN 

DOMB 

KOTA 

Figure P4. PCA of the log-transformed measurements, male specimens (n=37). 

2,5 

PCA log measurements ; females 

2,0 

PC 3 : EyeD, BD, ChD, DeArL, PhJL 

1,5 

1,0 

0,5 

0,0 

-0,5 

-1,0 

-1,5 

-2,0 

-2,5 -1,5 -0,5 0,5 1,5 2,5 

PC 2 : EyeD, DeArW, CPL, PhJW 

SWA 

CHIN 

DOMB 

Figure P5. PCA of the log-transformed measurements, female specimens (n=23). 

A PCA on the log-transformed measurements for the female specimens was done as 

well (Figure P5). Here we see that the SWA population is morphologically more diverse than 

209

the other two (there are obviously more observations from the SWA which explains part of 

the higher variability) and overlaps with both other populations. On PC3 the CHIN females 

are neatly separated from the DOMB females. Factor loadings on PC3 indicate that they 

mainly differ in eye diameter and body depth. 

Further PC analyses were made to compare all populations two-by-two. These 

analyses did not show clear or better results than the previous PC analyses. Large overlaps 

were observed for almost all PCA analyses when only two populations were included. 

Mann-Whitney U-Test. 

The Mann-Whitney U-test was used to analyse morphological differences between all 

populations. Data used for these analyses are the relative measurements (percentages). 

We used only specimens of the same sex for these tests. To avoid allometric inferences, 

specimens were selected on the basis of their standard length for each test between two 

populations, so that the p-value for the standard length was close to 0.5 or larger. So, for each 

test a different subset of specimens was used, and in most cases not all specimens from a 

given locality were included in the analyses. A first test was done including male specimens 

from all populations. 

Table P2 gives an overview of the results of the analysis between all four populations. 

Table P2. Comparison of all four populations using the Mann-Whitney U-Test. Above 

diagonal the number of characters for which significant differences were found, p 

values as follows : * indicates p < 0,05; ** p < 0.005; *** p < 0.0005. Below diagonal p 

value for SL and number of specimens from each population used for the comparison. 

Characters for which significant differences were found are given and discussed in the 

text. 

SWA DOMB KOTA CHIN 

SWA 2 * 3 * 

1 ** 

4 * 

3 ** 

DOMB 

KOTA 

CHIN 

p=0.52 for SL 

n SWA 19 

n DOMB 10 

p=0.86 for SL 

n SWA 9 

n KOTA 9 

p=0.46 for SL 

n SWA 18 

n CHIN 16 

p=0.56 for SL 

n DOMB 8 

n KOTA 9 

p=0.93 

n DOMB 6 

n CHIN 12 

1 * 

1 ** 

p=0.52 

n KOTA 10 

n CHIN 14 

3 * 

1 ** 

1 ** 

1 *** 

The number of characters that differ between populations increases with increasing distance 

between the populations. In addition, the morphological differences found between 

populations become more significant with increasing distance. 

210

The characters that were found to be significantly different between the four populations are: 

SWA-DOMB. 

SWA-KOTA. 

SWA-CHIN. 

DOMB-KOTA. 

DOMB-CHIN. 

KOTA-CHIN. 

* : EyeD/HL and DeArW/PhJW 

* : HL/SL, PrD/SL and DeArW/PhJW 

** : PhJL/HL. 

* : LacD/HL, SnL/HL, LJL/HL and CPD/SL 

** : HL/SL, PrD/SL and PhJL/HL 

* : EyeD/HL 

** : DeArW/PhJW 

* : LJL/HL, ChD/HL, IOW/HL and DeArW/PhJW 

** : LacD/HL 

*** : ChD/HL 

The table and short overview show that all populations differ from each other for at least two 

characters. Some populations differ in a particular character from all or most other 

populations. For the dentigerous area width of the pharyngeal jaw the SWA population is 

significantly different from the DOMB and KOTA populations but not from the CHIN 

population. Moreover, for this character the DOMB population differs from all other 

populations. For the eye diameter a significant difference was observed between the DOMB 

population and the SWA and KOTA, but again not when compared to the CHIN population. 

The CHIN population is furthermore significantly different from the SWA and KOTA 

populations but not from the DOMB Population for the lachrymal depth. No consistent 

differences were found between the Nkhotakota population and the three other populations. 

Differences between females from three populations were analysed as well. 

Unfortunately three out of four female specimens from the DOMB population were 

significantly larger than the females from the SWA and CHIN populations. So, no comparison 

between DOMB and the two other could be made. As a result, only the SWA and CHIN 

populations could be compared. Seven SWA and five CHIN female specimens were used in 

the analysis, the p value for SL was 0.68. We found only two significant differences (p0.005) for the ChD/HL and DeArL/PhJL. 

Meristics 

For only three characters (the number of outer teeth in the oral jaws and the number of 

dorsal fin spines) significant differences were found between the populations. 

Plots of the number of outer teeth in the upper and lower oral jaws are given. Specimens are 

categorised by locality and sex. Polygons are only given for male specimens (Figures P6 and 

P7). 

Both graphs show that female specimens cannot be reliably distinguished on the basis of this 

character. The SWA females overlap with all other populations for the number of outer teeth 

both in upper and lower jaws. The only difference was noted for the outer teeth in the upper 

jaw, the female DOMB specimens have a higher number than the female CHIN specimens 

(52-58 vs 43-52). 

211

64 

number of outer teeth in the upper oral jaw 

60 

56 

52 

48 

44 

40 

36 

110 115 120 125 130 135 140 145 150 

SL 

m SWA 

f SWA 

m CHIN 

f CHIN 

m DOMB 

f DOMB 

m KOTA 

Figure P6. Plot of the number of outer teeth in the upper oral jaw vs SL, male specimens 

marked with polygons (n=60). 

48 

number of outer teeth in the lower oral jaw 

44 

40 

36 

32 

28 

24 

110 115 120 125 130 135 140 145 150 

SL 

m SWA 

f SWA 

m CHIN 

f CHIN 

m DOMB 

f DOMB 

m KOTA 

Figure P7. Plot of the number of outer teeth in the lower oral jaw vs SL, male specimens 

marked with polygons (n=60). 

212

For the male specimens, the differences on teeth number are clearer. The SWA males are 

clearly different from the KOTA males, in having a lower number of outer teeth (42-50 vs 50- 

62 UJT and 28-40 vs 36-42 LJT respectively). For the outer teeth number in the upper oral 

jaw, the range of both other populations (DOMB and CHIN) is intermediate. For the outer 

teeth number in the lower jaw, the DOMB population has a relatively large range that 

overlaps completely with the KOTA population and is different from the SWA population. 

The CHIN males have the largest range and overlap with all other populations. 

12 

10 

male 

8 

6 

4 

No of observations 

2 

0 

12 

10 

female 

8 

6 

4 

2 

0 

14 15 16 17 18 19 

SWA 

14 15 16 17 18 19 

DOMB 

14 15 16 17 18 19 

KOTA 

14 15 16 17 18 19 

CHIN 

number of dorsal spines 

Figure P8. Histogram with the number of dorsal spines, categorised by sex and locality. 

(n=60). 

The histogram of the number of dorsal fin spines (Figure P8) shows that in males there was a 

slight increase in spine number along the south-north axis. This pattern was not visible in 

females. In the SWA population, the modal number of dorsal spines is 16 for males (about 2/3 

16, 1/3 17); for the DOMB males both numbers are about equal; for the KOTA and 

particularly the CHIN males the majority have 17 dorsal fin spines. 

These results show the presence of small but significant morphological differences 

between all M. anaphyrmus populations examined. The amount of difference is clearly linked 

to geographical distance. 

213

Table P3. Allelic frequencies observed in the four populations of M. 

anaphyrmus: South West Arm (SW), Domira Bay (DO), Nkhotakota 

(NK) and Chinteche (CH). 

SW DO NK CH 

Locus TMO M5 

(N) 21 15 14 4 

295 0.00 0.00 0.00 0.25 

313 0.05 0.00 0.00 0.00 

315 0.00 0.13 0.00 0.00 

317 0.02 0.10 0.00 0.00 

321 0.05 0.00 0.04 0.00 

323 0.43 0.20 0.00 0.00 

327 0.00 0.03 0.00 0.50 

329 0.12 0.03 0.04 0.25 

331 0.00 0.13 0.14 0.00 

333 0.00 0.03 0.21 0.00 

335 0.17 0.07 0.29 0.00 

337 0.02 0.17 0.21 0.00 

339 0.00 0.00 0.07 0.00 

353 0.02 0.00 0.00 0.00 

361 0.05 0.00 0.00 0.00 

363 0.00 0.07 0.00 0.00 

365 0.02 0.00 0.00 0.00 

367 0.02 0.00 0.00 0.00 

369 0.02 0.03 0.00 0.00 

H 0.79 0.90 0.83 0.71 

H obs. 0.48 0.67 0.50 0.50 

Locus TMO M27 

(N) 17 30 15 21 

236 0.00 0.02 0.00 0.05 

238 0.74 0.55 0.87 0.45 

240 0.00 0.08 0.03 0.05 

242 0.26 0.23 0.10 0.45 

244 0.00 0.12 0.00 0.00 

H 0.40 0.63 0.25 0.60 

H obs. 0.29 0.60 0.13 0.67 

H obs. 0.39 0.63 0.32 0.58 

A 7.00 8.00 5.00 3.50 

214

Genetic analysis 

Allelic frequencies obtained are summarised in Table P3 and P4 for Mylochromis 

anaphyrmus and Lethrinops gossei, respectively. 


Four populations were investigated: South West Arm (SWA), Domira Bay (DOMB), 

Nkhotakota (KOTA) and Chinteche (CHIN). 19 alleles were observed at locus TMO M5 and 

5 at locus TMO M27. Fis and Fst values are summarised in Table P5 and Table P6, 

respectively. 

Table P5. Fis Values in M. anaphyrmus populations 

for both TMO M5 and TMO M27 loci. 

Statistically significant values are underlined. 

Locus TMO M5 TMO M27 

South West Arm 0.3976 0.2727 

Domira Bay 0.2689 0.0526 

Nkhotokota 0.4052 0.4667 

Chinteche 0.323 -0.1133 

At locus TMO M5, three out of four Fis values were statistically significant indicating that 

there was an excess of homozygous. Only the Chinteche sample did not exhibit such 

homozygous excess but it was likely due to the low sample size (4). No significant values 

have been observed for locus TMO M27. 

Table P6. Fst Values in M. anaphyrmus populations for all loci 

combined, TMO M5 alone and TMO M27 alone. 

Statistically significant values are underlined. 

All loci DOMB KOTA CHIN 

SWA 0.038 0.1057 0.1563 

DOMB _ 0.0639 0.0968 

KOTA _ 0.2018 

Locus TMO M5 DOMB KOTA CHIN 

SWA 0.047 0.1348 0.1984 

DOMB _ 0.0453 0.1341 

KOTA _ 0.1851 

Locus TMO M27 DOMB KOTA CHIN 

SWA 0.035 0.0257 0.0824 

DOMB _ 0.0947 0.0378 

KOTA _ 0.2304 

215

Table P4. Allelic frequencies observed in the five populations of L. gossei: 

South West Arm (SW), Leopard Bay (LB), Kiramby Point (KP), Nkhotakota 

(NK) and Chinteche (CH). 

SW LB KP NK CH 

Locus TMO M5 

(N) 17 19 2 11 3 

237 0.00 0.00 0.00 0.05 0.00 

293 0.00 0.00 0.00 0.14 0.00 

299 0.00 0.00 0.00 0.18 0.00 

317 0.00 0.03 0.00 0.00 0.00 

319 0.00 0.03 0.00 0.00 0.00 

321 0.00 0.00 0.00 0.23 0.00 

323 0.00 0.00 0.50 0.00 0.00 

325 0.00 0.05 0.00 0.00 0.50 

327 0.09 0.05 0.25 0.05 0.00 

329 0.03 0.00 0.00 0.00 0.00 

331 0.06 0.18 0.25 0.00 0.00 

333 0.03 0.00 0.00 0.00 0.00 

335 0.12 0.16 0.00 0.00 0.17 

337 0.09 0.00 0.00 0.00 0.00 

339 0.03 0.13 0.00 0.00 0.33 

341 0.09 0.03 0.00 0.00 0.00 

343 0.03 0.05 0.00 0.00 0.00 

345 0.06 0.08 0.00 0.00 0.00 

347 0.03 0.00 0.00 0.00 0.00 

349 0.00 0.03 0.00 0.00 0.00 

351 0.00 0.03 0.00 0.00 0.00 

353 0.00 0.05 0.00 0.00 0.00 

355 0.12 0.00 0.00 0.23 0.00 

357 0.06 0.03 0.00 0.00 0.00 

359 0.00 0.03 0.00 0.00 0.00 

361 0.00 0.00 0.00 0.14 0.00 

365 0.03 0.03 0.00 0.00 0.00 

369 0.12 0.00 0.00 0.00 0.00 

373 0.00 0.03 0.00 0.00 0.00 

379 0.03 0.00 0.00 0.00 0.00 

H 0.94 0.92 0.83 0.86 0.73 

Hobs. 0.88 0.63 0.50 0.64 0.33 

Locus TMO M27 

(N) 27 29 24 27 6 

219 0.04 0.00 0.00 0.00 0.00 

225 0.02 0.00 0.00 0.00 0.00 

229 0.00 0.00 0.04 0.00 0.00 

232 0.00 0.02 0.00 0.00 0.00 

236 0.00 0.00 0.08 0.04 0.00 

238 0.78 0.86 0.56 0.81 0.83 

240 0.02 0.05 0.00 0.00 0.00 

242 0.15 0.07 0.29 0.15 0.17 

244 0.00 0.00 0.02 0.00 0.00 

Hn.b. 0.38 0.25 0.60 0.32 0.30 

Hobs. 0.26 0.24 0.54 0.22 0.33 

Hobs 0.57 0.44 0.52 0.43 0.33 

A 10.5 10.5 4.0 5.0 2.5 

216

When using the two loci, all the populations were differentiated with statistically significant 

Fst values. All the populations but one were differentiated with locus TMO M5 alone and all 

but two with locus TMO M27 alone. 

Mantel test (using Fst and geographical distances matrices) obtained with locus TMO M5 

indicated a positive correlation between Fst and the geographical distances between 

populations (no Z’ value superior to Z) (Table P7). 

Table P7. Mantel matrices for Fst values (upper matrices) 

using all loci, TMO M5 alone or TMO M27 alone, and for 

geographical distances between M. anaphyrmus populations 

(lower matrices) 

All loci SWA DOMB KOTA CHIN 

SWA 0.038 0.1057 0.1563 

DOMB 60 _ 0.0639 0.0968 

KOTA 135 75 _ 0.2018 

CHIN 275 215 140 _ 

Z=226.81, 24 permutations, 5 values ≥ Z 

Locus TMO M5 SWA DOMB KOTA CHIN 

SWA 0.047 0.1348 0.1984 

DOMB 60 _ 0.0453 0.1341 

KOTA 135 75 _ 0.1851 

CHIN 275 215 140 _ 


Locus TMO M27 SWA DOMB KOTA CHIN 

SWA 0.035 0.0257 0.0824 

DOMB 60 _ 0.0947 0.0378 

KOTA 135 75 _ 0.2304 

CHIN 275 215 140 _ 


217


Five populations were investigated: South West Arm (SWA), Leopard Bay (LB) 

Kiramby Point (KP), Nkhotakota (KOTA) and Chinteche (CHIN). 30 alleles have been 

observed at locus TMO M5 and 9 at locus TMO M27. 

Fis and Fst values are summarised in Table P8 and Table P9, respectively. 

Table P8. Fis Values in L. gossei populations for both 

TMO M5 and TMO M27 loci. Statistically significant 

values are underlined. 

Locus TMO M5 TMO M27 

South West Arm 0.0698 0.318 

Leopard Bay 0.3229 0.0485 

Kiramby Point NA 0.120 

Nkhotokota 0.2708 0.3067 

Chinteche NA -0.1111 

Significant Fis values were observed at Leopard Bay and Nkhotakota for locus TMO M5 and 

at South West Arm for locus TMO M27. 

Table P9. Fst values in L. gossei populations for all loci, TMO M5 alone 

and TMO M27 alone. Statistically significant values are underlined. 

All loci LB KP KOTA CHIN 

SWA 0.0092 0.0485 0.0403 0.0595 

LB _ 0.0514 0.0727 0.0175 

KP _ 0.0886 0.07133 

KOTA _ 0.1080 

Locus TMO M5 LB KP KOTA CHIN 

SWA 0.0108 0.0462 0.0608 0.0999 

LB _ 0.0061 0.0932 0.0294 

KP 0.0983 0.0938 

KOTA _ 0.1584 

Locus TMO M27 LB KP KOTA CHIN 

SWA 0.0046 0.0540 -0.0178 -0.0600 

LB _ 0.1404 0.0012 -0.0263 

KP _ 0.0693 0.0321 

KOTA _ -0.0636 

218

Fst values were statistically significant between KOTA and SWA, KOTA and LB and 

between SWA and CHIN, mainly due to locus TMO M5. 

All Mantel tests did not indicate any correlation between Fst and geographical distances 

matrices for L. gossei populations (Table P10). 

Table P10. Mantel matrices for Fst values (upper matrices) using all loci, TMO M5 

alone or TMO M27 alone, and for geographical distances between L. gossei 

populations (lower matrices). 

All locus SWA LB KP KOTA CHIN 

SWA _ 0.009 0.049 0.040 0.060 

LB 31 _ 0.051 0.073 0.018 

KP 63 32 _ 0.089 0.071 

KOTA 123 92 60 _ 0108 

CHIN 243 212 180 120 _ 

Z =131.82, 120 permutations, 75 values ≥ Z 

Locus TMO M5 SWA LB KP KOTA CHIN 

SWA _ 0.011 0.046 0.061 0.100 

LB 31 _ 0.006 0.093 0.029 

KP 63 32 _ 0.098 0.094 

KOTA 123 92 60 _ 0.158 

CHIN 243 212 180 120 _ 

Z =183.51, 120 permutations, 39 values ≥ Z 

Locus TMO M27 SWA LB KP KOTA CHIN 

SWA _ 0.005 0.054 -0.018 -0.060 

LB 31 _ 0.140 0.001 -0.026 

KP 63 32 _ 0.069 0.032 

KOTA 123 92 60 _ -0.064 

CHIN 243 212 180 120 _ 

Z =-23.82, 120 permutations, 117 values ≥ Z 

Despite the high number of specimens analysed at each location for each species, 

some populations are represented by a very low number of specimens due to technical 

problems. Both loci are polymorph and it is evident that the polymorphism of low sampled 

populations is underestimated (for every locus and species, the number of alleles increases 

with the sample size). Nevertheless, these data can be analysed keeping in mind that the 

results only indicate tendencies and have to be confirmed with more specimens and even 

more loci. 

219

120 

SWA 

6 

SWA 

100 

5 

80 

4 

60 

3 

40 

2 

20 

1 

0 

0 20 40 60 80 100 120 140 160 

0 

3,50 4,00 4,50 5,00 5,50 

120 

Domira Bay 

5 

Domira Bay 

Weight (g) 

100 

80 

60 

40 

20 

ln weight 

4 

3 

2 

1 

0 

0 20 40 60 80 100 120 140 160 

0 

3,50 4,00 4,50 5,00 5,50 

120 

Nkhotakota 

5 

Nkhotakota 

100 

80 

60 

40 

4 

3 

2 

20 

1 

0 

0 20 40 60 80 100 120 140 160 

0 

3,50 4,00 4,50 5,00 5,50 

120 

Chinteche 

5 

Chinteche 

100 

80 

60 

40 

20 

4 

3 

2 

1 

0 

0 20 40 60 80 100 120 140 160 

Length (mm) 

0 

3,50 4,00 4,50 5,00 5,50 

ln length 

Figure P9. Length-weight relationships for Mylochromis anaphyrmus females at four 

locations in Lake Malawi. 

220

In both species we found some significant Fis values. It is very likely that, to some extent, a 

non-random mating exists. Under the neutral model, heterozygous deficits are expected either 

when there is a mixture of different reproductive units inside a sample (“classical” Wahlund 

effect) or when there is family structuring or inbreeding within this sample. Typically when 

two or more populations are mixed in a sample, linkage desequilibrium (non randomly 

association between alleles of different loci) is observed as well as heterozygous deficits. 

Linkage desequilibrium can also be caused by random drift in small size population or by 

strong selection, two hypotheses that can be rejected here. For both species and loci, no 

linkage desequilibrium has been detected, hence local inbreeding might explain the observed 

results. 

We also have to pay attention to possible null alleles, which can create artificial heterozygous 

deficits: in this case also, no linkage desequilibrium was observed. 

Statistically significant Fst values were observed in both species for both loci and populations 

of M. anaphyrmus were more differentiated than those of L. gossei. It is then possible that the 

different populations of both species are genetically differentiated. 

The results of the Mantel test indicated a correlation between Fst values and the geographical 

distances between all pairs of populations in M. anaphyrmus. This could correspond to an 

isolation by distance model, a model in which populations exchange few migrants, mainly 

with close populations. 

Life history traits analysis 

Mylochromis anaphyrmus: 

At every sample site, fish were collected in February 99, which corresponded with the 

beginning of the breeding season in the SWA. The timing of breeding season for the species is 

unknown in Domira Bay, Nkhotakota and Chinteche. However, the percentages of ripe 

females were not significantly different among sites (One way ANOVA on ranks Kruskal- 

Wallis, H=2.916, 3 df, p=0.405): 6% in SWA, 4% in Domira Bay, 9% in Nkhotakota and 6% 

in Chinteche. 

Length-weight relationships at each location are given in Table P11. Comparison of 

length-weight relationships after Log transformation (Figure P9) revealed significant 

differences among populations (F 3,535 =6.145, p

Table P11. Comparison of Mylochromis anaphyrmus populations. Total number of fish (N), 

length (L)_weight (W) relationships, determination coefficient (R²), number of fish used for 

fecundity calculation (n), mean relative fecundity (MRF) ± SEM. 

Location N Relationships R² n MRF 

SWA 126 W = 0.00001 × L 3.2353 

ln(W) = 3.2373 × ln(L) – 11.4997 

Domira Bay 107 W = 0.00001 × L 3.1446 

ln(W) = 3.1446 × ln(L) – 11.1180 

Nkhotakota 122 W = 0.00003 × L 3.0204 

ln(W) = 3.0204 × ln(L) – 10.5690 

Chinteche 185 W = 0.000007 × L 3.2946 

ln(W) = 3.2946 × ln(L) – 11.8430 

0.984 7 2391 ± 94 

0.957 3 3469 ± 273 

0.971 11 2494 ± 121 

0.984 11 2281 ± 77 

Table P12. Comparison of length-weight relationships (ln transformed) among populations of 

Mylochromis anaphyrmus. ns : non significant, *: significant slope differences (p

Table P13. Comparison of relative fecundity among populations of Mylochromis anaphyrmus. 

ns : non significant, *: significant differences (p

SWA 

120 

100 

80 

60 

40 

20 

0 

0 50 100 150 

SWA 

5 

4 

3 

2 

1 

0 

3,50 4,00 4,50 5,00 5,50 

120 

100 

80 

60 

40 

20 

0 

Leopard Bay 

0 50 100 150 

Leopard Bay 

5 

4 

3 

2 

1 

0 

3,50 4,00 4,50 5,00 5,50 

Weight (g) 

Kiramby Point 

120 

100 

80 

60 

40 

20 

0 

0 50 100 150 

ln weight (g) 

Kiramby Point 

5 

4 

3 

2 

1 

0 

3,50 4,00 4,50 5,00 5,50 

Nkhotakota 

120 

100 

80 

60 

40 

20 

0 

0 50 100 150 

Nkhotakota 

5 

4 

3 

2 

1 

0 

3,50 4,00 4,50 5,00 5,50 

120 

100 

80 

60 

40 

20 

0 

Chinteche 

0 50 100 150 

Length (mm) 

5 

4 

3 

2 

1 

Chinteche 

0 

3,50 4,00 4,50 5,00 5,50 

ln length (mm) 

Figure P10. Length-weight relationships for Lethrinops gossei females at five 

locations in Lake Malawi. 

224

Table P15. Comparison of Lethrinops gossei populations. Total number of fish (N), length 

(L)_weight (W) relationships, determination coefficient (R²), number of fish used for 

fecundity calculation (n), mean relative fecundity (MRF) ± SEM. 

Location N Relationships R² n MRF 

SWA 287 W = 0.00004 × L 2.9598 

ln(W) = 2.9598 × ln(L) – 10.107 

Leopard Bay 192 W = 0.00003 × L 3.0107 

ln(W) = 3.0441 × ln(L) – 10.502 

Kiramby Point 89 W = 0.00002 × L 3.0749 

ln(W) = 3.0749 × ln(L) – 10.687 

Nkhotakota 281 W = 0.00002 × L 3.1267 

ln(W) = 3.1267 × ln(L) – 10.933 

Chinteche 13 W = 0.00005 × L 2.9097 

ln(W) = 2.9097 × ln(L) – 9.8257 

0.971 67 1999 ± 46 

0.866 0 

0.960 11 2204 ± 103 

0.959 27 2054 ± 57 

0.967 4 1823 ± 85 

Comparison of length-weight relationships after Log transformation (Figure P10) revealed 

significant differences among populations (F 4,857 =2.505, p=0.041) even when Chinteche was 

removed from analysis (F 3,845 =3.291, p=0.02). Multiple comparison procedure showed either 

slope or intercept differences between SWA and Kiramby Point, SWA and Nkhotakota, 

Leopard bay and Nkhotakota, Kiramby Point and Chinteche, Nkhotakota and Chinteche 

(Table P16). 

Table P16. Comparison of length-weight relationships (ln transformed) among populations of 

Lethrinops gossei. ns : non significant, *: significant slope differences (p

The number of eggs per kg body weight (relative fecundity Table P15) produced by 

females did not differ among populations for L. gossei (F=1.589, p=0.196). 

Except for the percentage of ripe females, which differed significantly among L. 

gossei populations only, differences among populations were more intense (see significance 

levels of statistic tests) and more numerous for M. anaphyrmus than for L. gossei. Indeed, 

there was no fecundity difference among populations for L. gossei, and the significance level 

for length-weight relationships differences was just below 5% (p=0.041) while it was highly 

significant between M. anaphyrmus populations (p

Chapter 6: 

The potential influence of fluvial 

sediments on rock-dwelling fish 

communities

Chapter 6: The potential influence of fluvial sediments on rockdwelling 

fish communities 


Introduction 

In Malawi, the increasing demographic pressure (Ferguson et al. 1993, Kalipeni 1996) 

has recently accelerated the unsustainable land use practices around the lakeshore and its 

catchments. As a result of deforestation, burning of vegetation, destruction of wet lands in the 

catchments for agricultural purposes and the cultivation of marginal areas such as steep slopes 

of hills (Mkanda 1999), massive quantities of sediment eroded from clear-cut watersheds are 

discharged in the rivers and eventually in the lake (Bootsma & Hecky 1993). The negative 

impact of excess sedimentation and water turbidity on the diversity and ecology of aquatic 

communities has been reported for other Great Lakes (Waters 1995, Evans et al. 1996) and 

Lake Tanganyika particularly (review by Patterson & Makin 1998). In Lake Tanganyika, 

species richness of crustacean ostracodes was found much lower at highly disturbed sites than 

at less disturbed sites, reductions ranging from 40 to 62%. The same pattern, though not 

statistically tested, was observed for fish, with reductions in species richness ranging from 35 

to 65% at highly disturbed sites (Cohen et al. 1993a, Cohen et al. 1996). Lower fish diversity 

was also reported in areas that have become turbid as a result of recent eutrophication in Lake 

Victoria, where increased turbidity was recognised to be partly responsible for the decline in 

cichlid diversity (Seehausen et al. 1997). In Lake Malawi, the impact of sediment discharge is 

not yet permanent and remains associated to the rainy season. However, intensification of 

unsustainable land use practices increases the sediment loads and is cause of major concern 

(Bootsma & Hecky 1993). The limnology team of the SADC/GEF Project identified the 

steadily increasing sediment and nutrient loads arriving both from rivers and atmosphere as 

the main threat to the water quality and therefore to the stability of the lake ecosystem as a 

whole (Bootsma & Hecky 1999). As an example, within three days following a strong rain 

event in February 1998, the Linthipe River, located in the highly populated southern part of 

the lake, delivered about 700 000 tonnes of suspended sediment to Lake Malawi (McCullough 

1999). Much of the suspended river load is made of sand and coarse silt which settled out 

quickly (McCullough 1999). Despite the rapid deposition of most of these sediments, from 

January to March-April, a sediment plume settles more or less permanently, depending on 

wind and currents, around the Maleri Islands in front of the river mouth. Among the potential 

effects of increased sediment loads on aquatic communities (listed in Patterson & Makin 

1998), the reduction of light penetration affecting photosynthetic rates or sexual mate choice 

(Seehausen et al. 1997), the reduction of habitat complexity and destruction of spawning 

grounds are of direct importance for fish (Waters 1995, Evans et al. 1996, Lévêque 1997). As 

an example, over-fishing and siltation resulting from deforestation have strongly diminished 

the abundance of potadromous fish species in Lake Malawi (Tweddle 1992, Turner 1994b). 

For the littoral rocky shore, which food web is based on benthic algae growing on rocks 

(Worthington & Lowe-McConnell 1994), the blanketing of benthic algae by deposited 

sediments primarily affects the specialised aufwuchs eaters (Patterson & Makin 1998). Lake 

Malawi rock-dwelling mbuna, whose communities directly rely on the algal carpet covering 

the rocks (Fryer 1959, Ribbink et al. 1983, Marsh & Ribbink 1986, Reinthal 1990) and whose 

mobility and migration capacity are very restricted (Ribbink et al. 1983, McElroy & Kornfield 

227

N 

Senga Bay 

Linthipe river 

Maleri islands 

Chipoka 

Cape Mclear 

SWA 

SEA 


MALERI ISLANDS 

Nankoma 

Maleri 

Nsh 

Nakantenga 

Nex 

Figure S1. Southern Lake Malawi, with a detail of the Linthipe river delta and the Maleri 

islands. Arrows indicate the two sampling sites on Nakantenga island, Nsheltered (Nsh) 

and Nexposed (Nex). Detailed region redrawn from P. Cooley (1999). 

228

1990, Van Oppen et al. 1997) should therefore be directly impacted by the increasing 

sediment discharges. Ribbink et al (1983) found a mantle of mud was deposited by plumes on 

the rocks of the Maleri Islands and postulated that it would affect the fishes access to benthic 

algae. Occasional observations of fish migrations from deep water to the shallows in the 

rocky shores of the Maleri Islands when the sediment plume was settled were reported (H. 

Bootsma, pers. com.). During the plume influence, the light penetration is greatly reduced and 

affects the benthic algal productivity on which most of the mbuna depend. Higgins et al. 

(1999) indicated how this would cause the fish to move upwards to the shallows to 

compensate for the shortage of food in the deeper waters. The present study, carried out in 

collaboration with the limnology team of the project, was designed to test for that hypothesis 

and to monitor the potential influence of suspended sediments on the diversity, abundance, 

condition and some life history traits of rock-dwelling fish communities. 


The rainy season in the Lake Malawi/Nyasa basin usually starts in November and ends 

in March (Eccles 1974, Ribbink 1994). In order to assess the potential influence of suspended 

sediment on rock-dwelling fish communities, we decided to sample during undisturbed 

condition (i.e. before the rains), during the disturbed situation (i.e. the rainy season) and after 

a few months of recovered undisturbed situation (i.e. after the end of rains). Hence, from 

October 1998 to May 1999, 2 sites impacted by fluvial sediments and 2 control sites were 

sampled monthly. 

The impacted sites were located on Nakantenga Island (Figure S1), off the Linthipe river 

mouth. The Linthipe is the largest river of the southern part of the lake and its catchment is 

one of the most densely populated around the lake (Mkanda 1999, Kingdon et al. 1999, 

Bootsma & Hecky 1999). The Linthipe catchment has been selected as a model for the study 

of land use, soil erosion and sedimentation in Lake Malawi watershed (Mkanda F.X. PhD 

thesis in prep., McCullough G., PhD thesis in prep.). Two sites were sampled on Nakantenga 

Island: 

- a sheltered bay (Nsh) located on the northern shore of the island and protected from the 

south-easterly trade wind (Mwera), mainly blowing from June to September, by a natural 

rock barrier, 

- an exposed bay (Nex) located on the other side of the rock barrier, hence submitted to the 

trade wind. 

The control sites were located in a sediment free area, in Thumbi West Island at Cape Maclear 

(Figure S2). T13 is a little protected bay on the north side of the island, whereas T8 is a non 

protected rocky area on the northeast side of the island. 

Fish species were named according to Ribbink et al. (1983). Every month from 

October 98 to April 99, the following sampling was carried out at each site: 

• Fish diversity and abundance were estimated using underwater visual censuses at 10 

m, 6 m and 2 m depth, following the protocol described in Ribbink et al. (1983). Transects 

were demarcated by two 6 mm diameter nylon cords, 25 m in length, held 2 m apart by a 

galvanised pipe at each end so that an area of 50 m² was sampled. At each site and depth, the 

4 corners of the 50 m² areas were permanently marked to ensure that the censuses were done 

exactly at the same place every month. Fishes within the demarcated areas were counted after 

waiting at least 3 min for them to recover from any diver disturbance. For most species, 

sexually active males, in breeding dress and apparently defending territories, were counted. 

For each species, the number of individuals was the mean number of two consecutive 

censuses along the transect. 

229

N 

Senga Bay 


Chipoka 

Maleri islands 

Cape Mclear 

SWA 

SEA 

T 13 

T 8 

Thumby west 

CAPE MCLEAR 

Otter point 

Figure S2. Southern Lake Malawi, with a detail of Cape Mclear showing the sampling sites 

on Thumby west island, T13 and T8. Detailed region redrawn from P. Cooley (1999). 

230

Owing to the very bad visibility conditions at both sites in Nakantenga island in February and 

March 1999 (Secchi disk measurements < 2 m), visual censuses were not possible. Owing to 

technical problems, none of the sites was sampled in November 1998, and Thumbi sites were 

not sampled in October. 

• At each site and depth, a 30 m length × 1.5 m height monofilament nylon gill net, 

constituted of three 10 m panels of 0.5, 1 and 1.5 inch mesh size, was set for two hours. Gill 

nets were carefully set apart from the areas used for visual censuses. The depth at the 

beginning and the end of setting was checked with a manual depth sounder (Echotest, 

Plastimo) and verified with scuba shortly after. After two hours, the fish were removed from 

the net, sorted to species, labelled and placed in 10% formalin for later examination. Gill nets 

were found to fluctuate too much in their catches to be used for diversity and abundance 

estimations. Indeed, parts of the net were sometimes torn by rocks either during setting or 

hauling, which modified the catch per unit effort (CPUE). Some fish species tended to follow 

the diver during the checking of the net depth, entangling themselves more than without the 

diver's disturbance, which also modified the CPUE. Therefore, gill nets were not used to 

estimate fish diversity and abundance. They were used only to assess the potential 

modification of life history traits during the plume, such as variations in the % of ripe 

females, in the condition factor, the fecundity and the egg size. 

Both visual censuses and gill net samplings were always done in the morning between 8.00 

and 12.00 to avoid potential diel fluctuations in fish distribution and abundance. 

• Concurrently, benthic algae samples were collected by scraping at the same depths on 

top of flat horizontal rocks, to monitor the benthic algae biomass. Algae samples were 

processed by the limnology team of the project. The scrapings were collected using a thickwalled 

acrylic tube with an inside diameter of 4.5 cm². The base of the tube was fitted with a 

neoprene skirt to ensure a good seal with the rock. A plunger fitted with a stiff wire brush 

within the tube was used to scrape the algae off of the rock. After scraping, the sample was 

drawn into a syringe that was attached near the base of the scraping tube via a small tube. 

After collection, the sample was poured into a bottle and diluted to 250 ml. After thoroughly 

mixing, a subsample of this was filtered on a Whatman glass fibre GF/F filter (nominal pore 

size = 0.7 µm). The filter was placed in a mixture of methanol and acetone for at least 24 

hours to extract algal pigments, after which chlorophyll a was measured on a Turner 

fluorometer. Based on the volume of extract, the filtration volume, and the area of rock that 

was scraped, the final measured volumetric chlorophyll a concentration was converted to an 

areal measurement of µg/cm². Samples were usually collected in triplicate. 

• Secchi disk measurements were also carried out to monitor the water transparency. 

All fish preserved in formalin were measured (SL) to the nearest 1 mm and weighed to 

the nearest 0.1 g. Their maturity stage was determined and the gonads in stage 4 were 

weighed for Gonado-Somatic Index (GSI) calculation (gonad weight/total body weight × 100) 

then preserved in 5% formalin for fecundity and mean oocyte weight calculation. 

The maturity stage of female gonads was macroscopically determined using the slightly 

modified scale of Legendre & Ecoutin (1989) (Duponchelle 1997). 

Stage 1: immature. The gonad looks like two short transparent cylinders. No oocytes are 

visible to the naked eyes. As a comparison, immature testicle is much longer and thinner, like 

two long tinny silver filaments. 

Stage 2: beginning maturation. The ovaries are slightly larger and little whitish oocytes and 

apparent. 

Stage 3: maturing. The ovaries continue to grow in length and thickness and are full of 

yellowish oocytes in early vitellogenesis. 

Stage 4: final maturation. The ovaries occupy a large part of the abdominal cavity and are full 

of large uniform sized oocytes in late vitellogenesis. 

231

Stage 5: ripe. Ovulation occurred, oocytes can be expelled by a gentle pressure on the 

abdomen. This stage is ephemera. 

Stage 6: spent. The ovaries look like large bloody empty bags with remaining large sized 

atretic follicles. Small whitish oocytes are visible. 

Stage 6-2: resting. The general aspect of the gonad recall a stage 2, but the ovarian wall is 

thicker, the gonad is larger, often reddish with an aspect of empty bag. This stage is 

distinctive of resting females, which have spawned during the past breeding season. 

Stage 6-3: recovering post-spawning females. The general aspect of the gonad is like a stage 3 

but with empty rooms, remaining large-sized atretic follicles and the blood vessels are still 

well apparent. This stage is characteristic of post-spawning females initiating another cycle of 

vitellogenesis. 

Males were only recorded as being either in "breeding colour" or not. 

The average size at first maturation (L 50 ) is defined as the standard length at which 

50% of the females are at an advanced stage of the first sexual cycle during the breeding 

season. In practice, this is the size at which 50% of the females have reached the stage 3 of the 

maturity scale (Legendre & Ecoutin 1996, Duponchelle & Panfili 1998). For the estimation of 

L 50 , only the fish sampled during the height of the breeding season were considered. 

Fecundity is defined here as the number of oocytes to be released at the next spawn, 

and correspond to the absolute fecundity. It is estimated, from gonads in the final maturation 

stage (stage 4), by the number of oocytes belonging to the largest diameter modal group. This 

oocyte group is clearly separated from the rest of the oocytes to the naked eye and 

corresponds approximately to oocytes that are going to be released (Duponchelle 1997, 

Duponchelle et al. 2000). 

Oocyte weight measurements were all carried out on samples preserved in 5% 

formalin. The average oocyte weight per female, was determined by weighing 50 oocytes 

(Peters 1963) belonging to those considered for fecundity estimates. 

In order to compare mean oocyte weight and diameter among the different species, the 

measurements need to be made on oocytes in a similar vitellogenic stage, then on oocytes 

whose growth is completed. A simplified version of the method applied by Duponchelle 

(1997) was used to determine the GSI threshold above which the oocyte weight do no longer 

increase significantly. For each species, the individual oocyte weights were plotted against the 

GSI. The GSI corresponded to the beginning of the asymptotic part of the curve was visually 

determined and the fish whose GSI was inferior to the defined GSI were removed. The final 

GSI threshold was reached when no correlation subsisted between the mean oocyte weight 

and the GSI. 

The Fulton's condition factor was calculated as CF = 100 000 × W/L 3 , where W is the 

wet weight and L the standard length (Anderson & Gutreuter 1983). Given the well known 

trend of CF to increase with the length of fish (Anderson & Gutreuter 1983), the absence of 

relationship between length and CF over the size range sampled was checked for every 

species using Pearson correlation. 

Statistics 

Comparisons of among months diversity and abundance of fish species were 

performed with one way repeated measures ANOVA using SigmaStat software (Jandel 

Scientific). 

The temporal variation of the condition factor for every species was assessed using 

univariate general linear model with month, site and sex as factors and CF as the dependant 

232

Nakatenga sheltered (Nsh) 

16 

14 

12 

10 

8 

6 

4 

2 

0 

2 m 

16 

14 

12 

10 

8 

6 

4 

2 

0 

6 m 

16 

14 

12 

10 

8 

6 

4 

2 

0 

10 m 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Nakatenga exposed (Nex) 

16 

14 

12 

10 

8 

6 

4 

2 

0 

2 m 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

16 

14 

12 

10 

8 

6 

4 

2 

0 

6 m 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

16 

14 

12 

10 

8 

6 

4 

2 

0 

10 m 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Thumby west 13 (T13) 

25 

2 m 

25 

6 m 

25 

10 m 

20 

20 

20 

15 

15 

15 

10 

10 

10 

5 

5 

5 

0 

0 

0 

Thumby west 8 (T8) 

25 

2 m 

20 

15 

10 

5 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

25 

6 m 

20 

15 

10 

5 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

25 

10 m 

20 

15 

10 

5 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Figure S4. Monthly progression of benthic chlorophyll a biomass (µg/cm²) (black dots) and of light 

penetration estimated by Secchi disk measurements (m) (white triangles) at 2 m, 6 m and 10 m depth 

at each sampled site, from October 1998 to May 1999. Note the different y axis scale between 

Nakantenga and Thumby sites. Algae sampling was done on top of flat horizontal rocks using a 

scrapper. Values are averages of three replicates. 

233

Results 

variable, using SPSS software. Given the low sample number at some sites for some species, 

the significance level for interaction between factors was fixed at 1%. 

Although rainfall patterns might be slightly different in Cape Maclear and Maleri 

Islands than in Senga bay, rains effectively started at the end of November 1998 (the 23 rd ) and 

stopped the 11 th of April 1999 at Senga Bay station (Figure S3). Despite an effective start of 

the rainy season in late November, the Linthipe river discharge really began in January to 

reach a peak in March (Figure S3). As the sediment plume is linked to the river discharge 

rather than to the rains directly, therefore the period under consideration when referring herein 

to the rainy season is "January to May". 

In February and March 1999, visibility was so poor in Nakantenga sites that visual 

censuses were impossible to carry out. However, while one of the two divers was scraping for 

benthic algae samples, the other diver was doing fish observations. Although this required to 

be less than 50 cm away from the fish and that it was often impossible to identify the species, 

behavioural observations were still possible. Fish were virtually immobile with fins held erect 

a they do at night. They stayed in or very close to the entrance to their hideaways among the 

rocks, or rested on the rocks. There was virtually no feeding, territorial or courtship 

behaviour. 

Chlorophyll a biomass 

Mean monthly benthic chlorophyll a biomass for each depth and site are presented in 

Figure S4. A similar temporal trend was evident at almost all sites and depths, suggesting that 

these data, though collected only once per month, represented real monthly trends. A 

chlorophyll peak was observed in December 1998 at every depth in Thumbi West sites, 

possibly due to an increased nutrient availability resulting from land runoff (H. Bootsma, 

pers. com.). Another smaller peak was observed in April-May, but apart from these peaks, 

chlorophyll concentration was rather similar between the dry and rainy seasons. At 

Nakantenga sites, there was also a slight increase in chlorophyll in December, though not 

apparent at all depths, but it was followed by chlorophyll concentrations that were lower than 

before the rainy season. This suggests a greater impact of siltation and reduced light 

penetration at Nakantenga. However, changes in chlorophyll concentration at the impacted 

sites (Nsh and Nex) were less important than expected. Indeed, from January-February to 

May 1999, the thickness of sediment (up to 2 cm sometimes) covering the rocks observed 

while diving at all depths (and especially at 6 and 10 m) should have almost completely 

prevent photosynthesis, even when the plume was temporary out of the island. Furthermore 

light penetration was severely diminished when the plume was settled, with Secchi disk 

measurements inferior to 4 m from February to April (Figure S4). Concern that benthic 

chlorophyll a concentration did not change as much as expected between the dry and the rainy 

season at Nakantenga matched observations on previous years at the nearby Maleri Island by 

the project's limnology team. The explanation probably lies in the phytoplankton production 

in the first few metres that had settled and mixed with the sediment (H. Bootsma, pers. com.). 

The question is whether this type of algae is still useful to the fish as a food source? 

234

Total rainfall (mm) 

250 

200 

150 

100 

50 

Rainfall 

Discharge 

800 

700 

600 

500 

400 

300 

200 

100 

Mean river discharge (m3/sec) 

0 

Oct Nov Dec Jan Feb Mar Apr May 

1998 1999 

0 

Figure S3. Total rainfall at Senga Bay station and mean Linthipe river 

discharge from October 1998 to April 1999. 

Table S1b. Visual censuses at 2 m, 6 m and 10 m depth at Nakantenga exposed (Nex) site. 

Month Oct-98 Dec-98 Jan-99 Apr-99 May-99 

Species name Depth (m) 2 6 10 2 6 10 2 6 10 2 6 10 2 6 10 

Pseudotropheus zebra red dorsal 2 6 68 0 11 77 2 10 39 2 8 38 3 9 11 

Pseudotropheus zebra yellow throat 15 8 0 13 6 0 19 6 0 18 4 0 17 7 0 

Pseudotropheus zebra black dorsal 0 2 3 0 0 1 0 0 3 0 2 3 0 3 3 

Pseudotropheus barlowi 0 9 3 0 11 1 0 4 1 0 9 1 0 6 3 

Pseudotropheus tropheops lilac maleri 4 2 0 4 1 0 4 1 0 8 1 0 5 1 0 

Pseudotropheus tropheops maleri blue 4 0 0 2 0 0 1 0 0 1 0 0 2 0 0 

Pseudotropheus tropheops orange chest 10 1 2 6 2 1 7 1 0 1 1 0 6 0 0 

Pseudotropheus williamsi 2 0 0 1 0 0 1 0 0 1 0 0 2 0 0 

Pseudotropheus elongatus brown 0 0 0 0 0 7 0 1 7 0 1 2 0 0 1 

Pseudotropheus aggressive yellow head 11 19 0 24 44 0 0 50 1 35 44 5 60 35 6 

Pseudotropheus aggressive blue 4 0 0 6 0 0 10 0 0 8 0 0 9 0 0 

Pseudotropheus aggressive zebra 15 0 0 6 1 0 30 4 0 4 1 0 5 0 0 

Pseudotropheus burrower 0 21 7 0 11 10 1 12 12 4 10 0 0 15 10 

Melanochromis auratus 2 3 2 3 2 1 2 8 8 8 2 1 5 2 0 

Melanochromis vermivorous 0 0 0 0 4 0 1 7 0 0 1 0 0 0 0 

Melanochromis melanopterus 3 5 3 0 0 0 0 0 1 1 0 3 0 2 1 

Melanochromis crabro 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 

Petrotilapia yellow chin 4 0 0 4 0 0 2 0 0 3 0 0 2 0 0 

Petrotilapia genalutea 2 0 0 3 0 0 2 1 0 2 0 0 2 0 0 

Petrotilapia fuscous 5 3 9 5 4 5 2 3 9 2 3 4 2 4 1 

Labidochromis vellicans 0 0 0 2 1 0 4 4 0 3 2 0 1 2 0 

Labidochromis pallidus 2 3 5 3 2 3 3 3 3 4 2 0 3 1 1 

Labeotropheus fuelleborni 6 0 2 6 2 2 5 1 0 8 0 5 9 0 0 

Labeotropheus trewavasae 1 0 1 2 1 0 0 0 1 0 0 0 0 1 0 

Genyochromis mento 0 2 1 0 0 1 1 1 1 0 1 0 0 1 0 

Aulonocara gold 0 0 0 0 1 0 0 0 1 0 0 0 0 0 2 

Protomelas taeniolatus 2 1 4 0 0 2 0 0 3 0 3 5 0 0 0 

Territorial Total 94 85 110 90 104 111 97 117 91 113 95 67 133 89 39 

235

Visual censuses 

It should be kept in mind that censuses were done exactly at the same place within 

each site. Variations in relative abundance of species were observed among months along the 

rainy season. At Nakantenga sheltered (Nsh) site (Table S1a), Ps. zebra 'red dorsal' tended to 

be more abundant at 6 m and especially at 10 m during the rainy season. Ps. zebra 'black 

dorsal' was also more abundant at 6 m during the rainy season. The same increasing trends at 

10 m were noticed for Ps. elongatus 'brown' and Petrotilapia fuscous, whereas Ps. barlowi 

abundance decreased at 10 m. Ps. 'aggressive yellow head' tended to be more abundant at 2 m 

but less at 6 and 10 m during the rainy months. However, diversity and abundance variations 

of species among months were not significant at any depth (one way repeated measures 

ANOVA, F=0.7 p=0.593 for 2m, F=0.653 p=0.626 for 6 m and F=0.697 p= 0.597 for 10 m 

depth). 

Table S1a. Visual censuses at 2 m, 6 m and 10 m depth at Nakantenga sheltered (Nsh) site. 

Month Oct-98 Dec-98 Jan-99 Apr-99 May-99 

Species name Depth (m) 2 6 10 2 6 10 2 6 10 2 6 10 2 6 10 

Pseudotropheus zebra 'red dorsal' 2 10 8 0 5 3 1 20 4 2 20 38 1 8 18 

Pseudotropheus zebra 'yellow throat' 8 0 0 14 0 0 15 0 0 22 0 0 13 0 0 

Pseudotropheus zebra 'black dorsal' 2 2 3 0 1 4 0 3 5 0 7 3 0 5 4 

Pseudotropheus barlowi 0 6 2 0 5 4 0 4 2 0 5 1 0 5 0 

Pseudotropheus tropheops 'lilac maleri' 7 0 0 3 1 0 4 1 0 3 0 0 3 0 0 

Pseudotropheus tropheops 'orange chest' 0 0 0 0 0 0 0 0 0 0 1 0 0 3 0 

Pseudotropheus williamsi 1 0 0 0 0 0 1 0 0 1 0 0 0 0 0 

Pseudotropheus elongatus 'brown' 1 1 0 2 0 0 0 6 1 1 1 2 0 4 5 

Pseudotropheus 'aggressive yellow head' 3 31 11 0 22 9 3 9 7 5 8 0 8 15 0 

Pseudotropheus 'aggressive zebra' 21 0 0 34 0 0 36 1 0 26 0 0 38 0 0 

Pseudotropheus burrower 10 15 14 5 18 7 6 23 13 6 16 5 4 16 10 

Melanochromis auratus 3 2 1 2 0 0 2 2 2 2 1 1 1 3 3 

Melanochromis vermivorous 0 0 0 4 0 0 3 3 0 2 1 0 1 0 0 

Melanochromis melanopterus 2 2 2 0 0 0 1 1 0 0 0 2 1 0 2 

Petrotilapia 'yellow chin' 0 0 0 1 0 0 1 0 0 1 0 0 3 0 0 

Petrotilapia genalutea 0 0 0 1 0 0 1 0 0 3 0 0 3 0 0 

Petrotilapia 'fuscous' 0 2 2 2 0 0 1 3 1 1 2 4 1 4 10 

Labidochromis vellicans 0 0 0 3 0 0 3 0 0 3 0 0 2 2 1 

Labidochromis pallidus 3 3 3 3 0 0 4 2 1 5 1 5 4 3 4 

Labeotropheus fuelleborni 1 0 0 1 0 0 1 0 0 3 0 0 1 0 0 

Labeotropheus trewavasae 1 1 0 1 0 0 0 1 0 1 0 0 0 1 0 

Genyochromis mento 1 1 0 2 0 0 1 0 0 1 1 0 1 2 0 

Aulonocara 'gold' 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 

Protomelas taeniolatus 0 0 3 0 0 0 0 0 0 0 7 5 0 0 0 

Territorial Total 66 76 49 78 52 27 84 79 37 88 71 66 85 71 57 

At Nakantenga exposed (Nex) site (Table S1b), Ps. zebra 'red dorsal' showed an opposite 

trend, being less abundant at 10 m during the rainy season, like Ps. tropheops 'orange chest' 

and Petrotilapia fuscous, which was also less abundant at 2 m. Labeotropheus fuelleborni and 

Melanochromis auratus tended to be more abundant at 2 m during the rainy months. Ps. 

'aggressive yellow head' abundance tended to increase at every depth during the rainy season. 

236

Table S2b. Visual censuses at 2 m, 6 m and 10 m depth at Thumby west T8 site. 

Month Dec-98 Jan-99 Feb-99 Mar-99 Apr-99 May-99 

Species name Depth (m) 2 6 10 2 6 10 2 6 10 2 6 10 2 6 10 2 6 10 

Pseudotropheus zebra 16 1 2 12 15 4 19 14 4 16 12 4 9 6 2 13 12 3 

Pseudotropheus zebra callainos 24 9 0 1 1 0 24 1 0 18 1 0 20 0 0 15 1 0 

Pseudotropheus aurora 0 0 2 9 8 2 0 11 3 0 12 3 0 8 2 0 10 3 

Pseudotropheus heteropictus 0 0 9 0 0 12 0 0 13 0 1 15 0 2 14 0 1 15 

Pseudotropheus livingstonii 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 

Pseudotropheus tropheops lilac 0 2 0 0 0 0 1 0 0 1 0 0 1 0 0 0 0 0 

Pseudotropheus tropheops red cheek 13 14 0 2 3 0 12 1 0 16 2 0 14 2 0 11 1 0 

Pseudotropheus tropheops orange chest 9 6 8 14 16 6 13 15 11 8 13 12 7 16 14 12 13 10 

Pseudotropheus tropheops intermediate 0 0 6 6 6 8 1 10 8 0 9 8 0 11 9 0 10 9 

Pseudotropheus tropheops gracilior 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 

Pseudotropheus williamsi 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 

Pseudotropheus elongatus brown 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 

Pseudotropheus elongatus slab 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 1 0 

Pseudotropheus aggressive brown 18 0 0 0 0 0 17 0 0 21 0 0 16 0 0 15 0 0 

Pseudotropheus tiny 0 0 4 0 0 1 0 0 4 0 0 5 0 0 5 0 0 5 

Melanochromis auratus 1 0 3 0 1 1 2 3 2 3 1 2 1 2 2 2 1 1 

Melanochromis melanopterus 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 

Melanochromis vermivorous 8 0 0 0 0 0 10 0 0 8 0 0 11 0 0 10 0 0 

Melanochromis joanjohnsonae 5 0 0 0 0 0 3 0 0 3 0 0 4 0 0 3 0 0 

Melanochromis crabro 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 

Petrotilapia mumbo blue 2 0 0 0 0 0 3 0 0 3 0 0 3 0 0 4 0 0 

Petrotilapia genalutea 3 6 0 0 0 0 2 0 0 3 0 0 6 0 0 2 0 0 

Petrotilapia nigra 1 0 0 6 4 3 2 8 1 1 5 1 2 7 2 1 3 1 

Labidochromis vellicans 0 0 0 0 1 0 5 0 0 3 1 0 4 2 0 3 0 0 

Labidochromis freibergi 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 

Labidochromis gigas 0 0 3 2 3 2 2 3 4 0 3 4 0 2 3 1 6 3 

Labidochromis blue bar 0 19 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 

Cynotilapia afra 13 1 21 19 26 26 26 26 24 20 24 26 8 5 6 28 28 21 

Labeotropheus fuelleborni 10 1 0 1 2 0 13 0 0 11 2 0 16 1 0 16 0 0 

Labeotropheus trewavasae 0 0 0 1 1 0 1 0 0 2 1 0 1 1 0 0 1 0 

Genyochromis mento 0 2 0 0 0 0 2 1 1 1 1 1 1 1 1 1 1 1 

Aulonocara blue 0 1 0 2 0 1 0 2 5 0 1 3 0 1 3 0 2 3 

Protomelas taeniolatus 0 65 0 1 2 1 0 3 0 0 3 1 0 2 1 0 0 1 

Territorial Total 123 205 59 77 90 69 158 99 80 138 92 86 124 69 64 137 91 76 

237

Again, the observed variations in species diversity and abundance were not significant at any 

depth (one way repeated measures ANOVA, F=0.453 p=0.770 for 2m, F=0.816 p=0.518 for 6 

m and F=1.220 p= 0.307 for 10 m depth). 

At Thumbi West T13 site (Table S2a), some species tended to be more abundant during the 

rainy season at 2 m (Ps. callainos, Melanochromis vermivorus, Labidochromis vellicans, L. 

fuelleborni), 6 m (L. fuelleborni) and 10 m (Ps. tropheops 'orange chest'). Ps. tropheops 

'intermediate' was less abundant at 6 m during the rainy months. However, these species 

diversity and abundance variations among months were not significant at any depth (one way 

repeated measures ANOVA, F=1.301 p= 0.265 for 2m, F=0.757 p=0.582 for 6 m and F=0.833 

p=0.527 for 10 m depth). 

Table S2a. Visual censuses at 2 m, 6 m and 10 m depth at Thumbi west T13 site. 

Month Dec-98 Jan-99 Feb-99 Mar-99 Apr-99 May-99 

Species name Depth (m) 2 6 10 2 6 10 2 6 10 2 6 10 2 6 10 2 6 10 

Pseudotropheus zebra 12 30 37 21 27 28 16 31 21 24 38 33 16 36 26 20 35 38 

Pseudotropheus zebra callainos 35 20 3 33 27 4 41 17 1 57 25 3 54 24 0 46 15 2 

Pseudotropheus aurora 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 

Pseudotropheus heteropictus 0 2 2 0 0 2 0 0 1 0 0 3 0 0 2 0 0 2 

Pseudotropheus tropheops lilac 0 0 0 0 0 0 1 0 0 1 0 0 1 0 0 1 0 0 

Pseudotropheus tropheops red cheek 4 4 0 4 7 1 4 5 0 6 7 0 4 4 0 3 4 0 

Pseudotropheus tropheops orange chest 2 6 8 5 13 8 3 16 9 5 13 10 4 17 12 4 11 13 

Pseudotropheus tropheops intermediate 0 9 5 0 3 10 0 4 8 0 2 8 0 2 7 0 1 7 

Pseudotropheus elongatus brown 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 

Pseudotropheus elongatus slab 0 0 0 0 1 1 0 1 3 0 2 3 0 1 3 0 1 1 

Pseudotropheus aggressive blue 10 0 0 6 0 0 6 0 0 6 0 0 6 0 0 5 0 0 

Pseudotropheus aggressive brown 32 0 0 19 0 0 26 0 0 19 0 0 23 0 0 22 0 0 

Pseudotropheus tiny 0 3 2 0 0 2 0 0 4 0 1 3 0 0 6 0 0 4 

Melanochromis auratus 2 2 3 3 2 2 0 1 2 1 2 1 2 2 3 2 3 3 

Melanochromis melanopterus 2 1 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 

Melanochromis chipokae 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 

Melanochromis vermivorous 8 2 0 14 3 0 21 1 0 10 3 2 12 2 1 16 4 1 

Melanochromis parallelus 3 2 1 2 3 2 3 2 3 2 0 1 2 3 1 2 1 2 

Melanochromis joanjohnsonae 6 0 0 4 1 0 6 0 0 2 0 0 5 0 0 5 0 0 

Petrotilapia mumbo blue 1 0 0 1 0 0 1 0 0 1 0 0 2 0 0 1 0 0 

Petrotilapia tridentiger 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 

Petrotilapia genalutea 2 0 0 4 0 0 1 0 0 1 0 0 2 0 0 1 0 0 

Petrotilapia nigra 1 1 2 0 3 3 1 6 2 1 2 2 0 4 3 0 4 2 

Labidochromis vellicans 1 2 1 4 4 4 2 4 1 5 3 0 7 2 0 5 3 1 

Cynotilapia afra 1 9 0 5 5 0 3 5 0 4 6 0 5 6 0 10 7 0 

Labeotropheus fuelleborni 2 4 0 8 18 0 16 13 0 11 15 0 21 18 0 14 11 0 

Labeotropheus trewavasae 1 2 1 1 2 3 2 4 2 1 3 4 2 3 6 2 4 4 

Genyochromis mento 0 0 2 0 0 0 1 1 1 1 1 0 2 1 3 1 2 1 

Aulonocara blue 0 3 0 0 0 2 0 1 1 0 0 1 0 0 1 0 0 2 

Protomelas taeniolatus 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 

Territorial Total 126 103 71 134 119 72 155 112 60 158 123 74 171 125 74 162 106 84 

At Thumbi West T8 site (Table S2b), Ps. callainos and Ps. tropheops 'red cheek' were found 

less abundant at 6 m during the rainy season. The opposite tendency was observed at 2 m for 

Labidochromis vellicans and L. fuelleborni, and at 6 m for Ps. aurora and Ps. tropheops 

238

'orange chest' . This last species was also more abundant at 10 m. As for T13 site, species 

diversity and abundance variations among months were not significant at any depth at T8 (one 

way repeated measures ANOVA, F=1.958 p= 0.086 for 2m, F=1.175 p=0.322 for 6 m and 

F=1.262 p=0.282 for 10 m depth). 

Life history traits and condition 

Life history traits such as fecundity, egg size and size at maturity require large fish 

samples to be determined. Between months comparisons within each site were not possible 

owing to insufficient sample numbers. As a consequence, they were compared only for the 

few species abundant enough and represented at both Nakantenga and Thumbi sites. Also 

owing to sample number, Nex and Nsh sites were pooled as were T13 and T8, for the 

Nakantenga versus Thumbi west comparisons of life history traits among populations. 

Potential condition variations were investigated for every species abundant enough at every 

sampled months. 

Species present at both Nakantenga and Thumbi west sites 

To be accurate, length at maturity has to be determined at the height of the breeding 

season. Determination of the breeding season was not possible as we did not sample during a 

complete annual cycle. However, most of the mbuna, including the two species below, breed 

throughout the year with a peak in August to October and one in February to March (Marsh et 

al. 1986). Our sampling period from October 1998 to May 1999, included part of the first 

breeding peak and completely the second one. Therefore, length at maturity were estimated 

from the complete sampling period. 

Labeotropheus fuelleborni 

The size at maturity of L. fuelleborni was slightly higher at Nakantenga (66 mm) than 

at Thumbi west (62 mm) (Figure S5). 

100 


50 

0 

Nakantenga 

Thumby 

52 57 62 67 72 77 82 87 92 97 


Figure S5. Mean length at first sexual maturity for Labeotropheus 

fuelleborni at Nakantenga and Thumbi west sites. 

239

Condition factor 

3,9 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

CF 

Chloro a 

Secchi 

16 

14 

12 

10 

8 

6 

4 

2 

0 

Chlorophyll a biomass (ug/cm²) 

Secchi disk depth (m) 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

a 30 24 27 25 20 5 19 32 


3,9 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

CF 

Secchi 

Chloro a 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 


Seeci disk depth (m) 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

b 37 18 50 23 22 24 54 32 

Figure S6. Monthly progression of Labeotropheus fuelleborni mean 

condition factor, Secchi disk measurements (m) and benthic chlorophyll 

a biomass (µg/cm²) at Nakantenga sites (a) and Thumby west sites (b). 

Numbers below the months are the corresponding sample numbers. 

240

As the relative fecundity was not correlated to the body weight for L. fuelleborni at 

either sites (Pearson correlation r=-0.225, p=0.506 and r=-0.298, p=0.140 for Nakantenga and 

Thumbi populations, respectively), comparison of fecundity were done on relative fecundity. 

There was no significant difference (Mann-Whitney rank sum test: T= 221, p=0.702) of 

relative fecundity between the populations of Nakantenga (N=11, mean relative 

fecundity=1725 ± 405 SD) and Thumbi west (N=26, mean relative fecundity=1675 ± 252 

SD). 

The GSI threshold above which oocyte weight did no longer increase significantly was 

3.8% for L. fuelleborni. At Nakantenga sites, only two females had a GSI superior to 3.8%, 

giving a mean oocyte weight of 31.9 mg ± 2.69. At Thumbi west sites, 4 females had a GSI 

superior to 3.8%, giving a mean oocyte weight of 29.5 mg ± 5.21. Owing to the low sample 

number at Nakantenga, statistical comparison was not possible, but the weak difference was 

likely to be insignificant. 

Analysis of monthly progression of mean condition factor (CF) were carried out 

separately at Nakantenga and Thumbi sites. 

At Nakantenga, no correlation was observed between CF and length of fish (Pearson 

correlation p>0.05), allowing to test for between months (October to May), sites (Nex and 

Nsh) and sex (females and males) differences using a general linear model with month, site 

and sex as factors and CF as dependant variable. The monthly progression of the mean 

condition factor, secchi disk measurements and benthic chlorophyll a biomass are presented in 

Figure S6a. The significant factors were month (F 7,156 =4.673, p

At Thumbi west, no correlation was observed between CF and length of fish (Pearson 

correlation p>0.05), allowing to search for between months, sites and sex differences. None of 

the factor had a significant effect on the mean condition factor, indicating that the rainy 

season did not significantly influence the condition of L. fuelleborni at Thumbi west. The 

monthly progression of the mean condition factor, Secchi disk measurements and benthic 

chlorophyll a biomass are presented in Figure S6b. No significant correlation was found 

between the mean CF, Secchi disk measurements and the chlorophyll a biomass (Pearson 

correlation p>0.05). 

Pseudotropheus tropheops 'orange chest' 

The mean size at maturity of Ps. tropheops 'orange chest' was the same (65 mm) in 

Nakantenga and Thumbi west sites (Figure S7). 

100 


50 

0 

Nakantenga 

Thumby 

47 52 57 62 67 72 77 82 87 92 97 


Figure S7. Mean length at first sexual maturity for Pseudotropheus 

tropheops 'orange chest' at Nakantenga and Thumbi west sites. 

As the relative fecundity was not correlated to the body weight for Ps. tropheops 

'orange chest' at either sites (Pearson correlation r=-0.21, p=0.489 and r=-0.28, p=0.134 for 

Nakantenga and Thumbi populations, respectively), comparisons of fecundity were done on 

relative fecundity. There was no significant difference (t-test: t= 1.889, 4 df, p=0.066) of 

relative fecundity between the populations of Nakantenga (N=13, mean relative 

fecundity=2451 ± 495 SD) and Thumbi west (N=30, mean relative fecundity=2183 ± 395 

SD). 

The GSI threshold above which oocyte weight did no longer increase significantly was 

4% for Ps. tropheops 'orange chest'. At Nakantenga sites, 7 females had a GSI superior to 

4%, giving a mean oocyte weight of 17.7 mg ± 2.73. At Thumbi west sites, 9 females had a 

GSI superior to 4%, giving a mean oocyte weight of 17.4 mg ± 2.68. The difference in oocyte 

weight between the females of two sites was not significant (t-test t=-0.534, 14 df, p=0.602). 

242

4,0 

16 


3,9 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

3,0 

CF 

Secchi 

Chloro a 

14 

12 

10 

8 

6 

4 

2 

0 



a 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

21 36 19 39 34 13 27 27 

Feb-99 

Mar-99 

Apr-99 

May-99 

4,0 

3,9 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

3,0 

CF 

Secchi 

Chloro a 

Females 

16 

14 

12 

10 

8 

6 

4 

2 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

b 8 16 4 19 10 8 17 10 

4,0 

3,9 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

3,0 

CF 

Secchi 

Chloro a 

Males 

16 

14 

12 

10 

8 

6 

4 

2 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

c 13 20 15 20 24 5 10 17 

Figure S8. Monthly progression of Pseudotropheus tropheops 'orange chest' mean 

condition factor, Secchi disk measurements (m) and benthic chlorophyll a 

biomass (µg/cm²) at Nakantenga sites for females and males pooled (a), 

females only (b) and males only (c). Numbers below the months are the 

corresponding sample numbers. 

243

Analysis of monthly progression of condition factor (CF) were carried out separately 

at Nakantenga and Thumbi sites. 

Nakantenga: The monthly progression of the mean condition factor, Secchi disk 

measurements and benthic chlorophyll a biomass are presented in Figure S8a. No correlation 

was observed between CF and length of fish (Pearson correlation p>0.05), allowing to test for 

between months (October to May), sites (Nex and Nsh) and sex (females and males) 

differences using a general linear model with month, site and sex as factors and CF as 

dependant variable. The only significant factors was month (F 7,189 =8.533, p0.05). Despite their similar monthly progression (Figure S8c), there 

was no significant correlation between the mean CF and Secchi disk measurements for males. 

Mean CF and chlorophyll a biomass were not correlated either. 

Once separated, both females and males showed significant differences of mean condition 

factor among months (F 7,79 =3.061, p=0.007 for females and F 7,110 =9.409, p


3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

3 

CF 

Secchi disk 

Chloro a 

18 

16 

14 

12 

10 

8 

6 

4 

2 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 



35 70 121 70 67 71 88 

Figure S9. Monthly progression of Pseudotropheus tropheops 'orange chest' 

mean condition factor, Secchi disk measurements (m) and benthic chlorophyll 

a biomass (µg/cm²) at Thumby west sites. Numbers below the months are the 


245

At Thumbi west, no correlation was observed between CF and length of fish (Pearson 

correlation p>0.05), allowing to search for between months, sites and sex differences. The 

only factor with a significant effect was month (F 7,496 =24.801, p0.05), allowing to test for between months (October to May), sites (Nex and Nsh) and sex 

(females and males) differences using a general linear model with month, site and sex as 

factors and CF as dependant variable. All three factors had a significant effect and interactions 

existed between all of them. Therefore, sites and sex had to be separated in analysis to test for 

between months differences. 

At Nakantenga sheltered site (Nsh), there were significant monthly differences of CF for both 

females (Figure S10b, F 7,499 =22.978, p


3,8 

3,6 

3,4 

3,2 

3,0 

2,8 

2,6 

Oct-98 

CF 

Secchi 

Chloro a 

Nov-98 

Dec-98 

Jan-99 

a 566 342 362 1037 216 419 381 333 

Feb-99 

Mar-99 

Apr-99 

May-99 

16 

14 

12 

10 

8 

6 

4 

2 

0 



3,8 

Nsh females 

16 

3,8 

Nsh males 

16 

3,6 

3,4 

14 

12 

10 

3,6 

3,4 

14 

12 

10 

3,2 

8 

3,2 

8 

3,0 

2,8 

6 

4 

2 

3,0 

2,8 

6 

4 

2 

2,6 

0 

2,6 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

b 

95 36 76 33 48 22 186 11 

c 

165 243 281 233 89 37 108 100 

3,8 

3,6 

3,4 

Nex females 

16 

14 

12 

10 

3,8 

3,6 

3,4 

Nex males 

16 

14 

12 

10 

3,2 

8 

3,2 

8 

3,0 

2,8 

6 

4 

2 

3,0 

2,8 

6 

4 

2 

2,6 

0 

2,6 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

d 

24 8 2 357 12 202 43 78 

e 

282 55 3 414 67 158 44 144 

Figure S10. Monthly progression of Pseudotropheus zebra 'red dorsal' mean condition factor, 

Secchi disk measurements (m) and benthic chlorophyll a biomass (µg/cm²) at Nakantenga for 

both sites, females and males pooled (a), at Nakantenga sheltered (Nsh) for females only (b), 

males only (c) and at Nakantenga exposed (Nex) for females only (d), males only (e). Numbers 

below the months are the corresponding sample numbers. 

247

At Nakantenga exposed site (Nex), there were also significant monthly differences of CF for 

both females (Figure S10d, F 7,718 =13.742, p


3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

CF 

Secchi 

Chloro a 

16 

14 

12 

10 

8 

6 

4 

2 



3,2 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

4 23 32 37 26 18 42 63 

Figure S11. Monthly progression of Pseudotropheus zebra 'yellow throat' 

mean condition factor, Secchi disk measurements (m) and benthic 

chlorophyll a biomass (µg/cm²) at Nakantenga for both sites. Numbers 


4,1 

16 


4,0 

3,9 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

CF 

Secchi 

Chloro a 

14 

12 

10 

8 

6 

4 

2 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

Chlorophylla biomass (ug/cm²) 


May-99 

57 35 11 9 16 21 21 17 

Figure S12. Monthly progression of Pseudotropheus zebra 'black dorsal' 


chlorophyll a biomass (µg/cm²) at Nakantenga for both sites. Numbers 


249

Pseudotropheus zebra 'black dorsal' 

The monthly progression of Ps. zebra 'black dorsal' mean condition factor, Secchi 

disk measurements and benthic chlorophyll a biomass at Nakantenga sites are presented in 

Figure S12. No correlation was observed between CF and length of fish (Pearson correlation 

p>0.05), allowing to test for between months (October to May), sites (Nex and Nsh) and sex 


factors and CF as dependant variable. The only factor with a significant effect was month 

(F 7,157 =6.220, p


4,0 

3,9 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

CF 

Secchi 

Chloro a 

16 

14 

12 

10 

8 

6 

4 

2 



3,1 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

14 14 5 9 43 20 16 56 

Figure S13. Monthly progression of Petrotilapia 'fuscous' mean condition 

factor, Secchi disk measurements (m) and benthic chlorophyll a biomass 

(µg/cm²) at Nakantenga for both sites. Numbers below the months are the 


3,8 

16 


3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

3,0 

2,9 

CF 

Secchi 

Chloro a 

14 

12 

10 

8 

6 

4 

2 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 



38 29 4 25 52 85 42 22 

Figure S14. Monthly progression of Pseudotropheus barlowi mean condition 


(µg/cm²) at Nakantenga for both sites. Numbers below the months are the 


251

versus February-May (Figure S13 and Table S7). Using these subsets, CF was significantly 

lower in February -May than in October- January (F 1,175 =46.847, p0.05). 

Pseudotropheus barlowi 

The monthly progression of Ps. barlowi mean condition factor, Secchi disk 

measurements and benthic chlorophyll a biomass at Nakantenga sites are presented in Figure 

S14. No correlation was observed between CF and length of fish (Pearson correlation 

p>0.05), allowing to test for between months (October to May), sites (Nex and Nsh) and sex 


factors and CF as dependant variable. The only factor with a significant effect was month 

(F 7,271 =7.967, p


4,0 

3,9 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

CF 

Secchi 

Chloro a 

16 

14 

12 

10 

8 

6 

4 

2 



3,2 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

15 2 10 42 2 43 73 

Figure S15. Monthly progression of Pseudotropheus williamsi mean 


biomass (µg/cm²) at Nakantenga for both sites. Numbers below the months 

are the corresponding sample numbers. 

3,5 

16 


3,4 

3,3 

3,2 

3,1 

3,0 

2,9 

2,8 

CF 

Secchi 

Chloro a 

14 

12 

10 

8 

6 

4 

2 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 



13 2 40 33 9 1 10 

Figure S16. Monthly progression of Pseudotropheus 'aggressive zebra' mean 


biomass (µg/cm²) at Nakantenga for both sites. Numbers below the months 

are the corresponding sample numbers. 

253

The only month differing from the others was April, when mean condition factor was slightly 

lower. However, the temporal variations of the mean CF did not show any clear seasonal 

pattern (Figure S15) and season did not seem to have a marked effect on CF for this species. 

The mean CF was correlated neither with the Secchi disk measurements, nor the chlorophyll a 

biomass (Pearson correlation p>0.05). 

Pseudotropheus 'aggressive zebra' 

The monthly progression of Ps. 'aggressive zebra' mean condition factor, Secchi disk 

measurements and benthic chlorophyll a biomass at Nakantenga sites are presented in Figure 

S16. No correlation was observed between CF and length of fish (Pearson correlation 

p>0.05), allowing to test for between months, sites and sex differences using a general linear 

model with month, site and sex as factors and CF as dependant variable. Despite the fact that 

the mean condition factor tended to decrease along the rainy season from November to May, 

none of the factors had a significant effect indicating that the observed differences were not 

significant. The mean CF was correlated neither with the Secchi disk measurements, nor the 

chlorophyll a biomass (Pearson correlation p>0.05). 

Species present only at Thumbi west sites 

Pseudotropheus zebra 

The monthly progression of Ps. zebra 'red dorsal' mean condition factor, Secchi disk 

measurements and benthic chlorophyll a biomass at Thumbi west sites are presented in Figure 

S17a. No correlation was observed between CF and length of fish (Pearson correlation 

p>0.05), allowing to test for between months (October to May), sites (T13 and T13) and sex 


factors and CF as dependant variable. All three factors had a significant effect and interactions 

existed between month and site, and between month and sex. Therefore, sites and sex had to 

be separated in analysis to test for monthly differences. 

At Thumbi T13, there were significant monthly differences of CF for both females (Figure 

S17b, F 7,283 =3.96, p


3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

CF 

Secchi 

Chloro a 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 



3,0 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

a 

381 117 59 481 273 233 125 300 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

3,0 

T13 females 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

3,0 

T13 males 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

b 

112 21 25 58 16 19 11 29 

c 

172 77 34 169 56 76 18 60 

3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

3,0 

T8 females 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

3,8 

3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

3,0 

T8 males 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

d 

35 7 114 82 60 26 104 

e 

62 12 140 119 78 70 107 

Figure S17. Monthly progression of Pseudotropheus zebra mean condition factor, Secchi disk 

measurements (m) and benthic chlorophyll a biomass (µg/cm²) at Thumby west for both sites, 

females and males pooled (a), at Thumby T13 for females only (b), males only (c) and at 

Thumby T8 for females only (d), males only (e). Numbers below the months are the 


255

Despite some significant differences among months, there was no evidence of seasonal effect 

on CF for females (Figure S17b), and only a slight decreasing trend for males (Figure S17c). 

The mean CF was correlated neither with the Secchi disk measurements, nor with the 

chlorophyll a biomass (Pearson correlation p>0.05) for both females and males. 

At Thumbi T8, there were significant monthly differences of CF for both females (Figure 

S17d, F 7,421 =21.746, p


4,0 

3,9 

3,8 

3,7 

3,6 

3,5 

3,4 

CF 

Secchi 

Chloro a 

18 

16 

14 

12 

10 

8 

6 

4 

2 



3,3 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

25 16 18 38 13 16 37 31 

Figure S18. Monthly progression of Pseudotropheus tropheops 'red cheek' 


chlorophyll a biomass (µg/cm²) at Thumby west for both sites. Numbers 



3,5 

3,4 

3,3 

3,2 

3,1 

3,0 

2,9 

CF 

Secchi 

Chloro a 

18 

16 

14 

12 

10 

8 

6 

4 

2 

2,8 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 



86 21 14 63 44 33 47 14 

Figure S19. Monthly progression of Pseudotropheus callainos mean 


biomass (µg/cm²) at Thumby west for both sites. Numbers below the 

months are the corresponding sample numbers. 

257

Despite the clear tendency of mean CF to decrease along the rainy season (Figure S18), it was 

not possible to group months (i.e. October-January versus February-May) owing to the 

significant difference between March and April and May (Table S10). However, separating 

April-May from the other months, the mean CF was significantly lower in April-May 

(F 1,192 =37.886, p0.05), allowing to test for between months (October to May), sites (T13 and T8) and sex 


factors and CF as dependant variable. Month had a significant effect (F 7,294 =4.050, p


3,7 

3,5 

3,3 

3,1 

2,9 

2,7 

CF 

Secchi 

Chloro a 

18 

16 

14 

12 

10 

8 

6 

4 

2 



2,5 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 

16 18 23 54 47 29 19 85 

Figure S20. Monthly progression of Pseudotropheus aurora mean condition 


(µg/cm²) at Thumby west for both sites. Numbers below the months are 

the corresponding sample numbers. 


3,7 

3,6 

3,5 

3,4 

3,3 

3,2 

3,1 

CF 

Secchi 

Chloro a 

18 

16 

14 

12 

10 

8 

6 

4 

2 

3 

0 

Oct-98 

Nov-98 

Dec-98 

Jan-99 

Feb-99 

Mar-99 

Apr-99 

May-99 



5 9 18 24 21 29 23 

Figure S21. Monthly progression of Petrotilapia nigra mean condition factor, 

Secchi disk measurements (m) and benthic chlorophyll a biomass (µg/cm²) 

at Thumby west for both sites. Numbers below the months are the 


259

A similar pattern as that of Ps. callainos was observed for Ps. aurora, with the mean 

condition factor decreasing from November to December, a plateau phase from December to 

March and then another decrease in April-May (Figure S20). During what we called the 

plateau phase, a sharp increase of the mean CF arose in January as for most of the species at 

Thumbi sites, though in a lesser extent. Statistical results confirmed that most of the observed 

differences were significant. The mean CF was significantly correlated to the Secchi disk 

measurements (r=0.829, p=0.011), but not with the chlorophyll a biomass. 

Petrotilapia nigra 

The monthly progression of P. nigra mean condition factor, Secchi disk measurements 

and benthic chlorophyll a biomass at Thumbi west sites are presented in Figure S21. No 

correlation was observed between CF and length of fish (Pearson correlation p>0.05), 

allowing to test for between months (October to May), sites (T13 and T8) and sex (females 

and males) differences using a general linear model with month, site and sex as factors and 

CF as dependant variable. None of the factors had a significant effect on mean condition 

factor, which translated the absence of seasonal effect on CF (Figure S21). The mean CF was 

correlated neither with the Secchi disk measurements nor the chlorophyll a biomass. 

Discussion 

Parallel to this study of rocky shore fish communities (September 1998 to April 1999), 

the impact of suspended sediments on the abundance and species richness of near-shore sandy 

fishes was investigated along a sedimentation gradient from Linthipe river mouth to 35 km 

northwards, during the course of a Master's degree (Sululu 2000). In Lake Tanganyika, Cohen 

et al. (1993a) compared the biodiversity and abundance of ostracods, diatoms and fish species 

in sites permanently characterised by diverse degrees of sedimentation disturbances, but for 

which no baseline data were available to assess the state of species communities before the 

apparition of the permanent disturbance. The originality of both Sululu's and ours studies, 

compared to that of Cohen et al. (1993a), lies in that we investigated the short term effects of 

a temporary sediment disturbance on sites for which pre-disturbance state was (previously) 

assessed. 

In Lake Tanganyika, sites highly impacted by sedimentation had a lower species 

richness and species abundance than non impacted sites, and sites characterised by an 

intermediate disturbance had intermediate species richness and abundance (Cohen et al. 

1993a). In Lake Malawi, significant variations in the abundance and species richness of sanddwelling 

fish were observed before and during the sedimentation period at the station closest 

to the Linthipe river mouth in gillnet catches, but not in beach seine catches (Sululu 2000). 

However, over all the stations sampled by beach seine, the mean abundance of cichlids rose 

during the rainy season, while species richness declined (Sululu 2000). Some fluctuations in 

the relative abundance of rock-dwelling species were observed among months at both 

Nakantenga and Thumbi sites. However, these variations were not significant. Thumbi west 

sites housed a slightly higher number of species than Nakantenga sites and on both islands 

exposed sites were richer than protected ones: 24, 27, 30 and 33 species for Nsh, Nex, T13 

and T8 respectively. However, within each site, the number of species did not changed along 

the rainy season, thus during the disturbance period. 

Close to the Linthipe river mouth, both species richness and abundance of sand-dwelling 

fishes showed overall significant negative correlation with the concentration of suspended 

sediments (Sululu 2000). The concentration of suspended sediment was assessed by Secchi 

disk measurements in our case. Correlation between species richness or abundance and 

260

concentration of suspended sediment was not investigated as species richness and abundance 

did not differ significantly. On the other hand, a striking pattern appeared when looking at the 

relationship between the fishes mean condition factor and the Secchi disk measurements or 

the mean chlorophyll a biomass. Despite often similar monthly progression, there was very 

few significant correlation between the species mean CF and these two parameters. However, 

at Nakantenga, the mean CF of species, when significant, was always positively correlated to 

the chlorophyll a biomass (r=0.744, p=0.034 for Ps. zebra 'black dorsal' and r=0.943, 

p=0.0005 for Ps. zebra 'yellow throat'), whereas it was always correlated with the Secchi disk 

depth when significant at Thumbi West (r=0.829, p=0.011 for Ps. aurora; r=0.870, p=0.005 

for Ps. callainos; r=0.768, p=0.026 for Ps. tropheops 'red cheek'; r=0.843, p=0.017 for 

females, r=0.842, p=0.017 for males Ps. zebra at T8). Chlorophyll a biomass was found to 

fluctuate very little in Thumbi West sites (Figure S4) apart from a temporary large peak 

resulting from land runoff. On the other hand, water clarity (assessed by Secchi disk 

measurements) decreased slightly but steadily from October to April-May (Figure S4). A 

decreased visibility might interfere with plankton foraging for the partly plankton feeding 

species of the Ps. zebra complex and account for a decreased condition. But it is more 

difficult to explain the influence of a slightly decreased water clarity on a preferentially 

aufwush feeder like Ps. tropheops 'red cheek', especially if the effect of water transparency 

does not affect algae biomass. Suspended particles, though not as abundant as at Nakantenga, 

might decrease the nutritional value of the benthic detritus and algae (Cohen et al. 1993b), 

accounting for the observed decreased condition of fish species. Next, if a slight decline of 

water clarity affects the fish condition, why for none of the species (not even species of the 

Ps. zebra complex) the CF was correlated to Secchi disk measurements at Nakantenga where 

the visibility strongly decreased? A likely explanation is that unlike the monthly mean 

chlorophyll biomass, which probably reflected real monthly trends (see results), monthly 

Secchi disk measurements, taken once a month, may not represent monthly trends at 

Nakantenga, but rather transient trends. Indeed, even though the island was most of the time 

surrounded by "cloudy waters" from February to April-May, the dense sediment plume 

resulting in very poor visibility was going back and forth according to wind and current 

direction. Given the observed steady sediment deposition on rocks from February to April- 

May at Nakantenga, and the subsequent decline of algal biomass, it appears logical that fish 

condition was rather correlated to algae biomass, even if this correlation was significant for a 

few species only. 

At Thumbi West sites, an sudden increase of the mean condition factor was observed for all 

the species in January. As this increase occurred for every species, it is likely to be a 

consequence of the algae biomass peak recorded in December at every site in Thumbi Island. 

The influence of suspended sediment on life history characteristics could only be 

assessed on two species present at both islands and abundant enough at both of them. For both 

species (L. fuelleborni and Ps. tropheops 'orange chest') no significant difference of size at 

maturity, fecundity or oocyte weight was observed between the site highly impacted by 

fluvial sediment (Nakantenga) and the control site (Thumbi West). 

The mean condition factor showed a clear decreasing trend along the rainy season (i.e. during 

the sediment plume's influence) for every species but one at Nakantenga sites. More 

surprising was the same tendency, though in a lesser extent, observed at Thumbi West sites, 

which were supposed to be the sediment free or control sites. Although the sediment plume 

was seen once surrounding Mumbo Island off Cape Maclear, it never reached Thumbi West. 

Also there is no major river at Cape Maclear and despite deforestation problems on the steep 

slopes of the Nankumba peninsula resulting in some silt deposition on the rocky shores 

(Bootsma 1992), Thumbi West is by far less impacted by sedimentation than Nakantenga. 

261

Table S12. Homogeneous months grouping for between-season comparison of mean condition factor per 

species at Nakantenga. White bars, mean CF does not differ from the dry season mean value. Black 

bars, mean CF significantly lower than the dry season mean value. Absence of bars means that season 

had no effect on the CF. 

Species Oct-98 Nov-98 Dec-98 Jan-99 Feb-99 Mar-99 Apr-99 May-99 

Ps. zebra 'red dorsal' 

Ps. zebra 'yellow throat' 

Ps. zebra 'black dorsal' 

Ps. 'aggressive zebra' 

Petrotilapia 'fuscous' 

Ps. barlowi 

Ps. williamsi 


Ps. tropheops 'orange chest' M 

F 

Table S13. Homogeneous months grouping for between-season comparison of mean condition factor 

per species at Thumby west. White bars: mean CF does not differ from the dry season mean value. 

Black bars: mean CF significantly lower than the dry season mean value. Grey bars: intermediate 

state. Absence of bars means that season had no effect on the CF. 

Species Oct-98 Nov-98 Dec-98 Jan-99 Feb-99 Mar-99 Apr-99 May-99 

Ps. zebra 

Ps. . tropheops 'red cheek 

Ps. callainos 

Ps. aurora 

Petrotilapia nigra 


Ps. tropheops 'orange chest' 

262

Benthic algal biomass and water clarity measurements along the course of this study clearly 

supported this statement. However, Thumbi West might possibly not be considered as a real 

sediment free site, which would account for the observed results. 

The decrease of mean condition factor along the rainy season did not start at the same time for 

every species at Nakantenga (Table S12). Depending on species, the date at which the mean 

CF started to significantly decrease varied from January to April, but for most of them it 

started in March (50%) and February (25%), which was the period when the Linthipe river 

discharge was maximum (Figure S3), the water clarity the worst and the algae biomass the 

lowest (Figure S4). A similar tendency was observed at Thumbi West site (Table S13), but the 

difference in CF appeared more progressively to become really significant later in the season 

for most of the species. It is interesting to note that for most of the species at Thumbi West, 

the lowest CF were recorded in April-May, when algae biomass and water clarity had already 

started to increase (Figure S4). 

The influence of season on the mean CF was also observed to vary between sites at each 

island. At Nsh, for Ps. zebra 'red dorsal' season had the same significant effect on CF for both 

females and males whereas no effect was detected at Nex for either sex. At Thumbi West T8, 

for Ps. zebra, season had the same significant effect on CF for both females and males 

whereas no effect was detected at site T13 for either sex. Interestingly, these species are 

ecological equivalent on the two islands and were the two most abundant species. 

The hypothesis that during the sediment plume, the fish would move upwards from the 

deep waters to the shallows to compensate for the shortage of food availability in the deeper 

waters, was not verified during this study. At Nsh site, only one species was found more 

abundant at 2 m during the rainy season. Two species were more abundant at 6 m and 3 at 10 

m, whereas only 2 species were less abundant at 10 m and one at 6 m. At Nex site, three 

species were found less abundant at 10 m and one at 6 m during the rainy season, whereas at 

each depth one species was found more abundant. At Thumbi West sites, only three species 

were less abundant at 6 m during the rainy season, whereas two were more abundant at 10 m, 

three at 6 m and six at 2 m. Furthermore, none of these trends were significant. However, it is 

likely that the monthly sampling frequency was not adequate to reveal such a phenomenon, 

which is probably rather transient. Indeed, during one sampling at Nakantenga exposed site in 

February 1999, we experienced an exceptional event. A dense sediment plume was settled 

around the island, water visibility being less than 2 m at the surface (Secchi disk depth: 1.5 

m), when a narrow band (about 20 m width) of clearer water drifted towards our site. We 

quickly equipped ourselves with scuba gears and slates, and jumped down the water to 

observe the fishes behaviour during that event. Before the arrival of the clearer water band, 

fishes were moving only very little, staying within a 50 cm radius from their hiding hole, 

some of them not moving at all, behaving as if it were night. When the clearer water band 

reached the site, within seconds all the fishes gathered in the water column near the surface in 

a huge crowd cropping frenetically upon plankton. A few minutes later, the clear water was 

gone and the fishes almost instantaneously returned down to their respective hiding holes, as 

if nothing had happened and it were night again (Ribbink & Duponchelle, unpublished data). 

This quick observation might actually be what happens when the wind and currents drive the 

plume away from the island. Opportunistic behaviour and ability to food switching of mbuna 

are well documented (Fryer & Iles 1972, McKaye & Marsh 1983, Ribbink et al. 1983, 

Reinthal 1990). Our hypothesis is that during darkness resulting from the dense plume's 

presence, mbuna would spend very little energy in territorial and sexual activity given the 

poor visibility and in feeding upon scarce and/or blanketed benthic material with diminished 

nutritional value. Instead of that, they would keep resting, expensing only limited amounts of 

263

energy, awaiting for improved environmental conditions. As soon as visibility improves, they 

all move in the water column towards the shallows, feeding upon the large amounts of 

plankton flourishing from the nutrients associated with suspended particles, making up for the 

food deprivation endured or building up reserves. This strategy would allow the highly 

stenotopic mbuna to pass through the still-temporary disturbance occasioned by the suspended 

sediment. It would explain why, unlike the easily-moving sand-dwelling species, no 

significant variations in species richness or abundance were observed and why only a 

decreased body condition was recorded for mbuna. The stable isotope samples taken during 

this study but unfortunately not available at the moment might shed some light and support or 

reject that hypothesis. 

This study is considered as preliminary. A clear decreased in body condition was observed for 

almost every species during the period of the sediment plume's influence. However, our 

results also suggest that a sampling design with much shorter sampling intervals is necessary 

to better understand the dynamics of rock-dwelling fish reaction to suspended sediment 

disturbance. A regular stomach content analysis before and during the sedimentation period 

would help testing our hypothesis. Multivariate analysis of the data, which would help 

clarifying the observed trends, was not possible given the time constraint over this report, but 

will be carried out ultimately. Despite the seasonal and temporary effects of suspended 

sediment, still restricted to the rainy season, impacts have already been detected on sandy 

(Sululu 2000) and rocky fish communities. Given the increase of anthropogenic activities that 

lead to habitat degradation and increased erosion around the Lake shore and the steadily 

increasing human population, this impact is very likely to considerably worsen in the coming 

years. 

264

References cited

References cited 

Albaret JJ. 1982. Reproduction et fécondité des poissons d'eau douce de Côte d'Ivoire. Rev. 

Hydrobiol. Trop. 15: 347-371. 

Allison E.H., K. Irvine, A.B. Thompson & B.P. Ngatunga 1996. Diets and food consumption 

rates of pelagic fish in Lake Malawi, Africa. Freshw. Biol. 35: 489-515. 

Anderson R.O. & S.J. Gutreuter 1983. Length, weight and associated structural indices, pp. 283- 

300. In Fisheries techniques (Nielsen L.A. & D.L. Johnson eds.). Am. Fisher. Soc. Bethesda, 

Maryland, 468p. 

Bagenal T.B. 1969. Relationship between egg size and fry survival in Brown Trout Salmo truta 

L.. J. Fish Biol. 1: 349-353. 

Bagenal T.B. 1978. Aspects of fish fecundity. In: Ecology of freshwater fish production, S.D. 

Gerking (ed.), pp 75-101. Blackwell Scientific Publications, Oxford. 

Banda M.C. & T. Tómasson 1996. Surveys of trawling grounds and demersal fish stocks in 

central Lake Malawi, from Dormira Bay to Nkhata Bay, in 1994 and 1995. Government of 

Malawi, Fisheries Department, Fish. Bul. 33. 34p. 

Banda M.C., T. Tómasson & D. Tweddle 1996. Assesment of the deep water trawl fisheries of 

the South East Arm of Lake Malawi using exploratory surveys and commercial catch data, pp. 

53-75. In Stock assessment in inland fisheries (I.G. Cowx ed.). Fishing New Books, Oxford. 

Belkhir K., P. Borsa, J. Goudet, L. Chikhi & F. Bonhomme 1996 - 1998. GENETIX, logiciel sous 

Windows TM pour la génétique des populations. Laboratoire Génome et Populations, CNRS 

UPR 9060, Université de Montpellier II, Montpellier (France). 

Bootsma H.A. 1992. Lake Malawi National Park: an overview. Mitt. Internat. Verein. Limnol. 23: 

125-128. 

Bootsma H.A. & R.E. Hecky 1993. Conservation of the African Great Lakes: a limnological 

perspective. Conserv. Biol. 7(3): 644-656. 

Bootsma H.A. & R.E. Hecky 1999. Water Quality Report, Lake Malawi/Nyasa Biodiversity 

Conservation Project. 276pp. 

Bootsma H.A., R. E. Hecky, R.H. Hesslein & G.F. Turner, 1996. Food partitioning among Lake 

Malawi nearshore fishes as revealed by stable isotope analyses. Ecology 77 (4): 1286-1290. 

Bouton N, O. Seehausen & J.J.M. van Alphen 1997. Resource partitioning among rock-dwelling 

haplochromines (Pisces: cichlidae) from Lake Victoria. Ecol. Freshwat. Fish 6: 225-240. 

Cambray J.A. & M.N. Bruton 1994. Evolutionary trade-off between egg size and egg number in a 

sister species pair of redfin ninnows, Pseudobarbus afer and P. asper (Osteichthyes: 

Cyprinidae). Ichthyol. Explor. Freshwaters 5: 305-320. 

Casselman J.M. 1987. Determination of age and growth, pp. 209-242. In: The Biology of Fish 

Growth (Weatherley A.H. & H.S. Gill eds.). Academic Press, 443 p. 

Cohen A.S., S. Kaufman & R. Oguttu-Ohwayo 1996. Anthropogenic threats, impacts and 

conservation strategies in the African Great Lakes: a review pp. 576-614. In The limnology, 

climatology and paleolimnology of the Great African Lakes (Johnson T.C. & E.O. Odada 

eds.). Gordon & Breach publishers. 

Cohen A.S., R. Bills, C.Z. Cocquyt & A.G. Caljon 1993a. The impact of sediment pollution on 

biodiversity in Lake Tanganyika. Conserv. Biol. 7: 667-677. 

Cohen A.S., R. Bills, M. Gashagaza, E. Michel, J-J. Tiercelin, K. Martens, M. Shoreghan, P. 

Coveliers, K. West, G. Ntakimazil, S. Mboko, & K. Sona 1993b. Preliminary observations of 

sedimentation impacts on benthic environments and biodiversity using an ROV submersible in 

265

Lake Tanganyika. In Abstracts and Programme of PARADI (Poissons, resources, diversité), 

Sénégal, November 1993. 

Cooley P.M. 1999. Coastal and Littoral Habitat. In CEOS Data Report in Support of the Lake 

Malawi Biodiversity Conservation Project, Malawi, Africa: Volume II (Cooley, P. M., G. K. 

McCullough, & F.X. Mkanda eds.). Centre for Earth Observation Science Data Report 99-9-1. 

99 p. 

Coulter G.W. 1991. Pelagic fish. In: Lake Tanganyika and its life (Coulter G.W. ed.), pp. 111- 

138. British Museum of Natural History/Oxford University Press, London. 

Craig J.F. 1978. A note on ageing in field with special reference to perch, Perca fluviatilis L.. 

Verh. Internat. Verei. Limn. 20: 2060-2064. 

Darwall W.R.T. 1999. Diet analysis. In "The trophic ecology of the demersal fish community of 

Lake Malawi/Niasa, central Africa", First Annual Report. INCO-DC: International 

Cooperation with Developing Countries. 

Day M. Cpt. 1999. NR International Ltd's consultant for SADC/GEF Lake Malawi Biodiversity 

Conservation Project. Final Report. 

De Silva S.S. 1986. Reproductive biology of Oreochromis mossambicus populations of manmade 

lakes in Sri Lanka: a comparative study. Aquacul. Fisher. Manag.17: 31-38. 

Duarte C.M. & M. Alcaraz 1989. To produce many small or few large eggs: a size-independante 

tactic of fish. Oceologia 80: 401-404. 

Duponchelle, F. 1997. Reproduction du tilapia (Pisces, Cichlidae) Oreochromis niloticus 

(Linnaeus, 1758) dans les retenues artificielles de Côte d'Ivoire : analyse comparative des 

modalités de reproduction et approche expérimentale de leur déterminisme. PhD Thesis, 

Université de Bretagne Occidentale. 

Duponchelle F. & Panfili J. (1998). Variations in age and size at maturity of female Nile tilapia, 

Oreochromis niloticus, populations from man-made lakes of Côte d'Ivoire. Envir. Biol. Fish. 

52: 453-465. 

Duponchelle F. P. Cecchi, D. Corbin, J. Nunez, & M. Legendre (1999). Spawning season 

variations of female Nile tilapia, Oreochromis niloticus, populations from man-made lakes of 

Côte d'Ivoire. Envir. Biol. Fish. 56: 377-389. 

Duponchelle F., P. Cecchi, D. Corbin, J. Nunez, & M. Legendre (2000). Variations in fecundity 

and egg size of female Nile tilapia, Oreochromis niloticus, populations from man-made lakes 

of Côte d'Ivoire. Envir. Biol. Fish. 57: 155-170. 

Eccles D.H. (1974). An outline of the physical limnology of Lake Malawi (Lake Nyasa). Limnol. 

Oceanogr. 19: 730-742. 

Eccles D.H. & D.S.C. Lewis (1981). Midwater spawning in Haplochromis chrysonotus 

(Boulenger) (Teleostei: Cichlidae) in Lake Malawi. Env. Biol. Fish. 6: 201-202. 

Eccles D.H. & E. Trewavas 1989. Malawian cichlid fishes. Lake Fish Movies, Herten, Germany, 

335 p. 

Elgar M.A. 1990. Evolutionary compromise between a few large and many small eggs: 

comparative evidence in teleost fish. Oikos 59: 283-287. 

Evans D.O., K.H. Nicholls, Y.C. Allen & M.J. Memurty 1996. Historical land use, phosphorus 

loading and loss of fish habitat in Lake Simcoe, Canada. Can J. Fish. Aquat. Sci. 53:194-218. 

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 

Technical Paper 21, 113p. 

Fergusson A., B. Dermanm & R. Mkandwire 1993. The new development rhetoric and Lake 

Malawi. Africa 63: 1-18. 

Fishelson L. 1966. Cichlidae of the genus Tilapia in Israel. Bamidgeh 18: 67-80. 

Fry B. P.L. Mumford, F. Tam, D.D. Fox, G.L. Warren, K.E. Havens & A. D. Steinman 1999. 

Trophic position and individual feeding histories of fish from Lake Okeechobee, Florida. Can. 

J. Fish. Aquat. Sci. 56: 590-600. 

266

Fryer G. 1959. The trophic interrelationships and ecology of some littoral communities of Lake 

Nyasa, with especial reference to the fishes. Proc. Zool. Soc. Lond. 132: 153-281. 

Fryer G. 1972. Conservation of the Great Lakes of Africa: a lesson and a warning. Biol. Conserv. 

4(4): 256-262. 

Fryer G. 1984. The conservation and rational exploitation of the biota of Africa's Great Lakes. S. 

Afr. Nat. Sci. Prog. Rep. 92: 135-154. 

Fryer G. & T.D. Iles 1972. The Cichlid Fishes of the Great Lakes of Africa. Oliver & Boyd, 

London, 641 p. 

Gannes L.Z., D.M. O'Brien and C.D.M. Rio 1997. Stable isotopes in animal ecology: 

assumptions, caveats, and a call for more laboratory experiments. Ecology 78: 1271-1276. 

Gauthier J.Y., M.A. Richard-Yris, B. Lefaucheux & M. Foraste 1996. Plasticité du cycle parental 

chez Oreochromis niloticus. In: Proceedings of The Third International Symposiumon Tilapia 

in Aquaculture (Pullin R. S. V., J. Lazard, M. Legendre, J.-B. Amon Kothias & D. Pauly eds.), 

pp. 340-346. ICLARM Conf. Proc. 41, 575 p. 

Gayanilo F.C. Jr., P. Sparre & D. Pauly 1996. The FAO-ICLARM Stock Assessment Tool 

(FiSAT) user's guide. FAO Comput. Info. Ser. (Fish.) 7, FAO, Rome. 124pp. 

Gayanilo F.C. Jr. & D. Pauly 1997. The FAO-ICLARM Stock Assessment Tool (FiSAT) 

Reference manual. FAO Comput. Info. Ser. (Fish.) 8, FAO, Rome. 262 p. 

Genner M.J., G.F. Turner, S. Barker & S.J. Hawkins 1999a. Niche segregation among Lake 

Malawi cichlid fishes? Evidence from stable isotope signatures. Ecol. Let. 2: 185-190. 

Genner M.J., G.F. Turner & S.J. Hawkins 1999b. Resource control by territorial male cichlid fish 

in Lake Malawi. J. Anim. Ecol. 68: 522-529. 

Genner M.J., G.F. Turner & S.J. Hawkins 1999c. Foraging of rocky habitat cichlid fishes in Lake 

Malawi: coexistence through niche partitioning? Oecologia 121: 283-292. 

Gorokhova E. & S. Hansson 1999. An experimental study on variations in stable carbon and 

nitrogen isotope fractionation during growth of Mysis mixta and Neomysis integer. Can. J. 

Fish. Aquat. Sci. 56: 2203-2210. 

Greenwood P.H. 1984. What is a species flock?, pp. 13-19. In Evolution of fish species flocks 

(Echelle A.A. & I. Kornfield eds.). University of Maine at Orono Press. 

Hecky R.E. & H.A. Bootsma 1999. Introduction, pp. 18-28. In Water Quality Report (Bootsma 

H.A. & R.E. Hecky eds.), SADC/GEF Lake Malawi / Nyasa Biodiversity Conservation 

Project. 276 p. 

Hecky R.E., H.J. Kling, T.C. Johnson, H.A. Bootsma & P. Wilkinson 1999. Algal and 

sedimentary evidence for recent changes in the water quality and limnology of Lake 

Malawi/Nyasa, pp. 191-212. In Water Quality Report (Bootsma H.A. & R.E. Hecky eds.), 

SADC/GEF Lake Malawi / Nyasa Biodiversity Conservation Project. 276 p. 

Hesslein R.H., K.A. Hallard & P. Ramlal 1993. Replacement of sulfur, carbon and nitrogen in 

tissue of growing broad whitefish (Coregonus nasus) in response to a change in diet traced by 

δ 34 S, δ 13 C and δ 15 N. Can. J. Fish. Aquat. Sci. 50: 2071-2076. 

Hesslein R.H., M.J. Capel, D.E. Fox & K.A. Hallard 1991. Stable isotopes of sulfur, carbon, and 

nitrogen as indicators of trophic level and fish migration n the lower Mackenzie river basin, 

Canada. Can. J. Fish. Aquat. Sci. 48: 2258-2265. 

Higgins S., H. Bootsma, R. Hecky, H. Kling, J. Mwita & B. Mwichande 1999. Conserving 

biodiversity in the rocky littoral zones of Lake Malawi/Nyasa: a limnological perspective, pp. 

46-50. In SADC/GEF, Lake Malawi/Nyasa/Niassa Biodiversity Conservation Project, 

Extended abstracts, Senga Bay Conference (A.J. & A.C. Ribbink eds.). 4-5 March 1999, 

Senga Bay, Malawi. 179 p. 

Hynes H. 1950. The food of freshwater sticklebacks (Gasterosteus aculeatus and Pygosteus 

pungitius), with a review of the methods used in the studies of the food of fishes. J. Animal 

Ecol. 19: 6-58. 

267

Iles T.D. 1960. A group of zooplankton feeders of the genus Haplochromis (Cichlidae) in Lake 

Nyasa. Ann. Mag. Nat. Hist. 13: 257-280. 

Iles T.D. 1971. Ecological aspects of growth in African cichlid fishes. J. Cons. int. Explor. Mer. 

33: 363-385. 

Irvine K. 1995. Standing biomasses, production, spatial and temporal distributions of the 

crustacean zooplankton. pp. 85-108. In The fisheries potential and productivity of the pelagic 

zone of Lake Malawi/Niassa (A. Menz, ed.), SADC/ODA Final Report, 386 p. 

Jackson P.B.M., T.D. Iles, D. Harding & G. Fryer 1963. Report on the survey of northern lake 

Nyasa. Government Printers, Zomba. 171 pp. 

James N.P.E. & M.N. Bruton 1992. Alternative life history traits associated with reproduction in 

Oreochromis mossambicus (Pisces: Cichlidae) in smallwater bodies of the eastern Cape, South 

Africa. Envir. Biol. Fish. 34: 379-392. 

Kalipeni E. 1996. Demographic response to environmental pressure in Malawi. Pop. Environ. 17: 

285-308. 

King M (ed.) 1995. Fisheries Biology, Assesment and Management. Fishing New Books, Oxford, 

England, 341 p. 

Kingdon M.J., H.A Bootsma, J. Mwita, B. Mwichande & R.E. Hecky 1999. River discharge and 

water quality, pp. 29-69. In Water Quality Report, Lake Malawi/Nyasa Biodiversity 

Conservation Project (Bootsma H.A. & R.E. Hecky eds.). 276 p. 

Konings A. 1995. Malawi cichlids in their natural habitat. 2 nd Edition. Cichlid Press, Germany. 

352 p. 

Legendre M. 1992. Potentialités aquacoles des Cichlidae (Sarotherodon melanotheron, Tilapia 

guineensis) et Clariidae (Heterobranchus longifilis) autochtones des lagunes ivoiriennes. 

Ph.D. Thesis, Université Montpellier II, Montpellier. 

Legendre, M. & J.M. Ecoutin. 1989. Suitability of brackish water tilapia species from the Ivory 

Coast for lagoon aquaculture. I - Reproduction. Aquat. Living Resour. 2: 71-79. 

Legendre, M. & J.M. Ecoutin. 1996. Aspects of the reproductive strategy of Sarotherodon 

melanotheron: comparison between a natural population (Ebrie lagoon, Côte d'Ivoire) and 

different cultured populations. pp.326-338. In: R.S.V. Pullin, J. Lazard, M. Legendre, J.-B. 

Amon Kothias & D. Pauly (ed.), The Third International Symposium on Tilapia in 

Aquaculture. ICLARM Conf. Proc. 41, Abidjan. 

Lévêque C. 1997. Biodiversity dynamics and conservation: the freshwater fish of tropical Africa. 

Cambridge University Press. 438p. 

Lewis D.S.C. 1982. Problems of species definition in Lake Malawi cichlid fishes (Pisces: 

Cichlidae). J.L.B. Smith Institute of Ichthyology Special Publication 23: 1-5. 

Lewis D.S.C. & D. Tweddle 1990. Breeding seasonality in some commercially important trawlcaught 

species in the southern Lake Malawi. pp. 23-35. In: Collected Reports on Fisheries 

Research in Malawi, Occasional Papers Vol. 1 (Pitcher T. J. & C. R. Hollingworth eds.). 

ODA, Overseas Development Administration, London. 

Liem K.F. 1980. Adaptative significance of intra- and interspecific differences in the feeding 

repertoires of cichlid fishes. Amer. Zool. 20: 295-314. 

Liem K.F. 1991. Functional morphology, pp. 129-150. In Cichlid Fishes: behaviour, ecology and 

evolution (Keenleyside M.H.A. ed.). Chapman & Hall, 378p. 

Lowe R.H. 1953. Notes on the ecology and evolution of Nyasa fishes of the genus Tilapia, with a 

description of Tilapia saka, Lowe. Proc. Zool. Soc. Lond. 122: 1035-1041. 

Lowe R.H. 1955. The fecundity of tilapia species. East Afr. Agricult. J. 21: 45-52. 

Lowe-McConnell R.H. 1979. Ecological aspects of seasonality in fishes of tropical waters. Symp. 

Zool. Soc. Lond. 44, 219-241. 

Lowe-McConnell R.H. 1982. Tilapias in fish communities. In The biology and culture of tilapias, 

Pullin R.S.V. and R.H. Lowe-McConnell (eds), pp 83-113. ICLARM Conference Proceedings 

7, Manila, Philippines, 432 p. 

268

Lowe-McConnell R.H. 1987. Ecological studies in tropical fish communities. Cambridge 

University Press, Cambridge, 382 p. 

Lowe-McConnell R.H. 1993. Fish fauna of the African Great Lakes: origins, diversity, and 

vulnerability. Conserv. Biol. 7 (3): 634-643. 

Lowe-McConnell R.H. 1994. The role of ecological and behaviour studies of cichlids in 

understanding fish diversity and speciation in African Great Lakes: a review. Arch. Hydrobiol. 

Beih. Ergebn. Limnol. 44: 335-345. 

Lowe-McConnell R.H., F.C. Roest, G. Ntakimazi & L. Risch 1994. The African Great Lakes. In 

Biological diversity in African fresh and brackish water fishes (Teugels et al. eds.). 

Symposium PARADI. Ann. Mus. r. Afr. Centr., Zool. 275: 87-94. 

McCullough G. 1999. Transport of Linthipe River suspended sediments in Lake malawi/Nyasa, 

pp. 71-84. In Water Quality Report, Lake Malawi/Nyasa Biodiversity Conservation Project 

(Bootsma H.A. & R.E. Hecky eds.). 276 p. 

McKaye K.R. & A. Marsh 1983. Food switching by two specialized algae-scraping cichlid fishes 

in Lake Malawi, Africa. Oecologia 56: 245-248. 

Mann R.H.K. & C.A. Mills 1979. Demographic aspects of fish fecundity. Symp. Zool. Soc. 

London 44: 161-177. 

Mantel N. 1967. The detection of disease clustering and a generalized regression approach. 

Cancer Res. 27: 209-220. 

Marsh E. 1986. Effects of egg size on offspring fitness and maternal fecundity in the orangethroat 

darter, Etheostoma spectabile (Pisces, Percidae). Copeia 1986: 18-30. 

Marsh A.C. & A.J. Ribbink 1985. Feeding-site utilization in three sympatric species of 

Petrotilapia (Pisces: Cichlidae) from Lake Malawi. Biol. J. Lin. Soc. 25: 331-338. 

Marsh A.C. & A.J. Ribbink 1986. Feeding schools among Lake Malawi cichlid fishes. Env. Biol. 

Fish. 15: 75-79. 

Marsh B. A., A. C. Marsh & A. J. Ribbink 1986. Reproductive seasonality in a group of rockfrequenting 

cichlid fishes in Lake Malawi. J. Zool., Lond. (A) 209: 9-20. 

Marshall B. 1993. Biology of the African clupeid Limnothrissa miodon with reference to its small 

size in artificial lakes. Rev. Fish Biol. Fisheries 3: 17-38. 

McElroy D.M. & I. Kornfield 1990. Sexual selection, reproductive behavior, and speciation in the 

mbuna species flock of Lake Malawi (Pisces: Cichlidae). Env. Biol. Fish. 28: 273-284. 

McKaye K.R. 1983. Ecology and breeding behaviour of a cichlid fish Cyrtocara eucinostomus on 

a large lek in Lake Malawi. Env. Biol. Fish. 8(2): 81-96. 

McKaye K.R. & J.R. Stauffer 1988. Seasonality, depth and habitat distribution of breeding males 

of Oreochromis spp., 'chambo' in Lake Malawi National Park. J. Fish Biol. 33: 825-834. 

Melard C., E. Baras & D. Desprez 1997. Compensatory growth of Nile Tilapia, Oreochromis 

niloticus., pp. 178-185. In: Proceedings of the Fourth International Symposium on Tilapia in 

Aquaculture (Fitzsimmons ed.). 9-12 November 1997, Orlando, USA. 

Merona de B. 1983. Modèle d'estimation rapide de la croissance des poissons. Application aux 

poissons d'eau douce d'Afrique. Rev. Hydrobiol. trop. 16(1): 103-113. 

Merona de B., T. Hecht & J. Moreau 1988. Growth of African freshwater fishes, pp. 191-219. In: 

Biology and Ecology of African Freshwater Fishes (Lévêque C., M.N. Bruton & G.W. 

Ssentongo eds.). ORSTOM editions, 508 p. 

Meyer A. 1993. Phylogenetic relationships and evolutionary processes in East African cichlid 

fishes. Trends Ecol. Evol. 8: 279-284. 

Meyer A., T.D. Kocher, P. Basasibwaki & A.C. Wilson 1990. Monophyletic origin of Lake 

Victoria cichlids suggested by mitochondrial DNA sequences. Nature 347: 550-553. 

Mkanda F.X. 1999. Land use, soil erosion and sedimentation in the Lake Malawi watershed. In 

CEOS Data Report in Support of the Lake Malawi Biodiversity Conservation Project, Malawi, 

Africa: Volume II (Cooley, P. M., G. K. McCullough, & F.X. Mkanda eds.). Centre for Earth 

Observation Science Data Report 99-9-1. 99 p. 

269

Moran P. I. Kornfield & P.N. Reinthal 1994. Molecular systematics and radiation of the 

haplochromine cichlids (Teleostei: Perciformes) of Lake Malawi. Copeia 1994: 274-288. 

Moreau J. & B. Nyakageni 1992. Luciolates stapperssi in Lake Tanganyika. Demographical 

status and possible recent variations assessed by length frequency distributions. Hydrobiologia 

232: 57-64. 

Moreau J., Bambino C. & D. Pauly 1986. Indices of overall growth performance of 100 tilapia 

(Cichlidae) populations, pp. 201-206. In: The First Asian Fisheries Forum (Mclean J., L.B. 

Dizon & L.V. Hosillos eds.). Asian Fisheries Society, Manila, Philippines. 

Moreau J., J. Munyandorero & B. Nyakageni 1991. Evaluation des paramètres émographiques 

chez Stolothrissa tanganyikae et Limnothrissa myodon du lac Tanganyika. Verh. Internet. 

Verein. Limnol. 24: 2552-2558. 

Moreau J., M.L.D. Palomares, F.S.B. Torres Jr. & D. Pauly 1995. Atlas démographique des 

populations de poissons d'eau douce d'Afrique. ICLARM Rapp. Tech. 45, 140 p. 

Msukwa A.V. & A.J. Ribbink 1997. The potential role of sanctuary areas for biological control of 

schistosomiasis in lake Malawi national Park, pp. 305-317. In Proceeding of "Workshop on 

Medical Malacology in Africa", Harare, Zimbabwe, September 22-26. 

Ngatunga B.P. & E.H. Allison 1996. Food consumption/biomass ratios of the pelagic fish 

community of Lake Malawi/Nyassa. Fishbyte: 36-42. 

van Oppen M.J.H., G.F. Turner, C. Ricco, J.C. Deutsch, K.M. Ibrahim, R.L. Robinson & G.M. 

Hewitt 1997. Unuasually fine-scale structuring found in rapidly speciating Malawi cichlid 

fishes. Proc. R. Soc. Lond. B 264: 1803-1812. 

van Oppen M.J.H., G.F. Turner, C. Ricco, R.L. Robinson, J.C. Deutsch, M.J. Genner & G.M. 

Hewitt 1998. Assortative mating among rock-dwelling cichlid fishes supports high estimates 

of species richness from Lake Malawi. Molecular Ecology 7: 991-1001. 

Owen R.B., R. Crossley, T.C. Johnson, D. Tweddle, I. Kornfield, S. Davison, D.H. Eccles & D.E. 

Engstrom 1990. Major low levels of Lake Malawi and their implications for speciation rates in 

cichlid fishes. Proc. R. Soc. Lond. B. 240: 519-553. 

Patterson G. & J. Makin (eds.) 1998. The state of Biodiversity in Lake Tanganyika - A Literature 

Review. Chatham, UK: Natural Resources Institute. 134pp. 

PaulyD. 1980. On the interrelationship between natural mortality, growth parameters and mean 

environmental temperature in 175 fish stocks. J. Cons. Int. Explor. Mer. 39(2): 175-192. 

Pauly D. 1983. The length converted catch curves. A powerful tool for fisheries research in the 

tropics. Fishbyte 1: 9-13. 

PaulyD. 1987. A review of the ELEFAN system for analysis of length frequency data in fish and 

aquatic invertebrates, pp. 7-34. In: Length-based methods in fisheries research (Pauly D. & 

G.R. Morgan eds.). ICLARM Conf. Proc. 13, 468 p. 

Peters, H.M. 1963. Fecundity, egg weight and oocyte development in tilapias (Cichlidae, 

Teleostei). Translated and edited by D. Pauly, 1983. ICLARM Translation 2, 28 p. 

Peterson B.J. & B. Fry 1987. Stable isotopes in ecosystem studies. Ann. Rev. Ecol. Syst. 18: 293- 

320. 

Pitcher T.J. & P.J.B. Hart 1982. Fisheries Ecology. Chapman & Hall, London. 414p. 

Reinthal P.N. 1990. The feeding habits of a group of herbivorous rock-dwelling cichlid fishes 

(Cichlidae: Perciformes) from Lake Malawi, Africa. Env. Biol. Fish. 27: 215-233. 

Reznick D.A. & J.A. Endler 1982. The impact of predation on life history evolution in 

Trinidadian guppies (Poecilia reticulata). Evolution 36: 160-177. 

Ribbink A.J. 1987. African lakes and their fishes: conservation scenarios and suggestions. 

Environ. Biol. Fish. 19: 3-26. 

Ribbink A.J. 1988. Evolution and speciation of African cichlids, pp. 35-51. In Biology and 

ecology of African freshwater fishes (Lévêque C., M.N. Bruton & G.W. Ssentongo eds.). 508 

p. 

270

Ribbink A.J. 1990. Alternative life history styles of some African cichlid fishes. Environ. Biol. 

Fish. 28: 87-100. 

Ribbink A.J. 1991. Distribution and ecology of the cichlids of the African Great Lakes, pp. 37-59. 

In Cichlid fishes: behaviour, ecology and evolution (M.H.A. Keenleyside ed.). Chapman & 

Hall, 378 p. 

Ribbink A.J. 1994. Lake Malawi. Arch. Hydrobiol. Beih. Ergebn. Lmnol. 44: 27-33, Stuttgart. 

Ribbink A.J. & D.H. Eccles 1988. Fish communities in the East African Great Lakes, pp. 277- 

301. In Biology and ecology of African freshwater fishes (Lévêque C., M.N. Bruton & G.W. 

Ssentongo eds.). 508 p. 

Ribbink A.J., B.J. Marsh, A.C. Marsh, A.C. Ribbink & B.J. Sharp 1983. A preliminary survey of 

the cichlid fishes of the rocky habitats of Lake Malawi. S. Afr. J. Zool. 18: 149-310. 

Rico C., I. Rico & G. Hewitt 1996. 470 Million years of conservation of microsatellite loci among 

fish species. Proc. R. Acad. Lond. B. 263: 549-557. 

Rico C., D. Zadworny, U. Kuhnlein & J.G. Fitzgerald 1993. Characterisation of microsatellites 

loci in the three-spine stickleback Gasterosteus aculeatus. Molec. Ecol. 2, 271-272. 

Ricker W.E. 1975. Computation and interpretation of biological statistics of fish populations. 

Bull. Fish. Res. Bd. Can. 191: 1-382. 

Sargent R.C., P.D. Taylor & M.R. Gross 1987. Parental care and the evolution of egg size in 

fishes. Am. Nat. 129: 32-46. 

Scherrer, B. 1984. Biostatistique. Gaëtan Morin, Boucherville. 850 p. 

Seehausen O., J.J.M. van Alphen & F. Witte 1997. Cichlid fish diversity threatened by 

eutrophication that curbs sexual selection. Science 277: 1808-1811. 

Snoeks J. 1994. The Haplochromines (Teleostei, Cichlidae) of Lake Kivu (East Africa). Ann. 

Mus. Roy. Afr. Centr., Sc. Zool, 270: 221 p. 

Schnute J. 1981. A versatile growth model with statistically stable parameters. Can. J. Fisher. 

Aquat. Sci. 38: 1128-1140. 

Stiassny M. & A. Meyer 1999. La naissance des espèces. Pour la Science 258: 70-75. 

Stauffer J.R. Jr. & K.R. McKaye 1985. Cyrtocara macrocleithrum, a deep-water cichlid 

(Teleostei: cichlidae) from Lake Malawi, Africa. Copeia 3: 591-596. 

Stearns S.C. 1983. A natural experiment in life history evolution: field data on the introduction of 

mosquito fish (Gambusia affinis) to Hawaii. Evolution 37: 601-617. 

Stewart K.M. 1988. Change in condition and maturation of the Oreochromis niloticus population 

of Ferguson’s Gulf, Lake Turkana, Kenya. J. Fish Biol. 33: 181-188. 

Sturmbauer C., W. Mark & R. Dallinger 1992. Ecophysiology of aufwuchs eating cichlids in 

Lake Tanganyika: niche separation by trophic specialisation. Env. Biol. Fish. 35: 283-290. 

Sululu R.S.M. 2000. Impact of suspended sediments on the abundance and species richness of the 

sandy-shore dwelling haplochromine cichlids in southwestern Lake Malawi / Nyassa (East 

Africa). MSc's Thesis, Waterloo University, Ontario, Canada. 

Tacon P., P. Ndiaye, C. Cauty, F. Le Menn & B. Jalabert 1996. Relationship between the 

expression of maternal behavior and ovarian development in the mouthbrooding cichlid fish, 

Oreochromis niloticus. Aquaculture 146: 261-275. 

Thompson A. B., E. H. Allison & B. P. Ngatunga 1996. Distribution and breeding biology of 

offshore cichlids in Lake Malawi/Niassa. Env. Biol. Fish. 47: 235-254. 

Thompson A. B., E. H. Allison, B. P. Ngatunga & A. Bulirani 1995. Fish growth and breeding 

biology, pp. 279-306. In The fisheries potential and productivity of the pelagic zone of Lake 

Malawi/Niassa (A. Menz, ed.), SADC/ODA Final Report, 386 p. 

Tómasson T. & M.C. Banda, 1996. Depth distribution of fish species in southern Lake Malawi. 

Implications for fisheries management. Government of Malawi, Fisheries Department, 

Fisheries Bulletin 34. 28 p. 

Tomassone R. C. Dervin & J.P. Masson 1993. Biométrie: modélisation de phénomènes 

biologiques. Masson, Paris. 553 pp. 

271

Trewavas E. 1983. Tilapiine fishes of the genus Sarotherodon, Oreochromis and Danakilia. 

British Museum (Natural History), London. 583 p. 

Turner G. F. 1994a. Description of a commercially important pelagic species of the genus 

Diplotaxodon (Pisces: Cichlidae) from Lake Malawi, Africa. J. Fish Biol. 44: 799-807. 

Turner G.F. 1994b. Fishing and conservation of the endemic fishes of Lake Malawi. Arch. 

Hydrobiol. Beih. Ergebin. Limnol. 44: 481-494. 

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. Pitcher T.J. & P.J.B. Hart 

(eds.). Chapman & Hall publishers. pp. 365-395. 

Turner G.F. 1996. Offshore cichlids of Lake Malawi. Cichlid Press. 240 p. 

Turner G.F. 1998. Explosive speciation in African cichlid fishes. In Evolution of Biological 

Diversity (Magurran A.E. & R.M. May eds.). Oxford University Press, Oxford, UK. 

Turner G.F. D. Tweddle & R.D. Makwinja, 1995. Changes in demersal cichlid communities as a 

result of trawling in southern Lake Malawi, pp. 397-412.. In: The impact of species changes in 

African lakes. Pitcher T.J. & P.J.B. Hart (eds.). Chapman & Hall publishers. 

Turner G.F. A.S. Grimm, O.K. Mhonet, R.L. Robinson & T.J. Pitcher 1991a. Reproductive 

isolation and the nest sites of Lake Malawi chambo, Oreochromis (Nyasalapia) spp.. J. Fish 

Biol. 39: 775-782. 

Turner G.F. A.S. Grimm, O.K. Mhonet, R.L. Robinson & T.J. Pitcher 1991b. The diet of 

Oreochromis lidole (Trewavas) and other chambo species in Lakes Malawi and Malombe. J. 

Fish Biol. 39: 15-24. 

Turner J.L., 1977a. Some effects of demersal trawling in Lake Malawi (Lake Nyasa) from 1968 

to 1974. J. Fish Biol. 10: 261-272. 

Turner J.L., 1977b. Changes in size structure of cichlid populations of Lake Malawi resulting 

from bottom trawling. J. Fish. Res. Bd. Can. 34: 232-238. 

Tweddle D. 1991. Twenty years of fisheries research in Lake Malawi. Fisheries Department, 

Ministry of Forestry & Natural Resources, Malawi, Fish. Bulletin 7, 43 p. 

Tweddle D. 1992. Conservation and threats to the resources of Lake Malawi. Mitt. Internat. 

Verein. Limnol. 23: 17-24. 

Tweddle D. & J.L. Turner 1977. Age, growth and natural mortality rates of some cichlid fishes of 

Lake Malawi. J. Fish Biol. 10: 385-398. 

Tweddle D. & J.H. Magasa 1989. Assessment of multispecies cichlid fisheries of the Southeast 

Arm of Lake Malawi, Africa. J. cons. Int. Explor. Mer. 45: 209-222. 

Ware D.M. 1977. Spawning time and egg size of Atlantic Mackerel, Scomber scombrus, in 

relation to the plankton. J. Fish. Res. Board Can. 34: 2308-2315. 

Waters T.F. 1995. Sediment in streams: sources, biological effects and control. Monograph 7. 

Am. Fisher. Soc. Bethesda, Maryland. 

Weir B.S.& C.C. Cokerham 1984. Estimating F statistics for the analysis of population structure. 

Evolution 38: 1358-1370. 

Witte F. & W.L.T. van Densen (eds.) 1995.Fish Stocks and Fisheries of Lake Victoria: A 

handbook for field observations. Samara Publishing Ltd., Cardigan (UK). 404 p. 

Wootton R.J. 1979. Energy cost of egg production and environmental determinants of fecundity 

in teleost fishes. Symp. Zool. Soc. Lond. 44: 133-159. 

Wootton R.J. 1990. The ecology of teleost fishes. Chapman & Hall, London. 404 pp. 

Wootton R.J. 1994. Life histories as sampling devices: optimum egg size in pelagic fishes. J. Fish 

Biol. 45: 1067-1077. 

Wootton R.J. & G.W. Evans 1976. Cost of egg production in the three-spined stickleback 

(Gasterosteus aculeatus L.). J. Fish Biol 8: 385-395. 

Worthington E.B & R. Lowe-McConnell 1994. African Lakes Reviewed: Creation and 

destruction of Biodiversity. Environmental Conservation 21(3): 199-213. 

272

Wright S. 1969. Evolution and the genetics of populations. Vol2: the theory of gene frequencies. 

University of Chicago press, Chicago. 

Yamaoka K. 1991. Feeding relationships, pp. 151-172. In Cichlid Fishes: behaviour, ecology and 

evolution (Keenleyside M.H.A. ed.). Chapman & Hall, 378p. 

Zardoya R., D.M. Vollmer, C. Craddock, J. Streelman, S. Karl & A. Meyer 1996. Evolutionary 

conservation of microsatellite flanking regions and their use in resolving the phylogeny of 

cichlid fishes (Pisces: Perciformes). Proc. R. Acad. Lond. B. 263: 1589-1598. 

273

Appendixes

Appendix 1. Date, GPS positioning and depth of each trawl transect (TBM), CTD cast (CTD) and Grab sample (GRB) in the SWA between June 1998 and May 1999. 

Cruise Sample Gear Date Time Time Latitude (deg) Latitude (min) Latitude (deg) Latitude (min) Longitude (deg) Longitude (min) Longitude (deg) Longitude (min) Bottom depth 

Setting Hauling Setting Setting Hauling Hauling Setting Setting Hauling Hauling 

38 1 TBM 15/6/98 09:30 09:50 14 0,557 14 1,37 34 35,564 34 35,482 50,5 

38 2 GRB 15/6/98 10:23 14 0,83 34 35,607 51 

38 3 CTD 15/6/98 10:31 14 0,843 34 35,57 50,8 

38 4 TBM 15/6/98 15:45 16:05 13 52,755 13 51,744 34 51,067 34 50,806 128 

38 5 GRB 15/6/98 16:46 13 52,121 34 50,938 131 

38 6 CTD 15/6/98 17:02 13 52,06 34 50,999 132 

38 7 TBM 16/6/98 06:58 07:18 13 57,321 13 56,63 34 43,992 34 43,46 105 

38 8 CTD 16/6/98 11:00 13 57,031 34 43,84 105 

38 9 GRB 16/6/98 11:12 13 56,9 34 43,813 105 

38 10 TBM 16/6/98 13:08 13:32 14 2,061 14 1,02 34 33,966 34 33,84 25,6 

38 11 CTD 16/6/98 14:10 14 1,576 34 33,876 31 

38 12 GRB 16/6/98 14:17 14 1,639 34 33,807 30,4 

38 13 TBM 17/6/98 07:19 07:39 14 0,1 13 59,388 34 37,47 34 36,997 77,6 

38 14 CTD 17/6/98 11:30 13 59,905 34 37,393 77 

38 15 GRB 17/6/98 11:41 13 59,822 34 37,378 76,9 

38 16 TBM 17/6/98 12:37 12:57 14 1,952 14 2,729 34 32,815 34 32,86 10,9 

38 17 CTD 17/6/98 13:30 14 2,1 34 32,787 10,7 

38 18 GRB 17/6/98 13:40 14 2,095 34 32,757 10,1 

39 1 TBM 9/7/98 11:15 11:35 14 1,07 14 2,01 34 33,86 34 33,92 29,8 

39 2 GRB 9/7/98 12:00 14 1,53 34 34,05 32,5 

39 3 TBM 9/7/98 14:42 15:02 13 32,7 13 51,7 34 51,23 34 50,9 127 

39 4 GRB 9/7/98 15:32 13 52,21 34 50,93 130 

39 5 TBM 10/7/98 08:16 08:36 13 57,26 13 56,54 34 43,95 34 43,5 103 

39 6 GRB 10/7/98 09:05 13 56,97 34 43,73 103 

39 7 TBM 10/7/98 11:22 11:42 13 59,35 14 0,29 34 36,97 34 37,67 71,5 

39 8 GRB 10/7/98 12:45 13 59,7 34 37,25 74 

39 9 CTD 10/7/98 12:15 13 59,78 34 37,28 74 

39 10 TBM 10/7/98 14:48 15:08 14 0,34 14 1,56 34 35,61 34 35,46 50 

39 11 CTD 10/7/98 15:35 14 0,91 34 35,52 14 

39 12 GRB 10/7/98 15:45 14 0,86 34 35,53 49 

39 13 CTD 10/7/98 14 1,57 34 33,95 31 

39 14 CTD 10/7/98 17:30 13 56,92 34 43,74 10,3 

39 16 TBM 11/7/98 11:38 11:58 14 1,77 14 2,71 34 32,83 34 33,01 9 

39 17 GRB 11/7/98 12:16 14 2,35 34 32,94 10,3 

40 1 TBM 12/8/98 09:43 10:03 14 87 14 203 34 3379 34 3398 29,4 

40 2 CTD 12/8/98 10:33 14 150 34 3404 32,1 

40 3 GRB 12/8/98 10:40 14 145 34 3403 31,5 

40 4 TBM 12/8/98 12:59 13:19 13 5750 13 5650 34 4411 34 4350 103 

40 5 CTD 12/8/98 13:49 13 5697 34 4376 103 

40 6 GRB 12/8/98 13:56 13 5679 34 4376 103 

40 7 TBM 12/8/98 15:31 15:21 13 5290 13 5186 34 5122 34 5091 126



40 8 CTD 12/8/98 16:27 13 5217 34 5100 130 

40 9 GRB 12/8/98 16:36 13 5212 34 5106 129 

40 10 TBM 13/8/98 09:00 09:20 14 18 13 5925 34 3762 34 3685 70 

40 11 CTD 13/8/98 09:45 13 5987 34 3735 74 

40 12 GRB 13/8/98 09:56 13 5976 34 3734 74 

40 13 TBM 13/8/98 11:52 12:12 14 32 14 147 34 3560 34 3551 50 

40 14 CTD 13/8/98 12:35 14 88 35 3556 49 

40 15 GRB 13/8/98 12:42 14 86 34 3558 49 

40 16 TBM 13/8/98 15:11 15:31 14 194 14 304 34 3287 34 3303 10 

40 17 CTD 13/8/98 15:55 14 248 34 3288 10 

40 18 GRB 13/8/98 16:02 14 247 34 3289 10 

43 1 TBM 7/10/98 08:04 08:24 14 0,936 14 1,87 34 32,981 34 33,104 11,3 

43 2 CTD 7/10/98 08:48 14 1,822 34 33,092 12,9 

43 3 GRB 7/10/98 08:56 14 1,813 34 33,078 12,8 

43 4 TBM 7/10/98 10:11 10:31 14 1,366 14 2,238 34 33,874 34 33,736 30,8 

43 5 CTD 7/10/98 11:11 14 1,361 34 33,823 29,8 

43 6 GRB 7/10/98 11:17 14 1,337 34 33,74 29,7 

43 7 TBM 7/10/98 13:45 14:10 14 0,257 14 1,054 34 35,097 34 36,085 43 

43 8 CTD 7/10/98 14:50 14 0,741 34 35,773 52,6 

43 9 GRB 7/10/98 14:57 14 0,691 34 35,791 52,8 

43 10 TBM 7/10/98 15:20 15:40 14 0,603 14 0,349 34 34,325 34 34,088 37 

43 11 TBM 8/10/98 06:41 07:01 13 53,379 13 52,737 34 51,465 34 51,692 125 

43 12 CTD 8/10/98 07:44 13 52,805 34 51,851 124 

43 13 GRB 8/10/98 07:57 13 52,836 34 51,908 124 

43 14 TBM 8/10/98 10:23 10:43 13 56,913 13 57,902 34 43,256 34 43,211 103 

43 15 TBM 8/10/98 11:22 11:42 13 57,953 13 57,103 34 43,196 34 42,692 99 

43 16 CTD 8/10/98 12:49 13 57,529 34 42,968 100 

43 17 GRB 8/10/98 12:59 13 57,514 34 42,997 99 

43 18 TBM 8/10/98 14:22 14:43 14 0,08 13 59,241 34 37,935 34 37,365 78 

43 19 CTD 8/10/98 15:24 13 59,274 34 37,32 75 

43 20 GRB 8/10/98 15:26 13 59,258 34 37,342 75,1 

49 1 TBM 24/11/98 12:15 12:39 14 0,816 14 2,015 34 33,28 34 33,101 12 

49 2 CTD 24/11/98 13:16 14 1,883 34 33,113 13,3 

49 3 GRB 24/11/98 13:22 14 1,836 34 33,067 12,1 

49 4 GRB 24/11/98 13:35 14 1,764 34 33,055 11,5 

49 5 GRB 24/11/98 13:38 14 1,729 34 33,044 11,5 

49 6 GRB 24/11/98 13:40 14 1,714 34 33,048 11,5 

49 7 GRB 24/11/98 13:42 14 1,706 34 33,035 11,4 

49 8 TBM 24/11/98 15:43 16:03 14 1,497 14 2,424 34 34,012 34 34,131 31,9 

49 9 CTD 24/11/98 16:38 14 1,772 34 34,11 32,5 

49 10 GRB 24/11/98 16:48 14 1,691 34 34,087 32,9 

49 11 GRB 24/11/98 16:55 14 1,678 34 34,055 32,5 

49 12 GRB 24/11/98 17:00 14 1,652 34 34,028 32,4



49 13 GRB 24/11/98 17:04 14 1,633 34 34,022 32,3 

49 14 GRB 24/11/98 17:07 14 1,629 34 34,015 32,2 

49 15 TBM 25/11/98 06:50 07:11 13 54,576 13 53,607 34 50,421 34 50,081 120 

49 16 CTD 25/11/98 07:57 13 53,848 34 50,151 122 

49 17 GRB 25/11/98 08:10 13 53,966 34 50,091 121 

49 18 GRB 25/11/98 08:21 13 53,107 34 50,053 120 

49 19 GRB 25/11/98 08:32 13 54,157 34 50,043 120 

49 20 GRB 25/11/98 08:39 13 54,196 34 50,029 120 

49 21 GRB 25/11/98 08:46 13 54,257 34 50,014 120 

49 32 TBM 26/11/98 09:47 10:09 13 56,718 13 57,521 34 42,626 34 43,056 99 

49 33 CTD 26/11/98 10:55 13 57,154 34 43,123 100 

49 34 GRB 26/11/98 11:05 13 57,254 34 43,247 101 

49 35 GRB 26/11/98 11:13 13 57,311 34 43,33 101 

49 36 GRB 26/11/98 11:28 13 57,442 34 43,426 101 

49 37 GRB 26/11/98 11:39 13 57,549 34 43,538 101 

49 38 GRB 26/11/98 11:45 13 57,553 34 43,579 101 

49 39 TBM 26/11/98 13:45 14:07 13 59,547 14 0,487 34 37,449 34 37,308 75 

49 40 CTD 26/11/98 14:52 14 0,129 34 37,254 74 

49 41 GRB 26/11/98 15:00 14 0,149 34 37,164 72 

49 42 GRB 26/11/98 15:07 14 0,125 34 37,15 72 

49 43 GRB 26/11/98 15:11 14 0,077 34 37,176 72 

49 44 GRB 26/11/98 15:19 14 0,095 34 37,202 72 

49 45 GRB 26/11/98 15:24 14 0,082 34 37,217 72 

49 46 TBM 26/11/98 15:54 16:14 14 0,269 14 0,031 34 35,827 34 35,489 53 

49 47 CTD 26/11/98 16:47 14 0,706 34 35,548 49 

49 48 GRB 26/11/98 16:51 14 0,724 34 35,524 49 

51 1 TBM 15/12/98 08:00 08:20 14 1,776 14 2,645 34 33,236 34 33,383 14 

51 2 CTD 15/12/98 08:50 14 2,042 34 33,286 14 

51 3 GRB 15/12/98 08:56 14 1,979 34 33,263 14 

51 4 TBM 15/12/98 10:45 11:05 13 57,484 13 56,129 34 43,095 34 43,246 100 

51 5 CTD 15/12/98 11:46 13 57,025 34 43,228 103 

51 6 GRB 15/12/98 11:58 13 56,894 34 43,192 104 

51 7 TBM 15/12/98 14:15 14:35 13 53,256 13 52 34 50,123 34 50,183 125 

51 8 CTD 15/12/98 15:21 13 52,9 34 50,26 126 

51 9 GRB 15/12/98 15:36 13 52,983 34 50,346 126 

51 10 TBM 16/12/98 05:50 06:10 14 0,125 13 59,455 34 39,117 34 38,533 80 

51 11 CTD 16/12/98 06:57 13 59,911 34 38,873 80 

51 12 GRB 16/12/98 07:00 13 59,91 34 38,867 80 

51 13 TBM 16/12/98 09:42 10:02 14 0,734 14 1,32 34 35,882 34 35,188 53 

51 14 CTD 16/12/98 10:34 14 0,807 34 35,369 49 

51 15 GRB 16/12/98 10:40 14 0,866 34 35,348 49 

51 16 TBM 16/12/98 12:16 12:36 14 1,198 14 1,532 34 34,105 34 34,511 31 

51 17 CTD 16/12/98 14:06 14 1,199 34 34,049 31



51 18 GRB 16/12/98 14:12 14 1,233 34 33,911 31 

52 1 TBM 27/1/99 08:07 08:27 14 1,705 14 2,534 34 33,153 34 33,271 14 

52 2 ZOO 27/1/99 09:02 14 1,952 34 33,136 13 

52 3 GRB 27/1/99 09:07 14 2,014 34 33,132 13 

52 4 GRB 27/1/99 09:17 14 2,014 34 33,132 13 

52 5 GRB 27/1/99 09:20 14 2,014 34 33,132 13 

52 6 GRB 27/1/99 09:25 14 2,014 34 33,132 13 

52 7 GRB 27/1/99 09:30 14 2,014 34 33,132 13 

52 8 GRB 27/1/99 09:35 14 2,014 34 33,132 13 

52 9 CTD 27/1/99 09:40 14 2,014 34 33,132 13 

52 10 TBM 27/1/99 11:34 11:54 13 58,042 13 57,306 34 43,25 34 43,763 100 

52 11 CTD 27/1/99 12:50 13 57,644 34 43,538 101 

52 12 ZOO 27/1/99 12:59 13 57,567 34 43,494 100 

52 13 GRB 27/1/99 13:15 13 57,524 34 43,454 101 

52 14 GRB 27/1/99 13:24 13 57,504 34 43,392 101 

52 15 GRB 27/1/99 13:32 13 57,496 34 43,338 101 

52 16 GRB 27/1/99 13:39 13 57,469 34 43,293 101 

52 17 GRB 27/1/99 13:46 13 57,47 34 43,295 101 

52 18 GRB 27/1/99 13:51 13 57,486 34 43,191 101 

52 19 TBM 27/1/99 15:26 15:46 13 52,658 13 51,835 34 50,347 34 50,4 127 

52 20 CTD 27/1/99 16:31 13 52,235 34 50,324 128 

52 21 ZOO 27/1/99 16:48 13 52,202 34 50,284 129 

52 22 GRB 27/1/99 17:01 13 52,171 34 50,211 129 

52 23 GRB 27/1/99 17:14 13 52,216 34 50,158 128 

52 24 GRB 27/1/99 17:25 13 52,176 34 50,168 128 

52 25 GRB 27/1/99 17:32 13 52,156 34 50,155 128 

52 26 GRB 27/1/99 17:36 13 52,208 34 50,09 128 

52 27 GRB 27/1/99 17:42 13 52,205 34 50,053 128 

52 28 TBM 28/1/99 06:36 06:56 13 59,906 13 59,247 34 37,979 34 37,503 77 

52 29 CTD 28/1/99 07:36 13 59,615 34 37,781 78 

52 30 ZOO 28/1/99 07:44 13 59,721 34 37,87 77 

52 31 GRB 28/1/99 08:02 13 59,719 34 37,807 77 

52 32 GRB 28/1/99 08:07 13 59,754 34 37,801 77 

52 33 GRB 28/1/99 08:14 13 59,752 34 37,782 77 

52 34 GRB 28/1/99 08:20 13 59,752 33 37,73 77 

52 35 GRB 28/1/99 08:27 13 59,702 33 37,69 77 

52 36 TBM 28/1/99 09:22 09:42 14 1,896 14 0,895 34 34,055 34 33,851 31 

52 37 CTD 28/1/99 10:20 14 1,495 34 34,113 33 

52 38 ZOO 28/1/99 10:29 14 1,398 34 34,044 32 

52 39 GRB 28/1/99 10:37 14 1,317 34 34,054 32 

52 40 GRB 28/1/99 10:41 14 1,308 34 34,057 32 

52 41 GRB 28/1/99 10:45 14 1,284 34 34,281 31 

52 42 GRB 28/1/99 10:48 14 1,269 34 34,033 31



52 43 GRB 28/1/99 10:51 14 1,245 34 34,042 31 

52 44 GRB 28/1/99 10:54 14 1,224 34 34,009 31 

52 45 TBM 28/1/99 12:00 12:20 14 0,218 14 1,153 34 35,531 34 35,807 48 

52 46 CTD 28/1/99 12:59 14 0,931 34 35,607 49 

52 47 ZOO 28/1/99 13:07 14 0,891 34 35,527 49 

52 48 GRB 28/1/99 13:17 14 0,881 34 35,393 49 

52 49 GRB 28/1/99 13:23 14 0,878 34 35,342 49 

52 50 GRB 28/1/99 13:28 14 0,87 34 35,263 49 

52 51 GRB 28/1/99 13:31 14 0,858 34 35,244 47 

52 52 GRB 28/1/99 13:34 14 0,86 34 35,202 47 

52 53 GRB 28/1/99 13:38 14 0,865 34 35,2 47 

52 54 GRB 28/1/99 14:38 13 58,95 34 36,8 53 

52 55 GRB 28/1/99 14:48 13 55,957 34 36,649 52 

53 1 TBM 16/2/99 08:46 09:06 14 2,878 14 2,184 34 33,415 34 32,954 13 

53 2 CTD 16/2/99 09:44 14 2,604 34 32,221 13 

53 3 GRB 16/2/99 09:48 14 2,604 34 32,221 13 

53 4 GRB 16/2/99 09:51 14 2,604 34 32,221 13 

53 5 GRB 16/2/99 09:53 14 2,604 34 32,221 13 

53 6 GRB 16/2/99 09:57 14 2,604 34 32,221 13 

53 7 GRB 16/2/99 09:59 14 2,604 34 32,221 13 

53 8 TBM 16/2/99 11:46 12:06 14 2,162 14 1,321 34 34,101 34 33,982 29 

53 9 CTD 16/2/99 12:59 14 1,69 34 33,942 32 

53 10 GRB 16/2/99 13:07 14 1,602 34 33,953 32 

53 11 GRB 16/2/99 13:12 14 1,617 34 34,039 32 

53 12 GRB 16/2/99 13:14 14 1,575 34 34,045 32 

53 13 GRB 16/2/99 13:17 14 1,595 34 33,999 32 

53 14 GRB 16/2/99 13:21 14 1,554 34 33,987 32 

53 15 TBM 16/2/99 13:59 14:19 14 1,427 14 0,579 34 35,341 34 35,344 49 

53 16 CTD 16/2/99 14:53 14 1,096 34 35,38 48 

53 17 GRB 16/2/99 14:58 14 1,091 34 35,323 48 

53 18 TBM 17/2/99 06:27 06:47 13 54,036 13 53,341 34 51,227 34 51,394 122 

53 19 CTD 17/2/99 07:35 13 53,536 34 51,284 124 

53 20 GRB 17/2/99 07:43 13 53,579 34 51,177 124 

53 21 GRB 17/2/99 07:58 13 53,577 34 51,153 124 

53 22 GRB 17/2/99 08:07 13 53,581 34 51,09 124 

53 23 GRB 17/2/99 08:16 13 53,621 34 51,021 124 

53 25 TBM 17/2/99 09:59 10:19 13 57,167 13 56,289 34 43,671 34 43,253 101 

53 26 CTD 17/2/99 11:01 13 56,577 34 43,462 104 

53 27 GRB 17/2/99 11:10 13 56,562 34 43,459 104 

53 28 TBM 17/2/99 12:25 12:45 13 59,345 14 0,196 34 37,522 34 37,422 76 

53 29 CTD 17/2/99 13:28 13 59,812 34 37,468 75 

53 30 GRB 17/2/99 13:35 13 59,762 34 37,425 75 

53 31 GRB 17/2/99 13:42 13 59,714 34 37,413 75



53 32 GRB 17/2/99 13:46 13 59,681 34 37,383 75 

53 33 GRB 17/2/99 13:52 13 59,647 34 37,383 75 

53 34 GRB 17/2/99 13:58 13 59,668 34 37,383 75 

53 35 TBM 17/2/99 18:38 18:58 13 38,993 13 38,013 34 40,343 34 40,219 126 

55 1 TBM 19/3/99 10:45 11:05 14 0,416 14 1,148 34 33,158 34 33,159 16 

55 2 CTD 19/3/99 11:37 14 0,922 34 33,138 17 

55 3 GRB 19/3/99 11:41 14 0,861 34 33,068 15,7 

55 4 GRB 19/3/99 11:44 14 0,84 34 33,081 15,5 

55 5 TBM 19/3/99 13:36 13:56 13 57,936 13 57,112 34 43,265 34 43,549 101 

55 6 CTD 19/3/99 14:30 13 57,088 34 43,264 102 

55 7 GRB 19/3/99 14:43 13 57,119 34 43,159 102 

55 8 GRB 19/3/99 14:50 13 57,138 34 43,11 102 

55 9 TBM 19/3/99 16:22 16:42 13 53,576 13 52,972 34 49,441 34 49,692 122 

55 10 CTD 19/3/99 17:28 13 53,378 34 49,582 122 

55 11 GRB 19/3/99 17:40 13 53,489 34 49,52 122 

55 12 GRB 19/3/99 17:46 13 53,57 34 49,508 122 

55 13 TBM 20/3/99 05:55 06:15 14 0,494 13 59,743 34 38,168 34 37,736 79 

55 14 CTD 20/3/99 06:50 13 59,797 34 37,541 79 

55 15 GRB 20/3/99 07:03 13 59,8 34 37,433 78 

55 16 GRB 20/3/99 07:09 13 59,87 34 37,523 78,5 

55 17 TBM 20/3/99 08:31 08:51 14 1,119 14 0,246 34 35,708 34 35,555 51,1 

55 18 CTD 20/3/99 09:25 14 0,755 34 35,623 50,1 

55 19 GRB 20/3/99 09:32 14 0,686 34 35,576 50 

55 20 GRB 20/3/99 09:35 14 0,658 34 35,573 49,7 

55 21 TBM 20/3/99 11:37 11:57 14 2,193 14 1,455 34 34,066 34 33,821 27,8 

55 22 CTD 20/3/99 12:38 14 1,658 34 33,862 30,3 

55 23 GRB 20/3/99 12:42 14 1,648 34 33,804 29,8 

55 24 GRB 20/3/99 12:47 14 1,648 34 33,794 29,8 

57 1 TBM 14/4/99 09:00 09:20 14 0,589 14 1,372 34 33,157 34 33,062 19 

57 2 CTD 14/4/99 09:55 14 1,258 34 33,072 11 

57 3 GRB 14/4/99 09:59 14 1,258 34 33,072 11 

57 4 GRB 14/4/99 10:03 14 1,258 34 33,072 11 

57 5 GRB 14/4/99 10:06 14 1,258 34 33,072 11 

57 6 GRB 14/4/99 10:09 14 1,258 34 33,072 11 

57 7 GRB 14/4/99 10:12 14 1,258 34 33,072 11 

57 8 GRB 14/4/99 10:15 14 1,258 34 33,072 11 

57 9 TBM 14/4/99 11:32 11:52 14 0,29 13 59,505 34 37,584 34 37,536 77 

57 10 CTD 14/4/99 12:25 13 59,451 34 37,481 76 

57 11 GRB 14/4/99 12:37 13 59,376 34 37,367 76 

57 12 GRB 14/4/99 12:45 13 59,357 34 37,311 75 

57 13 GRB 14/4/99 12:51 13 59,321 34 37,269 75 

57 14 GRB 14/4/99 12:56 13 59,273 34 37,228 75 

57 15 GRB 14/4/99 13:01 13 59,227 34 37,177 75



57 16 GRB 14/4/99 13:09 13 59,145 34 37,115 74 

57 17 TBM 14/4/99 14:20 14:40 14 0,835 14 0,087 34 35,732 34 35,643 52 

57 18 CTD 14/4/99 15:10 14 0,404 34 35,59 51 

57 19 GRB 14/4/99 15:16 14 0,291 34 35,474 50 

57 20 GRB 14/4/99 15:20 14 0,276 34 35,455 49 

57 21 TBM 15/4/99 06:10 06:30 13 53,878 13 53,063 34 50,216 34 50,084 122 

57 22 TBM 15/4/99 09:30 09:50 13 57,584 13 56,776 34 43,46 34 43,555 102 

57 23 CTD 15/4/99 10:34 13 57,11 34 43,608 104 

57 24 GRB 15/4/99 10:40 13 57,104 34 43,617 104 

57 25 GRB 15/4/99 10:51 13 57,051 34 43,676 104 

57 26 TBM 15/4/99 12:45 13:05 14 1,934 14 1,116 34 33,983 34 33,593 32 

57 27 CTD 15/4/99 13:40 14 1,068 34 33,639 29 

57 28 GRB 15/4/99 13:43 14 1,068 34 33,639 29 

57 29 GRB 15/4/99 13:46 14 1,068 34 33,639 29 

57 30 GRB 15/4/99 13:52 14 1,068 34 33,639 29 

57 31 GRB 15/4/99 13:55 14 1,068 34 33,639 29 

57 32 GRB 15/4/99 14:00 14 1,068 34 33,639 29 

57 33 GRB 15/4/99 14:03 14 1,068 34 33,639 29 

58 1 TBM 20/5/99 07:38 07:59 14 0,135 14 0,973 34 33,096 34 33,056 14 

58 2 CTD 20/5/99 08:25 14 0,545 34 33,001 13 

58 3 GRB 20/5/99 08:29 14 0,545 34 33,002 13 

58 4 GRB 20/5/99 08:38 14 0,545 34 33,002 13 

58 5 GRB 20/5/99 08:40 14 0,545 34 33,002 13 

58 6 GRB 20/5/99 08:45 14 0,545 34 33,002 13 

58 7 GRB 20/5/99 08:47 14 0,545 34 33,002 13 

58 8 TBM 20/5/99 10:03 10:23 14 0,5 14 1,443 34 35,674 34 35,719 51 

58 9 CTD 20/5/99 10:49 14 1,007 34 35,468 51 

58 10 GRB 20/5/99 10:58 14 0,865 34 35,418 51 

58 11 TBM 20/5/99 12:52 13:12 13 57,24 13 56,471 34 43,255 34 43,66 102 

58 12 CTD 20/5/99 13:50 13 56,908 34 43,591 105 

58 13 GRB 20/5/99 14:00 13 56,895 34 43,675 106 

58 14 TBM 20/5/99 15:26 15:46 13 53,242 13 52,565 34 51,348 34 51,762 127 

58 15 CTD 20/5/99 16:28 13 52,742 34 51,764 125 

58 16 GRB 20/5/99 16:42 13 52,702 34 51,861 124 

58 17 GRB 20/5/99 16:53 13 52,698 34 51,938 124 

58 18 GRB 20/5/99 17:03 13 52,741 34 52,014 123 

58 19 GRB 20/5/99 17:26 13 52,675 34 52,163 122 

58 20 TBM 21/5/99 06:07 06:27 14 0,117 13 59,401 34 37,765 34 37,342 78 

58 21 CTD 21/5/99 07:00 13 59,391 34 37,454 76 

58 22 GRB 21/5/99 07:08 13 59,288 34 37,478 77 

58 23 GRB 21/5/99 07:15 13 59,232 34 37,466 77 

58 24 GRB 21/5/99 07:21 13 59,223 34 37,449 77 

58 25 GRB 21/5/99 07:27 13 59,176 34 37,404 76



58 26 GRB 21/5/99 07:33 13 59,176 34 37,381 76 

58 27 TBM 21/5/99 08:29 08:49 14 1,707 14 0,901 34 34,036 34 33,537 33 

58 28 CTD 21/5/99 09:19 14 0,983 34 33,663 30 

58 29 GRB 21/5/99 09:25 14 0,967 34 33,642 30 

58 30 GRB 21/5/99 09:28 14 0,923 34 33,731 30 

58 31 GRB 21/5/99 09:31 14 0,875 34 33,72 29 

58 32 GRB 21/5/99 09:34 14 0,86 34 33,783 29 

58 33 GRB 21/5/99 09:42 14 0,851 34 33,794 29

Appendix 2. List of fish species caught by trawling in the SWA. 

Alticorpus 'geoffreyi' 



Alticorpus pectinatum 


Aristochromis christyi 

Aulonocara 'blue orange' 

Aulonocara 'copper' 

Aulonocara guentheri 

Aulonocara 'long' 

Aulonocara 'cf. macrochir' 

Aulonocara 'minutus' 

Aulonocara rostratum 

Aulonocara 'rostratum deep' 


Bagrus meridionalis 

Barbus eurystomus 

Barbus johnstonii 

Barbus litamba 

Bathyclarias spp. 

Buccochromis lepturus 

Buccochromis nototaenia 

Buccochromis rhoadesi 

Buccochromis 'small' 

Caprichromis liemi 

Champsochromis caeruleus 

Chilotilapia rhoadesi 

Copadichromis inornatus 

Copadichromis quadrimaculatus 

Copadichromis trimaculatus 



Corematodus taeniatus 

Ctenopharynx nitidus 

Ctenopharynx pictus 

Dimidiochromis sp. 

Diplotaxodon apogon 

Diplotaxodon argenteus 


Diplotaxodon 'brevimaxillaris' 

Diplotaxodon greenwoodi 


Diplotaxodon 'similis' 

Diplotaxodon spp. 

Docimodus johnstonii 

Engraulicypris sardella 

Exocochromis anagenis 

Haplochromis 'sp.' 

Hemitaeniochromis 'insignis' 

Hemitaeniochromis urotaenia 

Hemitilapia oxyrhynchus 

Lethrinops albus 

Lethrinops altus 


Lethrinops 'blue orange' 

Lethrinops 'cf. auritus' 

Lethrinops christyi 

Lethrinops dark 

Lethrinops 'deep water albus' 

Lethrinops 'deep water altus' 

Lethrinops 'cf. furcifer' 


Lethrinops 'grey' 

Lethrinops lethrinus 

Lethrinops longimanus 

Lethrinops longipinnis 

Lethrinops macrochir 

Lethrinops 'macrostoma' 

Lethrinops 'matumbae' 

Lethrinops microdon 

Lethrinops 'minutus' 

Lethrinops mylodon 

Lethrinops 'oliveri' 

Lethrinops 'cf. parvidens' 

Lethrinops 'pink head' 

Lethrinops polli 

Lethrinops stridei 

Lethrinops 'yellow chin' 


Mormyrus longirostris 


Mylochromis formosus 

Mylochromis gracilis 

Mylochromis melanonotus 

Mylochromis sphaerodon 

Mylochromis spilostichus 

Mylochromis 'torpedo' 


Nevochromis chrysogaster 

Nimbochromis livingstonii 

Nimbochromis polystigma 

Nimbochromis venustus 

Nyassachromis argyrosoma 

Nyassachromis eucynostomus 


Opsaridium microcephallus 

Opsaridium microlepis 

Oreochromis spp. 

Otopharynx argyrosoma 

Otopharynx auromarginatus 

Otopharynx brooksi 

Otopharynx 'productus' 

Otopharynx decorus 

Otopharynx speciosus 

Otopharynx spp. 

Pallidochromis tokolosh 

Placidochromis 'acuticeps' 

Placidochromis "flatjaws" 

Placidochromis 'hennydaviesae III' 

Placidochromis 'hennydaviesae IV' 

Placidochromis johnstonii 

Placidochromis long 

Placidochromis 'macrognathus' 


Placidochromis 'cf. subocularis' 


Protomelas triaenodon 

Protomelas spilopterus 

Pseudotropheus elegans 

Pseudotropheus lanisticola 

Pseudotropheus livingstonii 

Pseudotropheus spp. 

Rhamphochromis spp. 

Sciaenochromis alhi 

Sciaenochromis benthicola 

Sciaenochromis psammophilus 

Sciaenochromis spp. 

Serranochromis robustus 

Stigmatochromis pholidophorus 

Stigmatochromis woodi 

Stigmatochromis 'guttatus' 

Synodontis njassae 

Taeniochromis holotaenia 

Taeniolethrinops furcicauda 

Taeniolethrinops laticeps 

Taeniolethrinops praeorbitalis 

Tramitichromis lituris 

Trematocranus brevirostris 

Trematocranus macrostoma 

Trematocranus placodon 

Unknown spp.

Appendix 3. Mean CPUE (kg / 20min pull) per depth for each species over the 

sampling period (July-1998 to May-1999). 

Species 10m 30m 50m 75m 100m 125m 

Alticorpus spp. - 0,2 0,1 1,0 0,9 1,2 

Alticorpus 'geoffreyi' - - 0,3 21,4 3,5 6,1 

Alticorpus macrocleithrum - - - 0,9 2,5 0,4 

Alticorpus mentale - 0,0 0,7 12,2 21,9 13,4 

Alticorpus pectinatum - - - 2,8 2,8 1,7 

Aristochromis christyi 0,1 0,0 - - - - 

Aulonocara 'cf. macrochir' 0,0 0,1 1,7 - - - 

Aulonocara spp. 0,2 - 0,2 0,5 0,0 1,1 

Aulonocara 'blue orange' 5,6 5,6 0,5 - - - 

Aulonocara 'copper' - - - 0,4 - - 

Aulonocara guentheri 0,9 0,1 - - - - 

Aulonocara 'long' 0,0 - 0,0 0,2 0,1 0,2 

Aulonocara 'minutus' - - - 1,4 0,9 1,7 

Aulonocara rostratum 0,0 - - - - - 

Aulonocara 'rostratum deep' - - 0,1 1,0 0,1 0,3 

Bagrus meridionalis 9,6 12,9 17,8 13,0 5,6 4,3 

Barbus eurystomus 0,0 0,1 - - - - 

Barbus johnstonii - 0,2 - - - - 

Barbus litamba - - 0,1 - - - 

Bathyclarias spp. 11,3 10,4 49,8 23,1 19,7 9,3 

Buccochromis lepturus 3,7 0,6 - - - - 

Buccochromis nototaenia 1,0 2,3 0,1 - - - 

Buccochromis rhoadesi 0,5 0,1 - - - - 

Buccochromis 'small' 0,0 - - - - - 

Caprichromis liemi - 0,0 0,0 - - - 

Champsochromis caeruleus 0,1 0,1 - - - - 

Chilotilapia rhoadesi 1,4 1,1 - - - - 

Copadichromis inornatus 0,1 - - - - - 

Copadichromis quadrimaculatus 0,9 2,6 1,1 0,1 - - 

Copadichromis spp. 0,1 0,0 - - - - 

Copadichromis trimaculatus - - - - - 0,0 

Copadichromis virginalis 0,7 16,7 47,7 - 0,1 - 

Corematodus taeniatus 0,0 0,0 0,0 0,0 - - 

Ctenopharynx nitidus 0,2 0,1 - - - - 

Ctenopharynx pictus - - 0,1 - - - 

Dimidiochromis sp. 0,0 - - - - - 

Diplotaxodon apogon - - - 5,0 5,4 4,0 

Diplotaxodon argenteus - - 1,0 4,4 2,9 2,2 

Diplotaxodon spp. - - - 0,3 0,5 0,5 

Diplotaxodon 'brevimaxillaris' - - 0,0 0,1 0,2 0,6 

Diplotaxodon greenwoodi - - - 0,1 0,3 0,2 

Diplotaxodon limnothrissa 0,1 - 9,7 13,4 9,3 2,7 

Diplotaxodon macrops - - - 7,3 18,9 16,0 

Diplotaxodon 'similis' - - - 0,0 - 0,1 

Docimodus johnstonii 0,1 0,0 0,3 - - - 

Engraulicypris sardella 0,0 0,0 - 0,0 0,0 - 

Exocochromis anagenis - - 0,0 - - - 

Haplochromis 'sp.' - - 0,0 - - - 

Hemitaeniochromis 'insignis' - - 0,0 0,1 0,0 0,1 

Hemitaeniochromis urotaenia 0,1 - 0,0 - - - 

Hemitilapia oxyrhynchus - - - - - - 

Lethrinops christyi 0,3 0,0 1,0 - - - 

Lethrinops 'matumbae' - 1,0 0,0 - - - 

Lethrinops 'deep water albus' - - 1,4 3,5 0,0 4,8


Lethrinops albus - - 0,2 0,0 - 0,6 

Lethrinops altus 0,2 2,6 2,1 1,8 1,4 4,6 

Lethrinops spp. 1,0 0,1 2,0 0,5 0,1 0,4 

Lethrinops 'blue orange' - 0,6 - - - - 

Lethrinops 'cf. auritus' 0,0 - - - - - 

Lethrinops dark 2,6 0,3 4,1 0,3 0,3 1,5 

Lethrinops 'deep water altus' - - - 1,0 4,7 2,2 

Lethrinops 'cf. furcifer' 1,1 0,1 - - - - 

Lethrinops gossei - - 0,2 32,0 40,0 34,4 

Lethrinops 'grey' - - - - 1,0 - 

Lethrinops lethrinus 0,2 - - - - - 

Lethrinops longimanus - 0,5 7,3 0,2 1,1 0,1 

Lethrinops argenteus 20,3 23,9 40,7 0,1 0,0 0,2 

Lethrinops macrochir 3,5 - - - - - 

Lethrinops 'macrostoma' - - - 0,0 - - 

Lethrinops microdon 1,0 - 0,1 0,0 0,0 - 

Lethrinops 'minutus' 0,0 - 3,5 - - - 

Lethrinops mylodon - 0,2 0,1 - 0,2 - 

Lethrinops 'oliveri' - - - 18,2 8,3 3,6 

Lethrinops 'cf. parvidens' 0,4 0,0 0,1 - - - 

Lethrinops 'pink head' 0,2 - - - - - 

Lethrinops polli - - - 7,0 1,2 0,2 

Lethrinops stridei - - - - 0,0 - 

Lethrinops 'yellow chin' - - 0,9 - - - 

Mormyrus longirostris - - - - - 0,1 

Mylochromis anaphyrmus 4,6 8,6 2,3 0,0 - - 

Mylochromis formosus 0,1 0,1 0,2 - - - 

Mylochromis gracilis - 0,2 0,2 0,5 - - 

Mylochromis spp. 0,2 0,1 - - - - 

Mylochromis melanonotus 0,3 0,2 - - - - 

Mylochromis sphaerodon 0,0 0,0 - - - - 

Mylochromis spilostichus 0,4 0,5 7,4 - - - 

Mylochromis 'torpedo' 0,0 - - - - - 

Nevochromis chrysogaster 0,0 - - - - - 

Nimbochromis livingstonii 0,1 0,0 0,2 0,1 - - 

Nimbochromis venustus 0,0 - - - - - 

Nyassachromis argyrosoma 28,6 31,7 0,9 - - - 

Nyassachromis spp. 0,5 0,3 - - - - 

Nyassachromis eucynostomus 0,5 0,1 - - - - 

Nimbochromis polystigma 0,0 - - - - - 

Opsaridium microcephallus - 0,0 - - - - 

Opsaridium microlepis - 0,0 0,7 0,2 - - 

Oreochromis spp. 27,7 1,1 7,8 0,1 - - 

Otopharynx argyrosoma 2,2 2,2 0,1 - - - 

Otopharynx auromarginatus - - - - - - 

Otopharynx brooksi - - - 0,5 0,0 0,0 

Otopharynx 'productus' 0,7 0,0 - - - - 

Otopharynx decorus 0,4 0,4 - - - - 

Otopharynx spp. 0,2 - - - - 0,0 

Otopharynx speciosus - 0,8 2,1 0,1 - - 

Pallidochromis tokolosh - - 0,1 1,8 0,6 2,9 

Placidochromis 'acuticeps' - - - - - 0,1 

Placidochromis "flatjaws" - - - 0,0 1,3 0,3 

Placidochromis spp. - - 0,0 0,3 0,0 - 

Placidochromis 'macrognathus' - 0,0 0,0 0,0 0,0 0,1 

Placidochromis 'hennydaviesae III' - - - - - 0,1 

Placidochromis 'hennydaviesae IV' - - - - - 0,1 

Placidochromis johnstonii - 0,0 - - - -


Placidochromis 'long' - 0,4 1,7 - - - 

Placidochromis 'platyrhynchos' - - - 0,0 1,1 4,1 

Placidochromis 'cf. subocularis' 0,1 0,0 - - - - 

Protomelas spilopterus 0,0 0,0 - - - - 

Protomelas triaenodon 0,0 - - - - - 

Pseudotropheus elegans 0,1 0,1 - - - - 

Pseudotropheus lanisticola - - - - - - 

Pseudotropheus livingstonii 1,1 0,1 - - - - 

Pseudotropheus spp. - 0,0 - - - - 

Rhamphochromis spp. 0,6 3,9 8,5 4,2 0,7 0,6 

Sciaenochromis spp. 0,1 0,0 0,0 - - - 

Sciaenochromis alhi 0,4 0,1 0,1 0,3 0,2 0,2 

Sciaenochromis benthicola 0,1 0,5 3,1 1,4 0,2 0,0 

Sciaenochromis psammophilus - - - 0,1 - - 

Serranochromis robustus 0,1 - - - - - 

Stigmatochromis pholidophorus 0,0 - - - - - 

Stigmatochromis woodi 0,0 0,0 0,1 0,0 - 0,0 

Stigmatochromis 'guttatus' 0,0 - 0,5 0,3 0,0 0,0 

Synodontis njassae 2,5 7,2 8,8 6,4 11,8 11,7 

Taeniochromis holotaenia 0,0 0,0 - - - - 

Taeniolethrinops furcicauda 1,3 0,0 - - - - 

Taeniolethrinops laticeps - 0,2 - - - - 

Taeniolethrinops praeorbitalis 1,1 0,3 - - - - 

Tramitichromis lituris 0,8 - - - - - 

Trematocranus brevirostris - 0,0 6,6 - - - 

Trematocranus macrostoma 0,0 - - - - - 

Trematocranus placodon 1,3 - - - - - 

Minimum number of species per depth 80 71 66 58 47 48

fish ecology report - The Cichlid Fishes of Lake Malawi, Africa

Create successful ePaper yourself

Delete template?

Save as template?