Cotutelle Ph.D. THESIS For the obtaining of JOINT DOCTORAT of the UNIVERSITY Delivered by: L'Université Toulouse III Paul Sabatier (France) & Ubon Ratchathani University (Thailand) Discipline/Speciality: Aquatic Ecology Presented & Defended by: Apinun SUVARNARAKSHA At Ubon Ratchathanee University (Thailand), 21 July 2011 Biology of two keystone fish species and fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand JURY MEMBERS Sovan LEK Tuantong JUTAGATE Prof. University of Toulouse, France Assoc. Prof. Ubon Ratchathani University, Thailand Sithan ANG-LEK Assoc. Prof. University of Toulouse, France Sena S. DE SILVA Prof. Deakin University (Australia) and Director Network of Aquaculture Centres in Asia-Pacific (Asia)   Emili GARCÍA-BERTHOU Prof. Girona University, Spain     Michio FUKUSHIMA National Institute for Environmental Studies, Japan Kriengkrai SATAPORNWANIT Kasetsart University, Thailand     Director of thesis Co-director of thesis Co-director of thesis Referee Referee Referee Referee   Ecole doctorale: Science Ecologiques, Vétérinaires, Agronomiques et Bioigenieries (SEVAB-Toulouse) Research Units: Evolution & Diversité Biologique (Toulouse) & Ubon Ratchathani University (Ubon Ratchathani) Directors of Thesis: Sovan LEK & Tuantong JUTAGATE Acknowledgments I wish to express my gratitude to Prof. Sovan Lek Ph.D. and Assoc. Prof. Tuantong Jutagate Ph.D. for opportunity to undertake my Ph. D. under their supervisions. Thanks for their continuous interest and encouragement and for many of our discussions, which led me improve the understanding of various aspects in the project, and for the large amount of the time they had spent helping with the editing of my thesis. Many thanks also to Assoc. Prof. Sithan Ang-Lek Ph.D. and the “LekFamily”, who took care my beautiful moment in France. I also grateful to my jury members: Prof. Sena S. De Silva, Prof. Emili GarcíaBerthou, Assoc. Prof. Sithan Ang-Lek Ph.D., Assist. Prof. Praneet Ngamsanae, Ph.D. and Kriengkrai Satapornwanit, Ph.D. for their time to correct my work and examine over my thesis defending. I am grateful to Assoc. Prof. Gaël Grenouillet, Ph.D. for sharing his knowledge in mathematics, Prof. Sébastien Brosse, Ph.D. for sharing his experiences in aquatic ecology. Thanks also go to my colleagues at the Laboratoire Evolution & Diversité Biologique (Toulouse); Loïc Tudesque, Ph.D. (my French navigator), Muriel Gevrey, Ph.D. (thank you dinner), Cândida Shinn, Ph.D. (pretty Portuguese girl), Clément Tisseuil, Ph.D. (mathematics super freak), Assist. Prof. Géraldine Loot, Ph.D. (the dog lover), Laetitia Buisson, Ph.D. (“red cheeks” lady), Simon Blanchet, Ph.D. (a clever guy), Roseline Etienne (friendly smile), Dominique Galy (for her hospitality warm welcome), and Dominique Pantalacci and, finally, my best & beautiful Chinese friends, Yongfeng HE, Ph.D. and Chen Lin, Ph.D. I also many thanks to my colleague at the Faculty of Fisheries Technology and Aquatic Resources, Maejo University and anonymous guy for their help. I am grateful to the Royal Golden Jubilee Program of the Thailand Research Fund (TRF-RGJ) for supporting my Ph.D. study (Grant PHD/0290/2549). Grants for the field survey were by the National Research Council of Thailand (Grant MJ. 1-43037, MJ. 1-44-008, MJ. 1-45-017.1, MJ. 1-46-008.1 and MJ. 1-47-004.1) and the Nagao Natural Environment Foundation (Grant Nagao-51-001, Nagao-52-001). Partial support was received from the Franco-Thai Academic Collaboration (Grant PHC 16598RJ) and the French Embassy to Thailand (Grant CNOUS: 2009/2349), which made it possible for me to work at the Laboratoire Evolution & Diversité Biologié under the convention for the joint supervision of theses between Ubon Ratchathani University and Université Paul Sabatier (Toulouse III). Last but not least, I wish to express my special thanks to my parents, my brother, my sister and my family for their love, patience, encouragement and support. Résumé La région Indo-Birmane est un formidable hotspot de diversité biologique, mais il existe un manque évident de connaissances fiables sur la diversité des poissons, la biologie et l'histoire de vie des communautés, ainsi que des approches de modélisation des données. Ce travail de thèse apporte des informations sur la diversité des poissons et de la distribution dans une zone de montagne de haute et de basse altitude dans la partie supérieure du bassin du fleuve Chao Phraya, en Thaïlande. Des données de terrain ont été collectées sur quatorze années entre Janvier 1996 et avril 2009, couvrant 272 enquêtes dans 10 sous-bassins hydrographiques fournissant la richesse spécifique et des indices de diversité. Cette thèse a été divisée en 3 niveaux principaux : le niveau taxonomique (niveau descriptif), la biologie des poissons (niveau descriptif et prédictif), et la diversité des assemblages de poissons en fonction des facteurs environnementaux (niveau prédictif). Tout d'abord, concernant l’étude de la diversité des poissons (publication 1, P1): la raréfaction a été utilisée pour extrapoler la richesse spécifique et le nombre optimal d'espèces dans le bassin supérieur du fleuve Chao Phraya. Deux cent une espèces réparties dans 104 genres et 34 familles ont été collectées, dont 16 espèces exotiques. Les poissons sont dominés par la famille des Cyprinidae, suivie par les Balitoridae et les Cobitidae, caractéristiques de la zone de haute altitude. Le taux d'endémisme global dans la zone a été estimé à environ 10%. La plupart des espèces de poissons est particulièrement caractéristique des habitats rhithroniques. Ensuite, nous avons étudié la dynamique de population des espèces de poissons clefs de la zone d'étude à savoir, (1) l'histoire de vie d’un cyprinidae Henicorhynchus siamensis (Sauvage, 1881) d’un petit réservoir (Publication 2; P2) et (2) la biologie de la reproduction et la conservation de l’espèce vulnérable des cours d’eau thaïlandais Oreoglanis siamensis aux contreforts de l’Himalaya (Publication 3; P3). Les deux espèces sont des représentants de l'état écologique des écosystèmes lentiques et lotiques. H. siamensis est une espèce riverine migratrice qui s’adapte bien à des conditions de réservoir (eau stagnante) et c’est un poisson économiquement important en apportant une source de protéines à des populations rurales de la région. La reproduction, le régime alimentaire et la croissance de H. siamensis ont été étudiés. Par exemple, on sait que la ponte est en saison humide, la taille de maturité est d’environ 200 mm, et l’espèce se nourrit de phytoplancton, etc. Par contre, O. siamensis est une espèce vulnérable et endémique, qui vit dans eaux rapides froides de haute montagne. La période de frai est la saison sèche. La taille de maturité est de 68,9mm pour mâles et de 82,4mm pour les femelles et le taux de fécondité est d’environ 31 œufs de grande taille (I # 3 mm). Enfin, nous avons étudié les relations entre paramètres biologiques et paramètres environnementaux, visant à expliquer les assemblages des poissons dans la zone d’étude (Publication 4; P4). Les patrons de diversité des assemblages de poissons dans la zone amont du bassin de la rivière Ping-Wang ont été étudiés. Des outils mathématiques (par exemple, SOM, ANN) ont été utilisés pour analyser les relations entre paramètres environnementaux (physico-chimiques et paramètres géomorphologiques dans le bassin de la rivière longitudinal et la diversité des poissons. Les arbres de classification et de régression (CART) ont montré que les paramètres géo-morphologiques ont été plus importants dans le modèle de prédiction à la fois pour la richesse spécifique et l'indice de diversité de Shannon. Les paramètres physico-chimiques sont moins importants, et exprimés surtout par l'altitude. Les poissons ont été classés dans 4 groupes d'assemblage à savoir, montagne, piémont, zone de transition et de plaine. Enfin, les effets de barrage sur les assemblages de poissons de rivière ont été montrés dans la Publication 5 (P5). Le SOM (selforganizing map) a été utilisé pour classer les communautés de poissons. Trois communautés de poissons ont été obtenues, à savoir de réservoir, de ruisseau et de la zone intermédiaire. Les communautés des réservoirs caractérisées par des espèces adaptées aux conditions lentiques sont par exemple Labiobarbus lineatus, (Sauvage, 1878) et Puntioplites proctozysron (Bleeker, 1865), alors que les espèces rhéophiles sont par exemple Rasbora paviana Tirant, 1885 et Channa gachua (Hamilton, 1822). La communauté de la zone intermédiaire contenait un mélange d'espèces des deux autres communautés. Le pourcentage global de bonne prédiction par le modèle a été de 66,0% : le modèle a correctement prédit 100% des communautés de réservoir, mais très peu de communautés rhéophiles (40%). Les communautés de poissons dans la zone d’étude sont menacées par la déforestation, la collecte des poissons d'aquarium, et la présence des espèces exotiques dans la partie supérieure. La présence des espèces évadées de l'aquaculture devrait être un facteur important en termes d'hybridation génétique. Toutefois, dans le bassin du fleuve Chao Phraya, les travaux sur l'écologie aquatique et la diversité des poissons sont peu nombreux et plus d’études scientifiques sont nécessaires pour atteindre le but ultime de l'utilisation rationnelle et durable des ressources aquatiques de cette région. Abstract Indo-Burma hotspot is an incredibly rich biological diversity area, but lack of reliable fish diversity, biology and life history, fish assemblage, and modeling approaches data. This present works on fish diversity and distribution in a unique high altitude mountain to lowland area in the upper part of the Chao Phraya river basin, Thailand. Fourteen years of field dataset in the basin were used, collected between January 1996 and April 2009, covering 272 surveys of 10 sub-river basins to produce species richness and diversity indices. This thesis was divided into 3 main levels viz., taxonomic level (descriptive level), biology and life history of fishes (descriptive level to predictive level), and assemblages of fish diversity as function of environmental factors (predictive level). Firstly, fish diversity study (Publication 1; P1): the rarefaction was employed to extrapolate species richness and optimum species numbers in the upper Chaophraya river basin. Two hundred and one species in 104 genera and 34 families were collected, including 16 exotic species. Cyprinidae fish family was dominated, followed by Balitoridae and Cobitidae, implying the characteristic of high altitude area. The overall endemism in the area was found to be about 10%. Most of the fish communities were especially characterized by rhithronic habitants. Second, there were studies investigating life history and population dynamics of the keystone fish species in the study area i.e., (1) life history of riverine cyprinid Henicorhynchus siamensis (Sauvage, 1881) in a small reservoir (Publication 2; P2) and (2) reproductive biology and conservation approaches of a vulnerable species Siamese Freshwater batfish (Oreoglanis siamensis) from foothill Himalayan, Thailand (Publication 3; P3), both species were the representative of lentic and lotic ecosystem conditions. H. siamensis has a well adaptation from riverine species to reservoir conditions (stagnant water) and it was an important economic fish providing protein source to rural people around the reservoir. The reproductive, feeding aspects and growth of H. siamensis were studied e.g. spawning season in wet season, the length of 50% maturity (about 200 mm), and feed on phytoplankton, etc. Meanwhile, O. siamensis is a vulnerable species and endemic species, which inhabits cold swift of high mountain streams. The spawning time occurred in dry season. Meanwhile, the length of 50% maturity were 68.9 (males) and 82.4 (females) mm and it was a few fecundity (31.41 ± 7.67 eggs) and large eggs (I # 3 mm). Thirdly, there were studies about the relationships between biological parameters and environmental parameters which were also beneficial to investigate fish diversity and assemblage patterns in the studied area (Publication 4; P4). Fish diversity and assemblage patterns in the rhitral environment of the Ping-Wang river basin were investigated. Mathematics tool models (e.g. SOM, ANN) were used for analysing of the relationship between environmental parameters (physicochemical and geo-morphological parameters and fish diversity in longitudinal in the river basin, and the prediction of its diversity. The classification and regression trees showed that the geo-morphological parameters were more significant in controlling and predicting both species richness and Shannon diversity index than the physicochemical parameters, in which altitude was the most significant. The fish assemblages were organized into 4 assemblage patterns viz., mountainous, piedmont, transitory and lowland species. And lastly, the investigation of the effects of dam to the riverine fish assemblages was showed in Publication 5 (P5). A self-organizing map (SOM) was used to cluster the fish community; three fish communities were obtained characterizing reservoir-, stream- and intermediate- communities. The reservoir communities were characterized by “lentic-adapted” fish i.e. Labiobarbus lineatus, (Sauvage, 1878) and Puntioplites proctozysron (Bleeker, 1865), whereas rheophilic species, i.e. Rasbora paviana Tirant, 1885 and Channa gachua (Hamilton, 1822), were dominant in the stream community. The intermediate community contained a mixture of species from both the other communities. The overall percentage of successful prediction by the model was 66.0 %: the model was 100% accurate for the prediction of the reservoir community but very low for the stream community (40%). Threats to fish communities were deforestation, collection for aquarium fish, and the distribution of the exotic species in the upper reaches. Meanwhile distribution of aquaculture escapes should be a concerned in terms of genetic hybridization. However, in the Chao Phraya river basin, research on the aquatic ecology and fish diversity are few and need more scientific information to reach the ultimate goal of wise and sustained uses of aquatic resources. Part 1: Synthesis 1. General Introduction………………………………………………….…………..2 1.1 Background of the study…………………………………………………..2 1.2 The Chao Phraya River Basin……………………………………………..5 1.3 Diversity of freshwater fish in the Chao Phraya River ……………...……6 1.4 Biology and life history traits of the tropical freshwater fishes…………20 1.5 Freshwater ecological study and fish assemblage………………………..23 1.6 Objectives of this Thesis…………………..……………………………..24 2. General materials and methods…………………………………………………26 2.1 Studied sites and data collection…………………………………………26 2.2 Fish sampling…………………………………………………………….29 2.3 Data analyses……………………………………………………………..33 2.3.1 Diversity and abundance (P1, P4 and P5)…….………………..33 2.3.2 Biology aspects of life history, and population dynamics (P2 and P3) ………………...…………………………………………...33 2.3.3 Statistical analyses and modeling methods (P4 and P5)……….36 3. Main Results……………………………………………………………………...39 3.1 Fish diversity in the upper Chao Phraya river basin (P1) …………...…..39 3.2 Some aspects of life history and population dynamics of lotic and lentic tropical fish species (P2 and P3)…….…………..……………………….41 3.3 Fish assemblages and impacts of environmental factors (P4 and P5)…....46 4. General Discussion ……………………………………………………………....53 4.1 Fish diversity and assemblage patterns in a rhitral environment…….…..53 4.2 Life history facts, biology, and population dynamics of riverine key stone species.……………………………………………………………...…....54 4.3 Fish communities in highland tropical streams connected to a reservoir… …………………….…………………………………………………...…58 5. Conclusion………………………………………………………………………...63 6. References………………………………………………………………………...67 Part 2: Publications P1) Suvarnaraksha, A., S. Lek-Ang, S. Lek, and T. Jutagate. (2011) Fish diversity in the upper Chao Phraya river basin, Southeast Asia. Ichthyological Exploration of Freshwaters. (Submitted)…………………………….................................88 P2) Suvarnaraksha, A., Lek, S., Lek-Ang, S. and Jutagate, T. (2011) Life history of the riverine cyprinid Henicorhynchus siamensis (Sauvage, 1881) in a small reservoir. Journal of Applied Ichthyology, 27(4): 955-1000.………………113 P3) Suvarnaraksha, A., Lek, S., Lek-Ang, S. and Jutagate, T. (2011) Reproductive biology and conservation approaches of a vulnerable species Siamese Freshwater batfish (Oreoglanis siamensis) from foothill Himalayan, Thailand (in preparation)…………………………………………………………...…121 P4) Suvarnaraksha, A., S. Lek, S. Lek-Ang and T. Jutagate (2011) Fish diversity and assemblage patterns in a rhitral environment of Indo-Burma region (the Ping-Wang River Basin, Thailand). Hydrobiologia. (Revised)…….........…141 P5) Suvarnaraksha, A., S. Lek, S. Lek-Ang and T. Jutagate (2011) Fish communities in highland tropical streams connected to a reservoir. (in preparation) …………………………………………………….………..…173    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand   Part 1: Synthesis 1    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand 1. General introduction 1.1 Background of the study Tropical Southeast Asia (SEA) is among the diversity hotspots of the world, especially fishes e.g. Mekong river basin 773 species, Chao Phraya river basin 297 species and Salween river basin 147 species (Froese & Pauly, 2011). The exceptional diversity of fish in this region also supports a huge inland fishery, which is the basis of the livelihood and extremely important for food security among the rural poor people (Volbo-Jørgensen & Poulsen 2000), in which the best example from Lower Mekong basin, where an estimation of 2.2 million tonnes of wild fish are harvested annually (Hortle, 2009). However, there are very few scientific reports on the fish diversity and their related issues, such as life history of individual keystone species, patterns of fish assemblages as well as their relationships to environmental attributes of both biotic and abiotic. Figure 1 Conceptual frame of the thesis In the developed countries, there is today some evidences of a reduction in the rate of anthropogenic impacts on natural ecosystems due to declining fertility and birth rates, the emergence of environmental institutions and governance, as well as changing values and behaviors (Costanza et al., 2007; Hibbard et al., 2007). However, this situation has not yet likely to be occurred in the developing countries, such as 2    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand many countries in SEA, where the massive acceleration in plans for infrastructure development has become increased, especially to the river system (Hibbard et al., 2007; Dugan et al., 2010), such as land reclamation and hydropower development (Dugan et al., 2010). Allan & Flecker (1993) recognized six major threats to biodiversity in the river systems, which are directly and/or indirectly affected by a range of human disturbances as: 1) habitat loss and degradation caused by water infrastructure projects, land transformations and agricultures, which are consequent in modifications of river hydrology and connectivity as well as riparian-aquatic and instream habitat integrities; 2) species invasions; 3) over-harvesting; 4) secondary extinctions due to cascading effects, 5) chemical and organic pollutions; and 6) global climate changes. Figure 2 Flow diagram showing title of each main chapter that reflected to the conceptual frame of the study. The status of aquatic ecosystems and their responses to human impacts are commonly described by using indicators such as existences and conditions of keystone species, status of assemblage patterns of aquatic faunas and floras and ecological integrity of the focused area (Norris & Thoms 1999; Allan 2004), which 3    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand the modeling techniques could be further developed for the understanding in the larger scale and make any predictions aiming to manage natural resources and ecosystems (Guisan & Thuiller, 2005). Therefore, the major goals of this thesis were to bring various approaches from systematic (aka taxonomic study) to ecological modeling in aquatic science (Fig. 1) to make an understanding on the status of fish resources in the upper Chao Phraya River basin, where all the 6 threats described by Allan & Flecker (1993) are now becoming and scientific information is desired for better management. The keystone fish species are chosen from the possibility for the representative of the lotic and lentic ecosystem conditions in Thai waters. Henicorhynchus siamensis is a riverine species in mainland Southeast Asia, widely distributions (see Table 1). This species can well adapt to new environmental conditions. Moreover, it can reproduce and become a dominance species in the manmade reservoir and an important protein sources for local people (per se A. Suvarnaraksha). Meanwhile, the Oreoglanis siamensis is a ripids and shooting high mountain stream species (Smith, 1945), vulnerable species (Kottelat, 1996) and low fecundity (see Table 1). It needs our knowledge to prevent and protect them from the extinction in the near future. Then, both species could be the representatives of biology of keystone fish species and fish assemblage patterns in the lotic and lentic adaptation ecosystem conditions tropical river basin Souteast Asia. The content of the thesis is divided into 5 main topics and accordingly to be 5 publications (Fig. 2). Publication 1 illustrates how much the upper Chao Phraya River Basin is fruitful with fish diversity, in which taxonomy approach is applied for making baseline information of existence species in each sub-basin. Publications 2 and 3 are the results of the studies on biology and life history traits of the two keystone species, i.e. the endanger species Oreoglanis siamensis (Publication 2) and the riverine species Henicorhynchus siameneis that occupied the lentic environment (Publication 3). Predictive models, which showing the assemblage patterns of fishes along the river course and their relationships to environmental parameters, is presented in Publication 4. Finally, Publication 5 demonstrates the fluctuation in fish assemblage patterns induced by human disturbance, i.e. river damming. 4    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 1 General comparison details of the keystone fish species of tropical Southeast Asia in this thesis. Henicorhynchus siamensis (Sauvage, 1881) 1 Oreoglanis siamensis (Smith, 1933) High population/wide distribution Endemic species (Kottelat, 1996; Vidthayanon, (Rainboth, 1996; Lim et. al., 1999; Doi, 2005; Vidthayanon et al., 2009) 1997; Roberts, 1997) 2 3 Dominance species (Rainboth, 1996; Lim et. Threatened species (Kottelat, 1996; al., 1999; Doi, 1997) Vidthayanon, 2005) Well adaptation from lotic to lentic Difficult to survived in lowland and lentic condition (per se A. Suvarnaraksha) condition (Rainboth, 1996; per se A. Suvarnaraksha) 4 Inhabits in large river (Rainboth, 1996; Lim Inhabits in brook stream (Smith, 1945) et. al., 1999; Doi, 1997; Roberts, 1997; Viravong, 2006) 5 High fecundity (high impact of recruitment) Low fecundity (per se A. Suvarnaraksha) (Sokheng, et al., 1999; Viravong, 2006) 6 7 8 9 Important to natural food web (Sokheng, et Consumer in the small stream (per se A. al., 1999) Suvarnaraksha) Commercial species (Roberts & Warren, Conservation proposed (Kottelat, 1996; 1994; Sokheng, et al., 1999) Vidthayanon, 2005) Phytoplankton and periphyton (Rainboth, Aquatic insect and specific food (Vidthayanon, 1996) 2005) Migratory species (Singhanouvong et al., Non-migration species (?) 1996a; Singhanouvong et al., 1996b; Sokheng, et al., 1999) 1.2 The Chao Phraya River Basin The Chao Phraya Basin is the most important basin in Thailand. The Basin covers 30% of Thailand's land area, is home to 40% of the country's population, employs 78% of its work force, and generates 66% of its Gross Domestic Product (GDP) (Office of Natural Water Resources Committee of Thailand, 2003). The basin lies in the central of Thailand, covers an area approximately of 160,000 km2, in which covers almost one-third of the country’s geographical area and is divided into upper, middle and lower basin. The Chao Phraya River per se is about 365 km long and the headwater of the Basin originates from the mountainous terrain in the northern part of the country at 2,565 m in elevation. There are four large tributaries in the upper parts 5    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand (i.e. the Ping, Wang, Yom and Nan rivers (Fig. 3B)) that flow southward joining together at Nakornsawan Province to form the main Chao Phraya River (Fig. 3A).The Chao Phraya mainstem, then, flows southward through a large alluvial plain and splits into four channels (i.e. the Tha Chin, Noi, Lopburi and Chao Phraya per se) and enter to the Gulf of Thailand (Davikar et al., 2011). It supplies water and supports navigation, fisheries, and recreation. Figure 3 (A) Map of the Chao Phraya river basin with the main tributaries (B) The Ping-Wang river basin with its tributaries, reservoirs, and connected provinces. 1.3 Diversity of freshwater fishes in the Chao Phraya River The freshwater fishes are ecologically classified into three groups (Berra, 2001): 1) the principal freshwater species, which they can complete their life cycle in freshwaters; 2) the marine vagrants, which are the fish that are found in freshwaters but also spend time in brackish and/or marine waters and 3) the diadromous fishes, 6    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand which are undertake extensive migrations between freshwater and brackish or marine waters. In the basin, monsoon weather dominates, with the rainy season lasting from May to October and supplementary rain from occasional westward storm depressions originating in the Pacific. The average annual precipitation in the basin ranges from a minimum of 1,000 mm in the western part to about 1,400 mm in the headwaters and up to 2,000 mm in the eastern Chao Phraya delta (Office of Natural Water Resources Committee of Thailand, 2003). Temperature ranges from 15°C in December to 40°C in April, except in high altitude locations. The basin can be classified as a tropical rainforest with high biodiversity. Southeast Asia contains the highest mean proportion of country-endemic bird (9%) and mammal species (11%) and the second-highest proportion of country-endemic vascular plant species (25%) compared to the other tropical regions of Meso-America, South America, and SubSaharan Africa (UNEP–WCMC, 2004). Especially freshwater fishes, in which 222 species from 36 families were recorded (Nguyen & De Silva 2006). Meanwhile the most update data in the reference global fish database, FishBase (www.fishbase.org; Froese & Pauly 2010),) ranked the Chaophrya River the seventh in term of freshwater species in the world with 318 species. At the global scale, the freshwater fishes belong to 207 families, 2,513 genera and estimated up to 32,500 species, in which the 11,952 species are strictly to freshwater environment, and using freshwater 12,457 species (Nelson, 2006). The fish diversity in the tropical Asia is considerably higher than that of African and Latin American (Lundberg et al., 2000), where the East, South and Southeast Asia have a cumulative total freshwater species is at 7,447 species (Kottelat & Whitten, 1996; Gleick, 2000) and dominated by fishes in Families Cyprinidae (about 1,000 species), Balitoridae and Cobitidae , (together about 400 species), Gobiidae (about 300 species) and then followed by Siluridae, and Bagridae (Kottelat & Whitten, 1996; Vidthayanon, 2005; Lévêque et al. 2008). In Thailand, fish species are reported at 836 species (Froese & Pauly 2010), in which 318 species are in Chao Phraya River Basin and 15.3% of the indigenous finfish were endemic to the basin (De Silva et al., 2007). They can be found in various habitats such as the highland streams, caves, lakes, river mainstem and estuary. It is 7    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand generally accepted that the Ping-Wang River Basin contains more than three-fourths of the freshwater fishes known from Chao Phraya River Basin (Vidthayanon et al. 1997).The summary of fish species in these areas is shown in Table 2. Table 3 presents the list the species found in the Chao Phraya River Basin compared to the two adjacent river basins, i.e. the Mekong and the Salween river basins. Yap (2002) found that the Mekong fish were most similar to that of the Chao Phraya, and also found the fauna of the Mekong, mid-Mekong, lower Mekong, and Chao Phraya are equally similar to each other, reflecting recent or continuing connections and can reach the conclusion that the upper Mekong formed part of the Chao Phraya basin in the past. Valbo-JØrgensen et al., 2010). Kottelat (1989) also found that the Mekong and Chao Phraya had more than 50% of their fish fauna in common. The overall conservation status of endemic finfish in Asia was satisfactory in that only 92 species were in some state of vulnerability, of which 37 species (6.6%) are endangered or critically endangered (De Silva et al., 2007). Four threatened species were reported from Chao Phraya River basin viz., Oreoglanis siamensis, Scleropages formosus, Yasuhikotaki asidthimunki and Datnioides microlepis, which are already officially protected (Kottelat, 1996; Vidthayanon, 2005) including Pangasianodon gigas, which is listed in the Appendices I of CITES and Convention on the Conservation of Migratory Species of Wild Animals (http://www.cms.int/ documents/appendix/cms_app1_2.htm #appendix _II). 8 (CMS)    a Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Phraya, Mekong river, Salween river and Ping-Wang river. Abbreviations: F=Families, G=Genera, S=Species, *=base on Fish base (2010) Class Holocephali Cephalaspidomorphi Elasmobranchii Sarcopterygii Actinopterygii Order 9 Chimaeriformes Petromyzontiformes Carcharhiniformes Orectolobiformes Pristiformes Pristiophoriformes Rajiformes Myliobatiformes Ceratodontiformes Acipenseriformes Albuliformes Amiiformes Anguilliformes Atheriniformes Batrachoidiformes Beloniformes Characiformes Clupeiformes Cypriniformes Cyprinodontiformes Elopiformes Esociformes Gadiformes Gasterosteiformes Gobiesociformes Gonorynchiformes Gymnotiformes Hiodontiformes Lepisosteiformes Mugiliformes Ophidiiformes Osmeriformes Osteoglossiformes Perciformes Percopsiformes Pleuronectiformes Polypteriformes Salmoniformes Scorpaeniformes Siluriformes Synbranchiformes Syngnathiformes Tetraodontiformes Total a     World F 1 1 1 Fish base 2010 S 1 33 Nelson 2006 S 29 1 2 24 2 2 8 8 23 6 14 1 2 7 1 3 17 5 7 9 1 8 181 5 71 1794 72 3451 964 1 6 210 6 98 1674 79 3268 996 2 15 2 1 2 5 13 9 31 133 10 1 21 1 4 Asia* F 1 1 1 1 1 3 4 3 7 34 3 4 1 1 4 34 3 1 1 171 31 219 2402 9 23 16 161 75 2835 90 20 29 12740 Chao Phraya* G S 9 7 F Mekong* G S 1 1 1 1 1 1 1 2 6 F Salween* G S F Ping-Wang* G S 1 1 7 1 2 13 1 1 1 20 30 1 6 6 2 1 F S 12 7 4 1 1 2 2 1 1 3 3 1 1 5 5 1 2 7 2 4 2 2 63 2 2 170 2 1 3 4 1 1 12 96 1 1 19 412 1 2 5 10 1 1 1 1 2 2 2 2 3 3 1 49 1 109 1 1 4 1 1 53 3 1 115 3 1 3 1 4 1 5 1 12 2 15 2 22 2 3 6 2 1 22 31 134 2 6 1 5 1 1 1 1 7 1 82 218 2040 9 10 16 45 60 2740 96 2 4 17 295 1 11 2 20 3 31 2 20 3 70 4 118 1 10 2 2 3 2 3 12 9 3 2 1 40 21 5 2 1 124 64 9 4 3 297 11 3 33 5 136 13 6 3 13 3 24 4 7 2 20 3 42 5 1 60 4 247 15 764 18 72 147 1 34 1 104 1 201 14 11952 2 36 21 2 10 3 8 507 9 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 2. Comparison of families, genera and species number of freshwater fishes in the world by Fishbase (2010), Nelson (2006), Chao    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species found in Mekong, Chao Phraya and Salween rivers. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 Order/Family/Species Order Pristiformes Family Pristidae Pristis microdon Latham, 1794 Order Mylibatiformes Family Dasyatidae Dasyatis laosensis Roberts & Kanasuta, 1987 Himantura bleekeri (Blyth, 1860) Himantura Chao Phraya Monkolprasit & Roberts, 1990 Himantura krempfi (Chabanaud, 1923) Himantura signifer Compagno & Roberts, 1982 Pastinachus sephen (Forsskål, 1775) Order Osteoglossiformes Family Osteoglossidae Scleropages formosus (Schlegel & Müller, 1844) Family Notopteridae Chitala lopis (Bleeker, 1851) Chitala ornata (Gray, 1831) Chitala branci (Aubenton, 1965) Notopterus notopterus (Pallas, 1831) Order Elopiformes Family Megalopidae Megalops cyprinoides (Broussonet, 1782) Order Anguilliformes Family Anguillidae Anguilla bicolor M'Clelland, 1844 Anguilla marmorata Quoy & Gaimand, 1824 Anguilla bengalensis (Gray, 1831) Family Ophichthyidae Ophichthus rutidoderma (Bleeker, 1852) Pisodonophis boro (Hamilton, 1822) Pisodonophis cancrivorous (Richardson, 1844) Order Clupeiformes Family Clupeidae Clupeoides borneensis Bleeker, 1851 Corica laciniata Fowler, 1935 Clupeichthys aesarnensis Wongratana, 1983 Clupeichthys goniognathus Bleeker, 1855 Gudusia variegata (Day, 1870) Tenualosa ilisha (Hamilton, 1822) Tenualosa thibaudeaui (Durand, 1940) Tenualosa toli (Valenciennes, 1847) Anodontostoma chacunda (Hamilton, 1822) Anodontostoma thailandae Wongratana, 1983 Nematalosa nasus (Bloch, 1795) Family Pristigasteridae Ilisha megaloptera (Swainson, 1839) Ophisthopterus tardoore (Cuvier, 1829) Family Engraulidae Coilia dussumieri Valenciennes, 1848 Coilia lindmani Bleeker, 1858 Coilia macrognathus Bleeker, 1852 Coilia ramcarati (Hamilton, 1822) Lycothrissa crocodilus (Bleeker, 1851) Setipinna melanochir (Bleeker, 1849) Setipinna wheeleri Wongratana, 1983 Family Sundasalangidae Sundasalanx praecox Roberts, 1981 Sundasalanx mekongensis Britz & Kottelat, 1999 Order Gonorhynchiformes Family Chanidae Chanos chanos (Forsskål, 1775) Order Cypriniformes Family Cyprinidae Subfamily Alburninae Longiculter siahi Fowler, 1937 Paralaubuca barroni (Fowler, 1934) Paralaubuca harmandi Sauvage, 1883 Paralaubuca riveroi (Fowler, 1935) Paralaubuca stigmabrachium (Fowler, 1934) Paralaubuca typus Bleeker, 1864 Subfamily Danioninae Amblypharyngodon chulabhornae Vidthayanon & Kottelat, 1990 Amblypharyngodon mola (Hamilton, 1822) 10 Mekong Chao Phraya Salween Ref. X X - A X X X X - X X X X X X - B, D, E A B, E B, E B, E A X - X A, B, F X X X X X X X X B, E A, B, D, E A, B, E A, B, C, D, E, F X X - A, B X X - - X X B, F B, E F, H X X X - - B B B X X X X X X X X X X X X X X X X X X X X - B, E B, D B, E B F F B, E A, B A, B B A, B X X - - B B X X X X - - X X X F B A, B F A, B, E A, B, E F X X X - - B, E E X X - B X X X X X X X X X X X X - A, B A, B, D, E A, B, D, E A, B, D, E A, B, D A, B, D, E X - X - X B, E C, F    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species of Mekong, Chao Phraya and Salween rivers (Continued). 51 52 53 54 55 56 57 58 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 Order/Family/Species Amblypharyngodon atkinsoni (Blyth, 1860) Aspidoparia morar (Hamilton, 1822) Barilius barila (Hamilton, 1822) Barilius koratensis (Smith, 1931) Barilius ornatus Sauvage, 1883 Barilius pulchellus (Smith, 1931) Boraras micros Kottelat & Vidthayanon, 1993 Boraras urophthalmoides (Kottelat, 1991) Chela caeruleostigmata Smith, 1931 Chela laubuca (Hamilton, 1822) Danio albolineatus (Blyth, 1860) Danio erythromicron (Annandale, 1918) Danio dangila (Hamilton, 1822) Danio kyathit Fang, 1998 Danio roseus Fang & Kottelat, 2000 Devario acrostomus Fang & Kottelat, 1999 Devario aequipinnatus M'Clelland, 1839 Devario annandalei Chaudhuri, 1908 Devario browni (Regan, 1907) Devario laoensis (Pellegrin & Fang, 1940) Devario regina (Fowler, 1934) Diptychus kaznakovi Nikolskii, 1933 Esomus longimanus (Lunel, 1881) Esomus metallicus Ahl, 1824 Inlecypris auropurpurea (Annandale, 1918) Leptobarbus hoevenii (Bleeker, 1851) Luciosoma bleekeri Staindachner, 1879 Luciosoma setigerum (Valenciennes, 1842) Macrochirichthys macrochirus Val., 1844 Microrasbora rubescens Annandale, 1918 Oxygaster anomarula van Hasselt, 1823 Oxygaster pointoni (Fowler, 1934) Parachela maculicauda (Smith, 1934) Parachela oxygastroides (Bleeker, 1852) Parachela siamensis (Günther, 1868) Parachela williaminae Fowler, 1934 Raiamas guttatus (Day, 1870) Rasbora argyrotaenia (Bleeker, 1850) Rasbora aurotaenia Tirant, 1842 Rasbora borapetensis Smith, 1934 Rasbora caudimaculata Volz, 1903 Rasbora daniconius (Hamilton, 1822) Rasbora dorsinotata Kottelat & Shu, 1987 Rasbora dusonensis (Bleeker, 1851) Rasbora hobelmani Kottelat, 1984 Rasbora myersi Brittan, 1954 Rasbora pauciperforata Weber & de Beaufort, 1916 Rasbora paucisquamis Ahl, 1935 Rasbora paviana (Tirant, 1885) Rasbora rasbora (Hamilton, 1822) Rasbora rubrodorsalis Donoso-Büchner & Schmidt, 1997 Rasbora spilocera Rainboth & Kottelat, 1987 Rasbora sumatrana (Bleeker, 1852) Rasbora tornieri Ahl, 1922 Rasbora trilineata Steindachner, 1870 Salmostoma sardinella (Valenciennes, 1844) Sawbwa resplendens Annandale, 1918 Trigonostigma espei (Meinken, 1967) Trigonostigma heteromorpha (Duncker, 1904) Thryssocypris tonlesapensis Roberts & Kottelat, 1984 Subfamily Leuciscinae Aaptosyax grypus Rainboth, 1991 Subfamily Hypophthalmichthyinae Hypophthalmichthys molitrix (Valenciennes, 1844) Hypophthalmichthys nobilis (Richardson, 1845) Subfamily Gobioninae Abbottina rivularis (Basilewsky, 1855) Subfamily Acheilognathinae Acheilognathus deignani Smith, 1945 Subfamily Cyprininae Albulichthys albuloides (Bleeker, 1855) Amblyrhynchichthys truncatus (Bleeker, 1850) Balantiocheilos melanopterus (Bleeker, 1850) 11 Mekong X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Chao Phraya X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Salween X X X X X X X X X X X X X X X X X X X X X X - X X X - X X - Ref. C, F A, F C, F A, B, D, E P A, B, D, E E B A, E A, D, E, F F A, F F F E E F F F B, E A, B, D, E, H F A, B, E A, B, D, E F A, B, E, F, H A, B, D, E A, B, D, E A, B, D, E F B A, B, D, E A, B, E A, B, D, E B, E A, B, D, E B, E, F A B, E A, B, D, E B B, D, E, F E B, E B, E A, B, D B B A, B, D, E A, F E B, E A B A, B, D, E F F B A B X - - B, E X X X X - E E X - - E X - - A, E X X X X X X - A, B, E A, B, D E A, B, D, E    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species of Mekong, Chao Phraya and Salween rivers (Continued). 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 Order/Family/Species Bangana behri (Fowler, 1937) Bangana devdevi (Hora, 1936) Bangana elegans Kottelat, 1998) Bangana lippus (Fowler, 1936) Barbichthys laevis (Sauvage, 1878) Barbichthys nitidus (Sauvage, 1878) Barbonymus altus (Günther, 1868) Barbonymus balleroides (Valenciennes, 1842) Barbonymus gonionotus (Bleeker, 1850) Barbonymus schwanenfeldii (Bleeker, 1853) Catla catla (Hamilton, 1822) Catlocarpio siamensis Boulenger, 1898 Chagunius baileyi Rainboth, 1986 Cirrhinus caudimaculatus (Fowler, 1934) Cirrhinus cirrhosus (Bloch, 1795) Cirrhinus jullieni Sauvage, 1878 Cirrhinus microlepis Sauvage, 1878 Cirrhinus molitorella (Valenciennes, 1844) Cirrhinus prosemion Fowler, 1934 Cirrhinus rubirostris Roberts, 1997 Cosmochilus harmandi Sauvage, 1878 Crossocheilus atrilimes Kottelat, 2000 Crossocheilus burmanicus Hora, 1936 Crossocheilus cobitis (Bleeker, 1853) Crossocheilus oblongus (Valenciennes, 1842) Crossocheilus reticulatus (Fowler, 1934) Crossocheilus siamensis (Smith, 1931) Ctenopharyngodon idella (Valenciennes, 1844) Cyclocheilichthys apogon (Valenciennes, 1842) Cyclocheilichthys armatus (Valenciennes, 1842) Cyclocheilichthys enoplos (Bleeker, 1850) Cyclocheilichthys furcatus Sontirat, 1985 Cyclocheilichthys heteronema (Bleeker, 1853) Cyclocheilichthys lagleri Sontirat, 1985 Cyclocheilichthys microlepis (Bleeker, 1851) Cyclocheilichthys repasson (Bleeker, 1853) Cyprinus intha Annandale, 1918 Cyprinus carpio Linnaeus, 1758 Discherodontus schroederi (Smith, 1945) Discherodontus ashmeadi Fowler, 1937 Discherodontus halei (Duncker, 1904) Epalzeorhynchos bicolor (Smith, 1931) Epalzeorhynchos frenatum (Fowler, 1934) Epalzeorhynchos kalopterus (Bleeker, 1851) Epalzeorhynchos munense (Smith, 1934) Garra cambodgiensis (Tirant, 1884) Garra fasciacauda Fowler, 1937 Garra fisheri (Fowler, 1937) Garra fuliginosa Fowler, 1934 Garra imberbis Vinciguerra, 1890 Garra nasuta McClelland, 1838 Garra notata Blyth, 1860 Garra salweenica Hora & Mukerji, 1934 Hampala dispar Smith, 1934 Hampala macrolepidota (Valenciennes, 1842) Hampala salweenensis Doi & Taki, 1994 Henicorhynchus caudimaculatus (Fowler, 1934) Henicorhynchus cryptopogon Fowler, 1934 Henicorhynchus lineatus (Smith, 1945) Henicorhynchus lobatus Smith, 1945 Henicorhynchus ornatipinnis (Roberts, 1997) Henicorhynchus siamensis (de Beaufort, 1937) Hypsibarbus lagleri Rainboth, 1996 Hypsibarbus pierrei (Sauvage, 1880) Hypsibarbus suvatti Rainboth, 1996 Hypsibarbus vernayi (Norman, 1925) Hypsibarbus wetmorei (Smith, 1931) Hypsibarbus salweenensis Rainboth, 1990 Labeo barbatulus (Sauvage, 1878) Labeo calbasu (Hamilton, 1822) Labeo chrysophekadion (Bleeker, 1850) Labeo dyocheilus (McClelland, 1839) Labeo erythropterus Valenciennes, 1842 Mekong X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X 12 Chao Phraya X X X X X X X X X X X X X X X X X X X X X X X X Salween X X X X X X X X - X X - X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Ref. B, E F E E, K A, B, D, E A, B, D, E B, E A, B, D, E A, B, D, E F A, B, D, E F AF E, F A, B, D, E A, B, D, E E E F A, B, D, E E F A, B, E B, E A, B, E B E B, E, F A, B, D, E A, B, D, E B, E B, E B, E B A, B, D, E F A, B, E A A, D, B J Q B, E A A, B, E A, B, D, E B, D, E A, B A, B C F F H A, B, D, E A, B, D, E, F F, K B B E A, E E A, B, E B, E B, E B B, E B, E F E F, H A, B, D, F, H D, F, H D    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species of Mekong, Chao Phraya and Salween rivers (Continued). 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 Order/Family/Species Labeo indramontri Smith, 1945 Labeo pierrei (Sauvage, 1880) Labeo rohita (Hamilton, 1822) Labeo yunnanensis Chuadhuri, 1911 Labiobarbus lineata (Sauvage, 1878) Labiobarbus leptocheila (Valenciennes, 1842) Lobocheilos bo (Popta, 1904) Lobocheilos cryptopogon (Fowler, 1935) Lobocheilos davisi (Fowler, 1937) Lobocheilos delacouri (Pellegrin & Fang, 1940) Lobocheilos fowleri (Pellegrin & Chevey, 1936) Lobocheilos gracilis (Fowler, 1937) Lobocheilos melanotaenia (Fowler, 1935) Lobocheilos nigrovittatus Smith, 1945 Lobocheilos quadrilineatus (Fowler, 1935) Lobocheilos rhabdoura (Fowler, 1934) Lobocheilos thavili (Smith, 1945) Mekongina erythrospila Fowler, 1937 Mystacoleucus argenteus (Day, 1888) Mystacoleucus atridorsalis Fowler, 1937 Mystacoleucus chilopterus Fowler, 1935 Mystacoleucus ectypus Kottelat, 2000 Mystacoleucus greenwayi Pellegrin & Fang, 1940 Mystacoleucus marginatus (Valenciennes, 1842) Neolissochilus blanci (Pellegrin & Fang, 1940) Neolissochilus dukai (Day, 1878) Neolissochilus soroides (Duncker, 1904) Neolissochilus stracheyi (Day, 1871) Neolissochilus vittatus (Smith, 1945) Onychostoma gerlachi (Peters, 1881) Oreichthys cosuatis (Hamilton, 1822) Oreichthys parvus Smith, 1933 Osteobrama belangeri (Valenciennes, 1844) Osteobrama feae Vinciguerra, 1890 Osteochilus enneaporos (Bleeker, 1852) Osteochilus hasseltii (Valenciennes, 1842) Osteochilus lini Fowler, 1935 Osteochilus melanopleurus (Bleeker, 1852) Osteochilus microcephalus (Valenciennes, 1842) Osteochilus schlegeli (Bleeker, 1851) Osteochilus waandersii (Bleeker, 1852) Poropuntius bantamensis (Rendahl, 1920) Poropuntius deauratus (Valenciennes, 1842) Poropuntius carinatus (Wu & Lin, 1977) Poropuntius chondrorhynchus (Fowler, 1934) Poropuntius consternans Kottelat, 2000 Poropuntius genyognathus Roberts, 1998 Poropuntius hathe Roberts, 1998 Poropuntius huguenini (Bleeker, 1853) Poropuntius kontumensis (Chevey, 1934) Poropuntius laoensis (Günther, 1868) Poropuntius malcolmi (Smith, 1945) Poropuntius normani Smith, 1931 Poropuntius scapanognathus Roberts, 1998 Probarbus jullieni Sauvage, 1880 Probarbus labeamajor Roberts, 1992 Probarbus labeaminor Roberts, 1992 Puntioplites bulu (Bleeker, 1851) Puntioplites falcifer Smith, 1929 Puntioplites proctozysron (Bleeker, 1865) Puntioplites wanndersi (Bleeker, 1858-59) Puntius aurotaeniatus (Tirant, 1885) Puntius binotatus (Valenciennes, 1842) Puntius brevis (Bleeker, 1860) Puntius chola (Hamilton, 1822) Puntius jacobusboehlkei (Fowler, 1958) Puntius masyai Smith, 1945 Puntius orphoides (Valenciennes, 1842) Puntius partipentazona Fowler, 1934 Puntius rhombeus Kottelat, 2000 Puntius sophore (Hamilton, 1822) Puntius stoliczkanus (Day, 1871) Puntius ticto (Hamilton, 1822) Mekong X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X 13 Chao Phraya X X X X X X X X X X X X X X X X X Salween X X - X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Ref. A E, F E E A, D, L A, B, E, F A A B B P A, B A, B, E A A, B A, B, E A A, B, D, E A, F, H A, B, E A, D, E E D, E A, B, D, E B A B A, B, E, F A G A, F E F F, H B A, D, E, F A, D, E A, D, E E A A, D, E A, D A, D E F E F, H H A, D B E E E F, H A, B, D, E B, E B, E A, Q E A, B, D, E B, E E A, D E F E A A, D, E E E F A, E, F F    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species of Mekong, Chao Phraya and Salween rivers (Continued). 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 334 335 336 337 Order/Family/Species Scaphiodonichthys acanthopterus (Fowler, 1934) Scaphiodonichthys burmanicus Vinciguerra, 1890 Scaphognathops bandanensis Boonyaratpalin & Srirungroj, 1971 Scaphognathops stejnegeri (Smith, 1931) Sikukia gudgeri (Smith, 1934) Sikukia stejnegeri Smith, 1931 Thynnichthys thynnoides (Bleeker, 1852) Tor brevifilis (Peters, 1881) Tor douronensis (Valenciennes, 1842) Tor soro (Cuvier & Valenciennes, 1842) Tor tambra (Valenciennes, 1842) Tor tambroides (Bleeker, 1854) Family Balitoridae Subfamily Balitorinae Annamia normani (Hora, 1930) Balitora annamitica Kottelat, 1988 Balitora brucei Gray, 1833-34 Balitora burmanica Hora, 1932 Balitora meridionalis Kottelat, 1988 Hemimyzon nanensis Doi & Kottelat, 1998 Homaloptera bilineata Blyth, 1860 Homaloptera confuzona Kottelat, 2000 Homaloptera indochinensis Silas, 1953 Homaloptera leonardi Hora, 1941 Homaloptera maxinae Fowler, 1937 Homaloptera orthogoniata Vaillant, 1902 Homaloptera smithi Hora, 1932 Homaloptera modesta (Vinciguerra, 1890) Homaloptera sexmaculata Fowler, 1934 Homaloptera tweediei, Herre, 1940 Homaloptera zollingeri Bleeker, 1853 Subfamily Nemacheilinae Acanthocobitis botia (Hamilton, 1822) Acanthocobitis zonalternans (Blyth, 1860) Acanthocobitis rubidipinnis (Blyth, 1860) Nemacheilus binotatus Smith, 1933 Nemacheilus longistriatus Kottelat, 1990 Nemacheilus masyae Smith, 1933 Nemacheilus pallidus Kottelat, 1990 Nemacheilus platiceps Kottelat, 1990 Neonoemacheilus labeosus (Kottelat, 1982) Physoschistura pseudobrunneana Kottelat, 1990 Schistura alticrista Kottelat, 1990 Schistura atriceps (Smith, 1945) Schistura bella Kottelat, 1990 Schistura breviceps (Smith, 1945) Schistura bucculenta (Smith, 1945) Schistura cincticauda (Blyth, 1860) Schistura daubentoni Kottelat, 1990 Schistura defectiva Kottelat, 2000 Schistura desmotes (Fowler, 1934) Schistura dubia Kottelat, 1990 Schistura geisleri Kottelat, 1990 Schistura kengtungensis (Fowler, 1936) Schistura kohchangensis (Smith, 1933) Schistura maepaiensis Kottelat, 1990 Schistura magnifluvis Kottelat, 1990 Schistura mahnerti Kottelat, 1990 Schistura menanensis (Smith, 1945) Schistura moeiensis Kottelat, 1990 Schistura nicholsi (Smith, 1933) Schistura laterimaculata Kottelat, 1990 Schistura oedipus Kottelat, 1989 Schistura paucicincta Kottelat, 1990 Schistura poculi (Smith, 1945) Schistura reidi (Smith, 1945) Schistura schultzi (Smith, 1945) Schistura sexcauda (Fowler, 1937) Schistura similis Kottelat, 1990 Schistura spilota (Fowler, 1934) Schistura vinciguerrae (Hora, 1935) Schistura waltoni Fowler, 1937 Sectoria atriceps Kottelat, 1990 14 Mekong X X X X X X X X X X X X Chao Phraya X X X X X X X X Salween X X X X Ref. A, D, E A, F, H D, E A, D, E A, B, D, E A, B, E A, B, D, E A, E, F A, D A A, D, E A, D, E, F, H X X X X X X X X X X X X X X X X X X X - X - D, E E A F B C F E B B B B A, B, D, E A A B, E A, B, E X X - X X X X X X X X X X X X X X X X X X X - X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X - A E F A E, L A, E, L A, B, E, L A, B, E, L F, H F, L F, L A L G, L A, D, E, L F, L L E A, L L L A, E, L A, L H, L L G, H, L A, L H, L A, H, L A, L L F, L A, B, E, H, L A, L A, E, L A, E, L L A, L F, H, L A, L L    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species of Mekong, Chao Phraya and Salween rivers (Continued). 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 382 381 383 384 385 386 387 388 389 390 391 392 394 393 395 396 397 398 399 400 401 Order/Family/Species Tuberoschistura baezigeri (Kottelat, 1983) Vailantella maassi Weber & de Beaufort, 1912 Yunnanilus brevis (Boulenger, 1893) Family Cobitidae Subfamily Botiinae Botia kubotai Kottelat, 2004 Botia histrionica Blyth, 1860 Botia rostrata Günther, 1868 Syncrossus beauforti (Smith, 1931) Syncrossus berdmorei Blyth, 1860 Syncrossus helodes (Sauvage, 1876) Yasuhikotakia caudipunctata (Taki and Doi, 1995) Yasuhikotakia eos Taki, 1972 Yasuhikotakia lecontei (Fowler, 1937) Yasuhikotakia longidorsalis (Taki and Doi, 1995) Yasuhikotakia modesta (Bleeker, 1864) Yasuhikotakia morleti (Tirant, 1885) Yasuhikotakia nigrolineata (Kottelat and Chu, 1987) Yasuhikotakia sidthimunki (Klausewitz, 1959) Yasuhikotakia splendida (Roberts, 1995) Subfamily Cobitinae Acanthopsoides delphax Siebert, 1991 Acanthopsoides gracilentus (Smith, 1945) Acanthopsoides gracilis Fowler, 1934 Acanthopsoides hapalias Siebert, 1991 Acantopsis choirorhynchos (Bleeker, 1854) Acantopsis dialuzona Van Hasselt, 1823 Acantopsis spectabilis Blyth, 1860 Acantopsis thiemmedhi Sontirat, 1999 Lepidocephalichthys berdmorei (Blyth, 1860) Lepidocephalichthys furcatus (de Beaufort, 1933) Lepidocephalichthys hasselti (Valenciennes, 1846) Lepidocephalichthys micropogon (Blyth, 1860) Pangio anguillaris (Vaillant, 1902) Pangio fusca (Blyth, 1860) Pangio kuhlii (Valenciennes, 1846) Pangio myersi (Harry, 1949) Pangio oblonga (Valenciennes, 1846) Pangio pangia (Hamilton, 1822) Pangio semicincta (Fraser-Brunner, 1940) Serpenticobitis octozona Roberts, 1997 Serpenticobitis zonata Kottelat, 1998 Family Gyrinocheilidae Gyrinocheilus aymonieri (Tirant, 1884) Gyrinocheilus pennocki (Fowler, 1937) Order Siluriformes Family Akysidae Acrochordonichthys rugosus (Bleeker, 1847) Akysis brachybarbatus Chen, 1981 Akysis macronemus Bleeker, 1860 Akysis maculipinnis Fowler, 1934 Akysis recavus Ng & Kottelat, 1998 Akysis subtilis Ng & Kottelat, 1998 Akysis varius Ng & Kottelat, 1998 Akysis variegatus (Bleeker, 1846) Family Amblycipitidae Amblyceps mucronatum Ng & Kottelat, 2000 Amblyceps platycephalus Ng & Kottelat, 2000 Amblyceps serratum Ng & Kottelat, 2000 Family Ariidae Arius arius (Hamilton, 1822) Arius acutirostris Day, 1877 Arius maculatus (Thunberg, 1792) Arius intermedius (Vinciguerra, 1880) Batrachocephalus mino (Hamilton, 1822) Cochlefelis burmanica (Day, 1870) Hemipimelodus borneensis (Bleeker, 1851) Hemipimelodus jatius (Hamilton, 1822) Ketangus typus Bleeker, 1847 Osteogeneisosus militaris (Linnaeus, 1758) Hemiarius stormii (Bleeker, 1858) Family Bagridae Bagrichthys macracanthus (Bleeker, 1854) 15 Mekong X - Chao Phraya X X - Salween X Ref. L L F, L X X X X X X X X X X X X X X X X X X X - X X X X - F F F A, B, D, E F A, B, D, E E A, B, D, E A, B, D, E E A, B, D, E A, B, D, E E A, D, E E X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X - X X X X X X X X - A, E A, E D, E A, E A, E, D, H P F G E, F E D, E F A, D, E A, B, E A E B, E A A E, R R X X X - - A, B, D, E A, B, D, E X X X X X X X X X X X - - S T A A T E, T E, T B X - - X X X E H E, F X X X X X X X X X X X X X - X X X X - F F B B A F A, B, E F A A E X X - A, D, E    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species of Mekong, Chao Phraya and Salween rivers (Continued). 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 Order/Family/Species Bagrichthys macropterus (Bleeker, 1853) Bagrichthys obscurus Ng, 1999 Batasio tengana (Hamilton, 1822) Batasio havmolleri (Smith, 1931) Batasio tigrinus Ng & Kottelat, 2001 Hemibagrus filamentus (Fang & Chaux, 1949) Hemibagrus microphthalmus (Day, 1877) Hemibagrus nemurus (Valenciennes, 1840) Hemibagrus planiceps (Valenciennes, 1840) Hemibagrus variegatus Ng & Ferraris, 2000 Hemibagrus wyckii (Bleeker, 1858) Hemibagrus wyckioides (Fang & Chaux, 1949) Mystus albolineatus Roberts, 1994 Mystus atrifasciatus Fowler, 1937 Mystus bocourti (Bleeker, 1864) Mystus cavasius (Hamilton, 1822) Mystus gulio (Hamilton, 1822) Mystus leucophasis (Blyth, 1860) Mystus multiradiatus Roberts, 1992 Mystus mysticetus Roberts, 1992 Mystus pulcher (Chaudhuri, 1911) Mystus singaringan (Bleeker, 1846) Mystus rhegma Fowler, 1935 Mystus wolffi (Bleeker, 1851) Pseudomystus leiacanthus (Weber & de Beaufort, 1912) Pseudomystus siamensis (Regan, 1913) Pseudomystus stenomus (Valenciennes, 1839) Rita sacerdotum Anderson, 1879 Sperata acicularis Ferraris & Runge, 1999 Family Clariidae Clarias batrachus (Linnaeus, 1785) Clarias gariepinus (Burchell, 1851) Clarias leiacanthus Bleeker, 1851 Clarias macrocephalus Günther, 1864 Clarias meladerma Bleeker, 1847 Clarias nieuhofi Valenciennes, 1840 Family Heteropneustidae Heteropneustes kemratensis (Fowler, 1937) Family Pangasiidae Helicophagus leptorhynchus Ng & Kottelat, 2000 Helicophagus waandersii Bleeker, 1858 Pangasianodon hypophthalmus (Sauvage, 1878) Pangasianodon gigas Chevey, 1930 Pangasius bocourti (Sauvage, 1880) Pangasius conchophilus Roberts & Vidthayanon, 1991 Pangasius djambal Bleeker, 1846 Pangasius krempfi Fang & Chaux, 1949 Pangasius larnaudii Bocourt, 1851 Pangasius macronema Bleeker, 1851 Pangasius micronemus Bleeker, 1847 Pangasius myanmar Roberts & Vidthayanon, 1991 Pangasius pangasius (Hamilton, 1822) Pangasius pleurotaenia (Sauvage, 1878) Pangasius polyuranodon Bleeker, 1852 Pangasius sanitwongsei Smith, 1931 Family Plotosidae Plotosus canius Hamilton, 1822 Plotosus lineatus (Thunberg, 1791) Family Schilbeidae Clupisoma prateri Hora, 1937 Clupisoma sinense (Huang, 1981) Eutropiichthys burmannicus Day, 1877 Laides longibarbis (Fowler, 1934) Laides hexanema (Bleeker, 1852) Proeutropiichthys taakree macropthalmos (Blyth, 1860) Family Siluridae Belodontichthys dinema (Bleeker, 1851) Belodontichthys truncatus Kottelat & Ng, 1999 Ceratoglanis pachynema Ng, 1999 Hemisilurus mekongensis Bonbusch & Lundberg, 1989 Kryptopterus bicirrhis (Valenciennes, 1840) Kryptopterus cheveyi Durand, 1940 Kryptopterus cryptopterus (Bleeker, 1851) 16 Mekong X X X X X Chao Phraya X X X X X X X X X X X X X X X X X X X X X X X - Salween X X X X X X X Ref. A, D, E E O A, O O A, B, E A, F A, B, D, E A F A, B, E A, B, E B, E B, E A, B, D, E D, F A, F F B, E A, B, E Q B, E, F A, B, D, E A, B, E A A, B, D, E A F, H F, H X X X X X X X X X X X X X X X X X X X X X X X X X X X - A, B, D, E, F B, E A A, B, D, E A, B, E A, B X X X A, B, D, E, F X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X - E A, B, D A, B, D, E A, B, D, E A, B, E E B E A, B, D, E A, E A, E F D, F A, B, E A, B, E A, B, D, E X - X X X - A, B, E, F A X X X - X X - X X X F E F D, E A, B, D, E F X X X X X X X X X X - X X X - A, B, D, E A, B, E B, E B, E A, B, D, E A, B, D, E A, B, D, E    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species of Mekong, Chao Phraya and Salween rivers (Continued). 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 Order/Family/Species Kryptopterus dissitus Ng, 2001 Kryptopterus geminus Ng, 2003 Kryptopterus limpok (Bleeker, 1852) Kryptopterus macrocephalus (Bleeker, 1858) Kryptopterus moorei Smith, 1945 Kryptopterus schilbeides (Bleeker, 1858) Micronema hexapterus (Bleeker, 1851) Ompok bimaculatus (Bloch, 1794) Ompok eugeneiatus (Vaillant, 1893) Ompok hypophthalmus (Bleeker, 1846) Ompok pabda (Hamilton, 1822) Ompok pabo (Hamilton, 1822) Phalacronotus apogon (Bleeker, 1851) Phalacronotus bleekeri (Günther, 1864) Phalacronotus micronemus (Bleeker, 1846) Pterocryptis cochinchinensis (Valenciennes, 1840) Pterocryptis bokorensis (Pellegrin & Chevey, 1937) Pterocryptis torrentis (Kobayakawa, 1989) Silurichthys hasselti Bleeker, 1858 Silurichthys phaiosoma (Bleeker, 1851) Wallago attu (Bloch & Schneider, 1801) Wallago leerii Bleeker, 1851 Family Sisoridae Bagarius bagarius (Hamilton, 1822) Bagarius suchus Roberts, 1983 Bagarius yarrelli Sykes, 1838 Caelatogranis zonatus Ng & Kottelat, 2005 Erethistes maesotensis Kottelat, 1983 Exostoma berdmorei Blyth, 1860 Gagata cenia (Hamilton, 1822) Gagata gasawyuh Roberts & Ferraris, 1998 Gagata melanopterus Roberts & Ferraris, 1998 Glyptothorax burmanicus Prashad & Mukerji, 1929 Glyptothorax buchanani Smith, 1945 Glyptothorax callopterus Smith, 1945 Glyptothorax dorsalis Vinciguerra, 1890 Glyptothorax fuscus Fowler, 1934 Glyptothorax lampris Fowler, 1934 Glyptothorax laosensis Fowler, 1934 Glyptothorax major (Boulenger, 1894) Glyptothorax minimaculatus Li 1984 Glyptothorax prashadi Murerji, 1932 Glyptothorax rugimentum Ng & Kottelat, 2008 Glyptothorax trilineatus Blyth, 1860 Glyptothorax zanaensis Wu, He & Chu, 1981 Oreoglanis colurus Vidthayanon, Saenjundaeng & Ng, 2009 Oreoglanis heteropogon Vidthayanon, Saenjundaeng & Ng, 2009 Oreoglanis laciniosus Vidthayanon, Saenjundaeng & Ng, 2009 Oreoglanis nakasathiani Vidthayanon, Saenjundaeng & Ng, 2009 Oreoglanis siamensis Smith, 1933 Oreoglanis sudarai Vidthayanon, Saenjundaeng & Ng, 2009 Oreoglanis suraswadii Vidthayanon, Saenjundaeng & Ng, 2009 Oreoglanis tenuicauda Vidthayanon, Saenjundaeng & Ng, 2009 Oreoglanis vicinus Vidthayanon, Saenjundaeng & Ng, 2009 Pareuchiloglanis feae (Vinciguerra, 1890) Pareuchiloglanis kamengensis (Jayaram, 1966) Pseudecheneis sulcata (McClelland, 1842) Order Atheriniformes Family Phallostethidae Phenacostethus smithi Myers, 1928 Neostethus siamensis Myers, 1937 Order Beloniformes Family Adrianichthyidae Oryzias javanicus Bleeker, 1854 Oryzias latipes (Temminck and Schlegel, 1846) Oryzias mekongensis Uwa & Magtoon, 1986 Oryzias minutillus Smith, 1945 Oryzias sinensis Chen, Uwa & Chu, 1989 Family Belonidae Xenentodon cancila (Hamilton, 1822) Xenentodon canciloides (Bleeker, 1853) Strongylura strongylura (van Hasselt, 1823) Family Hemirhamphidae 17 Mekong X X X X X X X X X X X X X X X X X X X X Chao Phraya X X X X X X X X X X X X X - X X Salween X X X X - Ref. U V A, B, E B, E A, B, E A, B, D A, B, E A, B, E, D, F A, B, E A, B, E H F A, B, D, E A, B, D, E A, B, E A, E B B A A A, B, D, E, F A, B, D, E X X X X X X X X X X X X - X X X X X X X X X X X X X X X X X - X X X X X X X X X X X X X X X X X X X X A, B, D, E E A, B, F F F F A F F H A A F, H A, B, E A, B E A H A H A, B, D, F E N N N N A, N N N N N W W F X X X X - A A X X X X X X X - X X X B, Q X E A E X X - X X X - A, B, F, G A, D, E A    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species of Mekong, Chao Phraya and Salween rivers (Continued). 537 538 539 540 541 542 543 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 Order/Family/Species Dermogenys pusilla van Hasselt, 1823 Hyporhamphus limbatus (Valenciennes, 1846) Zenachopterus buffonis (Valenciennes, 1845) Zenachopterus dunckeri Mohr, 1926 Zenachopterus ectuntio (Hamilton, 1822) Order Cyprinodontiformes Family Aplocheilidae Aplocheilus panchax (Hamilton, 1822) Family Poeciliidae Gambusia affinis (Baird and Girard, 1853) Gambusia holbrookii (Girard, 1859) Poecilia reticulata Peters, 1859 Order Gasterosteiformes Family Indostomidae Indostomus paradoxus Prasad & Mukerji, 1929 Indostomus spinosus Britz & Kottelat, 1999 Family Syngnathidae Doryichthys boaja (Bleeker, 1851) Doryichthys contiguus Kottelat, 2000 Doryichthys deokhatoides (Bleeker, 1853) Doryichthys martensii (Peters, 1868) Hipichthys spicifer (Rüppell, 1838) Ichthyocampus carce (Hamilton, 1822) Microphis brachyurus (Bleeker, 1853) Order Synbranchiformes Family Chaudhuridae Chaudhuria caudata Annandale, 1918 Family Mastacembelidae Macrognathus aral (Bloch & Schneider, 1801) Macrognathus aculeatus (Bloch, 1786) Macrognathus caudiocellatus (Boulenger, 1893) Macrognathus circumcinctus (Hora, 1924) Macrognathus maculatus Cuvier, 1831 Macrognathus semiocellatus Roberts, 1986 Macrognathus siamensis (Günther, 1861) Macrognathus taeniagaster (Fowler, 1835) Macrognathus zebrinus (Blyth, 1858) Mastacembelus alboguttatus Boulenger, 1893 Mastacembelus armatus (Lacepède, 1800) Mastacembelus erythrotaenia Bleeker, 1870 Mastacembelus favus Hora, 1823 Mastacembelus tinwini Britz, 2007 Family Synbranchidae Monopterus albus (Zuiew, 1793) Monopterus cuchia (Hamilton, 1822) Ophisternon bengalense (McClelland, 1845) Order Perciformes Family Ambassidae Ambassis buruensis Bleeker, 1856 Ambassis gymnocephalus (Lacepède, 1802) Ambassis kopsi Bleeker, 1851 Parambassis apogonoides (Bleeker, 1851) Parambassis lala (Hamilton, 1822) Parambassis ranga (Hamilton, 1822) Parambassis siamensis (Fowler, 1937) Parambassis vollmeri Roberts, 1995 Parambassis wolffii (Blyth, 1860) Family Centropomidae Lates calcarifer (Bloch, 1790) Family Polynemidae Polynemus longipectoralis Weber & de Beaufort, 1922 Polynemus multifilis Schlegel, 1845 Polynemus paradiseus Linnaeus, 1758 Family Scieanidae Boesemania microlepis (Bleeker, 1858-59) Johnius coitor  (Hamilton, 1822) Otolithoides pama (Hamilton, 1822) Otolithoides biauritus (Cantor, 1849) Family Toxotidae Toxotes chatareus (Hamilton, 1822) Toxotes microlepis (Günther, 1860) Family Lobotidae Datnioides microlepis Bleeker, 1853 18 Mekong X X X X X Chao Phraya X X X X X Salween - Ref. A, B, E A A A A - X X A, F X X X X X - A, E A E X X X - X - F E X X X X X X X X X X X X X - A, E E A A A A Y X X - E X X X X X X X X X - X X X X X X X - X X X X X X F A, D F A, E Q A, E A, E A F F A, B, D, E, F A A Z X X X - X X - D, E, F F Q X X X X X X X X X X X X - Q Q Q E F A, F A, B, D, E F A, B, E X X X A, F X X X X X X - E Q A X - X - X X X A, B, D, E F F F X X X X X - A, E, F A, D X X - A, D    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species of Mekong, Chao Phraya and Salween rivers (Continued). 594 595 596 597 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 641 641 642 643 644 645 646 647 468 649 650 651 652 653 Order/Family/Species Datnioides polota (Hamilton, 1822) Datnioides pulcher (Kottelat, 1998) Datnioides undecimradiatus (Roberts & Kottelat, 1994) Family Nandidae Badis ruber Schreitmüller, 1923 Dario hysginon Kullander & Britz, 2002 Nandus nandus (Bleeker, 1851) Nandus nebulosus (Gray, 1835) Nandus oxyrhynchus Ng, Vidthayanon & Ng, 1996 Pristolepis fasciata (Bleeker, 1851) Family Cichlidae Oreochromis niloticus (Linnaeus, 1758) Family Gobiidae Brachygobius mekongensis Larson & Vidthayanon, 2000 Calamiana aliceae (Smith, 1945) Eugnathogobius siamensis (Fowler, 1934) Glossogobius aureus Akihito & Meguro, 1975 Glossogobius giuris (Hamilton, 1822) Gobiopterus chuno (Hamilton, 1822) Papuligobius ocellatus (Fowler, 1937) Rhinogobius chiengmaiensis (Fowler, 1934) Rhinogobius mekongianus (Pellegrin & Fang, 1940) Family Eleotridae Oxyeleotris marmorata (Bleeker, 1852) Family Odontobutididae Neodontobutis aurarmus (Vidthayanon, 1995) Family Anabantidae Anabas testudineus (Bloch, 1792) Family Helostomatidae Helostoma temmincki Cuvier, 1831 Family Osphronemidae Betta imbellis Ladiges, 1975 Betta pi Tan, 1998 Betta prima Kottelat, 1994 Betta pugnax (Cantor, 1849) Betta simplex Kottelat, 1994 Betta smaragdina Schaller, 1986 Betta splendens Regan, 1909 Colisa labiosa (Day, 1877) Osphronemus exodon Roberts, 1994 Osphronemus goramy Lacepède, 1802 Paraspherichthys ocellatus Prashad & Mukerji, 1929 Parosphromenus paludicola Tweedie, 1952 Trichogaster leerii (Bleeker, 1852) Trichogaster microlepis (Günther, 1861) Trichogaster pectoralis Regan, 1909 Trichogaster trichopterus (Pallas, 1770) Trichopsis pumila (Arnold, 1937) Trichopsis schalleri Ladiges, 1962 Trichopsis vittata (Cuvier, 1831) Family Channidae Channa aurolineata Channa gachua (Hamilton, 1822) Channa harcourtbutleri (Annandale, 1918) Channa lucius (Cuvier, 1831) Channa marulius (Hamilton, 1822) Channa melasoma (Bleeker, 1851) Channa micropeltes (Cuvier, 1831) Channa oreintalis (Schneider, 1801) Channa striata (Bloch, 1795) Order Pleuronectiformes Family Soleidae Achiroides leucorhynchos Bleeker, 1851 Achiroides melanorhynchus (Bleeker, 1850) Brachirus harmandi (Sauvage, 1878) Brachirus orientalis (Schneider, 1801) Brachirus panoides (Bleeker, 1851) Brachirus siamensis (Sauvage, 1878) Family Cynoglossidae Cynoglossus feldmanni (Bleeker, 1853) Cynoglossus microlepis (Bleeker, 1851) Order Tetraodontiformes Family Tetraodontidae 19 Mekong X X X Chao Phraya X X - Salween - Ref. A E E X X X X X X X X X X X - E, F F A, D A, D E A, B, D, E X X X E, H X X X X X X X X X X X X X X X X X - E A A E F A, E A, E G E X X - A, B, D, E X - - E X X X A, E, F X X - A X X X X X X X X - X X X X - X X AA AB E AD E A F E A, B, D, E F X X X X X X X X X X X X X X X A, B E A, B, E A, B, E A, E E A, B, E X X X X X X X X X X X X X X X F A, B, D, E F A, B, D, E A, E A, D A, D, E K A, B, D, E, F X X X X X X X X X X - A A A, D, E A A A, E X X X X - E A, B, E    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Table 3 Fishes species of Mekong, Chao Phraya and Salween rivers (Continued). 654 655 656 657 658 659 660 661 662 663 664 665 666 667 Order/Family/Species Auriglobus nefastus Roberts, 1982 Carinotetraodon lorteti (Tirant, 1885) Tetraodon abei Roberts, 1998 Tetraodon baileyi Sontirat, 1989 Tetraodon biocellatus Tirant, 1885 Tetraodon cambodgiensis (Chabanaud, 1923) Tetraodon cutcutia Hamilton, 1822 Tetraodon cochinchinensis (Steindachner, 1866) Tetraodon fluviatilis (Hamilton, 1822) Tetraodon leiurus (Bleeker, 1851) Tetraodon nigroviridis (Procé, 1822) Tetraodon palembangensis Bleeker, 1852 Tetraodon suvattii Sontirat, 1989 Tetraodon turgidus (Kottelat, 2000) Total species Mekong X X X X X X X X X X X X X 509 Chao Phraya X X X X X X X X X X 420 Salween X 190 Ref. E E X E E E C, D, F E A, D A, D A, E A E E Note: For abbreviations A = Smith (1945); B = Rainboth (1996); C = Jayaram (1999); D = Taki (1974); E = Kottelat (2001); F = Vidthayanon et al. (2005); G=Suvarnaraksha et al. (2004); H = Suvarnaraksha et al. (2010); I = Last & Compagno (1999); J = Doi & Taki (1994); K = Zhang, Yue & Chen (2000); L = Kottelat (1990); M = Freyhof & Serov (2001); N = Vidthayanon, Saenjundaeng, & Ng (2009); O = Ng & Kottelat (2001); P = Doi (1997); Q = Monkolprasit et al. (1997); R = Kottelat (1998); S = Ng & Ng (2001); T = Ng & Kottelat (1998); U = Ng (2002); V = Ng (2003); W = He (1996); X = Roberts (1998); Y = Dawson (1985); Z = Britz (2007); AA = Kottelat et al. (1993); AB = Tan (1998); AD = Tan & Tan (1996); AE = Kottelat (1994); AF= Roberts (1997) 1.4 Biology and life history traits of the tropical freshwater fishes The freshwater ecosystems in tropical Asia are rich of fauna and flora species and there are very complexities, especially fishes. The diverse groups of fishes are also resulted in the wide range of morphological, behavioural, and life history attributes that characterise the constituent species, which is due to the fact that various habitats are embedded in inland waterbodies (Mims et al., 2010). The life history of recent fish species have evolved from basal ancestors to survive, feed, reproduce and die in a given ecological niche within a given aquatic ecosystem (Froese, 2005). Understanding the life history of individual fish species includes what it eats, how fast it grows and how old and large it gets when it matures and how successfully it reproduces, and other aspects of its biology (Matthews, 1998; Froese, 2005). 20    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand In the tropical river system, most fishes breed during the rainy season (AlkinsKoo, 2000; Ballesteros et al., 2009), however, a few breed during the dry season (Pusey et al., 2002; Torres-Mejia & Ramírez-Pinilla, 2008) or throughout the year (Alkins-Koo, 2000). Variation in reproductive seasonality has been associated with several factors, such as availability of nursery areas, availability of food for adults or juveniles, competition for breeding sites, phylogenetic inertia and hydrological cycle in the river system (Ballesteros et al., 2009). Generally, most fish in the river system cannot complete its in a single habitat, when requirements for reproduction and for feeding at different life stages cannot be met in the same place, then fishes have to move between places to survive (Baran & Jutagate 2010). One classification of fish species relates to the ability to complete their life cycle dependant on access to the riverine environments. Obligatory riverine species spawn only in the river corridor, while facultative (non-obligatory) riverine species can realize their life history strategy in both stagnant and flowing waters (Schiemer & Waidbacher 1992; Kruk & Penczak, 2002). Thus, almost all obligatory riverine fish species suffer severely from dam construction without effective fish paasages, including the local extirpation of many of them (Penczak & Kruk 2000; Kruk & Penczak, 2002), in which this problem is among the most concern issues in Thailand, where a numbers of damming project are proposed including in the Ping-Wang River basin (Jutagate et al., 2011). Food consumption studies in fish populations have received attention among aquatic ecologists and fisheries biologists, mainly to assess trophic relationships in aquatic ecosystems (Christensen & Pauly, 1993; Amarasinghe et al., 2010). Welcomme et al. (2006) mentioned that there is flexibility in diets of many freshwater fish, which may be related to fish size, season and location within the system or most likely a combination of all three (Pusey et al., 1995). Moreover, dietary composition of many tropical freshwater species also showed that they are mostly omnivorous (Guruge, 2002). The highest feeding activities of tropical fishes usually occur during the rainy seasons when the availability of prey is relatively higher (Prejs & Prejs, 1987; Ballesteros et al., 2009). Kramer (1978) proposed the theory that the reproductive season of tropical freshwater fish would be synchronized with food consumption, which could be confirmed on the importance of feeding to sustain the fish stock and renew the next generations. Therefore, numbers of individual in fish 21    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand stock would decline if critical food resources are limited or eliminated by any disturbances (Welcomme et al., 2006) Fishes have indefinite growth (i.e. the size is increasing continuously, albeit different rate, throughout their lives) and the maximum life span may be taken as the age corresponding to 95% of the asymptotic size of the von Bertalanffy growth function (Froese, 2005). The von Bertalanffy growth function (VBGF) is based on a bioenergetic expression of fish growth and VBGF is the most important model and widely used to describe the average “size-at-agea wide variety of aquatic organisms (Cailliet et al., 2006) and the function is generally expressed as Equation 1 Lt > @ Lf 1  e  ( K (t t0 ) ----------------------- (1) where Lt is length at time t, Lf is the asymptotic length, K is the growth coefficient and t 0 is the theoretical age at length zero. Moreover, if there are strong seasonal changes in temperature, the modified version of the VBGF (Equation 2) was used, which incorporates seasonal oscillation in growth (Herrmann et al., 2009). Two more parameters were incorporated into the VBGF, when seasonality was taken into account: firstly, C, which is between 0 and 1 indicates the magnitude of the seasonal growth pattern and secondly, t s , the time from birth to the start of growth oscillations Lt > @ Lf 1  e  ( K (t t0 )CK / 2S ){sin 2S (t t0 )sin 2S (t t0 )} ---------------------- (2) Froese & Binohlan (2000) mentioned that about 7,000 species of fishes are consumed by humans, knowing on life history traits on growth and maturity, which is essential for proper management of exploited populations, is available for only about 1,200 species, which could be hampered efforts to sustainable uses the fish stocks. For example, maximum sustainable yield and the fishing mortality rate that produces the maximum yields can be estimated by using the key life-history parameters of fish species such as growth coefficient (K), the length at sexual maturity relative to asymptotic length incorporated with length at captured and natural mortality rate and sometimes, the stock recruitment relationship (Beddington & Cooke, 1983; Kirkwood 22    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand et al., 1994; Beddington & Kirkwood, 2005). Meanwhile, life span and age at first maturity are two important parameters in conservation management (Froese & Binohlan, 2000). 1.5 Freshwater ecological study and fish assemblages River ecology of tropical Southeast Asia is dominated by flow seasonality imposed by monsoonal rains with profound consequences for fishes and zoobenthos (Dudgeon, 2000). Thus, fluctuations and changes in discharge patterns affect the abundance, species composition and viability of living aquatic resources resident in the river. Also, along the river gradient, the variations in geo-morphological characteristics of the river as well as environmental variables (both biotic and abiotic) are the major factors that govern fish communities both in terms of species richness and distribution of individual species (Orrego et al., 2009; Alexandre et al., 2010; Kimmel & Argent, 2010). Moreover, environment favour specific suites of traits, resulting in the evolution of life history strategies or tactics that enable a species to cope with a range of ecological problems (Froese, 2005; Mims et al., 2010). Under natural conditions, a river is characterized by either a continuous succession of fish species along the spatial gradient or a staggered succession (Orrego et al., 2009). In a fluvial ecosystem, species composition is highly influenced by parameters such as altitude, gradient, current velocity, and temperature (Campos, 1985; Orrego et al., 2009). Meanwhile, along the downstream gradient the River Continuum Concept (Vannote et al., 1980) relates community structure and river functional changes, with physical factors such as flow regime, temperature, food availability, and river morphological conditions (Orrego et al., 2009). Generally, fishes show high adaptability to their habitat environment, whereas their morphological and ecological characteristics change correspondingly (He et al., 2010). Meanwhile, the distribution range of fishes along an upstream–downstream gradient within a river basin is determined by the ecological requirements of each fish species (Ferreira & Petrere, 2009). Distinguishing fish assemblages along the river gradient is very difficult because prostine environment does not exist any more due to anthropogenic stresses and invasion of non-native species (Vannote et al., 1980; Kruk et al., 2007). 23    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Moreover, temporal variability in fish assemblages is also common and driven by similar processes that impact on fish population dynamics via immigration, emigration, spawning, recruitment and mortality (King et al., 2003; Balcombe et al., 2006). The global growing concern about pervasive impacts of human modifications to riverine ecosystems (Allan & Flecker 1993; Malmqvist & Rundle, 2002), has led to increasing recognition of the need for quantitative procedures for assessing aquatic ecosystem and monitoring biotic responses to remedial management. Many theoretical classifications of running waters, notably fish-based classifications, have been proposed since the end of the 19th century (e.g. Huet, 1959) and becoming much more concern because the deviation between the observed assemblage type and the one expected in undisturbed (theoretical) conditions provides an assessment of their ecological status (Lasne et al., 2007). Recently, Welcomme et al., (2006) proposed environmental guilds of freshwater fishes along the river gradient (using location in river system, reproductive, behavioural, and ecological traits) as a tool for riverine ecological assessment. To evaluate the status and any changes in fish assemblages in each section and/or time, diversity indices are commonly used and the commonest indicator is the number of species found, i.e. species richness (Oberdorff et al., 2002; de Thoisy et al., 2008; He, 2010). This indicator is an integrative descriptor of the animal community, influenced by a large number of natural environmental factors as well as anthropogenic disturbances, including the geological history of the area, environmental stability, ecosystem productivity and heterogeneity (He et al., 2010). It is suggested that if the physical aspects of the stream are relatively stable, they are responsible for the consistent pattern in biological community structure (Orrego et al., 2009) although there may some other factors could be influenced such as competition, predation as well as point and non-point pollution sources (Ibarra et al., 2005; Orrego et al., 2009). 1.6 Objectives of this Thesis Because of natural functioning aquatic ecosystems have important intrinsic values and also provide many goods, services and long-term benefits to human 24    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand society (Baron et al., 2002), hence their protection, remediation and restoration is of critical importance. However, in Chao Phraya River basin, research on the aquatic ecology and fish diversity are few and it needs more scientific information to reach the ultimate goal of wise and sustained uses of aquatic resources. This thesis was divided into 3 main levels viz., taxonomic level (descriptive level), biology and life history of fishes (descriptive level to predictive level), and assemblages of fish diversity and environmental factors (predictive level). The first level is the investigation fish diversity in the upper Chao Phraya River basin; a part of IndoBurma hotspot region (Publication 1; P1). At the second level, investigation of life history and population dynamics of the keystone fish species in the study area i.e., (1) life history of riverine cyprinid Henicorhynchus siamensis (Sauvage, 1881) in a small reservoir (Publication 2; P2) and (2) reproductive biology and conservation approaches of endanger species stream sisorids (Oreoglanis siamensis) (Publication 3; P3). The H. siamensis is a well adaptation from riverine species to reservoir conditions and it was an important economic fish for fisherman in this reservoir. Meanwhile, O. siamensis is a vulnerable species, which inhabits cold swift of highn mountain streams and attaches itself to rock surfaces facing the current. Both species were the representative of lentic and lotic ecosystem conditions. The lentic H. siamensis was a riverine species but it was well adapted to the reservoir. And the lotic O. siamensis was an endemic and vulnerable species, restrict to the habitat and high elevations. Finally, the third level, investigation of the relationships between biological parameters and environmental parameters which are also benefit to investigate fish diversity and assemblages patterns in the studied area (Publication 4; P4) and lastly the investigation of the effects of dam to the riverine fish assemblages (Publication 5; P5). 25    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand 2. GENERAL MATERIALS AND METHODS 2.1 Studied sites and data collection This study was conducted in the Ping - Wang river basin, located in upper Chao Phraya river basin (the largest river of Thailand). The Ping basin is one of the largest drainage basins of the Chao Phraya river basin with a total length of 658 km and draining 33,896 km and extends to 44,688 km if included the Wang river basin. The Wang river is 440 km long and has a catchment area 10,791 km2 (Takeuchi et al., 2005). The Wang river flows southwestward to join the lowland of Ping river in Tak province and they combine to form a large watershed area between 15q42’ and 19q48’ North and 98q04’ and 100q08’ East. The highest altitude of the sampling sites in this thesis is at 1,700 m ASL and connected to the lower Chao Phraya river basin at the altitude of 40 m ASL. Table 2 Descriptions of the sub-basins in the Ping-Wang River Basin and sampling protocols Sub-basin Geographic Coordinate Bottom types Elevation (m ASL) Distance from the sea (km) Water depth (m) Stream width (m) Collecting period Upping Ping (UP) 19°07’-19°48’ N 98°47’-99°17’ E G, P, R, S 684 r 228.3 1,026 r 24.1 0.4 r 0.2 7 r 0.5 Maetang (MT) 19°10’-19°45’ N 98°27’-98°55’ E G, P, R, S 756 r 166.2 1,067 r 36.0 0.6 r 0.4 13 r 11.2 2000-2001 2003-2004 The second Ping (SP) 18°31’-19°33’ N 98°24’-99°22’ E G, P, R, S 553 r 160.2 982 r 41.0 1.9 r 6.0 74 r 230.5 1996, 2003-2004, 2008 Maeklang (MK) 18°24’-18°35’ N 98°28’-98°41’ E G, P, R, S 1,070 r 213.4 877 r 4.6 0.3 r 0 11 r 5.4 2008 Maecheam (MC) 17°57’-19°09’ N 98°04’-98°37’ E G, P, R, S 627 r 207.3 927 r 53.9 0.7 r 0.4 21 r 17.4 2007-2008 The third Ping (TP) 17°48’-18°43’ N 98°14’-98°44’ E G, S, M 261 r 11.5 704 r 43.8 2.8 r 1.2 424 r 224.6 2005-2006, 2009 Maeteon (ME) 17°13’-18°02’ N 98°14’-98°34’ E G, P, R, S 804 r 229.2 847 r 43.0 0.5 r 0.2 8 r 3.8 2008 The forth Ping (FP) 15°50’-17°49’ N 98°39’-100°02’ E G, S, M 120 r 33.5 580 r 68.4 2.7 r 0.5 359r77.5 2009 Lower Ping (LP) 15°42’-16°10’ N 99°27’-100°08’ E G, S, M 48 r 8.0 425 r 16.0 3.2 r 1.3 258 r 27.7 2009 Wang river (WA) 17°07’-19°24’ N 99°00’-100°06’ E G, S, M 408 r 123.8 833 r 225.2 0.9 r 0.9 28 r 48.2 2009 2008 Note Bottom types: R = Rocky, G = Gravel, P = Pebble, S = Sandy, M = Muddy 26    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand The collection of fishes and environmental variables for P1 and P4 was conducted in the mainstem of Ping and Wang rivers as well as their asscociated tributaries. Various habitats found in the studied area are presented in Figure 4. There were a total of 272 sampling sites from 10 sub-basins (Fig. 5A) viz., upper Ping (UP), Maetang (MT), the second Ping (SP), Maeklang (MK), Maecheam (MC), the third Ping (TP), Maeteon (ME), the forth Ping (FP), lower Ping (LP), and Wang river (WA). The locations and fundamental geographical characteristics of each sub-basin are provided in Table 2. Figure 4 Various freshwater ecosystems found in Ping-Wang River basin. A: Waterfall and mountainous habitats, B: Shooting flow stream in first order stream, C: Secondary order stream with rock and gravel bed was located in mountainous stream, D: Secondary order stream with sandy bottom was located in lowland area, E: River mainstream located in lower part of PingWang rivers and F: Reservoir Note: Habitats A, B, C, D, and E were sampling area for P1 and P4; meanwhile data for P2 and P5 was from the reservoir. Habitats A and B were also the sampling area of P5. 27    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Figure 5 Maps of the studied area; A: The Ping-Wang basin and its sub-basins (P1 and P4), B: the Mae-ngad reservoir (P2 and P5), and C: Maecheam stream and the sampling sites (P3) Note: (a) abbreviations for the sub-basins: UP=upper Ping, MT=Maetang, SP=second Ping, MK=Meklang, MC=Maechaem, TP=the third Ping, ME=Maeteon, FP=the fourth Ping, LP=lower Ping, and WA=Wang river (b) abbreviations for the sampling sites in the Maengad reservoir: K= Huay Mekhod, P = Huay Mepang, S = Huay Mesoon, T = Huay Tontong, M = Huay Mekua, H = Huay Phakub, J = Huay Mejog, W = Huay Panwa, C = Huay Chompoo, L = lower part of the reservoir, U = upper part of the reservoir. 28    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand P2 and P5 were focused on the Mae-ngad reservoir, a small high land reservoir of upper Ping river in Chiangmai province (19º15.18 N, 099º 03.35 E to 19º 15.25 N, 099º 17.43 E, Fig. 5b). The dam is multi-purposes as hydropower and irrigation, and fisheries is a secondary benefit. Its elevation ranges from 412 to 425 m ASL with a catchment area of 1,309 km2, a water surface of 16 km2 and it can store water up to 265 million m3. It is dammed across the Mae-ngad stream, one of the first order stream tributaries of the Ping river basin. The maximum depth of the reservoir area is 30 m with a mixed clay and silt bottom. Meanwhile, the depth of the tributary streams, connected to the reservoir, ranges between 0.25-2.0 m and there are various bottom types (i.e. rock, gravel, sand, silt and mud) along the stream gradient. Fifteen sampling sites in the mountainous area of Maecheam first order stream (Fig. 5C) were selected for P3 to collect Oreoglanis siamensis. Maecheam stream locates in the west wing of Ping river and lies between 282 and 2,565 m from ASL. It is a major upper tributary sub-basin of the Ping river, which locates 117 km South-West from Chiangmai city. The Maechaem sub-basin is bounded by coordinates 18° 06’ - 19°10’ N and 98°04’ - 98°34’ E, and includes a total area of 3,853 km2. The climate of this mountainous basin is defined by large variations in seasonal and annual rainfall that are influenced by Pacific-born typhoons, superimposed on the south-west monsoon (Walker, 2002). The orographic effect induces an altitudinal increase of spatial rainfall distribution (Dairaku et al., 2000; Kuraji et al., 2001). The average annual temperature ranges from 20 to 34°C and the rainy season is from May to October. 2.2 Fish sampling For P1 and P4, the long-term database on fish distribution and environmental data was compiled during the ichthyological surveys in the Ping-Wang river-system between January 1996 and April 2009. The sampling sites were distributed among 10 sub-basins in the river-system, where a Digital Elevation Model (DEM) was used to define and divide the geographical range of the Ping-Wang river-system into subbasins by ArcView GIS 9.2, according to the catchment area and fish sample spots. Collections of fish samples were taken at every habitat types in every selected site. Samplings were done by various methods i.e. beach seine net, cast net, multi-mesh 29    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand gillnets as well as the electro-fishing with an AC shocker powered (Honda EM 650, DC 220 V 550BA 450VA, 1.5–2 A, 50 Hz), which was placed on the riverbank together with block nets and scoop nets. Sampling sites were chosen on the basis of accessibility, similarity in habitat types, and to maximize the diversity of habitat types (pools, cascade, falls, riffles, and stagnant water) at each sub-basin. The environmental parameters (Table 3) were measured by standard methods (APHA, 1991). All specimens were preserved in 10% formalin and then taxonomical classified, counted and measured at Maejo Aquatic Resources Natural Museum (MARNM), Chiangmai, Thailand. Table 3 Environmental parameters and methods of measurement in this study. No. Environmental Parameters Methods/Equipments Water qualities/Physicochemical parameters 1 Water temperature (WT; °C) YSI 556 (multi-probe system) 2 Conductivity (CON; mg/l) YSI 556 (multi-probe system) 3 Total dissolved solids (TDS; mg/l) YSI 556 (multi-probe system) 4 Dissolved oxygen (DO; mg/l) YSI 556 (multi-probe system) 5 Nitrite (NIT; mg/l) APHA (1989) protocols 6 Ammonia (AMM; mg/l) APHA (1989) protocols 7 Phosphorus (PHO; mg/l) APHA (1989) protocols 8 pH YSI 556 (multi-probe system) 9 Alkalinity (ALK; mg/l) APHA (1989) protocols 10 Hardness (HAR; mg/l) APHA (1989) protocols 11 Current velocity (CUR; m/s) Flow meter (G.O. Environmental model 1295) 12 Depth (DEP; m) Meter Tape 13 Width (WID; m) Meter Tape 3 14 Discharge (DIC; m /s) Q=AV; Area of channel X Average velocity of flow 15 Altitude (ALT; m ASL) GPS GarmineTrex VISTA Geo-morphometric parameters 16 Distance from the sea (DIS; km) 2 ArcView GIS 9.2 17 Watershed area (WSH; km ) ArcView GIS 9.2 18 Forest area (FOR; %) ArcView GIS 9.2 19 Agricultural area (AGR; %) ArcView GIS 9.2 20 Urban area (URB; %) ArcView GIS 9.2 For P2 and P5, data collection was conducted in the Mae-ngad reservoir. Fishes were sampled monthly from October 2002 to September 2003 from 10 sites in the 30    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand tributaries and 2 stations in the reservoir (Fig. 5B). Two stations in the reservoir were a littoral zone where most of fish occupied (Prchalová et al., 2003, Brosse et al., 2007). Meanwhile, the central area of the lake is a steep shore and very deep. Therefore, very few samples are expected. For P5, data was obtained by the 12 fishermans using gill nets and the targeted species was Henicorhynchus siamensis (Fig. 6). The gill net assemblies were composed of five 30 m2 nets (10 m long X 3 m deep) with stretched mesh sizes of 10-30 mm. The nets were surface-set at twelve sites, which were equally distributed over the coastal area of the reservoir, using one gill net assembly per sampling site. All the nets were set overnight between 16h00 and 18h00 and lifted between 06h00 and 08h00. At least 120 H. siamensis were randomly sampled monthly from July 2003 to June 2004 (1,364 fish in total). Individuals were measured for total length (L, to the nearest 1 mm) and weighed (W, to the nearest 0.1 g). For P5 Data collection was focused in the tributaries connected to the reservoir. Fish samplings were conducted by using electro-fishing, i.e. a gasoline-powered electroshocker (DC, 250 V, 1.5–2 A, 50 Hz), each sampling was done with two replications for 30 to 45 minutes interval and the area cover was about 100 m2. In addition, gill net (20 x 1.2 m2, mesh size 4 cm stretched mesh) was also concurrently conducted in reservoir during the night time. The water quality parameters (Table 3) were also recorded at each sampling station by the similar protocols as in P4. Lastly, P3, the study was conducted with Maechaem stream. Oreoglanis siamensis (Fig. 7) were sampled monthly from October 2007 to September 2008 from 15 sites in the East part tributaries of Maechaem stream. Fish samplings were conducted by electrofishing (Honda EM 650, DC 220 V 550BA 450VA, 1.5–2 A, 50 Hz) in the upper Maechaem river system. Each tributaries sampling site was done at 45 to 60 minutes intervals or the area covered was about 100 m2, I was collected with various microhabitat, substrate type i.e. rocky, sandy, and gravel, and habitat type (riffle, pool, and run) to cover all species distributions. The skin diving was carried out to observe the abundance and behavior of the fish. Fish captured in each part were kept separate after selected O. siamensis and fixed in 10% formalin and the life specimens was released to the their habitat after measurement and weight. Then, O. siamensis was identified and separated from the other species, sacrificed in a lethal 31    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Figure 6. The specimen of Henicorhynchus siamensis used to study in P2. (TL=215 mm) Figure 7. The specimen of Oreoglanis siamensis study in P3. A: Top view, B: lateral view, and C: Sucking mouth O. siamensis. (TL=108 mm). solution of anesthetic, and conditioned in ice for transportation. The process in evening at the rest room and the following data were obtained: (i) total length (TL) to the nearest 0.1 mm (ii) total weight (WT) to the nearest 0.01 g (iii) sex (iv) gonad 32    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand weight (GW) to the nearest 0.01 g. Gonads were removed from the visceral cavity, Prior to the preservation of the ovaries/testis were classified in a macroscopic scale of gonadal development, for both sexes; for female size and colour of oocytes was also registered and, for males sperm liberation when pressing the abdomen. According to these characteristics, the following classification was considered: females – 2nd stage, immature, mature, and ripe; and males – 2nd stage, immature, mature, and ripe. Thereafter, ovaries were fixed in Bouin solution for oocytes measurements and total ripe eggs counts. The specimens were fixed in 10% formalin and preserved in 70% ethanol. Specimens were deposited at the Maejo Aquatic Resources Natural Museum. 2.3 Data analyses 2.3.1 Diversity and abundance (P1, P4 and P5) Fishes were identified into species level by using various documents (e.g. Smith, 1945; Taki, 1974; Rainboth, 1996; Kottelat, 2001; Nelson, 2006). Ranks of individual species were presented as the percentages of relative abundance (%RA) and occurrence frequency (%OF). The diversity indices (Magurran, 2004) viz., species richness, Shannon-Wiener diversity index (H’-index) and evenness (J’) (Weaver & Shannon, 1949) were calculated for each sub-basin. Under the assumption that species richness increase with the sample size, I rarefied species richness to the same number of individuals and a rarefaction curve was used to estimate species richness in each sub-basin (Hulbert, 1971), and the rarefactions values (R) were computed by using EstimateS v. 8.2.0 (Colwell, 2009). R S ª i 1 ¬ § N  mi · § N · º ¸ / ¨ ¸ ---------- (3) n ¹ © n ¹»¼ ¦ «1  ¨© 2.3.2 Biology aspects, life history, and population dynamics (P2 and P3) The length (L) –weight (W) relationships W = aLb, of the two selected keystone species H. siamensis and O. siamensis, where estimated where a and b are specific values for each species. The relationship was done to examine whether the weight 33    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand increased proportionally with length, i.e. isometric growth, for each species. The length frequency distribution (LFD with 1.0 cm interval), for each species, was constructed for further analysis on the von Bertalnaffy’s growth function (VBGF). Reproductive biology was studied in the aspects of gonad development, gonadosomatic index (GSI), fecundity and length at 50% maturity (L50). Gonads (i.e. ovaries and testes) were collected monthly and the stages of gonad development were examined by mean of histological study, and fixed in 10% formalin/acetic acid/calcium chloride (FAACC) for 1 month before being embedded in paraffin and stained with haematoxylin-eosin. The samples were then cut into sections (7 Pm) and observed under a light microscope. The stages of maturity of the gonads were graded into 5 stages (I to V) (Bagenal & Braum, 1978), where fish that showed stage III and above were considered to be mature. Spawning season was estimared during the period following peak in GSI. GSI was calculated as (100 x Gonad- Weight » Body Weight). Stages IV and V ovaries were selected for fecundity examination by fixing in Gilson’s fluid, shaken vigorously and stored in the dark for at least a fortnight before the total egg numbers were estimated by sub-sampling using the gravimetric method (Bagenal & Brown, 1978). Then, the relationships between fecundity with length and weight were examined. L50 was estimated by using the logistic function (Chen & Paloheimo, 1994) as in Equation 4 P 1 1  e a bL ------------------ (4) where P is proportion of mature in each length clas; a and b are constants and when they were calculated, the percentage at 50% maturity was replaced in the equation (4) to obtain the length at 50% maturity. While, the condition factor (k) of the experimental fish was estimated from the relationship in the equation (5) (Williams, 2000): K  100W L3 .................................(5) 34    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand where K=condition factor, W= weight of fish (g), and L= length of fish (mm).   Fecundity (Bagenal & Braum, 1971) was determined after counting all vitellogenic oocytes from ripe ovaries and correlated to TL and TW in equation (6). F=aTLb, and F=aTWb ...................(6) where F=fecundity, TL=total length, TW=total weight, while a and b are specific constant values.  Feeeding of fish was studied by examining the stomach contents. Stomachs were dissected and opened longitudinally and the digestive tracts fixed in 10% formalin. The stomach contents were squeezed out and diluted to a 1 ml. The suspended matter was then placed into a Sedgewick rafter-counting cell and examined under light microscopy. The food items were identified to the lowest possible taxonomic unit. For diet preference analysis, the percentages of frequency of occurrence (O%), number (N%) and index of relative importance (IRI%; Equations 7 and 8) (Hyslop, 1980) were applied for each diet item (i). IRI i N % u O% ------------- (7) and IRI % § IRI ¨ i ¨ n IRI © ¦i 1 i · ¸ u 100 ---------- (8) ¸ ¹ Length frequency distribution (LFD) data was used for growth performance estimation. The von Bertalanffy growth function (VBGF) with seasonal oscillation (Equation 2) was used to express “size at age” of the two keystone species. Analyses were carried out by the free-package: Fish Stock Assessment Tools-II (FiSAT-II; (Gayanilo et al., 2002), which the steps for estimation were already described (Amarasinghe & De Silva, 1992). Theoretical age at length zero (t0) was derived from the equation (9) proposed by Pauly (1979): log10  t 0 0.392  0.275 log10 Lf  1.038 log10 K -------------- (9) 35    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand The age at the onset of the first growth oscillations (ts) was calculated as ts = WP 0.5, whereWP is the time of the year during which the growth rate is minimal, i.e. winter point (Gayanilo et al., 2002). The best-fitted growth curve was chosen on the basis of non-parametric scoring from the goodness of fit index (i.e. Rn value). 2.3.3 Statistical analyses and modeling methods Because of the non-normality of the data, the non-parametric Kruskal-Wallis test was used to test the significance of equality of medians among group of the diversity indices (P1). Relationships between diversity indices to the individual environmetal variables were examined by simple linear regression (P4), where environmental variables were treated as descriptors and the diversity indices were predictors. Moreover, in P4, the classification and regression tree (CART: Breiman et al., 1984), which is used to optimize set of environmental parameters and aimed at predicting diversity index, was also applied. For making CART, both response variables were log (x+1) transformed to stabilize variances. The optimal tree size was determined by R2-value and the complexity parameter. Generally, CART is called a classification tree if the response variable is qualitative (e.g. fish assemblages as in P5) and a regression tree if the response variable is quantitative (e.g. species richness as in P4) (He et al., 2010). Cluster analysis as the hierarchical agglomerative clustering by Ward's method (Ward, 1963) was used to classify sets of dissimilarities of the fish assemblages in sub-basins by using the number of individual species found in each sub-basin as inputs (P1). Two multivariate exploratory techniques were applied to explore the structure of categorical variables included in the studies and to identify systematic relations between variables. Firstly, a self-organizing map (SOM), which is an unsupervised algorithm of an artificial neural network (ANN) model (Kohonen, 2001) (Fig. 8). The SOM is widely applied in the last decade for solving problems in aquatic ecology, because it is capability of clustering, classification, estimation, prediction and data mining (Kalteh et al., 2008) Moreover, the SOM has proved to be an effective and powerful tool for describing species distributions and assemblages (Suryanarayana et al., 2008). The SOM consists of two layers viz. the input and 36    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Figure 8. The schematic figure showing the general modeling process in the studies of this thesis. output layers, which connected with the weight vectors. The input layer receives input values from the data matrix, whereas the output layer consists of output neurons, which displayed as a hexagonal lattice (Fig. 9) for better visualization. During the learning process, the SOM weights were modified to minimize the distance between weight and input vectors. The map (i.e. SOM, output layer) obtained after the learning process contains all the samples assignedto neurons. Generally, samples assigned to the same neurons, or to nearby neurons, are similar and samples assigned to distant neurons differ. Additionally, samples assigned to nearby neurons differ considerably if those neurons belonged to different clusters, which were identified with use of a hierarchical cluster analysis (Ward linkage, Euclidean distance). The detailed algorithm of the SOM can be found in (Lek and Guégan, 2000; Kohonen, 2001; Kalteh et al., 2008). The occurred probability of each species in each cluster can be approximately estimated during the learning process and seen in SOM, in which the gray scale gradient account for probabilities of occurrence, with dark corresponding to high probability and light vice versa (Park et al., 2005). The SOM was simulated and 37    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand performed by MATLAB (Ver. 6.1.0) by using SOM-toolbox, which developed by the Laboratory of Computer and Information Science (CIS), Helsinki University (http://www.cis.hut.fi/projects/  somtoolbox/). Figure 9 Representation of the non-supervised artificial neural network-SOM (Kohonen, 2001. The SOM was used for Factorial Correspondence Analysis (FCA) to investigate significant differences in the dietary components from the stomach contents between seasons (P2) and spatio-temporal variations in fish assemblage patterns in the connected tributaries to the reservoir (P5). Relationships between fish assemblages and environmental parameters (P4) were examined by Canonical Correspondence Analysis (CCA), an ordination technique designed for direct analysis of relationships between multivariate ecological data (Ter Braak, 1986). Statistical significance, for CCA, of the relationship between a set of environmental factors and fish species was taken using a Monte Carlo permutation test with 999 permutations and was accepted at P-value < 0.05. All the above analyses were analyzed by using Program R (R Development Core Team 2009) with various related packages, which were informed in each publication (i.e. P1-P5). 38    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand 3. MAIN RESULTS 3.1 Fish diversity and ecological parameters relationships in the upper Chao Phraya river basin (P1) The totals of 201 species were collected in 272 sampling sites in Ping-Wang river basin. The most dominants were Cypriniformes, Siluriformes and Perciformes, respectively. In terms of family, Cyprinidae was ranked first with 40.3 % (81 species), followed by Balitoridae and Cobitidae with 10.0 % and 6.5 % (20 and 13 species), respectively. Among the genera, Schistura in family Balitoridae was as most diverse in species. The number of genera and number of species ratio were found 1: 1.93. The five most abundant species accounted for 32.4 %RA of total fish collected. The highest %OF was found in Channa gachua (47.1 %). Some species that showed at a high level in number but low in %OF indicated their restricted distribution e.g. Devario maetangensis. Also the economic aquaculture fishes were escaped or releasing into the river and/or reservoir e.g. Oreochromis niloticus and Clarias hybrid. Figure 10. Species richness in each sub-basins of the Ping-Wang river basin. Abbreviation: UP=Upper Ping, MT=Maetang, SP=the Second Ping, MK=Maeklang, MC=Maecheam, TP=the third Ping, ME=Maeteon, FP=the forth Ping, LP=Lower Ping, and WA=Wang river. 39    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Large sub-basins showed high H’, i.e. LP, FP and ME and the abundance of individual species were in similar proportions (Fig. 10). Although lower in H’ index in the upper sub-basins (i.e. MK and UP), these sub-basins were characterized by the endemic species, in which the rate of species that are restricted to the basin was up to 10 % i.e. Devario maetangensis, Schistura pridii, Oreoglanis siamensis, and Rhinogobius chiengmaiensis. Five IUCN fish species were also collected, i.e. O. siamensis, Himantura signifier, H. Chao Phraya, P. gigas and Pangasius sanitwongsei. Sixteen exotic species were found in all sub-basins, except the upper reach of the Ping River (UP). Among them, Gambusia affinis was the highest % OF at 17.3 % followed by Oreochromis niloticus (| 9 %OF). Species richness gradually increased from the upper part to the lowland area and the Kruskal-Wallis test showed that the significantly differentiated among subbasins. All the ten rarefaction curves for the sub-basins showed signs of reaching asymptotic levels. Adequacy of sampling was assessed also by the rarefaction curve and the asymptote was reach at about 250 species, confirming that the number of sampling sites in this study was satisfactory (Fig. 11). Figure 11. Rarefaction curves plot by number of species with number of sampling sites for the Ping river basin. 40    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand 3.2 Some aspects of life history and population dynamics of lotic and lentic tropical fish species. 3.2.1 The life history of the riverine cyprinid Henicorhynchus Siamensis (Sauvage, 1881) in a small reservoir (P2). The riverine species, Henicorhynchus siamensis (Sauvage 1881), is an important source of protein and an economical fish for the rural population of inland Indochina. The moderate species was 290 mm in maximum sized. Investigated in the present study were the reproductive feeding aspects and growth of H. siamensis living in a small reservoir. The equation derived was W = 0.01L3.08 (r = 0.82) and the exponential value indicated that the growth was isometric. The histology of the gonads confirmed that H. siamensis has a synchronous ovary. The temporal changes in the gonadosomatic index (GSI) clearly showed a single peak in both sexes, which tended to increase in June, was highest in August (Fig. 12). The individuals were taken 1.5 years to attain the length of 50% maturity of female and male were 197.6 and 201.6 mm (Fig. 13). Fecundity ranged widely was 105,782±59,930 eggs; it was depended on the length. Relative fecundity was 1,034±116 eggs per gram of body weight; the relationship between length and fecundity (Fe) can be described by an empirical power equation: Fe = 21,141L3.087 (r = 0.762, n = 171). Figure 12. Gonadosomatic Index (GSI) of of male and female H. siamensis in Maengud reservoir (July 2003 to Jun 2004). 41    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand 100 Percentage Maturity 90 80 70 60 50 40 Male 30 Female 20 10 0 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 T otal Length (mm) Figure 13. Percentage maturities of male and female H. siamensis against of fish in Mae-ngud reservoir. Stomach contents were dominated by phytoplankton, such as Cyclotella sp., Melosira varians and Navicula sp. (IRI% = 21.55, 18.20 and 12.58, respectively). They were found to be the main dietary components. The factorial analysis of the temporal variation in the diet based on each month’s sampling could be clearly divided into three main groups. The first group (group I) was found during the early rainy season (June and July), and dominated by Chlorophyta e.g. Crucigenia crucifera (Cruc). Group II was during the winter (December), characterized by few dietary items and low species diversity, and dominated by Cryptophyta, e.g. Chilomonas sp. (Chil). The third group was the most complex of phytoplankton; dominated by Staurastrum sp. (Stau), and Cyclotella sp. (Cycl) (Fig. 14). The growth curve that gave the highest goodness of fit index was selected. A clear seasonally oscillating growth pattern implies that the species is sensitive to seasonal variation and that recruitment started in July. The winter point (WP) was 0.95, which signifies that growth slowed during December. The growth performance (Ø) index was 4.72. From the derived growth parameters, H. siamensis attains at least 50% of the asymptotic length of H. siamensis was 264.2 mm, with a 0.75 year1 42    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand growth coefficient and approaches Lv at about 3.5 years of age. The potential longevity, 3 » K (Pauly and Munro, 1984), of H. siamensis was estimated at 4 years. Figure 14. Composition and seasonal variability of H. siamensis feeding. 3.2.2 Conservation approaches and reproductive biology of vulnerable stream Sisorids (Oreoglanis siamensis) from foothill Himalayan, Thailand (P3). A vulnerable and an endemic Freshwater batfish (Oreoglanis siamensis) were studied in 2006-2007 in a high mountain stream in northern Thailand (18° 06’ 19°10’ N and 98°04’ - 98°34’ E). This species was examined for reproductive biology preferences. Spawning in freshwater batfish occurred in late dry-cool season to early dry-hot season (January to March) in the Maechaem river basin; at least 87.1-95.7% of female were in ripe or spawning condition in this season (Fig. 15), while the sperm of male was mature and ripe through the year (Fig. 16). Size at first maturity was 47 mm for males, and 53 for females. L50 estimates were 68.9 ± 1.765 mm (males) and 82.4 ± 1.369 mm (females). Maximum fecundity was 47 oocytes. Fecundity (F) varied from 18-47 (31.41 ± 7.67) for ripe females of 53-113 mm, respectively, 43    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand correlation between TL and F, and W and F followed a linear relationship (F = 7.14+0.38TL; r2 = 0.424; or F = 20.41+2.3W; r2 = 0.491; n = 71). O. siamensis is a large size of eggs (Fig. 17). Then, ripe oocytes have mean diameter of 2.96 ± 0.28 mm (range = 2.5-4.2 mm; n=30). Siamese bat catfish could not clearly express the secondary sexual characteristic, it was difficult to distinguishable except during the spawning season. The sex ratio ZDV Ȥ2-test, p<0.05). Figure 15. Percentage frequency of maturity egg stage of O. siamensis. Figure 16. A: Whole ovary (length 14 mm.), B: mature stage of ovary. Abrreviations: Nu = nucleaus, FE=follicle epithelial, YG=Yolk granule, and FV = follicle vesicle. 44    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand A B Figure 17. Histological appearance of ovary (A) and testis (B) maturation of O. siamensis (n=9 per month). Abbreviation (A): Nu=nucleus, FE=follicle epithelial, YG=Yolk vesicle; Note a) ripe and spent stage (dry-hot season), a1) late ripe stage, a2) spent stage, a3) spent stage, b) late spent, primary stage and immature stage (rainy season), b1) late spent stage, b23) primary stage and immature stage, and c) mature and ripe stage (drycool season), c1-2) mature stage, c3 ripe stage. Abbreviation (B): SP=spermatozoa, Note a) ripe and spent stage (dry-hot season), a1) mature stage, a2) mature stage, a3) mature stage, b) late mature, primary stage and immature stage (rainy season), b1) late mature stage, b2-3) primary stage and immature stage, and c) mature stage (dry-cool season), c1-3) mature stage, c3 ripe stage. 45    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand 3.3 Fish assemblages and impacts of environmental factors 3.3.1 Fish diversity and assemblage patterns in a rhitral environment of the Indo-Burma region (the Ping-Wang River Basin, Thailand) (P4). One hundred and ninety eight species within 11 orders, 33 families were collected in the P4. The most diverse family was Cyprinidae, followed by Balitoridae, and Cobitidae. The highest species richness, Shannon diversity index and species evenness were found in the lower part of the river-system meanwhile the minimum species richness was obtain at high altitude (Fig. 18A, B). But, the numbers of individual were scattered among sub-basin. Only six physicochemical parameters from 20 environmental parameters were obtained, i.e. DO, water temperature, pH, conductivity, phosphorus and alkalinity showed statistically significant in their relationships to diversity parameters. However, due to extensive and high variation of the obtained data, all the linear models showed low power in prediction. And five geo-morphological parameters, i.e. altitude, distance to the sea, discharge, depth, and width, showed highly statistical significances in their relationships to diversity parameters. The diversity index and species richness of tropical fishes was depending on altitude, water depth, stream width, and distance from the sea (Fig. 18). Altitude and distance to the sea showed strongly negative relationships and the relationships trended to be exponential for both indices, implying that higher diversity was found in the lower altitude, in which closed to the sea and then sharply decline as the altitude increase (Fig. 18). Species richness and Shannon diversity index of each individual were sampling ranged. They were fed to CART model as a response variable by using 20 environmental predictors. The geo-morphological parameters were the major factors in determining both diversity indices. For species richness, 3 parameters were included in the CART model and altitude was the major contributor in predicting species richness followed by width and distance from the sea (Fig. 19). 46    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Figure 18. Some of the relations between diversity parameters and environmenatal parameters. A and B: The relation between altitude and Shannon-Weiner Diversity Index and species richness. C and D: the relation between distance from the sea (km) and Shannon-Weiner Diversity Index and species richness. E and F: The relation between water depth and ShannonWeiner Diversity Index and species richness. The relationships of fish assemblages and environmental parameters were loaded fifty three fish species and twenty environmental variables in the CCA analysis. The first CCA environmental axis (CCA1) was described by altitude, distance from the sea, water depth, stream width and water temperature of the basin. The first two parameters were negative correlated to CCA1 meanwhile the remaining parameters were vice versa. The most important variable for the second CCA environmental axis (CCA2) was watershed area. Composition of individual fish species, which related to the environmental vectors loaded to CCA, was shown the 47    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand first five species with have strong positive loading to CCA1 and CCA2 e.g. Pristolepis fasciatus, Barbonymus altus and Lepidocephalichthys hasselti. Figure 19. CART model to predict species richness in The Ping-Wang river basin. Figure 20. Dendrogram of fish assemblages in the Ping-Wang River Basin. 48    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Distribution of fish species along the CCA axes can be classified into 4 main assemblage patterns. The first assemblage (quadrant I) was inhibited in the mountainous area of high altitude with relative low temperature and strong current velocity. The second assemblage was the shorter distance from CCA1 (quadrant II) indicated that the fish in this assemblage occupied in the lower altitude then those in the first assemblage. The remaining two assemblage patterns were positively correlated to CCA1 (quadrants III and IV) and implying that the fishes in these assemblages live in the lower portion of the river course, where the river width and depth were more than the previous two assemblages. Figure 21. Cluster dendrogram summarizing similarity among sub-basins based on their fish assemblages and environmental parameters (plot by CCA site constraints values (linear combination of constraining variables)). Using Ward model (Dendrogram General tree structures). Ward’s analysis was used to refine the habitat preference of individual species after the trends from CCA analysis. In quadrant I, inhabited in the small streams in high altitude area with low temperature, were grouped together and defined as “mountainous” species e.g. Oreoglanis siamensis (Osia), Devario regina (Dreg), Exostoma vinciguerrae (Evin). The remaining species in quadrant I, which located closed to CCA1 axis, and all species in quadrant II were grouped and defined as 49    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand “piedmont” species e.g. Lepidocephalichthys hasseltii (Lhas), and Channa gachua (Cgac). Species, which positively correlated to CCA1, were divided into two groups. Firstly, the species, which located closed to CCA1, were defined as “transitory” species e.g. Puntius orphoides (Porp), Mastacembelus armatus (Marm), Puntius brevis (Pbre) and Mystacoleucus marginatus (Mmar). Secondly, the group of species that showed the highest positives loading to CCA1 and defined as “lowland” species e.g. Pristolepis fasciatus (Pfas), Barbonymus altus (Balt), and Mystus singaringan (Msin) (Fig. 20). The sub-basins were shown the difference, based on their fish assemblages and environmental parameters (Fig. 21). 3.3.2 Fish communities in the highland tropical streams connected to a reservoir (P5). Species composition of fishes and community assemblages can be changed after the change of ecosystem. Sixty-six species were collected; and dominated by Cyprinidae (34.9 %), Balitoridae and Cobitidae (10.6 %). Invertivores, carnivores and herbivores dominated the trophic guilds, respectively. The highest percentage of relative abundance (%RA) were Henicorhynchus siamensis (Sauvage, 1881), Mystacoleucus marginatus (Valenciennes, 1842) and Puntioplites proctozysron (Bleeker, 1865). The highest percentages of occurrence frequency (%OF) were shown by M. marginatus, Oxyeleotris marmorata (Bleeker, 1852), and Hampala macrolepidota Kuhl & Van Hasselt, 1823. According to the nature of the surveys found in each community, the communities can be designated into reservoir community (RC), stream community (SC), and intermediate community (IC) (Fig. 22), in which there were highly significant variations in the community structures among communities (ANOSIM, R=0.757, P<0.001). The movements of fishes were migrated in difference of seasonal or the stage of life during the year. The distributions of occurrence probability (OP) of individual species of each species in each community can be expressed as the community characteristics was arbitrarily set to show the dominant species in each community but two species gave the highest OP of all communities i.e. M. marginatus and O. marmorata. The highest 50    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand OP in SC was Rasbora paviana (Tirant, 1885) and Channa gachua (Hamilton, 1822), IC was dominated by Cyclocheilichthys armatus (Valenciennes, 1842), Barilius koratensis (Smith, 1931), and Garra cambodgiensis (Tirant, 1883). It is also worthy to note that the dominant species in the SC and IC were either invertivores or carnivores. Meanwhile, the RC was dominated by a number of species that were mostly herbivores e.g. Labiobarbus lineatus, (Sauvage, 1878), P. proctozysron, and H. siamensis, except for Hampala macrolepidota is a carnivorous cyprinids (Fig. 23). Figure 22. Distribution of surveys based on the SOM map according to the similarity of fish composition. The prediction of community assemblages and the contribution of environmental variables were shown the average values of the physicochemical and geo-morphological variables, obtained from the three communities. Also, they were used as predictors in the CART model to discriminate the clusters of fish communities. Based on the communities and environmental variables were selected to predict the response variables. The major variables corresponding to assemblages were water depth, which separated the RC from the other communities. Meanwhile, 51    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand the overlaps between the IC and SC were distinguished by physicochemical variables such as hardness, ammonia, alkalinity, orthophosphate and nitrite. The overall predictive power of this model was successfully the model could predict the assigned survey to the right community. Figure 23. Community characteristics for each cluster as shown by the occurrence probability. Abbr. SC=stream community, IC=Intermediate community, RC= Reservoir community. 52    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand 4. GENERAL DISCUSSION 4.1 Fish diversity and assemblage patterns in a rhitral environment. The Ping-Wang river basin contained 201 fish species and the rarefaction curves showed that it was close to the maximum number of species occurring in the Basin. Thus, it is shown that the upper Chao Phrya river system is a high in fish species richness and the total number of species in Chao Phraya river is 420 species (Table 2), except the Mekong (1,200 species), the Yangtze (China, 320 species), the Cauvery river (India and Nepal, 265 species) and the Kapus (Indonesia, 250 species) (Nguyen & De Silva, 2006). Normally, dominance by the multi-species and ecologically diverse Cyprinidae is common in Southeast Asia, where they may contribute 40% or more of the species in a watershed (Taki 1978; Kottelat & Whitten, 1996; Ward-Campbell et al., 2005; Beamish et al., 2006; Beamish et al., 2008), and follow by the rhithronic species i.e. stream Sisorids (Oreoglanis siamensis, Exostoma vinciguerrae, and Glyptothorax spp.) and Balitorids (Schistura spp. and Homaloptera spp.) (Vidthayanon, 2003; Vidthayanon et al., 2009; Hu & Zhang, 2010; Ng, 2010). This is because the rhithronic species have evolved partially through highly adapted body forms and mouth structures so they occupy virtually all habitats throughout their distributions (Ward-Campbell et al., 2005). The species richness and the H’ index were increased from the upper to the lower part of the basin. Moreover, deeper water, wider rivers and more discharges downstream are factors to increase diversity parameters (Horwitz, 1978). Low species richness in the high altitude reflects the low variability of food supplies (Tongnunui & Beamish, 2009), sub-basin size, i.e. the larger the sub-basin, but not in the study of Nguyen & De Silva (2006), the higher the species richness and the richness of nutrients, which increased natural food sources, as well as flood pulse effect in the lowland (Junk & Wantzen, 2004). Altitude and distance from the sea were found to be among of the most key factors that govern the species richness and fish community structures in riverine ecosystem (Oberdorff et al., 1995; Welcomme et al., 2006). Residents in this area are always generally small in size and equipped with suckers or adhesive apparatuses 53    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand (Bhatt & Pande, 1991), and may also have streamlined, sucking mouth or flattened forms such as Schistura spp., G. cambodgiensis, and O. siamensis which were the dominated in this study and none of them were found in the lowland sub-basins. Two out of five of IUCN and endemic fish species i.e. Rhinogobius chiengmaiensis and O. siamensis have been found so far in the Ping-Wang river basin, the uniqueness of the area, and endemism of fish should be of concern. Deforestation for agricultural purpose, indiscriminate fish collection for aquarium fish in the upper basin, and urbanization in the lowland area are among other major threats for the rhithronic habitants. Moreover, since most of the rhithronic members are mostly insectivorous (Tongnunui & Beamish, 2009), habitat degradation in the rhitral area, could lead to a decline in exogenous food sources including insects as well as their larvae (Raghavan et al., 2008). Nile tilapia, Chinese carps, Indian major carps as an exotic species were introduced for food enhancement purposes found and no impacts are reported so far. The presence of the Poeciliids fish as Gambusia affinis, Xiphophorus helleri and Poecilia reticulata in the upstream part must be of concern since they prey on aquatic insect larvae or stream fish larvae and often with devastating consequences (Mills et al., 2004; Vitule et al., 2009). Gambusia affinis was an invasive species to local species in many areas (Rehage et al., 2005) and it is aggressive foragers, feeding on a variety of prey, including the eggs, fry and larvae of native biota (Goodell & Kats, 1999; Garcia-Berthou, 1999). Na-Nakorn et al. (2004) mentioned that Clarias macrocephalus and C. batrachus in the wild might be directly replaced with the C. hybrid, which have a higher growth rate. Thai traditional ceremonies were released e.g. Clarias hybrid, Pterygoplichthys spp., and many species into the main stream for their lucky life, but the fishes would be negative impacts to the native species (Chaichana et al., 2010). 4.2 Life history facts, biology, and population of riverine keystone species. The change of habitat from river to lake showed variations in their spawning characteristics Türkmen et al., (2002). It was also observed that the spawning characteristics of fish of the same length, living in places with different ecological 54    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand features, but belonging to the same species, had some variations. The Henicorhynchus siamensis in the Mekong mainstream was mature and reproduce within the first year of their life van Zalinge et al., (2004). In the current study, the length at 50% maturity of H. siamensis was attained after one and a half years (i.e. about 190 mm). The H. siamensis can develop their gonads as early as the late dry season around March to April and peak during May and June (Sokheng et al., 1999), but in this study the GSI of H. siamensis began to develop at the beginning of rainy season. Viravong (2006) reported that the H. siamensis population above the Great Khone Falls spawns earlier than the population below the falls. This is possible because of the floods and the rain occured earlier there. The average GSI increased from low values in the dry season to a maximum of about 20% immediately before spawning (July to August), but the GSI was slightly higher for the river-dwelling H. siamensis (Viravong, 2006). As a member of the littoral community, H. siamensis is known to be a mainly plant and detritus feeder with the trophic level range between 2.0-2.19. In a newly impounded condition, which is rich in nutrients and with a dominance of planktonic algae, H. siamensis was shown to be restricted to a phytoplankton feeder (i.e. trophic level range is equal 2.0: Thapanand et al., 2009). The presence of a few zooplankton in stomach contents of H. siamensis in this study would put its trophic level slightly higher than 2.0. Feeding mostly on phytoplankton and plant materials, which have a low energetic value, means that H. siamensis consumes a large quantity of food and has a long feeding period during the daytime (Amarasinghe et al., 2008). The growth of H. siamensis in this study provided excellent data that could be used for simple length-based analysis (Hoenig et al., 1987). Moreover, the modal groups detectable from the raw data with the apparent shifts in the modal length over time make the results of the study reliable (Ama-Abasi et al., 2004). The growth performance index (ĭ¶) is a species-specific parameter to indicate the unreliability in the accuracy of estimated growth parameters (Pauly & Munro, 1984). The ĭ¶ of the present study (4.19) was close to (F2 test, P-value > 0.05) the value from a large reservoir (4.75: Moreau et al., 2008), which meant that the estimated growth parameters were authentic. The high amplitude of oscillation (C = 0.8) of H. siamensis indicated that growth does not completely cease but slows down during the unfavorable period (i.e. during December), which could due to diet items, which were 55    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand the lowest in number and diversity, and also, low water temperature. Similar results were obtained from a large reservoir, where C = 0.6 and WP = 0.95 (Moreau et al., 2008). This situation is also likely to occur in the Mekong mainstream where there is a drastic decline in temperature during November to January (Prathumratana et al., 2008). H. siamensis has shown well adapted to the lentic system, the piscimetric values on its biological traits showed to be lower than those in the lotic system. This phenomenon would relate to different in the flood pulse between the two systems. Tonle Sap population condition factor was remarkably increased during the flood season (Lamberts, 2001) as like as other Cyprinids (De Graaf, 2003). Meanwhile there is less variation in flood pulse in the regulated lake (Wantzen, et al., 2008), where the hydrological regimes are almost entirely dependent on the rainfall in the catchment areas and the demand for water for primary uses (Nissanka et al., 2000). Nevertheless, Mattson & Kaunda (1997) mentioned that the small reservoir environment is similar to a river floodplain, with large fluctuations in temperature, oxygen concentration, turbidity and water level, which are suitable to enable fish of river origin to adapt to the new environment. Moreover, the reproductive traits of H. siamensis such as early maturity, high fecundity, single broods and rapid egg and larval development would help them be successful in unfavorable environments (Viravong, 2006). The r-strategist with foraging behavior also makes H. siamensis a good candidate for maintaining the population in higher trophic levels in the lake similar to the case of the Thai river sprat (Clupeichthys aesarnensis Wongratana, 1983) into numbers of reservoirs in the LMB (Jutagate et al., 2003). The northern Thai’s stream, like on many other tropical streams, are characterized by a steep topography, fast flow, rocky bottom, canopy cover, and high level of dissolved oxygen. Nevertheless, the fish still have to well adapt to the special habitat e.g. Homaloptera spp., Balitora spp., and Glyptothorax spp. (Kottelat, 2001). Also, O. siamensis was well adapted by flatten belly, adhesive maxillary barbel and pair fins; streamline body shape, and aerodynamic dorsal part. These characteristics were suitable to feed on the small invertebrate and aquatic insect larvae on the rocks (Vidthayanon, 2005). It could tolerate a low water temperature in high altitude might limit the growth of the O. siamensis food items (Han et. al., 2000). The environmental 56    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand condition of O. siamensis was abundance along the habitats in Maechaem Stream showed that O. siamensis inhabited the waters between 500 to 1200 m altitudes. The early stages such as eggs and larvae stages are the great important for fish, then the reproductive tactics in teleostean fish involving the allocation of a sizedependent reproductive effort between fecundity and egg size. The demersal species tend to produce large and few eggs, the larger eggs and the larvae hatching from them are more likely to survive than smaller ones, but Duarte & Alcaraz, 1989 reported no evidence of evolutionary trends towards greater eggs. They were reduce the variance in growing conditions, should be more dependent on the survival of the individual larvae, which increases as egg size increases. Also, O. siamensis is a demersal steam species, it was produce large oocytes and few numbers like some of parental care species e.g. Xynobagrus nigri (Olurin & Odeyemi, 2010) and Notopterus notopterus (per se A. Suvarnaraksha) or rainbow trout, Sea back trout, and brook trout (Serezli et al., 2010). While, their fecundities were very small number of eggs compare with the other glyptothorine species e.g. Glyptothorax madraspatanum (18 to 47 vs. 1640 to 6830) (Dobriyal & Singh, 1993). The fecundity and egg size were related, egg size is one of the important determinants of eggs and larval quantity as it is positively correlated with both survival of egg and larval and also of the growth of the larvae (Gall, 1975). The adults and juvenile were found in the same habitats, it is possible a non-migratory species. The O. siamensis was spawn in the late dry-cool to dry-hot season (January to April) in Thailand, it conformed to study of Unsrisong et al., 2005, but a little bit early. Meanwhile, it was different to other lowland tropical stream species reproduction according to rainfalls regiems (Alkins-Koo, 2000; Chellappa et al., 2009). In the dry season, reduced stream flow and a reduced spate frequency ensure a more benign physical environment than during the wet season, and specific food for larvae may also be more abundant at this time also. Moreover, wet-season primary production may be reduced because of a combination of increased cloud cover associated with the monsoonal wet season and high suspended sediment loads during periods of elevated discharge, both of which limit light availability for primary producers (Pusey et al., 2001). The main habitats were in the the high elevation and canopy cover. The spawning season sufficient data on seasonal freshwater fish egg 57    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand variations are not available, but the time of spawning does appear to be linked with the availability of food for the larvae in both lake and stream species (Bagenal, 1971). Then, the few number of eggs and restrict to habitat of O. siamensis was lead to endanger or extinct in near by future. 4.3 Fish assemblage in lotic and lentic ecosystems of Ping-Wang River Basins. The complexity and non-linearity of the relations between the fish communities and their environment are very common (Gevrey et al., 2003) as shown by the low relationship values from the linear correlations for all environmental variables to both diversity indices. For prediction of fish diversity, it was found that geo-morphological parameters were the good predictors for species richness and Shannon diversity index compared to those physicochemical parameters. Changes in both diversity indices follow the general longitudinal pattern of river fish distribution as the lowest levels tend to be found at high altitudes, and the highest levels at mid to low altitudes (Gaston & Blackburn, 2000; He et al., 2010). The larger watershed, which suggests larger areas of habitat generally contain more species than smaller areas (Angermeier & Schlosser 1989; Han et al., 2008), the effects of land uses on fish community structure (Orrego et al., 2009; Alexandre et al., 2010), and also, which shows robust positive relationships to species richness (Connor & McCoy, 1979; Angermeier & Karr, 1983). However, due to the fact that most of the areas in this basin is intact and less disturbed by urbanization, in this study, reflected the longitudinal river gradient, which  is closely related to the gradual change in habitat diversity (Ferreira & Petrere, 2009; He et al., 2010). The physicochemical parameters would be important to fish species richness and abundance in a relatively drainage system (Oberdorff et al., 1995; Guégan et al. 1998; Tongnunui & Beamish, 2009; Alexandre et al., 2010). The summary diagram of lotic environmental parameters and diversity parameters relationships were shown in the thesis e.g. distance from the sea, altitude, and dissolved oxygen were negative relationship to diversity parameters (Fig. 24). Meanwhile, the summary diagram of the lentic environmental parameters and diversity parameters relationships were shown in Figure 25. From this study, 58    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand environmental factors must have an optimal level for aquatic organisms e.g. the inverse sigmoid curve of Figure 24 was slowly decreased, and/or rapidly descend in particular session (non scale). Figure 24. Summary diagram of relationships of environmental parameters and diversity parameters in lotic condition of this thesis (non scale). Diversity parameters are H’=Shannon Weiner diversity index, S=species richness, D=Simpson Dominance index and J=Species evenness. Environmental parameters are WT, WID, WSH, pH, PHO, DEP, DIC, AGR, ALK, DFS, ALT, and DO. Abbreviation: 1. Negative relationship to diversity parameters group i.e. DFS=distance from the sea, ALT=altitude, DO=dissolved oxygen, 2. positive relationship to diversity parameters group; WID=width, DIC=discharge, DEP=depth, pH=per hydrogen, PHO=phosphorus, AGR=agricultural area, ALK=alkaline. Distinct patterns of fish assemblages along the longitudinal river gradient reflects the homogenous spatial units within the river basin (Welcomme et al., 2006; Ferreira & Petrere, 2009) and the results from ordination and classification showed four fish assemblage patterns from the headwater to lowland river reaches: mountainous, piedmont, transitory and lowland assemblages. The assemblage of mountainous species showed their restricted occurrence in a high altitude area, with associated riffles and rapids, there were adapted their morphological for survive in the strong 59    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand flow conditions (Casatti & Castro, 2006). Assemblage diversity in the piedmont could also be explained by the potentially large number of modes of exploitation of resources, corresponding with highly differentiated patterns of habitat use, i.e. Competitive Exclusion Principle, CEP (Herder & Freyhof, 2006). The CEP theory also supports the assemblage pattern of the “transitory” species, where various habitats are also found and rheophilous cyprinid always dominate (Allouche, 2002) and also, was used to describe the lowland assemblage (Rainboth 1996; Kottelat, 1998), where the lentic cyprinids and other limnophilic fishes dominated (Allouche, 2002; Beamish et al., 2006). However, upstream movement of some lowland species is sometimes observed especially for reproduction (Silva & Davies, 1986; Ferreira & Petrere, 2009; Tongnunui & Beamish, 2009). This phenomenon supports the pattern of species addition for the shifting in species composition (Huet, 1959; Petry & Schulz, 2006). Figure 25. Summary diagram of relationships of environmental parameters and diversity parameters in reservoir of this study (non scale). Environmental parameters are DO=dissolved oxygen, WID=width, AMM=ammonia, DEP=depth, pH=per hydrogen, and CON=conductivity, while; diversity parameters i.e. H’=Shannon Weiner diversity index, S=species richness, D=Simpson Dominance index and J=Species evenness.  Species composition of fishes (Özcan & Balik, 2009)   and community assemblages were changed after the change of ecosystem. The tropical Southeast Asia 60    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand river basins fish species are dominated by cyprinids followed by silurids (MatinSmith & Tan, 1998; Campbell, et al 2006; Nguyen & De Silva 2006) but being followed by the Balitoridae and Cobitidae, as in this study, is unique for the stream areas in the region (Kottelat, 1998). Henicorhynchus siamensis was the highest percentage of relative abundance. Differences in the observed communities can be provided as an assessment of the ecological status (Lasne et al. 2007). The three communities, i.e. stream- (SC), reservoir- (RC), and intermediate community (IC) were divided the fish groups, where the hydrological regime was the major factor controlling fish community patterns (Welcomme & Halls 2005). The fish communities were different in the dry cold and dry hot season, which coincided that made the difference between the RC and SC. The water surface of the reservoir was increase in during rainy season, this case also improves the connectivity between the reservoir and the tributaries and that increases the aquatic biodiversity (Amoros & Bornette, 2002; Falke & Gido, 2006) as seen in the results in the intermediate community (IC). The variation in the occurrence probability (OP) of individual species in each cluster indicated the preferred habitat of the species. In the SC, the members were mostly rheophilic species e.g. Barilius koratensis, and Garra cambodgiensis, commonly found in small to medium-sized streams in upland areas (Kottelat, 1998), they were sensitive to catastrophic and habitat flows (Welcomme et al., 2006). Meanwhile C. gachua lives in the backwaters of first order streams (Taki, 1978) and R. paviana is usually found in shallow and moderately flowing streams (Kottelat, 1998). In the RC, the species found were the lentic-adapted species e.g. Henicorhynchus siamensis and Labiobarbus lineatus, the so called “facultative reservoir species”: they are generally native to the lower portions of a river course (Falke and Gido, 2006). Also, variations and high overlaps among communities could be due to some species moving in and out of the tributaries during their life cycle (Borcherding et al., 2002). For example, Henicorhynchus siamensis migrate upstream annually to spawn on shallow gravel beds at the confluence or in small rivers during short periods in rainy season (Sokheng et al., 1999; de Graaf et al., 2005). This is why these fish also showed sample OP in the IC. Meanwhile, high OP in all communities 61    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand of O. marmorata and M. marginatus could be caused by movement either for feeding or spawning purposes (Kottelat, 1998). The community structure in the headwater depends on abiotic- rather than biotic- factors (Schlosser, 1987). Prchalová et al., (2009) mentioned that the complexity of species composition in a reservoir, increased heading towards the tributary and peaked close to or at the tributary part of reservoir, which agreed with our results obtained for the complexity of the OP in the IC. Other selected variables in the CART to discriminate between the SC and IC were related to the major nutrients in the ecosystem i.e. phosphorus and nitrogen, both nutrients always increase during the rainy season and are released from upstream to downstream as well as from the land to the water body and then stimulate primary productivity in the ecosystem (Allen, 2001; Wondie et al., 2007). This phenomenon is eventually made more complex in the fish community in the area, at least for feeding purpose (Hoeinghaus et al., 2008). The one hundred percent predictive power for the RC indicated that the community assemblages in that area were relatively stable, while the low predictive power for the SC implied the movement of downstream species into the stream (Grossman et al., 1990). 62    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand 5. CONCLUSION This study was the investigation of the taxonomic level, biology and life history, and ecological approaches of the keystone species in the lotic and lentic waters in the upper Chao Phraya river   basin. The lack of fish diversities was reported in case of the Indo-Burma hotspot (Southeast Asia), especially upper Chao Phraya river basin. Fourteen years of field dataset in the Basin were used, which covered 272 surveys of 10 sub-river basins, collected between January 1996 and April 2009 to perform species richness and diversity indices. Twenty physicochemical water quality- and geo-morphological- parameters were also examined at each sampling. Similarities among sub-basins were examined by Ward‘s agglomerative method. Rarefaction was employed to extrapolate species richness and optimum numbers of the surveys. The longitudinal distribution in lotic conditions of fish was presents information of fish diversity and distribution in a unique high altitude mountain (Inthanon highest point of Thailand 2,565 masl, Chiangmai province) to lowland area (Nakornsawan province, the end of Ping-Wang River Basin 40 masl). Two hundred and one species in 104 genera and 34 families were collected, including 16 exotic species. The Cyprinidae (76 species) was dominated families, followed by Balitoridae (20 species) and Cobitidae (13 species), implying the characteristic of high altitude area. The overall endemism in the area was found at about 10%. Ward‘s method showed distinct differences between the upper- and lowland sub-basins. The rarefaction curve of each sub-basin reached the asymptote indicating the actual numbers of species were close to the species collected in this study. The prediction of the structure of fish assemblages in rivers and reservoirs are very important goal in ecological research, both from a purely theoretical point of view and from an applied one. Moreover, it will be beter studies in the future of Southeast Asia. Estimation of the probability of presence/absence of fish species has been obtained so far using different approaches. Although conventional statistical tools (e.g. logistic regression) provided interesting results, the application of artificial neural networks (ANNs) has recently outperformed those techniques. ANNs are especially effective in reproducing the complex, non-linear relationships that link environmental variables to fish species presence and/or abundance. In this study some new developments in ANN training procedures will be presented, which are 63    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand specifically aimed at solving ecological problems related to the way the errors are computed in species composition models. The resulting improvements in species prediction involve not only the accuracy of the models, but also their ecological consistency. A case history about fish assemblages in the rivers of the Ping-Wang River Basin is presented to demonstrate how the enhanced modelling strategy improved the accuracy of the predictions about fish assemblages. Highest and lowest diversity indexes were obtained in the lower Ping and Maeklang sub-basins, respectively. Six physicochemical parameters (i.e. dissolved oxygen, water temperature, pH, conductivity, phosphorus and alkalinity) and six geo-morphological parameters (i.e. altitude, distance from the sea, discharge, depth and width) showed statistically significant in their relationships to diversity parameters (P-value < 0.05). Results from the classification and regression trees showed that the geomorphological parameters were more significant in controlling and predicting both species richness and Shannon diversity index than the physicochemical parameters, in which altitude was the most significant. Fifty-three dominant fish species from 220 samplings were patternized into 4 assemblage-patterns viz., mountainous-, piedmont-, transitory- and lowland- species. Any environmental changes in the rhitral environment will seriously impact to the mountainous- and piedmont- species since their specific distributions. Importance of geographical parameters i.e. altitude, distance from the sea, stream width, discharge, water depth, and watershed area and physicochemical parameters i.e. water temperature, dissolved oxygen, and conductivity as variable explaining variation in fish community structure along a river gradient in a large scale whole basin (Fig. 18). However, the contribution of the other variables, especially the physicochemical water quality parameters, should be considered in terms of point and non-point pollution sources over a small scale (Ibarra et al., 2005; Orrego et al., 2009). The delineation of fish assemblage patterns enhances the understanding fish zonation in this region. Knowing the representatives of each assemblage allows for the development of indicator species for assessing the integrity of each river course, in the force of human influences in particular. Aquatic ecosystem is influenced by the landscapes through which they flow (Hynes, 1975; Vannote et al., 1980), a fundamental link recognised in many of the 64    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand conceptual models describing the structure and functioning of natural river systems. The word ecology has attracted to its scientific diversity; a useful working definition is the scientific study of the interactions between organisms and their abiotic and biotic environmental that determine the distribution and abundance of the organisms. Also, in this thesis the relationship of the diversity parameters (DP) and environmental parameters (EP) may occupy by the lentic and lotic conditions. The lotic condition was divided into positive proportion a negative proportion, the positive proportion was increased of the EP (e.g. WT, WID, WSH, pH, etc.) to diversity parameters and the second group was a negative relationship viz., DFS, ALT and DO. While, the lentic condition (manmade reservoir) was shown the positive proportion i.e. pH, DEP, and CON and the negative proportion i.e. DO, WID, and AMM. The fish communities in highland tropical streams connected to a reservoir were dominated by cyprinids. Three communities of fish were found in this study i.e. the reservoir community (RC), the stream community (SC) and the intermediate community (IC). Water depth had the main impact on the change in the communities. The Henicorhynchus siamensis a riverine species has shown that it can establish population in the lentic system. Also, H. siamensis could invade the tributaries during a certain period in rainy season as shown in the IC and SC. Meanwhile, the species in the SC could be found in the IC but they were not found in the reservoir area. Nevertheless, the small reservoir environment is similar to a river floodplain, which is suitable to enable fish of river origin to adapt to the new environment. Moreover, the r-strategist reproductive traits of H. siamensis such as early maturity, high fecundity, single broods and rapid egg and larval development would help them be successful in unfavorable environments. Threats to fish communities were deforestation and collection for aquarium fish, especially the exotic Poeciliid fish in the upper reach, which is of the major concern. Meanwhile distribution of aquaculture escapees should be concerned in terms of genetic hybridization. Further studies on the function of individual species in each community are recommended. Moreover, an examination of the fish larvae and juveniles in the system should be also being considered since they also move and distribute in the reservoir. This would also provide information on species interaction and recruitment to the reservoir system. 65    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Upper Chao Phraya river basin has a rich aquatic fauna. These report 201 fish species had been confirmed to inhabit in Ping-Wang River basin, which is the largest tributaries of Chao Phraya river basin. Historically, fishes were very abundant in the basin, and since there were few humans and the fishing gear used local made, fish harvesting had little impact on fish stocks, even through fishes constituted the important protein source for local people. In contrast, the anthropological treat to aquatic environment, instead to loss the stream habitat in the upper reach and lower reach of the river basin. In recent decades, fish populations have apparently declined. Especially, O. siamensis was a vulnerable and endemic species to the Chao Phraya river basin. RECOMMENDATIONS FOR FURTHER RESEARCH Further study should be conducted in the areas from the mountainous high land to the sea in whole Chao Phraya river basin and main Southeast Asia rivers and/or marine area. The data set should be studied to achieve long term data (ecological parameters and diversity parameters) to predict the fish assemblages in Thailand and SEA. The study of biology and life history should encourage to study in various species e.g. treatened species, native species, commercial species, and invasive species etc. There should be a modern method to study the biology and life history e.g. hormone, cytology, and DNA. Encouragement on using modern methods to predict the changes of aquatic resources for commercial and conservation purpose should be done. The polution should be concerned on the relationships between diversity parameters and ecological parameters. And the study should be added to the prediction of climate change with the change of biological parameters and diversity parameters and aquaculture. 66    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand REFERENCES Allan, J.D. 2004. Influence of land use and landscape setting on the ecological status of rivers. Limnetica 23: 187-198. Allen, D.J. 2001. Stream Ecology. Kluwer Academic Publications, Massachusetts. Alexandre, C.V., K.E. Esteves and M.A.M. de Moura e Mello. 2010. Analysis of fish communities along a rural–urban gradient in a neotropical stream (Piracicaba River Basin, São Paulo, Brazil). Hydrobiologia 641: 97–114. Allouche, S. 2002. Nature and functions of cover for riverine fish. Bulletin Francais de la Peche et de la Pisciculture 365/366: 297-324. Alkins-Koo, M. 2000. Reproductive timing of fishes in a tropical intermittent stream. Environmental Biology of Fishes, 57(1): 49-66. Ama-Abasi, D., S.H. Holzloehner, and U. Enin. 2004. The dynamics of the exploited population of EthmalosD ¿PEULDWD (Bowdich, 1825, Clupeidae) in the Cross River Estuary and adjacent Gulf of Guinea. Fish. Res. 68: 225–235. Amarasinghe, U.S. and S.S. De Silva. 1992. Population dynamics of Oreochromis mossambicus and O. niloticus (Cichlidae) in two reservoirs in Sri Lanka. Asian Fisheries Science 5: 37-61. Amarasinghe, U.S., W.S. Weliange, M . Kakkaeo, M.C. Villanueva, and J. Moreau. 2008. Diel feeding pattern and food consumption of selected ¿sh populations in Asian reservoirs. In: Aquatic ecosystems and development: comparative Asian perspectives. F. Schiemer, D. Simon, U. S. Amarasinghe, J. Moreau (Eds). Backhuys Publisher, Leiden, the Netherlands, pp. 249–264. Amarasinghe, U.S., W.S.Weliange, M. Kakkaeo, M.C. Villanueva, and J. Moreau. 2010. Diel feeding pattern and food consumption of selected fish populations in Asian reservoirs In: Schiemer, F., D. Simon, U.S. Amarasinghe and J. Moreau (eds.) Aquatic ecosystems and development: Comparative Asian perspectives, pp. 249-264. Backhuys Publishers, Amsterdam. Amoros, C. and G. Bornette. 2002. Connectivity and biocomplexity in waterbodies of riverine floodplains. Freshwater Biol 47: 761-776. Angermeier, P.L. and J.R. Karr. 1983. Fish communities along environmental gradients in a system of tropical streams. Environmental Biology of Fishes 9: 117–135. 67    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Angermeier, P.L. and I.J. Schlosser. 1989. Species–area relationships for stream fishes. Ecology 70: 1450–1462. APHA, 1991. Standard Methods for the Examination of Water and Wastewater: 16th edition. American Public Health Association, Washington DC. Bagenal, T. B. 1971. The interrelation of the size of fish eggs, the date of spawning and the production cycle. Journal of Fish Biology, 3: 207–219. Bagenal, T. B., and E. Braum. 1978. Eggs and early life history. pages 166-198 In W. E.Ricker, editor. Methods for assessment of fish production in fresh waters. Blackwell Scientific Publications, Oxford and Edinburgh. Bagenal, T. B. and E. Brown. 1978. Eggs and early life history. In: Bagenal, T. B. (ed.).Methods for the Assessment of Fish Production in Fresh Water. BlackwellScientific Publications Ltd., Oxford. p. 165-201. Balcombe, S.R., A.H. Arthington, N.D. Foster, M.C. Thoms, G.A. Wilson, and S.E. Bunn. 2006. Fish assemblages of an Australian dryland river: abundance, assemblage structure and recruitment patterns in the Warrego River, MurrayDarling Basin. Marine and Freshwater Research 57: 619-633. Ballesteros T. M., M. Torres-Mejia, and M.P. Ramírez-Pinilla. 2009. How does diet influence the reproductive seasonality of tropical freshwater fish? A case study of a characin in a tropical mountain river. Neotropical Ichthyology, 7(4): 693700. Baran, E. and T. Jutagate. 2010. Would barriers to migration have the same effect on all fish species? In: Dugan P. and Barlow C. (eds.) Dams as barriers to fish migration in the Mekong, and possibilities for mitigation, MRC Technical Paper, pp. XX-XX. Mekong River Commission, Vientiane. To be published in 2011 Baron, J.S., N. L. Poff, P. L. Angermeier, C.N. Dahm, P.H. Gleick, N.G. Hairston, JR., R.B. Jackson, C.A. Johnston, B.D. Richter, and A.D. Steinman. Meeting ecological and societal needs for freshwater. Ecological Applications, 12(5): 1247–1260 Beamish, F.W.H., P. Sa-Ardrit and S. Tongnunui. 2006. Habitat characteristics of the Cyprinidae in small rivers in Central Thailand. Environmental Biology of Fishes, 76: 237-253. 68    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Beamish, F.W.H., P. Sa-ardrit and V. Cheevaporn. 2008. Habitat and abundance of Balitoridae in small rivers in central Thailand. Journal of Fish Biology 72: 2467–2484. Beddington, J. R. and J.G. Cooke. 1983. On the potential yield of fish stocks. FAO Fisheries Technical Paper 242. Beddington, J.R. and G.P. Kirkwood. 2005. The estimation of potential yield and stock status using life-history parameters. Phil. Trans. R. Soc. B. 360: 163– 170 Berra, T. M. 2001. Freshwater Fish Distribution. Academic Press, San Diego. 604 p. Bhatt, S.D. and K. Pande. 1991. Ecology of the mountain waters. Ashish Publishing House, New Delhi. Borcherding, J., M. Bauerfeld, D. Hinzten, and D. Neuman. 2002. Lateral migrations of fishes between floodplain lakes and their drainage channels at the Lower Rhine: diel and seasonal aspects. J. Fish Biol. 61: 1154–1170. Breiman, L., J.H. Friedman, R.A. Olshen and C.J. Stone. 1984. Classification and regression trees. Wadsworth Inc., Monterey. Britz, R. 2007. Two new species of Mastacembelus from Myanmar (Teleostei: Synbranchiformes: Mastacembelidae). Ichthyol. Explor. Freshwaters, 18(3): 257-268, 7 figs., 2 tabs., Brosse, S., G.D. Grossman, and S. Lek. 2007. Fish assemblage patterns in the littoral zone of a European reservoir. Freshwater Biol. 52: 448–458 Cailliet, G. M., W. D. Smith, H. F. Mollet, and K. J. Goldman. 2006. Age and growth studies of chondrichthyan fishes: the need for consistency in terminology, verification, validation, and growth function fitting. Environmental Biology of Fish. 77: 211-228. Campbell, I.C., C. Poole, W. Giesen, J. Valbo-Jorgensen. 2006. Species diversity and ecology of Tonle Sap Great Lake, Cambodia. Aquat. Sci. 68: 355–373. Campos, H. 1985. Distribution of fishes in the Andean Rivers in the south of Chile. Archiv fȨr Hydrobiologie 104: 169–191. Casatti, L., and R.M.C. Castro. 2006. Testing the ecomorphological hypothesis in a headwater riffles fish assemblage of the Rio São Francisco, southeastern Brazil. Neotrop Ichthyol 4: 203-214. 69    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Chaichana, R., S. Poungcharean, and R. Yoonpundh. 2010. Invasion of Pterygoplichthys pardalis, Aquatic Invasive Species, in Thailand. Proceedings of the 3rd International Conference on Southeast Asian Natural Resources and Environmental Management (SANREM) 2010 (ISBN: 9789832641599) Chellappa, S., R.M.X. Bueno, T. Chellappa, N.T. Chellappa and V.M.F. Almeida e Val. 2009. Reproductive seasonality of the fish fauna and limnoecology of semi-arid Brazilian reservoirs. Limnologica 39: 325–329 Chen, Y. and J.E. Paloheimo. 1994. Estimating fish length and age at 50% maturity using a logistic type model. Aquatic Science, 56: 206-219. Christensen, V. and D. Pauly. 1993. Trophic Models of Aquatic Ecosystems, ICLARM Conf. Proc. 26, International Centre for Living Aquatic Resources Management, Manila, Philippines. 390 pp. Colwell, R. K. 2009. EstimateS, Version 8.2: Statistical Estimation of Species Richness and Shared Species from Samples (Software and User's Guide). Freeware for Windows and Mac OS. Connor, E.F. and E.D. McCoy, 1979. The statistics and biology of the species–area relationship. American Naturalists 113: 791. Costanza, R., L. Graumlich, W. Steffen, C. Crumley, J. Dearing, K. Hibbard, R. Leemans, C. Redman, and D. Schimel. 2007. Sustainability or collapse: What can we learn from integrating the history of humans and the rest of nature? Ambio 36: 522–527. Dairaku, K., K. Kuraji, M. Suzuki, N. Tangtham, W. Jirasuktaveekul, and K. Punyatrong. 2000. The effect of rainfall duration and intensity on orographic rainfall enhancement in a mountainous area: a case study in the Mae Chaem watershed, Thailand. Journal of the Japan Society of Hydrology and Water Resources 13(1): 57-68. Dawson, C.E. 1985. Indo-Pacific pipefishes (Red Sea to the Americas). The Gulf Coast Research Laboratory Ocean Springs, Mississippi, USA. De Graaf, G.J. 2003. The flood pulse and growth of floodplain fish in Bangladesh, Fisheries Management and Ecology, 10: 241-247. De Graaf, M, E. Nentwich, F.A. Sibbing., and J.W. Osse. 2005. Lacustrine spawning, is this a new reproductive strategy among ‘large’ African cyprinid fishes? J 70    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Fish Biol. 66: 1214-1236. De Silva S. S., N.W. Abery and T.T.T. Nguyen. 2007. Endemic freshwater finfish of Asia: distribution and conservation status. Diversity Distrib.13: 172–184 De Silva, E.I.L. and R.W Davies. 1986. Movements of some indigenous riverine fish in Sri Lanka. Hydrobiologia 137: 265-270 De Thoisy, B., S. Brosse, and M.A. Dubois. 2008. Assessment of large-vertebrate species richness and relative abundance in Neotropical forest using line transect censuses: What is the minimal effort required? Biodiv Conserv 17: 2627–2644 De Thoisy, B., C. Richard-Hansen, B. Goguillon, P. Joubert, J. Obstancias, P. Winterton, and S. Brosse. 2010. Rapid evaluation of threats to biodiversity: human footprint score and large vertebrate species responses in French Guiana. Biodivers Conserv 19: 1567–1584 Divakar, L., M.S. Babel, S.R. Perret, and A.D. Gupta. 2011. Optimal allocation of bulk water supplies to competing use sectors basedon economic criterion – An application to the Chao Phraya River Basin, Thailand. Journal of Hydrology. in press doi:10.1016/ j.jhydrol.2011. 02.003 Dobriyal, A. K. and H. R. Singh. 1993, Reproductive biology of a hillstream catfish, Glyptothorax madraspatanum (Day), from the Garhwal, Central Himalaya, India. Aquaculture Research, 24: 699–706 Doi, A. 1997. A review of taxonomic studies of cypriniform fishes in Southeast Asia. Jap. J. Ichthyol. 44(1): 1-33. Duarte, C.M. and M. Alcaraz. 1989. To Produce Many Small or Few Large Eggs: A Size-Independent Reproductive Tactic of Fish. Oecologia. 80(3): 401-404 Dugan P.J., C. Barlow, A.A. Agostinho, E. Baran, G.F. Cada, D. Chen, I.G. Cowx, J.W. Ferguson, T. Jutagate, M. Mallen-Cooper, G. Marmu, J. Nestler, M. Petrere, R.L. Welcomme, and K.O. Winemiller. 2010. Fish migration, dams, and loss of ecosystem services in the Mekong basin. Ambio 39(4): 344-348. Dudgeon, D. 2000. The Ecology of Tropical Asian Rivers and Streams in Relation to Biodiversity Conservation. Annual Review of Ecology and Systematic, 31: 239-263. 71    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Falke, J.A., and K.B. Gido. 2006. Effects of reservoir connectivity on stream fish assemblages in the Great Plains. Can. J. Fish Aquat Sci. 63: 480-493. Ferreira, F.C. and M. Petrere. 2009. The fish zonation of the Itanhae´m river basin in the Atlantic Forest of southeast Brazil. Hydrobiologia. 636: 11–34 Freyhof, J. and D.V. Serov. 2001 Nemacheiline loaches from Central Vietnam with descriptions of a new genus and 14 new species (Cypriniformes: Balitoridae). Ichthyol. Explor. Freshwat. 12(2): 133-191. Froese, R. 2005. Life-History Strategies of Recent Fishes: a Meta-Analysis. Habilitationsschrift, Christian-Albrecht Universität zu Kiel. 209 p. Froese, R. and C. Biholan. 2000. Empirical relationships to estimate asymptotic length, length at first maturity and length at maximum yield per recruit in fishes, with a simple method to evaluate length frequency data. Journal of Fish Biology. 56: 758–773 Froese, R. and D. Pauly. (eds.). 2010. FishBase. World Wide Web Electronic Publication, <www.fishbase.org>, Froese, R. and D. Pauly. Editors. 2011.FishBase. World Wide Web electronic publication. www.fishbase.org, version (02/2011). Gall, G.A.E. 1975. Genetics of reproduction in domesticated rainbow trout. J. Anim. Sei., 40: 19-28. García-Berthou, E. 1999. Food of introduced mosquitofish: ontogenetic diet shift and prey selection. J. Fish Biol. 55: 135-147. Gaston, K.J. and T.M. Blackburn, 2000. Pattern and process in macroecology. Blackwell Publishing Ltd., Oxford.  Gayanilo, JR.F.C., P. Sparre, and D. Pauly. 2002. FiSAT II (ver. 1.2.0.). Food and Agriculture Organization of the United Nations (FAO) (www.fao.org/fi /statist/fi soft/fi sat/index.htm). Gevrey, M., Y.S. Park, T. Oberdorff, and S. Lek. 2003. Prediction of fish assemblages in France and evaluation of the influence of their environmental variables. In Lek, S., M. Scardi, P. Verdonshot & S. Jorgensen (eds), Modelling Community Structure in Freshwater Ecosystems. Springer Verlag, Berlin: 54-63.  72    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Gleick, P.H. 2000. The World's Water 2000-2001: The Biennial Report on Freshwater Resources. Island Press, Washington D.C. Goodsell, J.A. and L.B. Kats. 1999. Effect of introduced mosquitofish on pacific treefrogs and the role of alternative prey. Conservation Biology 13: 921-924. Grossman, G.D., J.F. Dowd, and M. Crawford. 1990. Assemblage stability in stream fish: a review. Env Manag 14: 661–671. Guégan, J.F., S. Lek, and T. Oberdorff. 1988. Energy availability and habitat heterogeneity predict global riverine fish diversity. Nature 391: 382–384. Guisan, A. and W. Thuiller. 2005. Predicting species distribution: offering more than simple habitat models. Ecol Lett 8: 993-1009. Guruge, W.A.H.P. 2002. A food and feeding habit of three co-occuring cyprinids fishes in shallow lowland reservoir in Sri Lanka: consequences for resource partitioning. IN Omar, R., Ali Rahman, Z., Latif, M.T., Lihan, T. and Adam J.H. (Eds.) Proceedings of the Regional Symposium on Environment and Natural Resources 10-11th April 2002, Hotel Renaissance Kuala Lumpur, Malaysia. Vol 1: 154-160 Han, M., M. Fukushima and T. Fukushima. 2008. Species richness of exotic and endangered fishes in Japan's reservoirs. Environmental Biology of Fish. 83: 409-416. He, S.P. 1996. The phylogeny of the glyptosternoid fishes (Teleostei: Siluriformes: Sisoridae). Cybium, 20: 115–159. He, Y. 2010. Structure of endemic fish assemblages in the upper Yangtze River basin and population differentiation of an endangered endemic fish (Gobiocypris rarus). Ph.D. Thesis L’Université Toulouse 3 Paul Sabatier. Toulouse. 190 p. He, Y., J. Wang, S. Lek-Ang, and S. Lek, 2010. Predicting assemblages and species richness of endemic fish in the upper Yangtze River. Science of the Total Environment 408: 4211 – 4220. Herder, F. and J. Freyhof. 2006. Resource partitioning in a tropical stream fish assemblage. Journal of Fish Biology 69: 571–589. Herrmann, M., C. Daniel, F. Sonke, L. Jurgen, P. Pablo, and E. Arntz. 2009. Population structure, growth, and production of the wedge clam Donax 73    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand hanleyanus (Bivalvia: Donacidae) from northern Argentinean beaches. Journal of Shellfish research 28(3): 511-526 Hibbard, K.A., P. Crutzen, E.F. Lambin, D. Liverman, N.J. Mantua, J.R. McNeill, B. Messerli, and W. Steffen. 2007. The great acceleration. In: Sustainability or collapse? An integrated history and future of people on earth, ed. R. Costanza, L.J. Graumlich, and W. Steffen, 341–378. Cambridge: MIT Press. Hoenig, J. M., J. Csirke, M.J. Sanders, A. Abella, M.G. Andreoli, D. Lev, S. Ragonese, M. Al-shoushani, and M.M. El-Musa. 1987. Data acquisition for length-based stock assessment report of working group 1. In: ICLARM conference proceedings on length- based methods LQ ¿VKHULHV research, Vol 13. D. Pauly, G.R. Morgans (Eds). ICLARM, Manila, the Philippines, pp. 343–352. Hoeinghaus, D.J., K.O. Winemiller, and A.A. Agostinho. 2008. Hydrogeomorphology and river impoundment affect food-chain length of diverse Neotropical food webs. Oikos 117: 984-995. Hortle, K.G. 2009. Fisheries of the Mekong river basin. In The Mekong. Biophysical environment of a transboundary river, ed. I.C. Campbell. New York: Elsevier. Horwitz, R.J. 1978. Temporal variability patterns and the distributional patterns of stream fishes. Ecology Monographs, 48: 307-312. http://www.cms.int/ documents/appendix/cms_app1_2.htm #appendix _II Hu, Y.T. and E. Zhang. 2010. Homatula pycnolepis, a new species of nemacheiline loach from the upper Mekong drainage, South China (Teleostei: Balitoridae). Ichthyological Exploration of Freshwaters, 21: 51-62. Huet, M. 1959. Profiles and biology of western European streams as related to fish management. Transactions of the American Fisheries Society 88, 115–163. Hurlbert, S.H. 1971. The nonconcept of species diversity: a critique and alternative parameters. Ecology 52: 577-586. Hynes, H.B.N. 1975. The stream and its valley. Verh. Int. Ver. Theor. Angew. Limnol. 19, 1–15. Hyslop, E.J. 1980. Stomach content analysis: a review of methods and their applications. J. Fish Biol., Southampton, 17(4): 411-429. 74    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Ibarra, A.A., F. Dauba and P. Lim. 2005. Influence of non-point source pollution on riverine fish assemblages in South West France. Ecotoxicology 14: 573–588. Jayaram, K.C. 1999. The Freshwater Fishes of the Indian Region, Narendra Publishing House, New Delhi, xxvii + 551, Pls. xviii. Junk, W.J., and K.M.Wantzen. 2004. The Flood Pulse Concept: New Aspects Approaches and Applications – an Update. pages 117-140, In R. L. Welcomme, and T. Petr (eds.): Proceedings of the Second International Symposium on the Management of Large Rivers for Fisheries: Vol. 2. Food and Agriculture Organization & Mekong River Commission. FAO Regional Office for Asia and the Pacific, Bangkok. RAP Publication 2004/16. Jutagate, T., S.S. de Silva, and N.S. Mattson. 2003. Yield, growth and mortality rate of Thai river sprat, Clupeichthys aesarnensis, in Sirinthorn Reservoir, Thailand. Fish. Manag. Ecol. 10: 221–232. Jutagate, T., J. Sa-nguansin, G. Deein and S. Udduang. Accepted. Fish contributions in a river basin in the lower Northern of Thailand and a strategy for conservation following river damming. Chiang Mai Journal of Science. Kalteh, A.M., P. Hjorth, and R. Berndtsson. 2008. Review of the self-organizing map (SOM) approach in water resources: Analysis, modelling and application. Environ. Model. & Softw. 23: 835– 845. Kimmel, W.G. and D.G. Argent. 2010. Stream fish community responses to a gradient of specific conductance. Water Air and Soil Pollution 206: 49–56. King, A.J., P. Humphries, and P.S. Lake. 2003. Fish recruitment on floodplains: the roles of patterns of flooding and life history characteristics. Canadian Journal of Fisheries and Aquatic Sciences 60: 773–786. Kirkwood, G.P., J.R. Beddington, and J.R. Rossouw. 1994. Harvesting species of different lifespans. In Large-scale ecology and conservation biology (ed. P. J. Edwards, R. May & N. R. Webb), pp. 199–227. Oxford: Blackwell Scientific. Kohonen T. 2001. Self-organizing Maps. Springer, Berlin. Kottelat, M. 1996. Oreoglanis siamensis. In: IUCN 2010. IUCN Red List of Threatened Species. Version 2010.4. <www.iucnredlist.org>. Downloaded on 18 April 2011. 75    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Kottelat, M. 1989. Zoogeography of the fishes from Indochinese inland waters with an annotated check-list. Bull. Zool. Mus. Univ. Amsterdam 12(1): 1-55. Kottelat, M. 1990. Indochinese neamacheilines, A Revision of Nemacheiline loaches (Pisces: Cypriniformes) of Thailand, Burma, Laos, Cambodia and southern Vietnam. Verlag, Dr. Friedrich Pfiel, Munchen, 262 pp. Kottelat, M. 1998. On the valid generic names for the Indian fishes usually referred to Salmostoma and Somileptes (Teleostei: Cyprinidae and Cobitidae). Journal of South Asian Natural History 3(1): 117-119 Kottelat, M. 2001. Fishes of Laos. WHT Publications (Pte) Ltd, Sri Lanka. 198 pp. Kottelat, M. and T. Whitten, 1996, Freshwater biodiversity in Asia, with special reference to fish., World Bank Tech. Pap. 343: 59 p. Kottelat, M., A. J. Whitten, S. N. Kartikasari, and S. Wirjoatmodjo. 1993. Freshwater fishes of Western Indonesia and Sulawesi. Periplus Editions, Ltd., Republic of Indonesia. 221 pp. (+ plates). Kramer, D. L. 1978. Reproductive seasonality in the fishes of a tropical stream. Ecology, 59(5): 976-985. Kruk, A., and T. Penczak. 2002. Impoundment impact on populations of facultative riverine fish Ecology, 39(3): 197-210. Kruk, A., S. Lek, Y.S. Park and T. Penczak. 2007 Fish assemblages in the large lowland Narew River system (Poland): application of the self-organizing map algorithm, Ecol. Modell. 203: 45–61. Kuraji, K., K. Punyatrong, and M. Suzuki. 2001. Altitudinal increase in rainfall in the Mae Chaem watershed, Thailand. Journal of the Meteorological Society of Japan. 79(1B): 353-363. Lamberts, D. 2001. Tonle Sap ¿sheries: a case study on Àoodplain gillnets ¿sheries. FAO Regional Oƥce for Asia-3DFL¿F Publication No. 2001 » 11. FAO, Bangkok, pp. 141. Lasne, E., B. Bergerot, S. Lek and P. Laffaille. 2007. Fish zonation and indicator species for the evaluation of the ecological status of rivers: example of the Loire basin (France). River Research and Applications 23: 877-890. Last, P.R. and L.J.V. Compagno. 1999. Dasyatidae. In: K.E. Carpenter & V.H. Niem 76    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand (eds), FAO species identification guide for fishery purposes. The living marine resources of the Western Central Pacific. Volume 3. Batoid fishes, chimaeras and bony fishes part 1 (Elopidae to Linophyrnidae). FAO, Rome, pp. 1479– 1505. Lek, S. and J.F. Guegan. 2000. Artificial Neural Networks: Application to Ecology and Evolution. Springer Verlag, Berlin. Lévêque, C., T. Oberdorff, D. Paugy, M.L.J. Stiassny, and P.A. Tedesco. 2008. Global diversity of fish (Pisces) in freshwater. Hydrobiologia 595: 545–567. Lim, P., S. Lek, S.T. Touch, S.-O. Mao and B. Chhouk. 1999. Diversity and spatial distribution of freshwater fish in Great Lake and Tonle Sap River (Cambodia, Southeast Asia). Aquat. Living Resour. 12(6): 379-386. Lundberg, J. G., M. Kottelat, G. R. Smith, M. Stiassny and T. Gill. 2000. So many fishes, so little time: an overview of recent ichthyological discoveries in fresh waters. Annals of the Missouri Botanical Garden 87: 26-62. Magurran, A.E. 2004. Measuring biological diversity. Blackwell Publishing Company, Oxford. Malmqvist, B. and S. Rundle. 2002. Threats to the running water ecosystems of the world. Environmental Conservation 29: 134-153. Martin-Smith, K.M. and H.H. Tan. 1998. Diversity of freshwater fishes from eastern Sabah: annotated checklist for Danum valley and a consideration of interand intra-catchment variability. Raffles Bull. Zool. 46(2): 573-604. Matthews, W.J. 1998. Patterns in freshwater fish ecology. Kluwer Academic Publishers, Boston, USA. 756 p. Mattson, N.S. and E.K.W.H. Kaunda. 1997. Population dynamics of Oreochromis shiranus in two small water bodies in Malawi. J. Fish Biol. 50: 592–607. Mills, M., R. Rader, and M. Belk 2004. Complex interactions between native and invasive fish. Oecologia 141: 713-721. Mims, M.C., J.D.Olden, Z.R. Shattuck, and N.L. Poff. 2010. Life history trait diversity of native freshwater fishes in North America. Ecology of Freshwater Fish.19: 390–400. Monkolprasit, S., S., Sontirat, S. Vimollohakarn, and T. Songsirikul. 1997. Checklist of fish in Thailand. Office of Environtmental Policy and Planning. Bangkok, 77    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Thailand, 353 pp. Moreau, J., U.S. Amarasinghe, B. Sricharoendham, P.K. Jayasinghe, and M.C. Villanueva. 2008. Population dynamics of commercially important ¿sh species in four Asian reservoirs. In: Aquatic ecosystems and development: comparative Asian perspectives. F. Schiemer, D. Simon, U. S. Amarasinghe, J. Moreau (Eds). Backhuys Publisher, Leiden, the Netherlands, pp. 295–304. Na-Nakorn. U., W. Kamonrat, and T. Ngamsiri. 2004. Genetic diversity of walking catfish, Clarias macrocephalus, in Thailand and evidence of genetic introgression from introduced farmed C. gariepinus. Aquaculture 240: 145–163 Nelson, J.S. 2006. Fishes of the World, 4th Edition. 624 p. Ng, H.H. 2002. Descriptions of two new species of Kryptopterus from Thailand and Borneo in the K. limpok species group (Teleostei: Siluridae). Ichthyol, Explor. Freshwat. 13(1): 67-72. Ng, H.H. 2003. Kryptopterus geminus, a new species of silurid catfish (Teleostei: Siluridae) from mainland Southeast Asia. Zootaxa 305: 1–11. Ng, H.H. and M. Kottelat. 1998. The catfish genus Akysis Bleeker (Teleostei: Akysidae) in Indochina, with descriptions of six new species. J. Nat. Hist. 32: 1057-1097. Ng, H.H. and P.K.L. Ng. 2001. A revision of the akysid catfish genus Acrochordonichthys Bleeker. J. Fish Biol. 58: 386-418. Ng, H.H. and M. Kottelat. 2001. A review of the genus Batasio (Teleostei: Bagridae) in Indochina, with the description of B. tigrinus sp. n. from Thailand. Rev. Suisse Zool. 108(3): 495-511. Nguyen, T. T.T. and S.S. de Silva. 2006. Freshwater finfish biodiversity and conservation: an asian perspective. Biodiversity and Conservation 15: 35433568 Nissanka, C., U.S. Amarasinghe, and S.S. De Silva. 2000. Yield predictive models for Sri Lankan reservoir ¿sheries. Fish. Manag. Ecol. 7: 425–436. Norris, R.H., and Thoms, M.C., 1999. What is river health? Freshwat. Biol. 41, 197209. Oberdorff, T., J.F. Guegan, and B. Hugueny. 1995. Global patterns of fish species 78    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand richness in rivers. Ecography 18: 345–352. Office of Natural Water Resources Committee of Thailand, 2003 Chapter 16 – Chao Phraya River Basin, Thailand in the 1st UN world water development report: Water for People, Water for Life. UNESCO, Paris Olurin K.B. and O.I. Odeyemi. 2010. The Reproductive Biology of the Fishes of Owa Stream, South - West Nigeria. Research Journal of Fisheries and Hydrobiology, 5(2): 81-84 Orrego, R., S.M. Adams, R. Barra, G. Chiang and J. F. Gavilan. 2009. Patterns of fish community composition along a river affected by agricultural and urban disturbance in south-central Chile. Hydrobiologia 620: 35–46. Özcan, G. and S. Balik. 2009. Age and growth of bassan barbel, Barbus pectoralis (Actinopterygii: Cypriniformes: Cyprinidae), under conditions of a dam reservoir. Acta Ichthyologica et Piscatoria. 39(1): 27–32 Park, Y.S., T. Oberdoff, and S. Lek. 2005. Patterning riverine fi sh assemblages using an unsupervised neural network; In Lek S., Scardi M., Verdonschot P.F.M., Descy J.P., Park Y.S., eds., Modelling Community Structure in Freshwater Ecosystems, Springer-Verlag: Heidelberg,: 43-53. Pauly, D. 1979. Theory and Management of Tropical Multispecies Stocks: a Review, with Emphasis on the Southeast Asian Demersal Fisheries. ICLARM Studies and Reviews 1, 35 p. Pauly, D. and J. L. Munro. 1984. Once more on the comparison of growth in fish and invertebrates. Fishbyte, 2(1): 21. Penczak, T. and A. Kruk. 2000. Threatened obligatory riverine fishes in humanmodified Polish rivers. Ecol. Fre s h w. Fish, 9: 109-117. Petry, A.C. and U.H. Schulz. 2006. Longitudinal changes and indicator species of the fish fauna in the subtropical Sinos River, Brazil. Journal of Fish Biology 69: 272-290. Prathumratana, L., S. Sthiannopkao, and K.W. Kim. 2008. The relationship of climatic and hydrological parameters to surface water quality in the lower Mekong River. Environ. Int. 34: 860–866. Prchalová, M., J. .XEHþND, and M. Vašek, J. Peterka, J. Sed’a, T. JĤza, M. ětKD, O. Jarolím, M. Tušer, M. Kratochvíl, M. ýech, V. Draštík, J. Frouzová and E. 79    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Hohausová. 2008. Distribution patterns of fishes in a canyon-shaped reservoir. J Fish Biol 73: 54–78. 3UFKDORYi0-.XEHþND0ýHFK-)URX]RYi9'UDãWtN(+RKDXVRYi7-Ĥ]D 0.UDWRFKYtO-0DWČQD-3HWHUND0ětKD07XãHU 09DãHN The effect of depth, distance from dam and habitat on spatial distribution of fish in artificial reservoir. Ecol. Freshwat. Fish., 18: 247-260. Prejs, A. and K. Prejs. 1987. Feeding of tropical freshwater fishes: seasonality in resource availability and resource use. Oecologia, 71: 397-404. Pusey, B. J., A. H. Arthington, P. G. Close and J. R. Bird. 2002. Larval fishes in rainforest streams: recruitment and microhabitat use. Proceedings of the Royal Society of Queensland, 110: 27-46. Pusey, B.J., M.G. Read, and A.H. Arthington. 1995. The feeding ecology of freshwater fishes in two rivers of the Australian Wet Tropics. Environmental Biology of Fishes 43: 85–103. Pusey, B.J., A.H. Arthington, J.R. Bird, P.G. Close. 2001. Reproduction in three species of rainbowfish (Melanotaeniidae) from rainforest streams in northern Queensland, Australia. Ecology of Freshwater Fish. 10: 75–87. R Development Core Team. 2009. R: A language and environment for statistical computing, reference index Version 2.10. R Foundation for Statistical Computing, Vienna. Raghavan, R., G. Prasad, P.H. Anvar-Ali, & B. Pereira. 2008. Fish fauna of Chalakudy River, part of Western Ghats biodiversity hotspot, Kerala, India: patterns of distribution, threats and conservation needs. Biodiversity & Conservation, 17: 3119 – 3131. Rainboth, W.J. 1996. Fishes of the Cambodian Mekong. FAO Species Identification Field Guide for Fishery Purposes. FAO, Rome, 265 p. Rehage, J.S., B.K. Barnett, and A. Sih. 2005. Behavioral responses to a novel predator and competitor of invasive mosquitofish and their non-invasive relatives (Gambusia sp.). Behav. Ecol Sociobiol 57: 256–266 Roberts, T.R. 1997. Systematic revision of the tropical Asian labeon cyprinid fish genus Cirrhinus, with descriptions of new species and biological observations on C. lobatus. Nat. Hist. Bull. Siam Soc. 45: 171-203. 80    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Roberts, T.R. 1998. Freshwater fugu or pufferfishes of the genus Tetraodon from the Mekong basin, with descriptions of two new species. Ichthyol. Res. 45(3): 225-234. Roberts, T.R. and T.J. Warren. 1994. Observations of fishes and fisheries in southern Laos and northeastern Cambodia, October 1993-Febuary 1994. Nat. Hist. Bull. Siam. Soc. 42: 87-115. Schlosser, I.J. 1987. A conceptual framework for fish communities in small warmwater streams. In: Mattews WJ, Heins DC (eds.) Community and evolutionary ecology of North American stream fishes. University of Oklahoma Press, Norman, pp. 17-24. Schiemer, F. and H. Waidbacher. 1992. Strategies of conservation of a Danubian fish fauna. Pages 365-382 In Boon P.J., Calow P. & Petts G.E. (eds). River Conservation and Management. John Wiley & Sons Ltd., London. Schmutz, S., M. Kaufmann, B. Vogel, M. Jungwith and S. Muhar. 2000. A multilevel concept for fish-based, river-type-specific assessment of ecological integrity. Hydrobiologia 422: 279–289. Serezli, R., S. Guzel and M. Kocabas. 2009. Fecundity and egg size of three salmonid species (Oncorhynchus mykiss, Salmo labrax, Salvelinus fontinalis) Cultured at the Same Farm Condition in North-Eastern, Turkey. J ournal of Anim al and Vet erin ar y Advances . 9(3): 576-58 0 Singhanouvong, D., C. Soulignavong, K. Vonghachak, B. Saadsy and T.J. Warren. 1996a. The main dry-season fish migrations of the Mekong mainstream at Hat Village, Muang Khong District, Hee Village, Muang Mouan District and Hatsalao Village, Paxse. Indigenous Fishery Develoment Project, Fisheries Ecology Technical Report no. 3. Lao People's Democratic Republic. 130 p. Singhanouvong, D., C. Soulignavong, K. Vonghachak, B. Saadsy and T.J. Warren. 1996b. The main wet-season migration through Hoo Som Yai, a steep-gradient channel at the Great fault line on the Mekong River, Champassack Province, Southern Lao PDR. Indigenous Fishery Development Project, Fisheries Ecology Technical Report No. 4. Technical Section, Dept. of LivestockFisheries, Ministry of Agriculture-Forestry, Lao People's Democratic Republic. 115 p. 81    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Smith, H.M. 1945. The fresh-water fishes of Siam, or Thailand. Bull. U.S. Natl. Mus. 188: 633 p. Sokheng, C., C.K. Chhea, S. Viravong, K. Bouakhamvongsa, U. Suntornratana, N. Yoorong, N.T. Tung, T.Q. Bao, A.F. Poulsen and J.V. Jørgensen. 1999. Fish migrations and spawning habits in the Mekong mainstream: a survey using local knowledge (basin-wide). Assessment of Mekong fisheries: Fish Migrations and Spawning and the Impact of Water Management Project (AMFC). AMFP Report 2/99. Vientiane, Lao, P.D.R. Suryanarayana, I., Braibanti, A., Rao R.S., Raman V.A., Sudarsan D., and G.N. Rao. 2008 Neural networks in fisheries research. Fish Res 92: 115–139. Suvarnaraksha, A. 2004. Fish Diversity in Mae Ngud Somboonchon Dam, Maetang District, Chiangmai Province. J. Agri. Research and Extension. 22(special Issue): 47-58 Suvarnaraksha, A. 2010. The diversity of fishes in Salween River basin on Thailand border. Fianal Report, Maejo University. Taki, Y. 1974. Fishes of the Lao Mekong Basin. United States Agency for International Development Mission to Laos Agriculture Division. 232 p. Taki, Y. 1978. An analytical study of the fish fauna of the Mekong Basin as a biological production system in nature. Research Institute for Evolutionary Biology, Tokyo. Tan, H.H. 1998. Description of two new species of the Betta waseri group (Teleostei: Osphronemidae). Ichthyol. Explor. Freshwat. 8(3): 281-287. Tan, H.H. and S.H. Tan. 1996. Redescription of the Malaysian fighting fish Betta pugnax (Teleostei: Belontiidae), and description of Betta pulchra, new species from Peninsular Malaysia. Raffles Bull. Zool. 44(2): 419-434. Ter Braak, C.J.F. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis Ecology 67: 1167-1179. Thapanand, T., T. Jutagate, P. Wongrat, T. Lekchonlayuta, C. Meksumpun, S. Janekitkarn, A. Rodloi, J. Moreau, and L. Wongrat. 2009. Trophic relationships and ecosystem characteristics in a newly impounded man-made lake in Thailand. Fish. Manag. Ecol. 16: 77–87. 82    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Tongnunui, S. and F.W.H. Beamish. 2009. Habitat and relative abundance of fishes in small rivers in eastern Thailand. Environmental Biology of Fishes, 85: 209– 220. Torres-Mejia, M. and M.P. Ramírez-Pinilla. 2008. Dry-season breeding of a characin in a tropical mountain river. Copeia, 2008(1): 99-104. Türkmen, M., O. (UGR÷DQA.<ÕOGÕUÕPI. Akyurt. 2002. Reproduction tactics, age and growth of Capoeta capoeta umbla +HFNHOIURPWKH$úNDOHUHJLRQRIWKH Karasu River, Turkey, Fisheries Research, 54: 317-328. Unsrisong, G., P. Pornsopin, S. Kantiyawong, B. De Lapeyre, S.Wesels, G. HorstgenSchwark. 2005. Induced Spawning of Batfish (Oreoglanis siamensis). “The Global food & Product Chaindynamics, Innoations, Conflicts Strategies” Deutscher Tropentag, October 11-13, 2005, Hohenheim. Valbo-Jørgensen, J., D. Coates, and K. Hortle. The Mekong Biophysical Environment of an International River Basin Edited by: Ian C. Campbell Chapter 8 - Fish Diversity in the Mekong River Basin, Pages 161-196 van Zalinge, N., P. Degen, C. Pongsri, S. Nuov, J. G. Jensen, V.H. Nguyen, X. Choulamany. 2004. The Mekong river system. In: Proc. 2nd Internat. Symp. Management of large rivers for ¿sheries, vol I. R. Welcomme, T. Petr (Eds). FAO, Bangkok, pp. 335–357. Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell and C. E. Cushing, 1980. The river continuum concept. Canadian Journal of Fisheries and Aquatic Sciences 37: 130–137. Vidthayanon, C., J. Karnasutra and J. Nabhitabhata. 1997. Diversity of freshwater fishes in Thailand. Office of Environment Policy and Planing, Bangkok. Vidthayanon, C. 2003. Schistura pridii, a new nemacheiline loach (Teleostei: Balitoridae) from Upper Chao Phraya drainage, northern Thailand. Ichthyological Exploration of Freshwaters, 14: 307-310. Vidthayanon, C., P.Saenjundaeng, and H. H. Ng. 2009. Eight new species of the torrent catfish genus Oreoglanis) Teleostei: Sisoridae) from Thailand. Ichthyological Exploration of Freshwaters. 20(2): 127-156. Viravong, S. 2006.   Observations of the   fisheries of the Mekong with notes on the life history strategies RI IRXU ¿VK species (Botia modesta, Henicorhynchus 83    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand siamensis, Helicophagus waandersii and Probarbus jullieni). PhD. Thesis, Hull Univ., Hull, United Kingdom. Vitule, J.R.S., C.A. Freire and D. Simberloff. 2009. Introduction of non-native freshwater fish can certainly be bad. Fish & Fisheries, 10: 98–108. Walker, A. 2003. Agricultural transformation and the politics of hydrology in northern Thailand Development and Change 34(5): 941-964. Wantzen, K.M., W.J. Junk, and K.O. Rothhaupt. 2008. An extension of the Àoodpulse concept (FPC) for lakes. Hydrobiologia 613: 151–170. Ward, J.H. 1963. Hierarchical Grouping to optimize an objective function. Journal of American Statistical Association 58: 236-244. Ward-Campbell, B.M.S, F.W.H. Beamish and C. Kongchaiya. 2005. Morphological characteristics in relation to diet in five co-existing Thai fish species. Journal of Fish Biology 67:1266–1279. Weaver, W. and C.E. Shannon. 1949. The mathematical theory of communication. University of Illinois, Urbana. Welcomme, R.L., and A. Halls. 2005. Dependence of tropical river fisheries on flow. In: Welcomme R.L., Petr T. (eds.) Proceedings of the Second International Symposium on the Management of Large Rivers for Fisheries Vol. 2, Mekong River Commission, Vientiane, pp. 267–284. Welcomme, R.L., K.O. Winemiller, and I.G. Cowx. 2006. Fish environmental guilds as a tool for assessment of ecological condition of rivers. River Research and Applications 22: 377–396. Williams, J. E. 2000. The Coefficient of Condition of Fish. Chapter 13 in Schneider, James C. (ed.) 2000. Manual of fisheries survey methods II: with periodic updates.Michigan Department of Natural Resources, Fisheries Special Report 25, Ann Arbor. Wondie, A., S. Mengistu, J. Vijverberg, and E. Dejen. 2007. Seasonal variation in primary production of a large high altitude tropical lake (Lake Tana, Ethiopia): effects of nutrient availability and water transparency. Aqua Ecol 41: 195-207. Yap, S. Y. 2002. On the distributional patterns of Southeast-East Asian freshwater fish and their history. J. Biogeogr. 29, 1187-1199. Zhang, E., P.Q. Yue, and J.X. Chen. 2000. Labeoninae [A]. In: Yue PQ (eds). Fauna 84    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Sinica (Osteichthyes: Cypriniformes III)[M]. Beijing: Science Press. 85    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand 86    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Part 2: Publications 87    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand P1 Fish diversity in the upper Chao Phraya river basin, Southeast Asia Suvarnaraksha, A., S. Lek-Ang, S. Lek, and T. Jutagate. (2011) Ichthyological Exploration of Freshwaters. (Submitted) 88    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Fish diversity in the Upper Chao Phraya River Basin, Southeast Asia Apinun Suvarnaraksha *, **, ***, Sithan Lek-Ang **, Sovan Lek** and Tuantong Jutagate* * Faculty of Agriculture, Ubon Ratchathani University, Warin Chamrab, Ubon Ratchathani, Thailand 34190 ** University of Toulouse III, Laboratoire Dynamique de la Biodiversité, UMR 5172, CNRS – UPS, 118 route de Narbonne, 31062 Toulouse Cedex 4, France *** Faculty of Fisheries Technology and Aquatic Resources, Maejo University, Sansai, Chiangmai, Thailand 50290 Correspondence: Tuantong Jutagate. Tel. +66-45-353500 Fax. +66-45-288373 Email: tjuta@agri.ubu.ac.th 89    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand ABSTRACT The lack of reliable data causes a dispute over fish diversity in the Indo-Burma hotspot (Southeast Asia). This study presents information on fish diversity and distribution in a unique high altitude mountain to lowland areas in the upper part of the Chao Phraya river basin, Thailand. Fourteen years of field dataset in the basin was used, covering 272 surveys of 10 sub-river basins to produce species richness and diversity indices. Similarities among sub-basins were examined by Ward‘s agglomerative method. Rarefaction was employed to extrapolate species richness and optimum numbers of the surveys. Two hundred and one species in 104 genera and 34 families were collected, including 16 exotic species. Fish of Family Cyprinidae dominated, followed by Balitoridae and Cobitidae, implying the characteristic of high altitude area. The overall endemism in the areas was found to be about 10%. Five IUCN-list fish viz., Oreoglanis siamensis, Himantura signifier, H. chaophraya, Pangasianodon gigas and Pangasius sanitwongsei were found. Ward‘s method showed distinct differences between the upper- and lowland sub-basins. The rarefaction curve of each sub-basin reached the asymptote indicating the actual numbers of species were close to the species collected in this study. In conclusion, the fish community is especially characterized by rhithronic habitants, including some species that are not yet taxonomically described. Threats to fish communities were deforestation, collection for aquarium fish, and the distribution of the exotic Poeciliid fish in the upper reaches. Meanwhile, distribution of aquaculture escapees should be concerned in terms of genetic hybridization. 90    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand INTRODUCTION The Indo-Burma region is considered the third largest global biodiversity hotspot after the Tropical Andes and Mesoamerica (Myers et al., 2000). Within this region, the number of freshwater fish has been documented at more than 1,260 species, compared to 2,345 fish species in the Oriental region (Lévêque et al., 2008) and more than 560 of these species are endemic (Conservation International, 2010). However, it is generally accepted that data and information on fish is very poor compared to other vertebrate groups (Myers et al., 2000), especially in this region (Sodhi et al., 2004) and this has led to a dispute about their distribution and conservation (e.g. De Silva et al., 2007; Darwall et al., 2008). Moreover, freshwater ecosystems, in general, have received much less focus in terms of conservation prioritization exercises (Taylor, 2010). Figure 1 Location and map of the Ping-Wang River Basin and its sub-basins 91    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand In the Indo-Burma region, Thailand ranked among the top in the diversity of freshwater fish species (i.e. 690 species, Kottelat & Whitten, 1996; Vidthayanon et al., 1997; Nguyen & De Silva, 2006), which are supported by an extensive inland water area of about 4.5 million ha in seven major river basins viz., Chao Phraya, Mekong, Eastern, Southern, Salween, Mekhlong and Tenasserim (Jutagate, 2010). All these basins, except for the Mekong, the studies on fish diversity are very few and less than desirable (Coates et al., 2003; Nguyen & De Silva, 2006). This is also the case in the Chao Phraya river basin, which is one of the 21 major river basins in East, South and Southeast Asian nations, and the fish diversity in this basin is recorded at about 222 species (Dudgeon, 2000), of which 34 species are endemic (De Silva et al., 2007). This basin covers about 30 percent of Thailand’s land area, and is a home to about 40 percent of the country’s total population. Therefore, it is clear that there is no sign, until now, of deceleration in anthropogenic stresses like other developed countries (Costanza et al., 2007). In recent years, there have been changes in the land uses on the river bank and within the watershed from agricultural uses to industrial and urban use (Mahujchariyawong & Ikeda, 2001), in which effluents cause polluted water (Meksumpun & Meksumpun 2008) and directly impact the fish community as well as the fisheries. In order to create an approach for the appropriate conservation of fish, basic information on the diversity and distribution of individual species in the basin is needed. This is exclusively in Southeast Asia, where research and knowledge about the ecological and taxonomic aspects of many fauna are relatively rare and they may face extinction before we even know of their existence (Bickford et al., 2010). Therefore, in this study, we presented the baseline information on fish diversity distributed within the Upper Chao Phraya River Basin (also called “the Ping-Wang River Basin”), which contains more than three fourths of the freshwater fish species known from Chao Phraya river basin (Vidthayanon et al., 1997) and exclusively characterized as a high altitude area (i.e. comprised of 1st and 2nd order streams). The data cover 14 years consecutive surveys within 10 sub-river basins. This information could help correct the meager data on spatial diversity and distribution of fish species in the region. 92    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand MATERIAL AND METHODS Study area The Ping and the Wang rivers are the main rivers of northern Thailand and merge with Nan River in the Central part to form the Chao Phraya river. The Ping river is 740 km long, with a catchment area of about 33,896 km2.The Wang river is 440 km long and has a catchment area of 10,791 km2 (Takeuchi et al., 2005). The Wang river flows southwestward to join the lowland leg of Ping river in Tak Province to form a large watershed area between 15q42’ and 19q48’ North and 98q04’ and 100q08’ East. The sampling area was divided into 10 sub-basins with various numbers of sampling stations (Fig. 1 and Table 1) viz., the upper Ping river (UP), Maetang (MT), the second part of the Ping river (SP), Maeklang (MK), Maecheam (MC), the third part the Ping river (TP), Maeteon (ME), the Wang river (WA), the fourth part of the Ping river (FP) and the lower Ping river (LP). The first 8 sub-basins lie in a relatively high altitude mountainous area, and the latter 2 stations are in the lowland area. Geomorphometric characteristics of each sub-basin are also shown in Table 1 Fish samplings Two hundred and seventy two samplings were conducted spatially over the whole basin during January 1996 to April 2009 (Table 1). The sampling sites were chosen on the basis of accessibility, similarity and diversity of habitat types in streams and rivers (pools, cascade, falls, riffles, and stagnant water). Collection of fish samples were taken at every habitat type in every selected site, in which each site was approximately 35 or 40 mean stream width in length. Samplings were done by various methods i.e. beach seine net, cast net, multi-mesh gillnets as well as the electrofishing with an AC shocker (Honda EM 650, DC 220 V 550BA 450VA, 1.5–2 A, 50 Hz), which was placed on the riverbank together with block nets and scoop nets. Each sampling site was sampled at least twice to represent dry and wet seasons. Live fish were roughly identified in the field, measured for total length (mm), counted, and then returned back to the water. Only a few samples of individual species were anaesthetized in dilute solution of benzocaine (50 mg/l) and kept separately according to species level. Specimens were fixed in 10% formalin for a 93    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand a Sub-basin Geographic Coordinate Upping Ping 19°07’-19°48’ N (UP) 98°47’-99°17’ E Maetang 19°10’-19°45’ N (MT) 98°27’-98°55’ E The second Ping (SP) 18°31’-19°33’ N 98°24’-99°22’ E 94 Maeklang 18°24’-18°35’ N (MK) 98°28’-98°41’ E Maecheam 17°57’-19°09’ N (MC) 98°04’-98°37’ E The third Ping (TP) 17°48’-18°43’ N 98°14’-98°44’ E Maeteon 17°13’-18°02’ N (ME) 98°14’-98°34’ E The forth Ping (FP) 15°50’-17°49’ N 98°39’-100°02’ E Lower Ping 15°42’-16°10’ N (LP) 99°27’-100°08’ E Wang river 17°07’-19°24’ N (WA) 99°00’-100°06’ E Elevation Distance from Water Stream Forest Agricultures Urban Collecting Collecting No. of (m ASL) the sea (km) depth (m) width (m) (%) (%) (%) period techniques station G, P, R, S 684 r 228.3 1,026 r 24.1 0.4 r 0.2 7 r 0.5 76.5 23.3 0.1 G, P, R, S 756 r 166.2 1,067 r 36.0 0.6 r 0.4 13 r 11.2 72.1 25.4 0.3 G, P, R, S 553 r 160.2 982 r 41.0 1.9 r 6.0 74 r 230.5 75.0 24.9 0.1 Bottom types 2008 2000-2001 2003-2004 E 6 E 48 E, N, T, C 98 1996, 2003-2004, 2008 G, P, R, S 1,070 r 213.4 877 r 4.6 0.3 r 0 11 r 5.4 88.4 11.5 0.1 2008 E 6 2007-2008 E, N 44 E, N 18 G, P, R, S 627 r 207.3 927 r 53.9 0.7 r 0.4 21 r 17.4 74.7 24.4 0.9 G, S, M 261 r 11.5 704 r 43.8 2.8 r 1.2 424 r 224.6 88.2 11.6 0.1 G, P, R, S 804 r 229.2 847 r 43.0 0.5 r 0.2 8 r 3.8 85.0 11.5 0.2 2008 E 24 G, S, M 120 r 33.5 580 r 68.4 2.7 r 0.5 359r77.5 67.7 29.6 2.8 2009 B, N, T 6 G, S, M 48 r 8.0 425 r 16.0 3.2 r 1.3 258 r 27.7 72.6 22.0 5.4 2009 B, N, T 4 G, S, M 408 r 123.8 833 r 225.2 0.9 r 0.9 28 r 48.2 76.5 23.3 0.1 2009 E 18 2005-2006, 2009 Note (i) Bottom types: R = Rocky, G = Gravel, P = Pebble, S = Sandy, M = Muddy; (ii) Collecting techniques: E = Electro fishing, N= Gill net, T = Trap, C = Cast net, B=Beach seine net a     94 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 1 Descriptions of the sub-basins in the Ping-Wang River Basin and sampling protocols    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand month and changed to ethanol 30%, 50% and finally preserved in 70% ethanol. Specimens were re-checked and taxonomically identified into species at the Maejo Aquatic Resources Natural Museum (MARNM). Data analyses Ranks of individual species were presented as the percentages of relative abundance (%RA) and occurrence frequency (%OF). The diversity indices (Magurran 2004) viz., Shannon diversity index (H’) and evenness (J’) were calculated for each sub-basin. Hierarchical agglomerative clustering by Ward's method was used to classify sets of dissimilarities of the fish assemblages in sub-basins and was analyzed by using Program R (R Development Core Team 2009). Kruskal-Wallis test was used to test significance of species richness. Under the assumption that species richness increases with the sample size, we rarefied species richness to the same number of individuals and a rarefaction curve was used to estimate species richness in each subbasin. Rarefaction (R) was based on the equation (Hulbert, 1971) R S ª i 1 ¬ § N  mi · § N · º ¸ /¨ ¸ n ¹ © n ¹»¼ ¦ «1  ¨© where, N is the total number of individuals, S is the total number of species, mi is the number of individuals of species i, and n is the number of individuals in sub-sample, i.e. sub-basin. The rarefaction values were computed by using EstimateS v. 8.2.0 (Colwell, 2006). RESULTS A total of 32,080 fishes were collected representing 34 families, 104 genera and 201 species (Table 2). Fish in Order Cypriniformes were the most dominant, both in terms of the number of individuals (77.7 %) and the number of species (56.9 %), followed by order Siluriformes (8.6 % and 20.8 %) and Perciformes (7.7 % and 10.9 %) (Fig. 2). In terms of Family, Cyprinidae dominated with 40.3 % (81 species), followed by Balitoridae and Cobitidae with 10.0 % and 6.5 % (20 and 13 species), respectively (Table. 2). Among the genera, Schistura in Family Balitoridae was the 95    a Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand Species 96 Garra cambodgiensis (Tirant, 1883) Schistura breviceps (Smith, 1945) Mystacoleucus marginatus (Valenciennes, 1842) Schistura poculi (Smith, 1945) Henicorhynchus siamensis (Sauvage, 1881) Barilius pulchellus (Smith, 1931) Gambusia affinis (Baird & Girard, 1853) (e) Rasbora paviana Tirant, 1885 Scaphiodonichthys burmanicus Vinciguerra, 1890 Channa gachua (Hamilton, 1822) Puntioplites proctozysron (Bleeker, 1865) Poropuntius deauratus (Valenciennes, 1842) Puntius stoliczkanus (Day, 1871) Oreoglanis siamensis Smith, 1933 (a, d) Paralaubuca riveroi (Fowler, 1935) Parambassis siamensis (Fowler, 1937) Schistura spilota (Fowler, 1934) Schistura geisleri (Kottelat 1990) Schistura obeini Kottelat, 1998 Schistura menanensis (Smith, 1945) Exostoma vincegerrae Regan, 1905 Puntius orphoides (Valenciennes, 1842) Hampala macrolepidota Kuhl & Van Hasselt, 1823 Pristolepis fasciatus (Bleeker, 1851) Discherodontus schroederi (Smith, 1945) Notopterus notopterus (Pallas, 1769) Schistura waltoni (Fowler, 1937) Barbonymus gonionotus (Bleeker, 1850) Devario maetangensis (Fang, 1997) (d) Hemibagrus nemurus (Valenciennes, 1840) Cyclocheilichthys armatus (Valenciennes, 1842) Mastacembelus armatus (Lacepède, 1800) Oxyeleotris marmorata (Bleeker, 1852) Schistura sexcauda (Fowler, 1937) Barilius koratensis (Smith, 1931) Devario regina (Fowler, 1934) Labiobarbus lineatus (Sauvage, 1878) Homaloptera smithi Hora, 1932 Mystus mysticetus Roberts, 1992 a     Family CYP BAL CYP BAL CYP CYP POE CYP CYP CHA CYP CYP CYP SIS CYP AMB BAL BAL BAL BAL SIS CYP CYP NAN CYP OST BAL CYP CYP BAG CYP MAS ELE BAL CYP CYP CYP BAL BAG Sub-basins N %RA %OF TL (mm) Guilds M 3354 2766 1529 1439 1330 1126 831 825 749 735 696 690 654 645 626 619 606 548 526 479 476 468 446 387 360 305 287 285 274 271 267 245 237 216 201 198 186 175 172 10.46 8.62 4.77 4.49 4.15 3.51 2.59 2.57 2.33 2.29 2.17 2.15 2.04 2.01 1.95 1.93 1.89 1.71 1.64 1.49 1.48 1.46 1.39 1.21 1.12 0.95 0.89 0.89 0.85 0.84 0.83 0.76 0.74 0.67 0.63 0.62 0.58 0.55 0.54 40.44 39.34 23.16 18.75 11.40 34.93 17.28 29.41 19.49 47.06 11.03 19.85 23.16 13.97 3.68 8.46 15.07 13.97 5.88 5.88 9.19 23.16 12.87 9.19 11.76 7.72 8.09 9.93 5.15 14.71 8.82 16.18 9.93 2.94 5.51 9.93 3.68 9.56 4.78 21-120 30-86 10-160 23-73 51-185 22-109 12-52 21-97 30-95 12-200 82-280 20-213 45-124 25-144 48-440 21-58 48-61 36-124 15-65 25-105 38-101 21-34 33-370 30-220 23-113 70-290 36-42 73-245 21-85 23-480 58-231 95-480 20-286 15-87 42-91 20-85 115-155 19-67 143-152 HER INV INV INV HER INV INV INV HER PIS HER HER INV INV INV INV INV INV INV INV INV INV PIS INV INV PIS INV HER INV PIS HER INV PIS INV INV INV HER INV PIS Y N Y N Y Y N ? Y Y Y Y N N Y Y N N N N N Y Y N Y Y N Y ? Y Y ? N N Y ? Y N Y 96 UP MT SP X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X MK X X X X X X X MC ME X X X X X X X X X X X X X X X X X X X X X TP WA FP LP X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 2 Species composition of fish in the Ping-Wang River Basin and their occurrence in each sub-basin    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand a Species 97 Trichogaster trichopterus (Pallas, 1770) Acantopsis choirorhynchos (Bleeker, 1854) Rasbora myseri Brittan, 1954 Esomus metallicus Ahl, 1923 Mystus singaringan (Bleeker, 1846) Cyclocheilichthys repasson (Bleeker, 1853) Oreochromis niloticus (Linnaeus, 1758) (e) Crossocheilus reticulatus (Fowler, 1934) Yasuhikotakia modesta (Bleeker, 1864) Puntius brevis (Bleeker, 1850) Pangasius pleurotaenia Sauvage, 1878 Schistura bucculentus (Smith, 1945) Monopterus albus (Zuiew, 1793) Rasbora dusonensis (Bleeker, 1851) Schistura magnifluvis Kottelat, 1990 Lobocheilos quadrilineatus (Fowler, 1935) Channa striata (Bloch, 1793) Trichopsis vittata (Cuvier, 1831) Homaloptera zollingeri Bleeker, 1853 Macrognathus siamensis (Günther, 1861) Osteochilus hasseltii (Valenciennes, 1842) Sikukia stejnegeri Smith, 1931 Garra fuliginosa Fowler, 1934 Dermogenys pusilla Kuhl & van Hasselt, 1823 Homaloptera leonardi Hora, 1941 Balitora brucei Gray, 1830 Rhinogobius chiengmaiensis Fowler, 1934 (d) Xenentodon cancila (Hamilton, 1822) Syncrossus helodes (Sauvage, 1876) Lepidocephalichthys hasselti (Valenciennes, 1846) Lepidocephalichthys berdmorei (Blyth, 1860) Schistura mahnerti Kottelat, 1990 Labiobarbus leptocheila (Valenciennes, 1842) Anabas testudineus (Bloch, 1795) Pseudomystus siamensis (Regan, 1913) Schistura pridii Vidthayanon, 2003 (d) Cyclocheilichthys apogon (Valenciennes, 1842) Kryptopterus cryptopterus (Bleeker, 1851) Glyptothorax lampris Fowler, 1934 Raiamas guttatus (Day, 1870) Glyptothorax trilineatus Blyth, 1860 a     Family OSP COB CYP CYP BAG BAG CIC CYP COB CYP PAN BAL SYN CYP BAL CYP CHA OSP BAL MAS CYP CYP CYP HEM BAL BAL GOB BEL COB COB COB BAL CYP ANA BAG BAL CYP SIL SIS CYP SIS N %RA %OF 160 159 156 149 145 133 129 123 122 108 106 106 105 105 103 101 97 95 95 90 90 90 89 85 85 84 83 83 78 77 75 73 70 68 67 67 65 64 58 58 57 0.50 0.50 0.49 0.46 0.45 0.41 0.40 0.38 0.38 0.34 0.33 0.33 0.33 0.33 0.32 0.31 0.30 0.30 0.30 0.28 0.28 0.28 0.28 0.26 0.26 0.26 0.26 0.26 0.24 0.24 0.23 0.23 0.22 0.21 0.21 0.21 0.20 0.20 0.18 0.18 0.18 8.46 8.09 1.84 9.19 6.99 1.84 9.93 2.21 5.15 9.93 2.21 2.94 9.93 2.94 2.21 0.74 15.44 6.25 6.25 5.15 6.25 3.31 3.31 7.72 6.99 4.04 2.57 8.09 5.88 9.19 4.41 5.15 4.41 9.56 4.41 5.88 2.21 3.68 6.62 6.62 8.82 TL (mm) 23-54 30-120 15-83 18-58 120-178 50-119 27-256 29-130 45-56 26-195 61-223 31-77 167-620 31-101 22-53 60-135 13-2880 650-712 27-54 105-162 35-246 38-64 82-160 18-67 40-102 32-80 13-163 32-40 56-91 17-93 39-40 25-105 82-182 59-127 32-168 23-43 80-195 124-173 31-77 50-108 67-168 97 Guilds M INV INV INV INV PIS HER HER HER INV INV HER INV INV INV INV HER PIS INV INV INV HER INV HER INV INV INV INV INV INV INV INV INV HER PIS PIS INV HER PIS PIS PIS PIS N ? ? N Y Y N Y Y Y Y N ? ? N Y N N N ? Y Y Y ? N N N ? Y N N N Y Y Y N Y Y ? Y ? UP MT X SP X X X X X MK X X X X X X X X Sub-basins MC ME X X X X X X X X X X X X X X X X X X WA X X FP X X LP X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X TP X X X X X X X X X X Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 2 (cont.) Species composition of fish in the Ping-Wang River Basin and their occurrence in each sub-basin    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand a Species 98 Neolissochilus stracheyi (Day, 1871) Rasbora atridorsalis Kottelat & Chu, 1987 Barbonymus altus (Günther, 1868) Phalacronotus bleekeri (Günther, 1864) Paralaubuca typus Bleeker, 1865 Hypsibarbus wetmorei (Smith, 1931) Osteochilus lini Fowler, 1935 Acanthocobitis botia (Hamilton, 1822) Toxotes chatareus (Hamilton, 1822) Cirrhinus cirrhosus (Bloch, 1795) (e) Tor tambroides (Bleeker, 1854) Pangasianodon hypophthalmus (Sauvage, 1878) Pangasius macronema Bleeker, 1851 Onychostoma gerlachi (Peters, 1881) Clupeoides borneensis Bleeker, 1851 Labeo chrysophekadion (Bleeker, 1850) Nemacheilus binotatus Smith, 1933 Tetraodon leiurus Bleeker, 1851 Barbonymus schwanenfeldii (Bleeker, 1853) Rasbora daniconius (Hamilton, 1822) Schistura desmotes (Fowler, 1934) Brachirus siamensis (Sauvage, 1878) Puntius partipentazona (Fowler, 1934) Hemibagrus wyckii (Bleeker, 1858) Osphronemus goramy Lacepède, 1801 Pangio anguillaris (Vaillant, 1902) Clarias batrachus (Linnaeus, 1758) Oreochromis hybrid (e) Mastacembelus cf. tinwini Britz, 2007 Osteochilus melanopleurus (Bleeker, 1852) Osteochilus microcephalus (Valenciennes, 1842) Yasuhikotakia morleti (Tirant, 1885) Devario aequipinnata (McClelland, 1839) Acantopsis thiemmedhi Sontirat, 1999 Rasbora borapetensis Smith, 1934 Paralaubuca harmandi Sauvage, 1883 Acanthocobitis zonalternans (Blyth, 1860) Micronema hexapterus (Bleeker, 1851) Osteochilus waandersii (Bleeker, 1852) Channa micropeltes (Cuvier, 1831) Chitala ornata (Gray, 1831) a     Family N %RA %OF CYP CYP CYP SIL CYP CYP CYP BAL TOX CYP CYP PAN PAN CYP CLU CYP BAL TET CYP CYP BAL SOL CYP BAG OSP COB CLA CIC MAS CYP CYP COB CYP COB CYP CYP COB SIL CYP CHA NOT 56 55 54 54 50 49 56 46 45 44 43 42 42 41 40 37 37 35 34 34 34 31 26 25 25 25 24 24 24 23 23 23 21 20 20 20 19 19 19 18 18 0.17 0.17 0.17 0.17 0.16 0.15 0.17 0.14 0.14 0.14 0.13 0.13 0.13 0.13 0.12 0.12 0.12 0.11 0.11 0.11 0.11 0.10 0.08 0.08 0.08 0.08 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.06 0.06 0.06 0.06 0.06 0.06 0.06 0.06 5.88 1.47 6.25 4.04 4.78 2.21 2.94 2.94 3.31 2.94 1.10 4.41 2.57 2.21 4.41 5.88 3.68 3.31 2.94 0.37 1.47 3.68 2.21 3.68 4.04 2.21 4.41 2.21 4.78 2.21 1.84 2.57 1.47 2.57 1.10 1.47 0.37 1.84 1.84 2.21 3.31 TL (mm) 31-252 20-55 67-145 115-410 67-221 98-218 65-121 40-68 98-165 165-235 101-312 390-753 198-251 25-100 31-52 94-705 27-42 87-132 43-215 40-84 20-49 98-122 20-61 154-320 97-365 36-37 90-132 40-144 55-375 189-275 82-93 37-170 22-66 145-183 32-108 102-134 20-78 121-235 75-89 390-475 305-415 98 Guilds M HER INV HER PIS INV HER HER INV INV HER HER HER INV HER INV HER INV PIS HER INV INV INV INV PIS HER INV PIS HER INV HER HER INV INV INV INV INV INV PIS HER PIS PIS Y ? Y Y Y Y Y ? N Y Y Y Y Y Y Y N N Y ? N ? ? Y N ? Y N ? Y Y Y ? ? ? Y ? Y Y Y Y UP MT X SP X X X X MK Sub-basins MC ME X X X X X X X X X TP X LP X X X X X X X X X X X X X X X X X X X X FP X X X WA X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 2 (cont.) Species composition of fish in the Ping-Wang River Basin and their occurrence in each sub-basin    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand a Species 99 Clarias hybrid (C. macrocephalus X C. gariepinus) (e) Danio albolineatus (Blyth, 1860) Devario malabaricus (Jerdon, 1849) Mystus albolineatus Roberts, 1994 Syncrossus beauforti (Smith, 1931) Luciosoma bleekeri Steindachner, 1878 Ompok bimaculatus (Bloch, 1794) Poropuntius bantamensis (Rendahl, 1920) Acanthopsoides delphax Siebert, 1991 Mystus multiradiatus Roberts, 1992 Barbichthys laevis (Valenciennes, 1842) Clarias macrocephalus Günther, 1864 Kryptopterus cheveyi Durand, 1940 Parambassis wolffii (Bleeker, 1851) Labeo rohita (Hamilton, 1822) (e) Trichogaster pectoralis (Regan, 1910) Mystacoleucus greenwayi Pellegrin & Fang, 1940 Sikukia gudgeri (Smith, 1934) Wallago attu (Bloch & Schneider, 1801) Belodontichthys truncatus Kottelat & Ng, 1999 Cyclocheilichthys enoplos (Bleeker, 1851) Leptobarbus hoevenii (Bleeker, 1851) Pangasius larnaudii Bocourt, 1866 Amblyrhynchichthys truncatus (Bleeker, 1851) Bagarius bagarius (Hamilton, 1822) Bagrichthys macracanthus (Bleeker, 1854) Cyprinus carpio Linnaeus, 1758 (e) Himantura signifer Compagno & Roberts, 1982 (b) Hypophthalmichthys molitrix (Valenciennes, 1844) (e) Glyptothorax fuscus Fowler, 1934 Mastacembelus favus Hora, 1924 Barbichthys nitidus Sauvage, 1878 Cosmochilus harmandi Sauvage, 1878 Cynoglossus microlepis (Bleeker, 1851) Hemibagrus wyckioides (Fang & Chaux, 1949) Phalacronotus apogon (Bleeker, 1851) Acanthopsoides gracilentus (Smith, 1945) Bagrichthys macropterus (Bleeker, 1853) Channa lucius (Cuvier, 1831) Pterygoplichthys disjunctivus (Weber, 1991) (e) a     Family N %RA %OF CLA CYP CYP BAG COB CYP SIL CYP COB BAG CYP CLA SIL AMB CYP OSP CYP CYP SIL SIL CYP CYP PAN CYP SIS BAG CYP DAS CYP SIS MAS CYP CYP CYN BAG SIL COB BAG CHA LOR 18 18 18 18 17 16 16 16 15 15 14 14 14 14 14 13 12 12 11 10 9 9 9 8 8 8 8 8 8 7 7 6 6 6 6 6 5 5 5 5 0.06 0.06 0.06 0.06 0.05 0.05 0.05 0.05 0.05 0.05 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.03 0.03 0.03 0.03 0.03 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 2.94 1.84 1.47 1.10 2.21 1.10 2.21 0.74 2.21 1.10 0.74 2.57 1.10 1.84 2.94 1.47 1.47 1.10 3.31 1.47 2.21 1.10 1.84 0.74 1.10 0.37 2.21 1.47 1.47 0.74 1.10 0.74 0.74 0.37 1.10 0.74 0.37 0.37 1.47 1.84 TL (mm) 154-200 21-48 59-73 129-143 128-200 125-158 78-147 97-179 32-45 126-139 86-155 64-277 133-167 46-53 396-537 26-105 35-48 36-98 320-430 350-423 231-435 257-463 541-563 79-162 198-465 176-211 321-389 398-410 450-752 46-75 75-126 157-185 312-323 154-213 220-435 278-432 59-77 154-187 210-315 87-206 99 Guilds M PIS INV INV PIS INV INV PIS HER INV PIS HER PIS PIS INV HER INV INV INV PIS PIS HER HER HER HER PIS PIS HER INV HER PIS INV HER HER INV PIS PIS INV PIS PIS HER Y ? Y Y Y Y ? Y ? Y Y Y Y Y Y N Y Y Y Y Y Y Y Y Y Y Y Y Y ? ? Y Y ? Y Y ? Y Y ? UP MT SP X X MK Sub-basins MC ME TP WA FP X LP X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 2 (cont.) Species composition of fish in the Ping-Wang River Basin and their occurrence in each sub-basin    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand a Species 100 Trichogaster microlepis (Günther, 1861) Boesemania microlepis (Bleeker, 1858) Clarias gariepinus (Burchell, 1822) (e) Cynoglossus feldmanni (Bleeker, 1853) Hypsibarbus vernayi (Norman, 1925) Puntius leiacanthus Bleeker, 1860 Schistura vinciguerrae (Hora, 1935) Tuberoschistura baenzigeri (Kottelat, 1983) Amblyceps foratum Ng & Kottelat, 2000 Amblyceps mucronatum Ng & Kottelat, 2000 Bagarius yarrelli (Sykes, 1839) Gyrinocheilus aymonieri Tirant, 1883 Pangasianodon gigas Chevey, 1931 (c, e) Pangasius conchophilus Roberts & Vidthayanon, 1991 Xiphophorus helleri Heckel, 1848 (e) Bangana sinkleri (Fowler, 1934) Brachirus aenea (Smith, 1931) Betta splendens Regan, 1910 Cirrhinus molitorella (Valenciennes, 1844) Helicophagus leptorhynchus Ng & Kottelat, 2000 Helostoma temminckii Cuvier & Valenciennes, 1831 Pangasius bocourti Sauvage, 1880 Parachela oxygastroides (Bleeker, 1852) Albulichthys albuloides (Bleeker, 1855) Brachirus harmandi (Sauvage, 1878) Crossocheilus cobitis (Bleeker, 1853) Ctenopharyngodon idellus (Valenciennes, 1844) (e) Himantura chaphraya Monkolprasit & Roberts, 1990 (a) Hypophthalmichthys nobilis (Richardson, 1845) (e) Labiobarbus siamensis (Sauvage, 1881) Lates calcarifer (Bloch, 1790) (e) Lobocheilos melanotaenia (Fowler, 1935) Pangasius sanitwongsei Smith, 1931 (c) Poecilia reticularis Peters, 1859 (e) Puntius binotatus (Valenciennes, 1842) Rasbora argyrotaenia (Bleeker, 1850) Solea ovata Richardson, 1846 Thynnichthys thynnoides (Bleeker, 1852) Tor douronensis (Valenciennes, 1842) Wallago leerii Bleeker, 1851 a     Family N %RA %OF OSP SCI CLA CYN CYP CYP BAL BAL AML AML SIS GYR PAN PAN PEO CYP SOL 5 4 4 4 4 4 4 4 3 3 3 3 3 3 3 2 0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.37 0.74 1.10 0.74 1.10 0.74 0.37 0.74 0.74 0.74 1.10 1.10 1.10 0.37 1.10 0.74 TL (mm) 32-34 254-302 210-271 65-97 33-66 23-64 15-91 21-24 90-93 45-102 87-260 136-173 1550-1774 675-634 60-190 71-192 2 3 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0.01 0.01 0.01 0.01 0.01 0.01 0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 0.37 0.37 0.37 0.37 0.74 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 120-143 34-55 158-287 305-312 131-147 564-632 59-120 124 120 90 364 681 480 98 458 76 1270 37 25 34 87 187 167 565 OSP CYP PAN HEL PAN CYP CYP SOL CYP CYP DAS CYP CYP CEN CYP PAN PEO CYP CYP SOL CYP CYP SIL 100 Guilds M INV PIS PIS INV HER INV INV INV INV INV PIS HER HER INV INV INV INV N Y Y ? Y Y N N N N Y Y Y Y N Y ? INV HER INV INV INV INV HER INV HER HER INV HER HER PIS HER HER INV INV INV INV HER HER PIS N Y Y Y Y Y Y ? Y Y Y Y Y Y Y Y N Y ? ? Y Y Y UP MT SP MK Sub-basins MC ME X TP WA FP X X X X X X X X LP X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 2 (cont.) Species composition of fish in the Ping-Wang River Basin and their occurrence in each sub-basin a   Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand i. X indicate that the species was found ii. UCN Red-list (as shown after scientific name): (a) vulnerable species, (b) endangered species and (d) critical endangered species and (e) exotic species iii. Family: DAS=Dasyatidae; NOT=Notopteridae; CLU=Clupeidae; BAL=Balitoridae, CYP=Cyprinidae, COB=Cobitidae, GYR=Gyrinocheilidae; AML=Amblycipitidae, BAG=Bagridae, CLA=Clariidae, LOR=Loricariidae, SIL=Siluridae, SIS=Sisoridae, PAN=Pangasiidae; HEM=Hemiramphidae, BEL=Belonidae; MAS=Mastacembelidae, SYN-Synbranchidae; POE=Poecillidae; AMB=Ambassidae, ANA=Anabantidae, CHA=Channidae, CEN=Centropomidae, ELE=Eleotridae, GOB=Gobiidae, HEL=Helostomidae, NAN=Nandidae, OSP=Osphronemidae, SCI=Sciaenidae, TOX=Toxotidae; SOL=Soleidae, CYN=Cynoglossidae; TET=Tetraodontidae. iv. %RA= Percentage of relative abundance; %OF = Percentage of occurrence frequency 101 v. Guilds: INV= invertivorous, HER = herbivorous, PIS = piscivorous vi. M=Migratory species (Y = yes, N=no, ? = not clear) Table 3 Aspects of diversity in fish in the Ping-Wang River Basin. Aspects Species richness Number of individuals Diversity index (H’) Species evenness (J’) Endemic species (%) Vulnerable species (%) Endangered species (%) Critical endangered species (%) Species Restricted to Ping-Wang River (%) Species Restricted to Chao Phraya river (%) Exotic species (%) a     UP 16 545 2.06 0.74 6.25 0 0 0 6.25 0.45 0 MT 56 5,607 2.82 0.70 5.36 1.79 1.79 0 5.36 1.80 1.79 SP 100 11,953 3.42 0.74 1 1 1 1 1 0.45 10.0 MK 11 474 1.46 0.61 9.09 0 0 0 9.09 0.45 9.09 101 Sub-basins MC TP 66 30 4,425 1,755 2.93 2.67 0.70 0.79 3.03 3.33 1.52 3.33 0 0 0 0 1.52 3.33 0.90 0.45 1.52 3.33 ME 90 2,639 3.51 0.78 1.11 0 1.11 0 1.11 0.45 2.22 WA 79 2,541 3.40 0.78 0 0 0 0 0 0.00 7.60 FP 78 1,060 3.94 0.90 0 0 0 0 0 0.00 8.97 LP 112 1,081 4.45 0.94 0 0.89 0.89 1.79 0 0.00 9.82 Summary 201 32,080 4.02 0.62 3.03 1.52 0.00 0.00 1.52 0.90 5.08 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Note    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand most diverse with 14 species, followed by Rasbora (Cyprinidae), Pangasius (Pangasiidae) and Puntius (Cyprinidae) with 7, 6, and 6 species, respectively. The number of genera and number of species ratio was found to be 1: 1.93. Figure 2. Relative abundances of major taxonomic groups in each sub-basins of the Ping-Wang River Basin. Results of %RA and %OF showed that only a few species had values beyond 4 % and 10 %, respectively, which implied a typical left skew of species abundance. The five most abundant species accounted for 32.4 %RA of total fish collected. The highest %OF was found in Channa gachua (47.1 %) and only 8 species showed %OF higher than 20 % (Table 2), in which 6 of them belong to the family Cyprinidae, i.e. Garra cambodgiensis, Mystacoleucus marginatus, Barilius pulchellus, Rasbora paviana, Puntius stoliczkanus and Puntius orphoides and the other fish was Schistura breviceps (Balitoridae). The species that were high in number (t 200 individuals) but low %OF (d 6 %) implied the restricted distribution. These fishes were two cyprinids Paralaubuca riveroi and Devario maetangensis and three balitorids Schistura obeini, Schistura menanensis and Schistura sexcauda. On the other hand, Channa striata and Anabas testudineus showed a wide distribution (%OF t 10 %), though the samples of each species were less than 100. Pangasianodon gigas, from the samples, was 102    Biology of two keystone fish species, fish assemblage patterns and modeling approaches in tropical river basin: Case study of Ping River Basin, Thailand expected from stocking, whereas Lates calcarifer and Clarias hybrid were the escapees from culture practices. Large sub-basins showed high H’, i.e. L (4.5), F (3.9) and P (3.5) and the abundance of individual species were in similar proportions, i.e. J’ > 0.75 (Table 3). Although lower in H’ in the upper sub-basins (i.e. U, T and K), these sub-basins were characterized by the endemic species and the rate of species that are restricted to the Basin was up to 10 % (Table 3). These endemic species were D. maetangensis, Schistura pridii, Oreoglanis siamensis, and Rhinogobius chiengmaiensis. Five IUCN fish species were also collected, i.e. O. siamensis, Himantura signifier, H. Chao Phraya, P. gigas and Pangasius sanitwongsei. Sixteen exotic species were found in all sub-basins, except the upper reach of the Ping River (UP). Among them, Gambusia affinis was the highest %OF at 17.3 % (Table 2) followed by Oreochromis niloticus (| 9 %OF). Figure 3. Species richness in each sub-basins of the Ping-Wang river basin. Species richness gradually increased from the upper part (UP) to the lowland area (LP: Fig. 3) and were significantly different (Kruskal-Wallis test; P < 0.001). Results of hierarchical clustering showed that each sub-basin had its own characteristic of fish species composition (Fig. 4). The lowland sub-basins (FP and LP) were the least similar to those sub-basins in the upper part of the Basin. Meanwhile within the upper part, the most upstream sub-basins (UP and WA) were 103    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Figure 4. Dendrogram of the cluster analysis results corresponding to the sub-basins of the Ping-Wang river basin. Figure 5. Rarefaction curves plot by sub-basins for the species richness of the PingWang river basin. 104    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Figure 6. Rarefaction curves plot by number of species with number of sampling sites Ping-Wang river basin. most similar. Numbers of expected species within the Basin was justified by the rarefaction curve because of the differences in numbers and kinds of habitats in each sub-basin. All the ten rarefaction curves for the sub-basins (Fig. 5) showed signs of reaching asymptotic levels. Adequacy of sampling was assessed also by the rarefaction curve and the asymptote was reach at about 250 (Fig. 6), confirming that the number of sampling sites in this study (272) was satisfactory. DISCUSSIONS The Ping-Wang river basin contained 201 fish species and the rarefaction curves showed that it was close to the maximum number of species occurring in the Basin. All the rarefaction curves reached asymptote implies that the species, which found in this study, covered almost all, if not all, taxonomically described species in the Basin. Thus, it is shown that the Chao Phrya River system is very high in fish species richness. Even in the upper part of the basin, as observed in this study, the accumulated species number was higher than most of the river basins in South, East and Southeast Asia, except the Mekong (1,200 species), the Yangtze (China, 320 105    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand species), the Cauvery river (India and Nepal, 265 species) and the Kapus (Indonesia, 250 species) (Nguyen & De Silva, 2006). Dominance by fish in family Cyprinidae is common in the Asian freshwater bodies (Kottelat & Whitten, 1996), where they may contribute 40% or more of the species in a watershed (Taki 1978; Beamish et al., 2006). They have evolved partially through highly adapted body forms and mouth structures so that they occupy virtually all habitats throughout their distribution (Ward-Campbell et al., 2005; Beamish et al., 2006), but prevalence of the rhithronic species such as O. siamensis, Exostoma vinciguerrae, Schistura spp. and Homaloptera spp. (Vidthayanon, 2003; Vidthayanon et al., 2009; Hu & Zhang, 2010) is an exclusive characteristic of the mountainous area of Upper Chao Phraya River Basin. This is also included the 3 species of the sisorid species (Ng, 2010) viz., Glyptothorax lampris, G. trilineatus and G. fuscus. The species richness increased from the upper (UP) to the lower part (LP) of the basin. Low species richness in the high altitude reflects the low variability of food supplies because of the fast turnover time of available food from the terrestrial inputs (Tongnunui & Beamish, 2009). Also the findings showed that the species richness was related to sub-basin size, i.e. the larger the sub-basin, the higher the species richness. However, on a large scale, Nguyen & De Silva (2006) reported that species richness did not necessarily correlate to river basin size since rivers with small basins show high diversity. The H’ index generally lies between 1.5 and 3.5 (Magurran, 2004), and the high value of 4.5 in the lower Ping river (LP) was because of the richness of nutrients, which increased natural food sources, as well as flood pulse effect in the lowland (Junk & Wantzen, 2004), which all supported high populations. Moreover, deep water, wider rivers and more discharges downstream are factors to increase diversity parameters (Horwitz, 1978). Altitude and distance from the sea are found to be among of the most key factors that govern the species richness and fish community structures in riverine ecosystem (Oberdorff et al., 1995). The sub-basins, at an altitude > 250 m, were grouped together and regarded as rhithronic community (Welcomme et al., 2006). Residents in this area are always generally small in size and equipped with suckers or adhesive apparatuses (Bhatt & Pande, 1991), and may also have streamlined or flattened forms such as G. cambodgiensis, O. siamensis and Schistura spp., which 106    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand were the dominated in this study and none of them were found in the lowland subbasins (i.e. FP and LP). Although only 5- ICUN fish species have been found so far in the Ping-Wang river basin, the uniqueness of the area, characterized by high mountains, and a high rate of endemism of fish should be of concern. There are also numbers of rhithronic species in this basin, and at least 3 species have not yet been taxonomically classified (A. Suvarnarhaksha, personal collection), in which the new taxonomically described, Schistura pridii was also from this basin (Vidthayanon, 2003). The threats for fish in the basin are becoming higher. As a consequent of urbanization in the lowland area, polluted water could be expected, and this could be harmful to many fish, even the generalist such as most of cyprinids, which are ubiquitous in the area. Deforestation for horticultures, i.e. cabbage, corn, and tomato, in the upper basin and indiscriminate fish collection for aquarium fish are among other major threats for the rhithronic habitants. Both issues not only affect the fish population per se but also the ecosystem, such as erosion from agricultural fields and habitat destruction by searching for aquarium target species. Although most of the rhithronic species have medium to high resilience and their minimum population doubling times are on average at 2 years (Froese & Pauly 2010), this would not be possible if their habitats were altered. Moreover, since most of the rhithronic members are mostly insectivorous (Tongnunui & Beamish, 2009), habitat degradation in the rhitral area, could lead to a decline in exogenous food sources including insects as well as their larvae (Raghavan et al., 2008). Most of exotic species found in this basin were introduced for food enhancement purposes such as Nile tilapia, Chinese and Indian major carps and no impacts are reported so far. The presence of the Poeciliid fish as G. affinis, Xiphophorus helleri and Poecilia reticulata in the upstream part must be of concern since they prey on many aquatic larvae and often with devastating consequences (Mills et al., 2004; Vitule et al., 2009), not only to fish but also to amphibians, which have high endemism in the area (Bickford et al., 2010). Escapees from aquaculture, such as Clarias gariepinus and hybrid walking catfishes, were found in the studied area and should also be of concern. Senanan et al. (2004) observed the introgression of C. gariepinus genes into native catfish, C. macrocephalus in wild populations 107    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand caused by the release/escape of hybrid catfish (C. macrocephalus x C. gariepinus). Na-Nakorn et al. (2004) mentioned that C. macrocephalus and C. batrachus in the wild may be directly replaced with the hybrid catfish, which have a higher growth rate. ACKNOWLEDGEMENT A. Suvarnaraksha is grateful to the Royal Golden Jubilee Program of the Thailand Research Fund for supporting his Ph.D. study (Grant PHD/0290/2549). The research was also supported by National Research Council of Thailand and the Nagao Natural Environment Foundation project. Partial support was received from the Franco-Thai Academic Collaboration (Grant PHC 16598RJ) and the French Embassy to Thailand (Grant CNOUS: 2009/2349), which made it possible for him to work at Laboratoire Evolution et Diversité Biologié under the convention for the joint supervision of theses between Ubon Ratchathani University and Université Paul Sabatier (Toulouse III). REFERENCES Beamish F.W.H., P. Sa-Ardrit & S. Tongnunui. 2006. Habitat characteristics of the Cyprinidae in small rivers in Central Thailand. Environmental Biology of Fishes, 76: 237-253. Bhatt, S.D. & K. Pande. 1991. Ecology of the mountain waters. Ashish Publishing House, New Delhi. Bickford, D., S.D. Howard, D.J.J. Ng & J.A. Sheridan. 2010. Impacts of climate change on the amphibians and reptiles of Southeast Asia. Biodiversity & Conservation, 19: 1043–1062. Coates, D., O. Poeu, U. Suntornratana, N.T. Tung & S. Viravong. 2003. Biodiversity and fisheries in the Lower Mekong Basin. Mekong Development Series No.2. Mekong River Commission, Phnom Penh. Conservation International 2001. Biodiversity hotspots. http://www.biodiversityhotspots.org Costanza, R., L. Graumlich, W. Steffen, C. Crumley, J. Dearing, K. Hibbard, R. Leemans, C. Redman & D. Schimel. 2007. Sustainability or collapse: what can 108    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand we learn from integrating the history of humans and the rest of nature? Ambio, 36: 522-­‐527. Cowell, R.K. 2006. EstimateS: Statistical estimation of species richness and shared species from samples: Version 8. http://www.purl.oclc.org/estimates. De Silva, S.S., N.W. Abery & T.T.T. Nguyen. 2007. Endemic freshwater finfish of Asia: distribution and conservation status. Diversity & Distribution, 13: 172184. Darwall, W.R., C.M. Pollock, K.G. Smith & J.C. Vie 2008. Response to De Silva et al. (2007): Endemic freshwater finfish of Asia: distribution and conservation status. Diversity & Distribution, 14: 159-160. Dudgeon, D. 2000. The Ecology of Tropical Asian Rivers and Streams in Relation to Biodiversity Conservation. Annual Review of Ecology and Systematic, 31: 239-263. Froese, R. & D. Pauly 2010. FishBase. http:// www.fishbase.org Horwitz, R.J. 1978. Temporal variability patterns and the distributional patterns of stream fishes. Ecology Monographs, 48: 307-312. Hu, Y.T. & E. Zhang. 2010. Homatula pycnolepis, a new species of nemacheiline loach from the upper Mekong drainage, South China (Teleostei: Balitoridae). Ichthyological Exploration of Freshwaters, 21: 51-62. Hurlbert, S.H. 1971. The non-concept of species diversity: a critique and alternative parameters. Ecology, 52: 577-585. Junk, W.J. & K.M. Wantzen. 2004. The flood pulse concept: new aspects, approaches and applications - an update. Pp. 335-357. in: Proceedings of the Second International Symposium on the Management of Large Rivers for Fisheries: Volume 2, FAO Regional Office for Asia and the Pacific, Bangkok. Jutagate, T. 2009. Reservoir fisheries in Thailand. Pp. 96-113. in: S. S. De Silva & U.S. Amarasinghe (eds.), Reservoir Fisheries in Asia and Pacific. Network for Aquaculture Centers in Asia & Pacific, Bangkok. Kottelat, M. & Whitten, T. 1996. Freshwater biodiversity in Asia, with special reference to fish. World Bank Technical Paper No. 343. World Bank, Washington. 109    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Lévêque, C., T. Oberdorff, D. Paugy, M. L. J. Stiassny & P. A. Tedesco. 2008. Global diversity of fish (Pisces) in freshwater. Hydrobiologia, 595: 545–567. Magurran, A.E. 2004. Measuring biological diversity. Blackwell Publishing Company, Oxford. Mahujchariyawong, J. & S. Ikeda 2001. Modelling of environmental phytoremediation in eutrophic river – the case of water hyacinth harvest in Tha-chin River, Thailand. Ecological Modelling, 142: 121-134. Meksumpun, C. & S. Meksumpun. 2008. Integration of aquatic ecology and biological oceanographic knowledge for development of area-based eutrophication assessment criteria leading to water resource remediation and utilization management: a case study in Tha Chin, the most eutrophic river in Thailand. Water Science and Technology, 58: 2303-2311. Mills, M.D., R.B. Rader & M.C. Belk. 2004. Complex interactions between native and invasive fish: the simultaneous effects of multiple negative interactions. Oecologia, 141: 713–721. Myers, N., R.A.Mittermeier, C.G.Mittermeier, G.A.B. da Fonseca, & J. Kent. 2000. Biodiversity hotspots for conservation priorities. Nature: 403, 853-858. Na-Nakorn, U., W. Kamonrat & T. Ngamsiri. 2004. Genetic diversity of walking catfish, Clarias macrocephalus, in Thailand and evidence of genetic introgression from introduced farmed C. gariepinus. Aquaculture, 240: 145– 163. Ng, H.H. 2010. The monophyly and composition of the Asian hillstream catfish family Sisoridae (Teleostei: Siluriformes): evidence from morphology. Ichthyological Exploration of Freshwaters, 21: 247-278. Nguyen, T.T.T. & S.S. De Silva. 2006. Freshwater finfish biodiversity and conservation: an Asian perspective. Biodiversity & Conservation, 15: 35433568. Oberdorff, T., J.F. Guégan & B. Hugueny. 1995. Global scale patterns of fish species richness in rivers. Ecography, 18: 345–352. R Development Core Team. 2009. R: A language and environment for statistical computing, reference index Version 2.10. R Foundation for Statistical Computing, Vienna. 110    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Raghavan, R., G. Prasad, P.H. Anvar-Ali, & B. Pereira. 2008. Fish fauna of Chalakudy River, part of Western Ghats biodiversity hotspot, Kerala, India: patterns of distribution, threats and conservation needs. Biodiversity & Conservation, 17: 3119 – 3131. Senanan, W., A.R. Kapuscinski, U. Na-Nakorn & L. Miller. 2004. Genetic impacts of hybrid catfish farming (Clarias macrocephalus x C. gariepinus) on native catfish populations in central Thailand. Aquaculture, 235: 167–184. Sodhi, N.S., L.P. Koh, B.W. Brook & P.K.L. Ng. 2004. Southeast Asian biodiversity: an impending disaster. Trends in Ecology and Evolution, 19: 654-660. Takeuchi, K., A.W. Jayawarderna, Y. Takahasi & B. Machbub. 2005. Catalogue of rivers for Southeast Asia and the Pacific vol. I and II (CD-rom). UNESCO, Jakarta. Taki, Y. 1978. An analytical study of the fish fauna of the Mekong Basin as a biological production system in nature. Research Institute for Evolutionary Biology, Tokyo. Taylor, E.B. 2010. Changes in taxonomy and species distributions and their influence on estimates of faunal homogenization and differentiation in freshwater fishes. Diversity & Distribution, 16: 676-689. Tongnunui, S. & F.W.H. Beamish. 2009. Habitat and relative abundance of fishes in small rivers in eastern Thailand. Environmental Biology of Fishes, 85: 209– 220. Vidthayanon, C. 2003. Schistura pridii, a new nemacheiline loach (Teleostei: Balitoridae) from Upper Chao Phraya drainage, northern Thailand. Ichthyological Exploration of Freshwaters, 14: 307-310. Vidthayanon, C., P. Saenjundaeng & H. H. Ng.2009. Eight new species of the torrent catfish genus Oreoglanis (Teleostei: Sisoridae) from Thailand. Ichthyological Exploration of Freshwaters, 20: 127-156. Vidthayanon, C., J. Karnasutra. & J. Nabhitabhata. 1997. Diversity of freshwater fishes in Thailand. Office of Environment Policy and Planing, Bangkok. Vitule, J.R.S., C.A. Freire & D. Simberloff. 2009. Introduction of non-native freshwater fish can certainly be bad. Fish & Fisheries, 10: 98–108. 111    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Ward-Campbell, B.M.S., F.W.H. Beamish & C. Kongchaiya. 2005. Morphological characteristics in relation to diet in five co-existing Thai fish species. Journal of Fish Biology, 67: 1266–1279 Welcomme, R.L., K.O. Winemiller & I.G. Cowx. 2006. Fish environmental guilds as a tool for assessment of ecological condition of rivers. River Research & Applications, 22: 377-396. 112    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand P2 Life history of the riverine cyprinid Henicorhynchus siamensis (Sauvage, 1881) in a small reservoir Suvarnaraksha, A., Lek, S., Lek-Ang, S. and Jutagate, T. (2011), Journal of Applied Ichthyology, 27(4): 955-1000 113    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand 114    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand 115    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand 116    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand 117    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand 118    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand 119    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand 120    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand P3 Reproductive biology and conservation approaches of a vulnerable species Siamese Freshwater batfish (Oreoglanis siamensis) from foothill Himalayan, Thailand Suvarnaraksha, A., Lek, S., Lek-Ang, S. and Jutagate, T. (2011), (in preparation) 121    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Reproductive biology and conservation approaches of a vulnerable species Siamese Freshwater batfish (Oreoglanis siamensis) from foothill Himalayan, Thailand Apinun SUVARNARAKSHA1, 2, 3, , Sovan LEK 2, Sithan LEK-ANG 2, Tuantong JUTAGATE1, 1. Faculty of Agriculture, Ubon Ratchathani University, Warin Chamrab, Ubon Ratchathani, Thailand 34190 2. University of Toulouse III, Laboratoire Dynamique de la Biodiversité, UMR 5172, CNRS – UPS, 118 route de Narbonne, 31062 Toulouse Cedex 4, France 3. Faculty of Fisheries Technology and Aquatic Resources, Maejo University, Sansai, Chiangmai, Thailand 50290  Correspondence: Tel. +66-53-873470 Fax. +66-53-498178 E-mail: apinun@mju.ac.th   122    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand ABSTRACT A vulnerable and an endemic Freshwater batfish (Oreoglanis siamensis) was studied in 2006-2007 in a high mountain stream in northern Thailand (18° 06’ 19°10’ N and 98°04’ - 98°34’ E). This species was examined for reproductive biology preferences. Spawning in freshwater batfish occurred in late dry-cool season to early dry-hot season (January to April) in the upper tributaries of Maechaem river basin; at least 87.1-95.7% of female were in ripe or spawning condition in this season, while the sperm of male was mature and ripe through the year. Size at first maturity was 47 mm for males, and 53 for females. L50 estimates were 68.9±1.765 mm (males) and 82.4 ±1.369 mm (females). Maximum fecundity was 47 oocytes. Fecundity (F) varied from 18-47 (31.41 ± 7.67) for ripe females of 53-113 mm, respectively, correlation between TL and F and W and F followed a linear relationship (F=7.14+0.38TL; r2=0.424; or F= 20.41+2.3W; r2=0.491; n=71). Ripe oocytes have mean diameter of 2.96±0.28 mm (range = 2.5-4.2 mm; n=30). 123    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand INTRODUCTION The Siamese bat catfish (Oreoglanis siamensis Smith, 1933) is a red list vulnerable benthic species (Kottelat, 1996) inhabiting endemic to Inthanon mountain in Chiangmai province of northern Thailand. Then, it is the important for understanding the biology, life history and conservation propose. The Maechaem watershed is located in the West of Inthanon mountain composes a large tributaries of the Ping river basin. It is located 117 km south-western of Chiangmai city. The Maechaem sub-basin is bounded by coordinates 18° 06’ - 19°10’ N and 98°04’ 98°34’ E, and it covers a total area of 3,853 km2. The climate of this mountainous basin is defined by large variations in seasonal and annual rainfall that are influenced by Pacific-born typhoons, superimposed on the south-west monsoon (Walker, 2002). There are freshwater resources utilizing for urban and agricultural purposes. This resulted in increased concern for the future of the vulnerable freshwater stream species. Thus, the O. siamensis is particularly sensitive to any anthropogenic perturbations which disrupt stream flows for extended period. Figure 1. Map of Maechaem watershed. The studies concerning with the reproduction of tropical freshwater fishes revealed a diversity of life-history and its relationship to discharge regime. Many lowland species reproduce during the wet season, when the inundation of lateral 124    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand floodplains ensures an expanded habitat and a greater array and abundance of food (Fernandes, 1997). As a consequence many species of tropical and sub-tropical stream-dwelling fishes spawn during the dry season (Milton & Arthington 1983; Wooton 1990). Most of those studies have been carried out in stream tropical waters, with only a few at limit of biological data of tropical stream species, especially this species is seriously lacking of biological information. Studies of the ecology of Southeast Asia tropical freshwater fishes are limited and none have examined the reproductive biology of tropical stream-dwelling fishes. Studies of temperate stream fishes also emphasized on the importance of localized productivity and the acquisition of energy in determining the timing of reproduction (Encino & Granado-Lorencio 1997), and these factors may be of importance in tropical stream fishes aswell (Roberts 1989). The members of the genus Oreoglanis are distributed from the upper part of the Salween river basin (Vidthayanon et al., 2009), Chao Phraya river basin (Suvarnaraksha, 2003), and Mekong river (Rainboth, 1996). O. siamensis, however, was reported to occur only in Inthanon mountain range (Smith, 1945). Because of this torrent stream species was occurs in montane brooks and small high-gradient streams (Rainboth, 1996). Inhabits cold swift mountain streams and high altitude 500-1,200 m asl (Suvarnaraksha, 2003). Attaches itself to rock surfaces facing the current (Smith, 1945) and feeds on crustaceans and insect larvae (Vidthayanon, 2005). However, some of the hill tribe residents used it as a protein sources, land use change, and fragmentation of streams by human being along the habitat of this fish has driven it to the edge of extinction. It has a vulnerable red list species in Thailand (Kottelat, 1996). Despite being a vulnerable species, O. siamensis was little studied on biology e.g. growth, reproduction, and fecundity. However, it has a few numbers of eggs (per se A. Suvarnaraksha). The life-history characteristics and restrict to habitat make them be sensitive to the intense exploitation. The conservation of natural population and exploitation of sustainable resources of O. siamensis have been become increasingly matters of concerns. Unfortunately, to the best knowledge no work has been done on the biology and the life of O. siamensis. Then, the first study program of O. siamensis was initiated with the aim to understand the reproduction biology and length weight relationship of this vulnerable fish. 125    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand MATERIALS AND METHODS Study Area The Maechaem river watershed is located in Chiang Mai province of northern Thailand (Fig. 1). It is a major upper tributary sub-basin of the Ping river, which in turn, is the largest tributary of central Thailand’s Chao Phraya river, it is located 117 km southwestern of Chiangmai city. A large part of Maecheam river drainage was covered by mountains and forests (74.73%). The Maechaem sub-basin is bounded by coordinates 18° 06’ - 19°10’ N and 98°04’ - 98°34’ E, and covers a total area of 3,853 km2, west of Inthanon highest spot of Thailand (2,565 m a.s.l.). The depth of the sampling sites ranges between 0.25-2.0 m. with various bottom types (i.e. rock, gravel, sand, silk and mud). There are some small hill tribe villages in the area. Temperatures from mid-November to January average between 13°C and 28°C; the hills are even colder. Temperatures in Chiang Mai begin to rise in February and in the hot season (March-May) ranges between 17°C and 36°C. In the rainy season (Junemid November) (Fig. 2A). The average annual temperature ranges from 20 to 34°C and the rainy season is from May to October. The climate of this mountainous basin is defined by large variations in seasonal and annual rainfall that are influenced by Pacific-born typhoons, superimposed on the south-west monsoon (Walker, 2002) (Fig. 2B). The orographic effect induces an altitudinal increase of spatial rainfall distribution (Dairaku et al., 2000; Kuraji et al., 2001). B A Figure. 2 A: Average of temperature; B: average of rainfall and hubidity in Maechaem watershed (1988-2009). 126    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Sample collection Fishes were electrofished (Honda EM 650, DC 220 V 550BA 450VA,1.5–2 A, 50 Hz) in the upper Maechaem river system, between November 2007 and October 2008, through monthly sampling. Each tributaries sampling site was done at 45 to 60 minutes intervals or the area covering about 100 m2, The datas were collected with various microhabitat, substrate type i.e. rocky, sandy, and gravel, and habitat type (riffle, pool, and run) to cover all species distributions. The skin diving was carried out to observe the abundance and behavior of the fish. Fish captured in each part was kept separate after selected O. siamensis and fixed in 10% formalin and the life specimens was released to the their habitat after measurement and weight. Then, O. siamensis was identified and separated from the other species, sacrificed in a lethal solution of anesthetic, and conditioned in ice for transportation. The processe in evening at the rest room and the following data obtained: (i) total length (TL) to the nearest 0.1 mm (ii) total weight (WT) to the nearest 0.01 g (iii) sex (iv) gonad weight (GW) to the nearest 0.01 g. Gonads were removed from the visceral cavity, Prior to the preservation of the ovaries/testis were classified in a macroscopic scale of gonadal development, for both sexes; for females size and colour of oocytes was also registered and, for males sperm liberation when pressing the abdomen. According to these characteristics the following classification was considered: females – 2nd stage, immature, mature, and ripe; and males – 2nd stage, immature, mature, and ripe. Thereafter, ovaries were fixed in Bouin solution for oocytes measurements and counts. The specimens were fixed in 10% formalin. After one month, we were series of ethanol from 30%, 50% and preserved in 70% ethanol. Specimens were deposited at the Maejo Aquatic Resources Natural Museum. Ovaries preserved in Gilson’s fluid were stored for two weeks and shaken periodically to promote oocyte release. Oocytes were then cleaned by subsequent alcohol change and removal of the ovarian walls, and stored in a 70° GL alcohol solution. Fecundity (Bagenal & Braum, 1971) was determined after counting all vitellogenic oocytes from ripe ovaries and correlated to TL and WT. Spawning type was evaluated according to the distribution of oocytes diameter (measured on subsamples of 10ml under a compound microscope -x4) from dissociated ovaries in different maturation stages (mature, and ripe). 127    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand The reproductive patterns of the O. siamensis were assessed by two methods: gonadosomatic index (GSI) and histology. GSI was calculated for each fish for both sexes to determine the spawning seasons using the equation: GSI = [testis/ovary weight/body weight] x 100). For histological examination of gonads, a subsample of gonadal tissue was removed from each fish. These gonad samples were weighed (±0.01g), placed in tissue cassettes, dehydrated and impregnated with wax. Histological sections were cut at 8 µm from each block using a tissue microtome, mounted on glass microscope slides and stained with Harris’s haematoxylin and eosin counter stain. Each histological section was scored by estimating the percentage that each of the gonad maturity stages occupied within the total area of the section. Female gonads were classified into maturity stages: stage II (previtellogenic oocytes); immature (yolk precursor or non staining (primary) yolk); mature (red-staining (secondary) yolk); and spent; and for males: stages I, (primary germ cells and spermatogonia); immature (spermatocytes and spermatids); mature (spermatozoa); and spent. As no observable difference in scoring was detected between replicate blocks from the same fish, only one subsample was taken from the mid-position on a randomly selected gonad lobe for the remaining O. siamensis samples. Size at first sexual maturity (LMAT) was determined from the minimum total length of fish with developed vitellogenic eggs (maturity stages IV or V) for females and spermatids (maturity stages V or VI) for males. Gonads were classed as ripe when the majority of the gonad was in maturity stages IV and V for females and stages V and VI for males. Fish were in spawning condition when the greatest proportion of their gonad was in stage V (females) and stage VI (males). To estimate the size of fish in the population where 50% of fish in a length class were mature (L50). Fish were grouped into 10 mm total length classes to increase sample sizes. The logistic function was defined as; P 1 1  e a bL ------------------ (1) where a and b are constants and when calculated, the percentage at 50% maturity was replaced in upper equation to obtain the length at 50% maturity. 128    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand The condition factor (k) of the experimental fish was estimated from the relationship (Williams, 2000): K 100W L3 ----------------------- (2) where K=condition factor, W= weight of fish (g), and L= length of fish (mm).  Fecundity (Bagenal & Braum, 1971) was determined after counting all vitellogenic oocytes from ripe ovaries and correlated to TL and TW in equation (3). F=aTLb, and F=aTWb -------- (3) The relationship between the length (TL) and weight (W) of fish was expressed by equation (Pauly, 1983): W=aLb --------------------------- (4) where W=Weight of fish in (g), L=Total length (TL) of fish in (mm), a Constant (intercept), and b=The length exponent (slope). The a and b values were obtained from a linear regression of the length and weight of fish. The correlation (r2), which is the degree of association between the length and weight, was computed from linear regression analysis: R=r2. RESULTS Environmental conditions The fish tended to stay in areas with clear, slightly alkaline water with high level of dissolved oxygen, water temperature was less than 20 °C and moderate fast flow (Table 1). Monthly water flow at riffles was in the dry season (dry-cool and dryhot season), then increased in rainy season. Flows in pools and runs were slower than at slope high slope. Bottom substrates at the stations were stone, rocks and gravel and surrounded by large rock. The stream canopy was cover by large three and high humidity, moss and fern were growing along the stream bank. Many of them were 129    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand inhabits at the creeks of the stream, they were lied on the rocky or stone bottom by flat ventral of body for feeding and against the water flow. The dorsal part of O. siamensis coloration was mimic to the rock color shelter. While, in the spawning season were found the sinking eggs in the pool with lower flow. Table 1. Environmental parameters at the sampling sites where O. siamensis were observed. Environmental parameters pH Dissolved Oxygen Temp Alkalinity Hardness Total Dissolve Solid Conductivity Flow Stream Depth Stream Width Nitrite Ammonia 7.74-8.20 5.5-8.4 mg/l 15.93-19.93 °C 50-76 ppm. 93.3-128.6 ppm. 40-160 ppm. 50.7-160.0 µS/cm. 19-100 cm/sec-1 17-60 m. 2-8 m. 0.002-0.003 ppm. 0.001-0.004 ppm. Reproductive conditions Siamese bat catfish could not clearly express the secondary sexual characteristic, it was difficult to distinguishable except during the spawning season. Female, the belly is enlarged, swelled and flat from top view, and large yellow egg can be seen. Genital papilla was enlarged and urogenital pore is magnified, round tip and reddish. Male, has a protrude genital papillae and urogenital pore is enlarged and smaller size than female. Of the total of 249 Oreoglanis siamensis studied, 170 (48%) ZHUHPDOHV  ZHUHIHPDOHV7KHVH[UDWLRZDV Ȥ2-test, p<0.05). The size at 50% maturity was 68.9 mm TL (SD 1.765) in males, and 82.4 mm TL (SD 1.369) in males (Fig. 6). Smaller females were first mature 53 mm TL, and the smallest mature male was 47 mm TL. In O. siamensis gonadal maturation followed a similar annual pattern (Fig. 4). Between late of dry-hot season to beginning of rainy season (May to August), the majority of the collected fish were in stage II (post-spawning) and immature stage (early preparatory periods). In the late of that rainy to early dry-cool season, the 130    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand mature individuals (mature stage) were more abundant in September to December (pre-spawning period). During January to March or/and April, high percentage of fish specimens collected were in the ripe stage (Fig. 4). Figure 3. Length at 50% maturity, female (circle) and male (diamond) O. siamensis, Maechaem w a t e r s h e d , as indicated by percentage of sample maturing in each length class (10 mm interval). Figure 4. Percent frequency of maturity stage of O. siamensis. 131    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Figure 5. Histological appearance of ovary maturation of O. siamensis (n=9 per month). Abbreviation: Nu=nucleus, FE=follicle epithelial, YG=Yolk vesicle; Note a) ripe and spent stage (dry-hot season), a1) late ripe stage, a2) spent stage, a3) spent stage, b) late spent, primary stage and immature stage (rainy season), b1) late spent stage, b2-3) primary stage and immature stage, and c) mature and ripe stage (dry-cool season), c1-2) mature stage, c3 ripe stage. Figure 6. Histological appearance of testis maturation of O. siamensis (n=9 per month). Abbreviation: SP=spermatozoa, Note a) ripe and spent stage (dry-hot season), a1) mature stage, a2) mature stage, a3) mature stage, b) late mature, primary stage and immature stage (rainy season), b1) late mature stage, b2-3) primary stage and immature stage, and c) mature stage (dry-cool season), c13) mature stage, c3 ripe stage.  132    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Microscopic study revealed similar characteristics in the gonadal tissue for the species. In Fig. 3, gonadal maturity stage in female are presented. Spent stage (after spawning) observed during April to June (dry-hot season) and July (starting to rainy season) (Fig. 5a1, 5a2, and 5a3). The ovaries in second stage and immature stage were observed during August to October (Rainy season) (Fig. 5b1, 5b2 and 5b3). While, they were started to mature stage in November to December (early dry-cool season) and ripe stage in January to February (late dry-cool season) (Fig. 5c1, 5c2 and 5c3). Also, some of specimens were ripe in March to early April. The late of the dry-cool season testis was in stage V (Fig. 6a1, 6a2 and 6a3). The testis in stage II and III (Fig. 6b1, 6b2 and 6b3) presented seminiferous tubules with cells at different stages of development. In mature stage V (Fig. 6c1, 6c2 and 6c3), testes showed the lumen of seminiferous tubules filled with spermatozoa (s). The spermatogonia and the seminiferous tubules were observed along the entire testis. Table 2. The condition factor (K) in O. siamensis. Month Jan Feb Mar Apr May Jun Condition factor 1.01±0.39 0.82±0.17 0.93±0.20 0.86±0.19 0.85±0.22 0.87±0.33 Month Jul Aug Sep Oct Nov Dec Condition factor 0.95±0.26 0.96±0.30 0.92±0.17 0.89±0.19 0.89±0.19 1.18±0.39 Of the total of 179 adult females sampled, only 96 ovaries were found to be suitable for an estimate of fecundity. The females studied ranged from 53 to 113 mm TL, and were captured along year round cycle. The condition factor ranged from 0.82 to 1.18±0.09 during the period studied, which maximum valued in dry-cool season (December to January) (Table 2). Fecundity (F) varied from 18-47 (31.41 ± 7.67) for ripe females of 53-113 mm, respectively, correlation between TL and F and W and F followed a linear relationship (F=7.14+0.38TL; r2=0.424; or F= 20.41+2.3W; r2=0.491; n=71). Egg character of Siamese bat catfish is rounded-shape, ripened egg is pale-yellowish color, transparent and glossy. The egg type is demersal but not 133    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand sticky. Ripe oocytes have mean diameter of 2.96±0.28 µm (range = 2.5-4.2 µm; n=30). Considering that the mean number of oocytes per gram weight is independent of fish size the mean number of oocytes per grams of body weight as 7 oocytes. A total of 532 specimens were analyzed, being the value obtained for the length-weight relationship showed that the O. siamensis was allometric in its growth. Ranging from 20-117 mm, 532 samples were used in the analysis. The relationship was derived from unsexed samples since there is no external sexual dimorphism. The equation derived was W=0.00005L2.738 (r2=0.947) (Fig. 7). Figure 7. Length-weight relationships of O. siamensis. DISCUSSIONS The streams in northern Thailand, like in many other tropical streams, are characterized by a steep topography, fast flow, rocky bottom, canopy cover, and high level of dissolved oxygen. Nevertheless, the fish still have to well adapted to the special habitat e.g. Homaloptera spp., Balitora spp., and Glyptothorax spp. (Kottelat, 2001). Also, O. siamensis was well adapted by flatten belly, adhesive maxillary barbel and pair fins, stream line body shape, and aerodynamic dorsal part. These characteristics were suitable for feeding on the small invertebrate and aquatic insect larvae on the rocks (Vidthayanon, 2005). It could tolerate a low water temperature in high altitude which might limit the growth of the O. siamensis food items (Han et. al., 2000). The environmental condition of O. siamensis was abundant along the habitats 134    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand in Maechaem Stream showed that O. siamensis inhabited the waters between 500 to 1200 m altitudes. The early stages such as eggs and larvae stages are of great important for fishes, then the reproductive tactics in teleostean fish involving the allocation of a size-dependent reproductive effort between fecundity and egg size. The demersal species tend to produce large and few eggs, the larger eggs and the larvae hatching from them are more likely to survive than smaller ones, but Duarte & Alcaraz, 1989 reported no evidence of evolutionary trends towards greater eggs. They were reduce the variance in growing conditions, should be more dependent on the survival of the individual larvae, which increases as egg size increases. Also, O. siamensis is a demersal steam species, it produces large oocytes and few numbers like some of parental care species (Paugy, 2002) e.g. Xynobagrus nigri (Olurin & Odeyemi, 2010) and Notopterus notopterus (per se A. Suvarnaraksha) or rainbow trout, Sea back trout, and brook trout (Serezli et al., 2010). While, their fecundities were very small number of eggs compare with the other glyptothorine species e.g. Glyptothorax madraspatanum (18 to 47 vs. 1640 to 6830) (Dobriyal & Singh, 1993) and a little bit fewer number of eggs than parental care species (Paugy, 2002). The fecundity and egg size was related, egg size is one of the important determinants of eggs and larval quantity as it is positively correlated with both survival of egg and larval and also of the growth of the larvae (Gall, 1975). But, Elger (1990) reported the product of clutch size and egg volume is not correlated with either clutch size or egg volume after removing the effects of body size. Furthermore, as larger eggs sizes often take longer to hatch than smaller eggs, they are at risk from predation or adverse abiotic conditions for longer periods of time (Miller et al., 1988); it was related to the report of Unsrisong et al., 2005. The adults and juvenile were found in the same habitats, it is possible a non-migratory species. The O. siamensis was spawn in the late dry-cool to dry-hot season (January to April) of Thailand, this conformed to a study of Unsrisong et al., 2005, but a liitle bit early. Meanwhile, it was difference with lowland tropical stream species reproduction according to rainfalls regiems (Alkins-Koo, 2000; Chellappa et al., 2009). In the dry season, reduced stream flow and a reduced spate frequency ensure a more benign physical environment than during the wet season, and food may also be more 135    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand abundant at this time as well. Moreover, the wet-season primary production may be reduced because of a combination of increased cloud cover associated with the monsoonal wet season and high suspended sediment loads during the period of elevated discharge, both of which limit light availability for primary producers (Pusey et al., 2001). The main habitats were in the high elevation and canopy cover. The spawning season sufficient data on seasonal freshwater fish egg variations are not available, but the time of spawning does appear to be linked with the availability of food for the larvae in both lake and stream species (Bagenal, 1971). Then, the few numbers of eggs restrict to the habitat of O. siamensis led to endanger or extinct in the near future. CONCLUTIONS The first report showed the dry-cool to dry-hot spawning season of mountainous vulnerable species in northern of Thailand and tropical Southeast Asia. The situation of low fecundity, restrict to the specific habitat, and anthropology disturbs were one of the chance to be extinct in the near future. Then, this vulnerable species should prevent aggression from human activities and more study of their life history and strategies for management. REFERENCE Alkins-Koo, M. 2000. Reproductive timing of fishes in a tropical intermittent stream. Environmental Biology of Fishes 57: 49–66, 2000. Encino, L. and Granado-Lorencia, C. 1997. Seasonal changes in condition, nutrition, gonad maturation and energy content in barbel, Barbus sclateri, inhabiting a fluctuating river. Environmental Biology of Fishes 50: 75–84. Bagenal, T. B. (1971), The interrelation of the size of fish eggs, the date of spawning and the production cycle. Journal of Fish Biology, 3: 207–219. Bagenal, T. B., and E. Braum. 1978. Eggs and early life history. Pages 166-198 in W. E.Ricker, editor. Methods for assessment of fish production in fresh waters. Blackwell Scientific Publications, Oxford and Edinburgh. Duarte, C.M. and M. Alcaraz. 1989. To Produce Many Small or Few Large Eggs: A Size-Independent Reproductive Tactic of Fish. Oecologia. 80(3): 401-404 136    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Chellappa, S., R.M.X. Bueno, T. Chellappa, N.T. Chellappa and V.M.F. Almeida e Val. 2009. Reproductive seasonality of the fish fauna and limnoecology of semi-arid Brazilian reservoirs. Limnologica 39: 325–329 Dairaku, K., K. Kuraji, M. Suzuki, N. Tangtham, W. Jirasuktaveekul, and K. Punyatrong. 2000. The effect of rainfall duration and intensity on orographic rainfall enhancement in a mountainous area: a case study in the Mae Chaem watershed, Thailand. Journal of the Japan Society of Hydrology and Water Resources 13(1): 57-68. Dobriyal, A. K. and Singh, H. R. (1993), Reproductive biology of a hillstream catfish, Glyptothorax madraspatanum (Day), from the Garhwal, Central Himalaya, India. Aquaculture Research, 24: 699–706 Elger, M.A. 1990. Evolutionary Compromise between a Few Large and Many Small Eggs: Comparative Evidence in Teleost Fish. Oikos. 59(2): 283-287 Fernandes, C.C. 1997. Lateral migration of fishes in Amazon floodplains. Ecology of Freshwater Fish 6: 36–44. Gall, G.A.E., 1975. Genetics of reproduction in domesticated rainbow trout. J. Anim. Sei., 40: 19-28. Han, C.-C., K.S. Tew, I-S. Chen, L.-Y. Su and L.-S. Fang. 2000. Environmental biology of an endemic cyprinid, Varicorhinus alticorpus, in a subtropical mountain stream of Taiwan. Environmental Biology of Fishes 59: 153–161, Kottelat, M. 1996. Oreoglanis siamensis. In: IUCN 2010. IUCN Red List of Threatened Species. Version 2010.4. <www.iucnredlist.org>. Downloaded on 16 March 2011. Kottelat, M. 2001. Fishes of Laos. WHT Publications Ltd., Colombo 5, Sri Lanka. 198 p.Kuraji, K., K. Punyatrong, and M. Suzuki. 2001. Altitudinal increase in rainfall in the Mae Chaem watershed, Thailand. Journal of the Meteorological Society of Japan. 79(1B): 353-363. Miller, T.J., Crowder, L.B., Rice, J.A. and Marschall, E.A. (1988) Larval size and recruitment mechanisms in ®shes: toward a conceptual framework. Can J. Fish. Aquat. Sci. 45, 1657±1670. Milton, D.A. and Arthington, A.H. 1983. Reproduction and growth of Craterocephalus marjoriae Whitley and C. stercusmuscarum (Gunther) 137    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand (Pisces: Atherinidae) in south-eastern Queensland, Australia. Freshwater Biology 13: 589–598. Olurin K.B. and O.I. Odeyemi. 2010. The Reproductive Biology of the Fishes of Owa Stream, South - West Nigeria. Research Journal of Fisheries and Hydrobiology, 5(2): 81-84 Paugy, D. 2002. Reproductie strategies of fish in a tropical temporary stream of the Upper Senegal basin: Baoulé River in Mali. Aquatic Living Resources, 15(1): 25-35 Pusey BJ, Arthington AH, Bird JR, Close PG. Reproduction in three species of rainbowfish (Melanotaeniidae) from rainforest streams in northern Queensland, Australia. Ecology of Freshwater Fish 2001: 10: 75–87. Rainboth, W.J. 1996. Fishes of the Cambodian Mekong. FAO Species Identification Field Guide for Fishery Purposes. FAO, Rome, 265 p. Roberts, T.R. 1989. The freshwater fishes of western Borneo (Kalimantan Barat, Indonesia). Memoirs of the Californian Academy of Sciences 14: 1–210. Riehl, R. and H.A. Baensch. 1991. Aquarien Atlas. Band. 1. Melle: Mergus, Verlag für Natur- und Heimtierkunde, Germany. 992 p. Serezli, R., S. Guzel and M. Kocabas. 2009. Fecundity and egg size of three salmonid species (Oncorhynchus mykiss, Salmo labrax, Salvelinus fontinalis) Cultured at the Same Farm Condition in North-Eastern, Turkey. J ournal of Animal and Vet erinar y Advances . 9(3): 576-580 Smith, H.M., 1945, The fresh-water fishes of Siam, or Thailand., Bull. U.S. Natl. Mus. 188: 633 p. Suvarnaraksha, A. 2003. Fish Diversity of Chiang Dao Wild Life and Sanctuary. Final report, Maejo University, Chiangmai. 62 p. Unsrisong, G., P. Pornsopin, S. Kantiyawong, B. De Lapeyre, S.Wesels, and G. Horstgen-Schwark. 2005. Induced Spawning of Batfish (Oreoglanis siamensis). “The Global food & Product Chaindynamics, Innoations, Conflicts Strategies” Deutscher Tropentag, October 11-13, 2005, Hohenheim. Vidthayanon, C. 2005. Thailand red data: fishes. Office of Natural Resources and Environmental Policy and Planning, Bangkok, Thailand. 108 p. Vidthayanon, C., Saenjundaeng, P., and Ng, H. H. 2009. Eight new species of the 138    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand torrent catfish genus Oreoglanis) Teleostei: Sisoridae) from Thailand. Ichthyological Exploration of Freshwaters. 20(2): 127-156. Walker, A. 2002. Agricultural transformation and the politics of hydrology in northern Thailand. Development and Change. 34(5): 941-964. Williams, J. E. 2000. The Coefficient of Condition of Fish. Chapter 13 in Schneider, James C. (ed.) 2000. Manual of fisheries survey methods II: with periodic updates. Michigan Department of Natural Resources, Fisheries Special Report 25, Ann Arbor. Wooton, R.J. 1990. Ecology of teleost fishes. London; Chapman and Hall. 404 pp. 139    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand 140    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand P4 Fish diversity and assemblages patterns along the longitudinal gradient of tropical river in the Indo-Burma hotspot region (the Ping-Wang river basin, Thailand) Suvarnaraksha, A., S. Lek, S. Lek-Ang and T. Jutagate Hydrobiologia (Revised) 141    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Fish diversity and assemblage patterns along the longitudinal gradient of tropical river in the Indo-Burma hotspot region (the Ping-Wang river basin, Thailand) Apinun SUVARNARAKSHA1, 2, 3 *, Sovan LEK 2, Sithan LEK-ANG 2 and Tuantong JUTAGATE1 1. Faculty of Agriculture, Ubon Ratchathani University, Warin Chamrab, Ubon Ratchathani, Thailand 34190 2. University of Toulouse III, Laboratoire Dynamique de la Biodiversité, UMR 5172, CNRS – UPS, 118 route de Narbonne, 31062 Toulouse Cedex 4, France 3. Faculty of Fisheries Technology and Aquatic Resources, Maejo University, Sansai, Chiangmai, Thailand 50290 Running title: Fish assemblages in a rhitral environment in Thailand * Tel: +66-53-498178 ext 401 Fax: +66-53-873470 ext 130 E-mail: apinun@mju.ac.th Current address: Faculty of Fisheries Technology and Aquatic Resources, Maejo University, Sansai, Chiangmai, Thailand 50290 142    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand ABSTRACT Fish diversity and assemblage patterns along the longitudinal gradient of the Ping-Wang river basin were investigated. Diversity study was based on data from 272 samplings, collected between January 1996 and April 2009. Sixteen physicochemical water quality- and geo-morphological- parameters were also examined at each sampling as well as area and the percentage of 3 types of land-uses of each sub-basin. One hundred and ninety two fish species were collected and the most diverse family was Cyprinidae (76 species) followed by Balitoridae (20 species) and Cobitidae (13 species). The highest and lowest diversity values were obtained in the “Maeklang” and “lower Ping” sub-basins, respectively. Six physicochemical parameters (i.e. DO, water temperature, pH, conductivity, phosphorus and alkalinity) and six geomorphological parameters (i.e. altitude, distance from the sea, discharge, depth and width) were statistically significant in their relationships to diversity parameters (Pvalue < 0.05). Results from the classification and regression trees showed that the geo-morphological parameters were more significant in controlling and predicting both species richness and Shannon diversity index than the physicochemical parameters, in which altitude was the most significant. Fifty-three fish species from 220 samplings were patternized into 4 assemblage patterns viz., mountainous, piedmont, transitory and lowland species. Any environmental changes in the rhitral environment will seriously impact to the distribution of species in the mountainous and piedmont assemblages. Keywords Fish diversity, Environmental variables, Prediction, Assemblage patterns, CCA, Indo-Burma Hotspot, Thailand 143    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand INTRODUCTION Variations in geomorphology characteristics of the river as well as environmental variables, both biotic and abiotic, are the major factors that govern riverine fish communities both in terms of species richness and distribution of individual species (Orrego et al., 2009; Alexandre et al., 2010; Kimmel and Argent, 2010). Knowledge on this issue has been widely reported both on regional and local scales but is still very poor for the Indo-Burma, the third largest global biodiversity hotspot (Myers et al., 2000), particularly on the species living exclusively in the headwater section (i.e. rhithral environment), where are difficult to access. So far, more than 1,260 freshwater fish species in the region (i.e. about 10 % of global freshwater fishes) have been reported and more than 560 of this species are endemic (Conservation International, 2010). The longitudinal gradient of river course can be divided into upper (i.e. rhithron), middle and lower (i.e. potamon) sectors, in which each area has its own characteristics of species assemblages, though overlapping to some degree (Schmutz et al., 2000). To evaluate the status and any changes of assemblages in each section over time, diversity indices are commonly used and the commonest indicator is the number of species found, i.e. species richness (Oberdorff et al., 2002; de Thoisy et al., 2008; He et al., 2010). This indicator is an integrative descriptor of the animal community, influenced by a large number of natural environmental factors as well as anthropogenic disturbances, including the geological history of the area, environmental stability, ecosystem productivity and heterogeneity (Lenat, 1988; Céréghino et al., 2003; He et al., 2010). It is suggested that if the physical aspects of the stream are relatively stable, they are responsible for the consistent pattern in biological community structure (Orrego et al., 2009) even though some other factors may have an influence, such as competition, predation, point and non-point pollution sources (Ibarra et al., 2005; Orrego et al., 2009) as well as hydraulic stress (Welcomme et al., 2006). The occupancy by species of particular sections throughout the length of a river depends on the extent that specific needs are supplied by the locally available resources, especially food and shelter (Tomanova et al., 2007; de Oliveira & Eterovick, 2009). The species that is exclusively present in a particular section, 144    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand incorporated with the studies on habitat disturbance gradient, could be the bioindicator to evaluate ecological integrity of that zone (Lasne et al., 2007). Many classifications of running waters, notably fish-based classifications, have been proposed since the end of the 19th century (e.g. Huet, 1959) and are becoming more important since the last decade, especially when the anthropogenic impacts are accelerated (e.g. Schmutz et al., 2000; Welcomme et al., 2006) because the deviation between the observed assemblage type and the one expected in undisturbed conditions provides an assessment of their ecological status (Lasne et al., 2007). Figure 1. Location and map of the Ping-Wang river basin (showing also the locations of 10 sub-basins) Due to the fact that most of the areas in this basin are almost intact and less disturbed by urbanization, fish community structure, in this study, reflected the longitudinal river gradient, which would closely related to the gradual change in habitat diversity (Ferreira & Petrere, 2009; He et al., 2010). In this study, we aimed 145    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand to draw out the perspectives of (a) the relationship between biotic and abiotic variables as descriptors to predict their influences on the fish community in terms of species diversity and (b) identification of the fish species community structures along the longitudinal gradient in a rhithral environment of a large scale of river system located in the Indo-Burma hot spot, i.e. the Ping-Wang River Basin, where high concentrations of endemic fish species are evident and are undergoing immense habitat loss, especially urbanization and infrastructure developments (Sodhi et al., 2004; De Silva et al., 2007; Dugan et al., 2010). Materials and methods Study area The Ping - Wang river-system is the major river-system of northern Thailand (Fig. 1) and located in the Chao Phraya river basin and a high altitude river basin in IndoBurma. The Ping river is 740 km long with a catchment area of about 33,896 km2. The Wang river is 440 km long and has a catchment area of 10,791 km2 (Takeuchi et al., 2005). The Wang river flows southwest ward to join the lowland of Ping river at Tak province to form a large watershed area lying between 15o42’ and 19o48’ North and 98 o04’ and 100o08’ East. The highest altitude of this river system is at 2,565 m ASL and connected to the lower Chao Phraya river basin at the altitude of 48 m ASL. Fish data The databases of fish samples were compiled during the ichthyological surveys in running water of the Ping-Wang River-system between January 1996 and April 2009 (A. Suvarnaraksha, own collected data) and no major changes in land-uses were observed during the sampling period. The total number of sampling sites was 272, which were selected to cover the main rivers and tributaries of the Ping-Wang riversystem. The sampling sites were distributed among 10 sub-basins in the river-system (Fig. 1), where a Digital Elevation Model (DEM) was used to define and divide the geographical range of the Ping-Wang river-system into sub-basins by ArcView GIS 9.2, according to the catchment area and fish sample spots. Each sub-basin was visited to cover both in dry and wet seasons. The total number of sampling sites was 272. 146    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand a each sub- basin Sub-basin Upping Ping (UP) Maetang (MT) The second Ping (SP) Maeklang(MK) 147 Maecheam (MC) The third Ping (TP) Maeteon (ME) The forth Ping (FP) Lower Ping (LP) Wang river (WA) Geographic Coordinate 19°07’-19°48’ N 98°47’-99°17’ E 19°10’-19°45’ N 98°27’-98°55’ E 18°31’-19°33’ N 98°24’-99°22’ E Bottom G, P, R, S G, P, R, S G, P, R, S 18°24’-18°35’ N 98°28’-98°41’ E 17°57’-19°09’ N 98°04’-98°37’ E 17°48’-18°43’ N 98°14’-98°44’ E G, P, R, S 17°13’-18°02’ N 98°14’-98°34’ E 15°50’-17°49’ N 98°39’-100°02’ E 15°42’-16°10’ N 99°27’-100°08’ E 17°07’-19°24’ N 99°00’-100°06’ E G, P, R, S G, P, R, S G, S, M G, S, M G, S, M G, S, M Categorical identities Fast flowing and clear water, rocky, gravel, pebble and sandy bottom, stream enclosed by forest canopy. Fast flowing and clear water. Rocky, gravel, pebble and sandy bottom, covered by forest canopy. Fast flowing and stagnant water. Clear or turbid water. Rocky, gravel, pebble, muddy and sandy bottom. Partially covered by forest canopy, agricultural area and urban. Fast flowing and clear water, rocky, gravel, pebble and sandy bottom, stream enclosed by forest canopy. Fast flowing and clear water, rocky, gravel, pebble and sandy bottom, stream enclosed by forest canopy. Slow flowing or stagnant water and turbid water. Gravel, muddy and sandy bottom. Partially covered by forest canopy, agricultural area and urban. Fast flowing and clear water, rocky, gravel, pebble and sandy bottom, stream enclosed by forest canopy. Slow flowing and turbid water. Gravel, muddy and sandy bottom. Partially covered by forest canopy, agricultural area and urban. Slow flowing and turbid water. Gravel, muddy and sandy bottom. Partially covered by forest canopy, agricultural area and urban. Fast flowing and clear water on upper reaches and slow flowing and turbid water in lower part of Wang river. Gravel, muddy and sandy bottom. Partially covered by forest canopy, agricultural area and urban. Note R = rocky, G = gravel, P = pebble, S = sandy and M = muddy a     147 Collection period 2008 No. of Stations 6 2000-2001 2003-2004 1996, 20032004, 2008 2008 48 2007-2008 44 2005-2006, 2009 18 2008 24 2009 6 2009 4 2009 18 98 6 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 1 Descriptions of the sub-basins in the Ping-Wang River Basin, collection period and number of stations in    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Each site was single visit and chosen on the basis of accessibility, similarity in habitat types, and to maximize the diversity of habitat types (pools, cascade, falls, riffles, and stagnant water) at each sub-basin (Table 1). To gather all species within sampling site, fish samples were collected by using various fishing methods such as small and large seines, cast-nets, gillnets of various mesh sizes, and traps (Table 1). The electro-fishing did supplement sampling with an AC shocker (Honda EM 650, DC 220 V 550BA 450VA, 1.5–2 A, 50 Hz), which was placed on the riverbank together with block nets and scoop nets. Live fishes were identified in the field, measured for total length (mm), counted, and then returned back to the water. Only a few samples of individual species were anaesthetized in dilute solution of benzocaine (50 mg/l) and kept separately according to species level for further taxonomical reference. Specimens were preserved in formalin, identified in the lab by several related publications e.g. Smith 1945, Taki, 1974, Kottelat 1985, 1998, 2001, Roberts 1993, 1994, Rainboth 1996, Vidthayanon et al., 1997 and others. And then, specimens were deposited in the Maejo Aquatic Resources Natural Museum (MARNM). Fish data was presented in terms of diversity parameters as species richness, Simpson Dominance index, species evenness and Shannon diversity index (Shannon, 1948). Environmental parameters The physicochemical water quality parameters were measured at each sampling, incOXGLQJZDWHUWHPSHUDWXUH :7ƒ& FRQGXFWLYLW\ &21ȝ6FP WRWDOGLVVROYHG solids (TDS; mg/l), dissolved oxygen (DO; mg/l), and pH, and were detected in situ by using a YSI 556 (MPS) multi-probe system. Water was sampled for laboratory analyses of nitrite (NIT; mg/l), ammonia (AMM; mg/l), total phosphorus (PO4+; mg/l), alkalinity (ALK; mg/l) and hardness (HAR; mg/l) following APHA (1989) protocols. The geo-morphological parameters were also obtained from each sampling. Water depth (DEP; m) and stream width (WID; m) were measured at the beginning, middle and end of the each sampling site. Velocity of the water flow was measured by flow-meter (G.O. Environmental model 1295, VEL; m/s) and measurement was conducted at least three times (i.e. at the middle of the stream and both the bank sides) 148    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand and the mean values were used. Discharge (DIC; m3/s) was calculated as Q = AV, where Q is discharge, A is the cross-sectional area of the channel and V is the average flow velocity. The altitude (ALT; m ASL) of the sampling site was provided by a GPS GarmineTrex VISTA. ArcView GIS 9.2 was used to estimate landscape position i.e. distance from the sea (DFS; km), watershed area (WSH; km2), and the land cover (i.e., forest area (FOR; %), agricultural area (AGR; %) and urban area (URB; %)). Statistical analyses A matrix data of numbers of fish captured in each species at each site was made for further analyses. The linear regression model (Cade & Noon, 2003) was used to examine the relationships between individual environmental parameters and species richness and also Shannon diversity index (Gutiérrez-Estrada et al., 2008). The classification and regression tree (CART: Breiman et al., 1984) was optimized from a set of environmental parameters and aimed at predicting species richness and Shannon diversity index by site. The cost-complexity pruning was used to prune the regression tree (Breiman et al., 1984). For making CARTs, species richness was log (x+1) transformed to stabilize variances (He et al., 2010). The optimal tree size was determined by r2-value and the complexity parameter. The data of 192 fish species from 272 samplings was rearranged by eliminating the species that occurred less than 5 % of total samplings and the samplings that contained less than 5 % of total species. Then, after less than 5% eliminating, a matrix of 53 fish species, according to 220 samplings, was employed and the data were transformed into presence/absence data. Relationships between fish assemblages and environmental parameters were examined by Canonical Correspondence Analysis (CCA), an ordination technique designed for direct analysis of relationships between multivariate ecological data (Ter Braak, 1986). Statistical significance, for CCA, of the relationship between a set of environmental factors and fish species was taken using a Monte Carlo permutation test with 999 permutations and was accepted at P-value < 0.05. Ward’s hierarchical agglomerative clustering was used to classify the group of fish species based on their similarity in occurrences (Ward, 1963). All statistical analyses were performed by using an R-statistical 149    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand software (Ihaka & Gentleman, 1996) using packages “stats” (R Development Core Team, 2010), “rpart” (Therneau & Atkinson, 2010) and “ade4” (Chessel et al., 2004). RESULTS A total of 192 species within 11 orders, 33 families were collected. The most diverse family was Cyprinidae (76 species) and followed by Balitoridae (20 species), Cobitidae (13 species) and Bagridae (10 species) and the remaining families were contained less than 10 species (Table 2). The greatest species richness was found in the lower portion of the river-system “Lower Ping” Sub-basin (112 species) while the minimum species richness was obtain at the highest altitude “Maeklang” Sub-basin (11 species) and there were similar trends for Shannon diversity index, Simpson Dominance index, species evenness and species richness (Table 3). However, numbers of individuals per 100 m2 and biomass (kg) per hectare were scattered among sub-basin (Table 3). Relationships between the environmental and diversity parameters Summary of values of the 20 environmental parameters in each sub-basin is presented in Table 4. Five geo-morphological parameters, i.e. altitude, distance to the sea, discharge, depth, and width, showed high statistical significance in their relationships to diversity parameters (P-value < 0.001, Fig. 2). The higher r2 values (i.e. strong relationships) of geo-morphological parameters and diversity indices were found (Fig. 2), when compared to those physicochemical parameters (Fig. 3). Altitude and distance to the sea showed strongly negative relationships for both indices, implying that higher diversity was found in the lower altitude, which was close to the sea and then declines as the altitude increases. It was also observed that diversity indices in the low levels of the three remaining parameters fluctuated widely and they all showed positive trend, i.e. the higher the value, the higher the diversity indices. The high altitude sub-basins (i.e. the Upper Ping, Maetang, Maeklang and Maeteon) showed the characteristics of low water temperature, high water current velocity and non-polluted area. Meanwhile, the high percentage of agricultural and urban area in the fourth Ping sub-basin would dedicate to the lower dissolved oxygen and higher in nitrite and ammonia compared to the other sub-basins. There were six 150    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand a Scientific name 151 Myliobatiformes/Dasyatidae Himantura chaophraya Monkolprasit and Roberts, 1990 Himantura signifer Compagno & Roberts, 1982 Osteoglossiformes/Notopteridae Chitala ornata (Gray, 1831) Notopterus notopterus(Pallas, 1769) Clupeiformes/Clupeidae Clupeoides borneensis Bleeker, 1851 Cypriniformes/Cyprinidae Albulichthys albuloides (Bleeker, 1855) Amblyrhynchichthys truncatus (Bleeker, 1851) Bangana sinkleri (Fowler, 1934) Barbichthys laevis (Valenciennes, 1842) Barbichthys nitidus Sauvage, 1878 Barbonymus altus (Günther, 1868) Barbonymus gonionotus (Bleeker, 1850) Barbonymus schwanenfeldii (Bleeker, 1853) Barilius koratensis (Smith, 1931) Barilius pulchellus (Smith, 1931) Cirrhinus cirrhosus (Bloch, 1795)* Cirrhinus molitorella (Valenciennes, 1844) Cosmochilus harmandi Sauvage, 1878 Crossocheilus cobitis (Bleeker, 1853) Crossocheilus reticulatus (Fowler, 1934) Ctenopharyngodon idellus (Valenciennes, 1844) Cyclocheilichthys apogon (Valenciennes, 1842) Cyclocheilichthys armatus (Valenciennes, 1842) Cyclocheilichthys enoplos (Bleeker, 1851) Cyclocheilichthys repasson (Bleeker, 1853) Cyprinus carpio Linnaeus, 1758* Danio albolineatus (Blyth, 1860) Devario aequipinnata (McClelland, 1839) Devario maetangensis (Fang, 1997) Devario malabaricus (Jerdon, 1849) Devario regina (Fowler, 1934) Discherodontus schroederi (Smith, 1945) Esomus metallicus Ahl, 1923 Garra cambodgiensis (Tirant, 1883) Garra fuliginosa Fowler, 1934 Hampala macrolepidota Kuhl & Van Hasselt, 1823 Henicorhynchus siamensis (Sauvage, 1881) a     Abbr. Hcha Hsig Corn Nnot Cbor Aalb Atru Bsin Blae Bmic Balt Bgon Bsch Bkor Bpul Ccir Cmol Char Ccob Cret Cide Capo Carm Ceno Crep Ccar Dalb Dequ Dmae Dmar Dreg Dsch Emet Gcam Gful Hmac Hsia Scientific name Hypsibarbus vernayi (Norman, 1925) Hypsibarbus wetmorei (Smith, 1931) Labeo chrysophekadion (Bleeker, 1850) Labeo rohita (Hamilton, 1822) Labiobarbus leptocheila (Valenciennes, 1842) Labiobarbus lineatus (Sauvage, 1878) Leptobarbus hoevenii (Bleeker, 1851) Lobocheilos melanotaenia (Fowler, 1935) Lobocheilos quadrilineatus (Fowler, 1935) Luciosoma bleekeri Steindachner, 1878 Mystacoleucus greenwayi Pellegrin & Fang, 1940 Mystacoleucus marginatus (Valenciennes, 1842) Neolissochilus stracheyi (Day, 1871) Onychostoma gerlachi (Peters, 1881) Osteochilus hasseltii (Valenciennes, 1842) Osteochilus lini Fowler, 1935 Osteochilus melanopleurus (Bleeker, 1852) Osteochilus microcephalus (Valenciennes, 1842) Osteochilus waandersii (Bleeker, 1852) Parachela oxygastroides (Bleeker, 1852) Paralaubuca harmandi Sauvage, 1883 Paralaubuca riveroi (Fowler, 1935) Paralaubuca typus Bleeker, 1865 Poropuntius bantamensis (Rendahl, 1920) Poropuntius deauratus (Valenciennes, 1842) Puntioplites proctozysron (Bleeker, 1865) Puntius brevis (Bleeker, 1850) Puntius orphoides (Valenciennes, 1842) Puntius partipentazona (Fowler, 1934) Puntius stoliczkanus (Day, 1871) Raiamas guttatus (Da3y, 1870) Rasbora atridorsalis Kottelat & Chu, 1987 Rasbora borapetensis Smith, 1934 Rasbora daniconius (Hamilton, 1822) Rasbora dusonensis (Bleeker, 1851) Rasbora myseri Brittan, 1954 Rasbora paviana Tirant, 1885 Scaphiodonichthys burmanicus Vinciguerra, 1890 Sikukia gudgeri (Smith, 1934) Sikukia stejnegeri Smith, 1931 Thynnichthys thynnoides (Bleeker, 1852) 151 Abbr. Hver Hwet Lchr Lroh Llep Llin Lhoe Lmel Lqua Lble Mgre Mmar Nstr Oger Ohas Olin Omel Omic Owaa Poxy Phar Priv Ptyp Pban Pdea Ppro Pbre Porp Ppar Psto Rgut Ratr Rbor Rdan Rdus Rmys Rpav Sbur Sgud Stej Tthy Scientific name Gyrinocheilus aymonieri Tirant, 1883 Balitoridae Balitora brucei Gray, 1830 Homaloptera smithi Hora, 1932 Homaloptera zollingeri Bleeker, 1853 Homaloptera leonardi Hora, 1941 Nemacheilus binotatus Smith, 1933 Schistura breviceps (Smith, 1945) Schistura bucculentus (Smith, 1945) Schistura desmotes (Fowler, 1934) Schistura geisleri (Kottelat 1990) Schistura magnifluvis Kottelat, 1990 Schistura mahnerti Kottelat, 1990 Schistura menanensis (Smith, 1945) Schistura obeini Kottelat, 1998 Schistura poculi (Smith, 1945) Schistura pridii Vidthayanon, 2003 Schistura sexcauda (Fowler, 1937) Schistura spilota (Fowler, 1934) Schistura vinciguerrae (Hora, 1935) Schistura waltoni (Fowler, 1937) Tuberoschistura baenzigeri (Kottelat, 1983) Cobitidae Acanthocobitis botia (Hamilton, 1822) Acanthocobitis zonalternans (Blyth, 1860) Acanthopsoides delphax Siebert, 1991 Acanthopsoides gracilentus (Smith, 1945) Acantopsis choirorhynchos (Bleeker, 1854) Acantopsis thiemmedhi Sontirat, 1999 Lepidocephalichthys berdmorei (Blyth, 1860) Lepidocephalichthys hasselti (Valenciennes, 1846) Pangio anguillaris (Vaillant, 1902) Syncrossus beauforti (Smith, 1931) Syncrossus helodes (Sauvage, 1876) Yasuhikotakia modesta (Bleeker, 1864) Yasuhikotakia morleti (Tirant, 1885) Siluriformes/Amblycipitidae Amblyceps mucronatumNg & Kottelat, 2000 Amblyceps foratum Ng & Kottelat, 2000 Bagridae Bagrichthys macracanthus (Bleeker, 1854) Abbr. Gaym Bbru Hsmi Hzol Hleo Nbin Sbre Sbuc Sdes Sgei Smag Smah Smen Sobe Spoc Spri Ssex Sspi Svin Swal Tbae Abot Azon Adel Agrl Acho Athi Lber Lhas Pang Sbea Shel Ymod Ymor Amuc Afor Bmac Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 2 Species list and its abbreviation (abbr.) of 198 fish species found during 1996-2009 in the Ping-Wang River Basin.    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand a 152 a   Scientific name Abbr. Hypophthalmichthys molitrix (Valenciennes, 1844) Hypophthalmichthys nobilis (Richardson, 1845) Hemibagrus wyckii (Bleeker, 1858) Hemibagrus wyckioides (Fang & Chaux, 1949) Mystus albolineatus Roberts, 1994 Mystus multiradiatus Roberts, 1992 Mystus mysticetus Roberts, 1992 Mystus singaringan (Bleeker, 1846) Pseudomystus siamensis (Regan, 1913) Clariidae Clarias batrachus (Linnaeus, 1758) Clarias gariepinus (Burchell, 1822) Clarias hybrid(C. macrocephalus X C. gariepinus) Clarias macrocephalus Günther, 1864 Loricariidae Pterygoplichthys disjunctivus (Weber, 1991) Pangasiidae Helicophagus leptorhynchus Ng and Kottelat, 2000 Pangasianodon gigas Chevey, 1931 Pangasianodon hypophthalmus (Sauvage, 1878) Pangasius bocourti Sauvage, 1880 Pangasius conchophilus Roberts & Vidthayanon, 1991 Pangasius larnaudii Bocourt, 1866 Pangasius macronema Bleeker, 1851 Pangasius pleurotaenia Sauvage, 1878 Pangasius sanitwongsei Smith, 1931 Siluridae Belodontichthys truncatus Kottelat & Ng, 1999 Kryptopterus cheveyi Durand, 1940 Kryptopterus cryptopterus (Bleeker, 1851) Micronema hexapterus (Bleeker, 1851) Ompok bimaculatus (Bloch, 1794) Phalacronotus apogon (Bleeker, 1851) Phalacronotus bleekeri (Günther, 1864) Hmol Hnob Hwyc Hwyk Malb Mmul Mmys Msin Psim   Cbat Cgar Chyb Cmac Pdis Help PGIG Phyp Pboc Pcon Plar Pmac Pple Psni Btru Kche Kcry Mhex Obin Papo Pble Scientific name Tor tambroides (Bleeker, 1854) Gyrinocheilidae Wallago attu (Bloch & Schneider, 1801) Wallago leerii Bleeker, 1851 Sisoridae Bagarius bagarius(Hamilton, 1822) Bagarius yarrelli (Sykes, 1839) Exostoma vincegerrae Regan, 1905 Glyptothorax lampris Fowler, 1934 Glyptothorax fuscus Glyptothorax trilineatus Blyth, 1860 Oreoglanis siamensis Smith, 1933 Cyprinodontiformes/Poeciliidae Gambusia affinis (Baird & Girard, 1853) Poecilia reticularis Peters, 1859 Xiphophorus helleri Heckel, 1848 Synbranchiformes/Synbranchidae Monopterus albus (Zuiew, 1793) Mastacembelidae Macrognathus siamensis (Günther, 1861) Mastacembelus armatus (Lacepède, 1800) Mastacembelus favus Hora, 1924 Mastacembelus cf. tinwiniBritz, 2007 Beloniformes/Belonidae Xenentodon cancila (Hamilton, 1822) Hemiramphidae Dermogenys pusilla Kuhl & van Hasselt, 1823 Perciformes/Ambassidae Parambassis siamensis (Fowler, 1937) Parambassis wolffii (Bleeker, 1851) Anabantidae Anabas testudineus (Bloch, 1795) Channidae Channa gachua (Hamilton, 1822) 152 Abbr. Ttam Watt Wlee Bbag Byar Evin Glam Gfus Gtri Osia Gaff Pret Xhal Malb Msia Marm Mfav Mtin Xcan Dpus Psia Pwol Ates Cgac Scientific name Bagrichthys macropterus (Bleeker, 1853) Hemibagrus nemurus (Valenciennes, 1840) Channa lucius (Cuvier, 1831) Channa micropeltes (Cuvier, 1831) Channa striata (Bloch, 1793) Cichlidae Oreochromis niloticus (Linnaeus, 1758)* Oreochromis hybrid* Eleotridae Oxyeleotris marmorata (Bleeker, 1852) Gobiidae Rhinogobius chiengmaiensis Fowler, 1934 Helostomidae Helostoma temminckii Cuvier & Valenciennes, 1831 Nandidae Pristolepis fasciatus (Bleeker, 1851) Osphronemidae Osphronemus goramy Lacepède, 1801 Trichogaster pectoralis (Regan, 1910) Trichogaster trichopterus (Pallas, 1770) Trichopsis vittata (Cuvier, 1831) Sciaenidae Boesemania microlepis (Bleeker, 1858) Toxotidae Toxotes chatareus (Hamilton, 1822) Pleuronectiformes/Cynoglossidae Cynoglossus microlepis (Bleeker, 1851) Cynoglossus feldmanni (Bleeker, 1853) Soleidae Brachirus harmandi (Sauvage, 1878) Brachirus siamensis (Sauvage, 1878) Solea ovata Richardson, 1846 Tetraodontiformes/Tetraodontidae Tetraodon leiurus Bleeker, 1851 Abbr. Bmar Hnem Cluc Cmic Cstr Onil Ohyb Omar Rchi Htem Pfas Ogor Tpec Ttri Tvit Bmic Tcha Cmio Cfel Bhar Bsia Sova Tlei Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 2 Species list and its abbreviation (abbr.) of 198 fish species found during 1996-2009 in the Ping-Wang River Basin (Cont.).    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand a basin Parameters Sub-basins UP MT SP MK MC TP ME FP LP WA Water temperature (°C) 23.4r1.5 23.4r1.4 23.7r2.9 24.8r1.4 22.1r1.3 27.2r2.4 21.8r0.6 29.0r1.5 30.8r0.3 27.2r3.9 Conductivity (mg/l) 64.0r16.7 72.1r26.1 84.6r49.9 65.0r13.8 78.6r73.9 77.2r28.5 51.7r39.5 103.3r8.2 100.0r8.2 84.4r11.5 126.0r107.4 79.6r34.6 93.8r58.7 66.7r17.5 90.9r66.9 148.9r85.5 95.4r43.5 81.7r7.5 80.0r8.2 98.9r16.0 6.06r1.0 6.25r0.8 7.08r1.3 5.52r0.4 6.18r0.7 5.09r0.9 5.93r0.5 4.17r1.1 4.90r0.5 4.72r2.0 0.03r0.03 0.08r0.02 0.010r0.54 0.0r0 0.02r0.01 0.08r0.02 0.08r0 0.03r0.01 Total dissolved solids (mg/l) Dissolved oxygen (mg/l) Nitrite(mg/l) 153 0.01r0 0.01r0 Ammonia(mg/l) 0.03r0.01 0.03r0.02 0.03r0.2 0.02r0.03 0.012r0.70 0.01r0 0.01r0.01 0.33r0.11 0.035r0.07 0.02r0.01 Phosphorus (mg/l) 0.216r0.1 0.116r0.1 0.060r0.1 0.071r0 0.106r0.1 0.076r0 0.068r0 0.122r0 0.203r0.1 0.097r0.1 7.3r0.4 7.2r0.4 7.5r0.6 7.0r0.3 8.0r0.2 7.1r0.5 6.6r0.4 8.4r0.2 8.5r0.3 7.6r0.9 Alkalinity (mg/l) 82.0r14.8 64.2r27.6 100.8r138.9 45.0r8.4 59.1r33.0 81.3r27.8 62.1r12.2 132.5r17.8 142.8r28.5 94.5r15.9 Hardness (mg/l) 92.0r22.8 103.3r23.6 84.3r50.3 51.7r7.5 105.2r27.2 101.8r25.8 53.8r7.1 80.2r20.3 103.5r28.5 60. 6r25.1 Current velocity (m/s) 0.804r0.4 0.519r.5 0.389r0.2 0.330r0.1 0.795r0.1 0.443r0.2 0.341r0.2 0.535r0.3 0.592r0.2 0.430r0.5 0.4r0.2 0.6r0.4 1.9r6.0 0.3r0 0.7r0.4 2.8r1.2 0.5r0.2 2.7r0.5 3.2r1.3 0.9r0.9 7r0.5 13r11.2 74r230.5 11r5.4 21r17.4 424r224.6 8r3.8 359r77.5 258r27.7 28r48.2 Discharge (m /s) 4.05r0.71 15.38r28.41 2.77r1.52 1.52r0.9 35.18r63.7 19.31r40.38 2.76r3.1 814.86r691.3 799.27r401.3 7.05r6.76 Altitude (m) 684r228.3 756r166.2 553r160.2 1,070r213.4 627r207.3 261r11.5 804r229.2 120r33.5 48r8.0 408r123.8 Distance from the sea (km) pH Depth (m) Width (m) 3 1,026r24.1 1,067r36.0 982r41.0 877r4.6 927r53.9 704r43.8 847r43.0 580r68.4 425r16.0 833r225.2 Watershed area (km2) 6,355 1,761 4,236 600 3,838 3,071 3,143 2,940 2,944 10,791 Forest area (%) 76.5 72.1 75 88.4 74.7 88.2 85 67.7 72.6 76.5 Agricultural area (%) 23.3 25.4 24.9 11.5 24.4 11.6 11.5 29.6 22 23.3 Urban area (%) 0.1 0.3 0.1 0.1 0.9 0.1 0.2 2.8 5.4 0.1 Note abbreviations of sub-basin as in Table 1 a     153 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 4 Average rSD) of physicochemical parameters and geo-morphological parameters in each sub-basin of the Ping-Wang River    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand a only the statistically significant parameter, P < 0.05) a     154 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand 154 Figure 2. Scattered plots between physicochemical water quality parameters and diversity indices, and their linearity trends (selected    a Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand statistically significant parameter, P < 0.05) a     155 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand 155 Figure 3. Scattered plots between geo-morphological parameters and diversity indices and their linearity trends (selected only the    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand physicochemical parameters, i.e. DO, water temperature, pH, conductivity, phosphorus and alkalinity, which showed statistical significance relationships to diversity parameters (P-value < 0.05, Fig. 3). Higher DO in the high altitude area (Table 4) made a negative relationship of this parameter to both species richness and Shannon diversity index but showed positive relationships to the other predictive parameters. However, due to extensive and high variation of the obtained data, all the linear models showed low power in prediction, i.e. r2 was less than 0.5. Scattered plots of temperature to both response variables showed that from the low temperature to 30 oC, the relationships tended to be non-linear. But if temperature was beyond 30 o C, the diversity trended to decline. The pH ranged from 5.5 to 8.7 and the diversity parameters were obviously low in acidic water and the relationship to species richness WHQGHGWREHDVLQSRZHUIXQFWLRQ$YHUDJHRIFRQGXFWLYLW\ZDV“ȝ6FP-1 and high fish diversity was observed around this range. Species richness and Shannon diversity index were slightly increased as alkalinity and phosphorus increased but non-statistical relationship was found between Shannon diversity index and phosphorus (P-value > 0.05). Table 3 Summary of fish diversity indices from the samplings during 1996-2009 in the Ping-Wang river basin. Sub-basins UP MT SP MK MC TP ME FP LP WA Total individuals 545 5607 10375 474 4425 1789 1755 1060 1081 2244 Shannon Diversity index 2.0628 2.8194 3.4195 1.4581 2.9265 3.5133 2.6698 3.9366 4.4506 3.3971 Simpson Dominance index 5.179 6.168 6.745 3.523 6.536 9.015 4.532 30.42 60.094 9.697 Species evenness 0.744 0.7004 0.7425 0.6081 0.6985 0.7808 0.785 0.9036 0.9432 0.7775 Species richness 16 56 98 11 66 84 30 78 112 77 Predicting of diversity parameters Species richness and Shannon diversity index of each individual sampling ranged from 11 to 112 species and 1.099 to 3.401, respectively. They were then logtransformed and fed to CART model as a response variable by using 20 environmental predictors. The geo-morphological parameters were the major factors in determining both diversity indices. For species richness, by the tree “pruning 156    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand “process and optimal tree selection, 3 parameters were included in the CART model and altitude (ALT) was the major contributor in predicting species richness followed by width (WID) and distance from the sea (DFS) (Fig. 4). Altitude was used in both of the first and the second splits, meanwhile the other parameters were used in the third split. The coefficient of determination, r2, of this model was 0.59 and showed that if the altitude was less than 93.5 m ASL, high species diversity was observed, i.e. about 60 species. The r2 of the model for Shannon diversity index was 0.76. Six parameters were accumulated and used as predictors viz., width, altitude, discharge, pH, agricultural area and alkalinity (Fig. 5). Width was used in the first split and showed that the index would not beyond 3. Altitude was used in the second split and showed the trend that the higher the altitude, the lower the index. The remaining 4 parameters were combined with altitude to make further splits for prediction. Relationships of fish assemblage and environmental parameters Fifty-three fish species and twenty environmental variables were loaded in the CCA analysis. Total model inertia (sum of unconstrained eigen values) was 4.232, and the sum of all canonical eigen values was 5.531, in which the species-environment correlation coefficients for the first and second axes of CCA accounted for 55.9 % and 17.62 %, respectively. Monte-Carlo permutation attested that both axes were significant (P < 0.001). The length of vector of a given variable on the CCA plots indicates the importance of that variable. The first CCA environmental axis (CCA1) was described by altitude, distance from the sea, water depth, stream width and water temperature of the basin. The first two parameters were negatively correlated to CCA1 while the remaining parameters were positively. The most important variable for the second CCA environmental axis (CCA2) was watershed area, meanwhile the others were correlated less than 0.5 (Table 5 and Fig. 6a). Composition of individual fish species, which related to the environmental vectors loaded to CCA, was shown in Fig. 6b and the first five species having strong loading to CCA1 and CCA2, either positive or negative correlation, are presented in Table 5. 157    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Figure 4. CART model to predict species richness in the Ping Wang River Basin. Distribution of fish species along the CCA axes can be classified into 4 main assemblage patterns (Fig. 6b). The first assemblage (quadrant I) was negatively correlated to both CCA1 and CCA2, implying that they inhabited in the mountainous area of high altitude with relative low temperature and strong current velocity. The second assemblage was negatively correlated to CCA1 and positively to CCA2. The shorter distance from CCA1 (quadrant II) indicated that the fish in this assemblage occupied a lower altitude than those in the first assemblage. The remaining two assemblage patterns were positively correlated to CCA1 (quadrants III and IV) and implying that the fishes in these assemblages live in the lower portion of the river course, where the river width and depth were more than the previous two assemblages. The last assemblage (quadrant IV, negatively to CCA2) inhabited a larger watershed close to agricultural and urban areas, which have high phosphorus loading. Meanwhile, species that distributed in the around the center of the bi-plot 158    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand (Fig. 6B) had very little differentiation among each other e.g. Discherodontus schoeroderi, Puntius brevis, Puntius orphoides, and Rasbora paviana. Figure 5. CART model to predict Shanon diversity index in the Ping Wang River Basin. The result from Ward’s analysis (Fig 6c) was used to refine the habitat preference of 53 individual species after the CCA analysis. Eight species in quadrant I, inhabiting the small streams in high altitude area with low temperature, were grouped together and defined as “mountainous” species They were Oreoglanis siamensis (Osia), Devario regina (Dreg), Exostoma vinciguerrae (Evin), Schistura pridii (Spri), Schistura waltoni (Swal), Scaphiodonichthys burmanicus (Sbur) Glyptothorax trilineatus (Gtri) and Devario maetangensis (Dmae). The remaining species in quadrant I, which located closely to CCA1 axis, and all species in quadrant II were grouped and defined as “piedmont” species (22 species). The examples in this group were Lepidocephalichthys hasseltii (Lhas), Dermogenys pusilla (Dpus), Monopterus albus (Malb), Channa gachua (Cgac) and Homaloptera leonardi (Hleo). Species positively correlated to CCA1, were divided into two groups. Firstly, the species located close to CCA1, were defined as “transitory” species, i.e. species that migrated between piedmont and lowland area, (8 species) such as Puntius orphoides (Porp), 159    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand a variables and the first five fish species that showed strong loading to CCA 160 Axis 1 Correlations of geo-morphological parameters loading with axis Altitude (m) -0.929 Distance from the sea (km)* -0.703 Current velocity (m/s)* -0.224 Width (m)* 0.583 Depth (m)* 0.591 Discharge (m3/s)* 0.430 Watershed area (km2)* 0.153 Forest area (km2) -0.228 Agricultural area (km2)* 0.105 Urban area (km2)* 0.232 Correlations of fish species with strong positive loadings on CCA1 Pristolepis fasciatus 1.398 Barbonymus altus 1.387 Mystus singaringan 1.387 Notopterus notopterus 1.360 Osteochilus hasselti 1.322 Correlations of fish species with strong positive loadings on CCA2 Lepidocephalichthys hasselti -0.546 Dermogenys pusilla -0.484 Monopterus albus 0.966 Channa gachua 0.038 Homaloptera leonardi -0.405 a     Axis 2 Axis 1 Correlations of physicochemical parameters loading with axis Water temperature (°C)* 0.583 Conductivity (µS/m)* 0.387 pH 0.238 Total Dissolved Solids (mg/l)* 0.219 Phosphorus (mg/l)* 0.191 Alkalinity (mg/l)* 0.202 Hardness (mg/l)* 0.082 Dissolved Oxygen (mg/l) -0.359 Nitrite (mg/l)* 0.105 Ammonia (mg/l)* 0.163 Correlations of fish species with strong negative loadings on CCA1 Oreoglanis siamensis -1.119 Devario regina -1.102 Exostoma vinciguerrae -1.090 Schistura pridii -0.898 Devario maetangensis -0.890 Correlations of fish species with strong negative loadings on CCA2 Exostoma vinciguerrae -1.090 Oreoglanis siamensis -1.119 Schistura pridii -0.898 Scaphiodonichthys burmanicus -0.873 Glyptothorax trilineatus -0.854 -0.142 0.196 -0.045 -0.284 -0.227 -0.242 -0.809 0.149 -0.142 -0.406 -0.364 -0.398 -0.528 -0.489 -0.331 1.344 0.731 0.654 0.592 0.559 160 Axis 2 0.064 0.149 0.150 0.024 -0.428 0.388 -0.052 0.377 0.167 -0.056 -0.889 -0.517 -1.064 -0.742 -0.465 -1.064 -0.889 -0.742 -0.735 -0.683 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 5 Statistics associated with the first two canonical axes from Canonical correspondences analysis (CCA) for 20 environmental    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Figure 6. Biplots CCA ordination with the composition of fish species related to the environmental vectors in the Ping-Wang River basin. a) environmental variables loading to CCA axes and b) species assemblages and c) Dendrogram of fish assemblages in the Ping-Wang River Basin. Note See text for abbreviation of environmental parameters and Table 2 for fish species. 161    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Mastacembelus armatus (Marm), Puntius brevis (Pbre) and Mystacoleucus marginatus (Mmar), and secondly, the group of species that showed the highest positive loading to CCA1 were defined as “lowland” species. Examples of this group were Pristolepis fasciatus (Pfas), Barbonymus altus (Balt), Mystus singaringan (Msin), Notopterus notopterus (Nnot) and Osteochilus hasselti (Ohas). DISCUSSION Understanding the diversity, abundance and coexistence of diverse fish species assemblages in river ecosystems are among the central goals in tropical ecological research (Herder & Freyhof, 2006). Dominance by the multi-species and ecologically diverse Cyprinidae is common in Southeast Asia, where they may contribute 40% or more of the species in a watershed (Taki 1978; Beamish et al., 2006). This is because cyprinids have evolved partially through highly adapted body forms and mouth structures so they occupy virtually all habitats throughout their distributions (WardCampbell et al., 2005). The second most dominant family of Balitoridae indicates the characteristic of high altitude mountainous area of this study since the fish in this family are significantly related with high elevation (Beamish et al., 2008). The complexity and non-linearity of the relations between the communities and their environment are very common (Gevrey et al., 2003) as shown in the results. Meanwhile, the low r2 values from the linear correlations found in some environmental variables to the diversity indices showed the weak relationships. For prediction of fish diversity, it was found that geo-morphological and landscape parameters were the good predictors for species richness and Shannon diversity index compared to those physicochemical parameters. Changes in both diversity indices follow the general longitudinal pattern of river fish distribution as the lowest levels tend to be found at high altitudes, and the highest levels at mid to low altitudes (Gaston & Blackburn, 2000; He et al., 2010). High values of both diversity indices in the lower altitude with larger watershed also supports the species–area relationships pattern, which suggests larger areas of habitat generally contain more species than smaller areas (Angermeier & Schlosser 1989; Han et al., 2008). The large watershed is also associated with deep and wide area, which shows robust positive relationships to species richness (Connor & McCoy, 1979; Angermeier & Karr, 1983). High 162    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand discharge implies a large volume of water for a number of fishes to occupy and increase in river flow results in more fish species richness because of greater heterogeneity of local fish habitats (Guégan et al., 1998; He et al., 2010). The physicochemical parameters would be important to fish species richness and abundance in a relatively small spatial scale or single drainage system (Tongnunui & Beamish, 2009; Alexandre et al., 2010). Low rate of correlation between these parameters to species richness implies that they had weak relationships. Killgore & Hoover (2001) reported on a second degree polynomial relationship between DO and species richness, implying there is an optimum level of DO to fish diversity, which as seen in the result that peak of species richness and diversity index ranged between 4 and 6 mg.l-1. Temperature reflects directly on metabolism and is recognized as a dominant factor in the control of species diversity (Oberdorff et al., 1995; Guégan et al., 1998), which the relationship as power function indicated the critical temperature at about 30 oC and few species can survive beyond this point. This kind of relationship is also the case for pH, which few species can inhabit the acidic condition and being optimized at range 7.5-8.5. Meanwhile the linear-trend positive relationships of the remaining three parameters, i.e. conductivity, phosphorus and alkalinity, to species diversity are widely reported (e.g. Johal et al., 2001; Shahnawaz et al., 2010). The effects of land uses on fish community structure have been widely investigated and proven to be the important determinants (e.g. Orrego et al., 2009; Alexandre et al., 2010). Distinct patterns of fish assemblage along the longitudinal river gradient reflects the homogenous spatial units within the river basin (Welcomme et al., 2006; Ferreira & Petrere, 2009) and the results from ordination and classification showed four fish assemblage patterns from the headwater to lowland river reaches. Multiple mechanisms can explain partitioning of fish assemblages along longitudinal gradients of the river such as resource availability, quality of habitats and adaptation of individual species (Matthews, 1998). The assemblage of mountainous species showed their restricted occurrence in a high altitude area, with associated riffles, pools and rapids. All the fish in this assemblage shows their morphological adaptation to survive in the strong flow conditions (Casatti & Castro, 2006; Welcomme et al., 2006). The interplay between strong currents and rocky substrates usually generates the 163    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand mountainous areas rich in food, such as patches of rapidly growing periphytic algae and the aquatic insect larvae that directly or indirectly fed by mountainous species (Casatti & Castro, 2006). Moreover, the cold water designation whole year round in the mountainous area also likely influence the fish community (Wanner et al., 2011). Any human activities that disturb the pool-riffle structure, such as changes to the flow regime, increases in sediment load and make and anoxic condition would affect this assemblage (Welcomme et al., 2006). Various microhabitats in the piedmont, such as main channel, backwaters and side channel anabranches as well as various bottom types support the richness of both the fluvial specialist and habitat generalist fish species (Freeman & Marcinek, 2006). Yet, large debris from forest area in the piedmont, which is characterized by extreme flooding and bank erosion during the rainy season, shows a positive effect on fish densities and diversity (Angermeier & Karr, 1983; Wright & Flecker, 2004). Residences in this assemblage also require relatively high dissolved oxygen levels and as such they are sensitive to reductions conditions are sensitive to reductions in water quality (Welcomme et al., 2006). Meanwhile, differences in food resources and habitats use among the fish species within the assemblages results in complexity in this assemblage. For example, Garra cambodgiensis, an algae eater, and an insectivore Schistura breviceps occupy the rocky and pebble bottoms (Rainboth, 1996; Kottelat, 1998; Ward-Campbell et al., 2005). The fluvial specialists, Barilius pulchellus and Homaloptera spp. inhabit the main channel. Although both of them are insectivores, B. pulchellus feeds on odonatan larvae whereas Homaloptera spp. feed on benthic insects (Rainboth, 1996). Meanwhile, the inhabitants in the backwater include Channa gachua, Clarias batrachus and Mastacembelus spp., the first two species being predators and the latter an insectivore (Rainboth, 1996; Kottelat, 1998). Complexity of fish community in the lowland river could be driven by a great amount of productive littoral zone due to the large watershed with deep and wide river channel (Angermeier & Karr, 1983; Han et al., 2008). Both rheophilous and limnophilous fishes were the common residents in the lowland river. However, some species were sub-divided to involve the transitory assemblage, in which rheophilous cyprinids always dominate (Allouche, 2002). Four rheophilous cyprinids (out of 8 species) were included in this assemblage viz., Puntius orphoides, Puntius brevis, 164    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Mystacoleucus marginatus and Esomus metallicus. The word “transitory” was used to describe this assemblage implying that, indeed, the fish could also occupy the lowland rivers (Rainboth 1996; Kottelat, 1998), where the lentic cyprinids and other limnophilic fishes dominated, i.e. lowland species (Allouche, 2002; Beamish et al., 2006). However, upstream movement of these lowland species is sometimes observed especially for reproduction (Silva & Davies, 1986; Ferreira & Petrere, 2009; Tongnunui & Beamish, 2009). This phenomenon supports the pattern of species addition for the shifting in species composition (Huet, 1959; Petry & Schulz, 2006). Meanwhile the pattern of species replacement is expected in mountainous regions, where abrupt transitions could be observed as well as physicochemical conditions being stressful and fewer fish species adapt to survive (Ferreira & Petrere, 2009; He et al., 2010). Damming of the river course upstream for irrigation purpose, which would likely to be taken place in the near future at the Ping-Wang River Basin, is inevitably affect the fish assemblages in the downstream river course. Damming alters the river flow, reduces nutrient loading from upstream and prevents fish migration (Welcomme et al., 2006), especially the transitory species. In conclusion, this study confirms the importance of geo-morphological i.e. altitude, stream width, and distance from the sea as a variable explaining variation in fish community structure along a river gradient (Esselman et al., 2006; Sullivan et al., 2006; Grossman et al., 2010) in a large scale whole basin. However, the contribution of the other variables, especially some of the physicochemical water quality parameters, should be considered in terms of point and non-point pollution sources over a small scale (Ibarra et al., 2005; Orrego et al., 2009). The delineation of fish assemblage patterns enhances the understanding of fish zonation in this region. The patterns in fish assemblage structure of this large-scale Ping-Wang river basin seemed to be influenced by species-specific responses to dominant environmental gradients. Meanwhile, further study is needed to examine the role of individual species within each zone for better assessment of the impacts of human disturbances in each zone in the future. ACKNOWLEDGEMENT 165    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand A. Suvarnaraksha is grateful to the Royal Golden Jubilee Program of the Thailand Research Fund for supporting his Ph.D. study (Grant PHD/0290/2549). Grant for the field survey was by National Research Council of Thailand and the Nagao Natural Environment Foundation. Partial support by Franco-Thai Academic Collaboration (Grant PHC 16598RJ) and the French Embassy to Thailand (Grant CNOUS: 2009/2349) supported him to work at Laboratoire Evolution et Diversité Biologié under the convention for the joint supervision of thesis between Ubon Ratchathani University and Université Paul Sabatier (Toulouse III). REFERENCES APHA, 1989. Standard methods for the examination of water and wastewater. American Public Health Association, Washington. Alexandre, C.V., K.E. Esteves & M.A.M. de Moura e Mello, 2010. Analysis of fish communities along a rural–urban gradient in a neotropical stream (Piracicaba River Basin, São Paulo, Brazil). Hydrobiologia 641: 97–114. Allouche, S., 2002. Nature and functions of cover for riverine fish. Bulletin Francais de la Peche et de la Pisciculture 365/366: 297-324. Angermeier, P.L. & J.R. Karr, 1983. Fish communities along environmental gradients in a system of tropical streams. Environmental Biology of Fishes 9:117–135. Angermeier, P.L. & I.J. Schlosser, 1989. Species–area relationships for stream fishes. Ecology 70:1450–1462. Beamish, F.W.H., P. Sa-ardrit & V. Cheevaporn, 2008. Habitat and abundance of Balitoridae in small rivers in central Thailand. Journal of Fish Biology 72: 2467–2484. Beamish, F.W.H., P. Sa-Ardrit & S. Tongnunui, 2006. Habitat characteristics of the cyprinidae in small rivers in Central Thailand. Environmental Biology of Fishes 76, 237-253. Breiman, L., J.H. Friedman, R.A. Olshen & C.J. Stone, 1984. Classification and regression trees. Wadsworth Inc., Monterey. Cade, B.S. & B.R. Noon, 2003. A gentle introduction to quantile regression for ecologists. Front Ecol Environ, 1(8): 412-420. Casatti, L. & R.M.C. Castro, 2006. Testing the ecomorphological hypothesis in a 166    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand headwater riffles fish assemblage of the Rio São Francisco, Southeastern Brazil. Neotropical Ichthyology, 4 (2): 203-214. Céréghino, R. Y.-S. Park, A. Compin and S. Lek, 2003. Predicting the species richness of aquatic insects in streams using a limited number of environmental variables. J. North Am. Benthol. Soc. 22: 442_/456. Chessel, D., A.B. Dufour & J. Thioulouse, 2004. The ade4 package-I. R News 4:5-10 Connor, E.F. & E.D. McCoy, 1979. The statistics and biology of the species–area relationship. American Naturalists 113: 791. Conservation International, 2001. Biodiversity hotspots. http://www.biodiversityhotspots.org de Oliveira, F.F.R. & P.C. Eterovick, 2009. The role of river longitudinal gradients, local and regional attributes in shaping frog assemblages. Acta Oecologica 35: 727–738. De Silva, S.S., N.W. Abery, & T.T.T. Nguyen, 2007. Endemic freshwater finfish of Asia: distribution and conservation status. Diversity and Distribution: 13, 172184. de Thoisy, B., S. Brosse & M.A. Dubois, 2008. Assessment of large-vertebrate species richness and relative abundance in Neotropical forest using line-transect censuses: what is the minimal effort required? Biodiversity Conservation 17:2627–2644. Dugan, P.J., C. Barlow, A.A. Agostinho, E. Baran, G.F. Cada, D. Chen, I.G. Cowx, J.W. Ferguson, T. Jutagate, M. Mallen-Cooper, G. Marmula, J. Nestler, M. Petrere, R.L. Welcomme & K.O. Winemiller, 2010. Fish migration, dams, and loss of ecosystem services in the Mekong basin. Ambio 39: 344-348. Esselman, P.C., M.C. Freeman & C.M. Pringle, 2006. Fish-assemblage variation between geologically defined regions and across a longitudinal gradient in the Monkey River Basin, Belize. J. N. Am. Benthol. Soc., 25(1): 142–156. Ferreira, F.C. & M. Petrere Jr., 2009. The fish zonation of the Itanhae´m river basin in the Atlantic Forest of southeast Brazil. Hydrobiologia 636: 11–34. Freeman, M.C. & P.A. Marcinek, 2006. Fish assemblage responses to water withdrawals and water supply reservoirs in Piedmont streams. Environmental Management 38: 435-450. 167    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Gaston, K.J. & T.M. Blackburn, 2000. Pattern and process in macroecology. Blackwell Publishing Ltd., Oxford.  Gevrey, M., Y.S. Park, T. Oberdorff, & S. Lek, 2003. Prediction of fish assemblages in France and evaluation of the influence of their environmental variables. In Lek, S., M. Scardi, P. Verdonshot & S. Jorgensen (eds), Modelling Community Structure in Freshwater Ecosystems. Springer Verlag, Berlin: 54-63. Grossman G.D., R.E. Ratajczak Jr, M.D. Farr, C.M. Wagner, & J.T. Petty, 2010. Why there are fewer fish upstream. American Fisheries Society Symposium. 73: 63– 81. Guégan, J.F., S. Lek, & T. Oberdorff, 1988. Energy availability and habitat heterogeneity predict global riverine fish diversity. Nature 391: 382–384. Gutiérrez-Estrada, J., C. R. Vasconcelos & M. J. Costa, 2008. Estimating fish community diversity from environmental features in the Tagus estuary (Portugal): Multiple Linear Regression and Artificial Neural Network approaches. J. Appl. Ichthyol. 24: 150–162. Han, M., M. Fukushima & T. Fukushima, 2008. Species richness of exotic and endangered fishes in Japan's reservoirs Environmental Biology of Fish 83: 409416. He, Y., J. Wang, S. Lek-Ang, & S. Lek, 2010. Predicting assemblages and species richness of endemic fish in the upper Yangtze River. Science of the Total Environment 408: 4211 – 4220. Herder, F. & J. Freyhof, 2006. Resource partitioning in a tropical stream fish assemblage. Journal of Fish Biology 69: 571–589. Huet, M., 1959. Profiles and biology of Western European streams as related to fish management. Transactions of the American Fisheries Society 88:155–63. Ibarra, A.A., F. Dauba & P. Lim, 2005. Influence of non-point source pollution on riverine fish assemblages in South West France. Ecotoxicology 14: 573–588. Ihaka, R. and R. Gentleman, (1996). "R: A language for data analysis and graphics," Journal of Computational and Graphical Statistics 5: 299-314. Johal, M.S., K.K. Tandon, Y.K. Rawal, A.K Tylor, H.S. Banyal & H.S. Rumana, 2001. Species richness of fish in relation to environmental factors. Current Science, 80: 499-502. 168    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Killgore, K. J. & J. J. Hoover. 2001. Effects of hypoxia on fish assemblages in a vegetated waterbody. J. Aquat. Plant Manage., 39: 40–44. Kimmel, W.G. & D.G. Argent, 2010. Stream fish community responses to a gradient of specific conductance. Water Air and Soil Pollution 206: 49–56. Kottelat, M., 1985. Fresh-water fishes of Kampuchea. Hydrobiologia 121: 249-279. Kottelat, M., 1998 Fishes of the Nam Theun and Xe Bangfai basins, Laos, with diagnoses of twenty-two new species (Teleostei: Cyprinidae, Balitoridae, Cobitidae, Coiidae and Odontobutidae). Ichthyol. Explor. Freshwat. 9(1): 1-128. Kottelat, M., 2001. Fishes of Laos. WHT Publications Ltd., Colombo 5, Sri Lanka. 198 pp. Lasne, E., B. Bergerot, S. Lek & P. Laffaille, 2007. Fish zonation and indicator species for the evaluation of the ecological status of rivers: example of the Loire basin (France). River Research and Applications 23: 877-890. Lenat, D.R., 1988. Water quality assessment using a qualitative collection method for benthic macroinvertebrates. J.N. Am. Benthological Soc. 7: 222-233. Matthews, W. J., 1998. Patterns in freshwater fish ecology. Chapman and Hall, New York. Myers, N., R.A. Mittermeier, C.G. Mittermeier, G.A.B. da Fonseca, & J. Kent, 2000 Biodiversity hotspots for conservation priorities. Nature 403: 853-858. Oberdorff, T., J.F. Guégan & B. Hugueny, 1995. Global scale patterns of fish species richness in rivers. Ecography 18: 345–352. Oberdorff, T., D. Pont, B. Hugueny, J. Belliard, R.B.D. Thomas,& J.P. Porcher, 2002. Development and validation of a fish-based index (FBI) for the assessment of ‘‘river health’’ in France. Freshw Biol 47: 1720–1734. Orrego, R., S.M. Adams, R. Barra, G. Chiang & J. F. Gavilan, 2009. Patterns of fish community composition along a river affected by agricultural and urban disturbance in south-central Chile. Hydrobiologia 620: 35–46. Petry, A.C. & U.H. Schulz, 2006. Longitudinal changes and indicator species of the fish fauna in the subtropical Sinos River, Brazil. Journal of Fish Biology 69: 272-290. R Development Core Team, 2010. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. http://www.r- 169    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand project.org Rainboth,W.J., 1996. Fishes of the Cambodian Mekong. Food and agriculture organization of the united nations, Rome. Roberts, T. R., 1993. Systematic revision of the southeast Asian cyprinid fish genus Labiobarbus (Teleostei: Cyprinidae). Raffles Bull. Zool., 41: 315-329. Roberts, T. R., 1994. Systematic revision of Asian bagrid catfishes of the genus Mystus sensu stricto, with a new species from Thailand and Cambodia. Ichthyol. Explor. Freshwat., 5: 241-256. Shahnawaz, A., M. Venkateshwarlu, D.S. Somashekar & K. Santos, 2010. Fish diversity with relation to water quality of Bhadra River of Western Ghats (INDIA). Environmental Monitoring & Assessment 161: 83–91. Shannon, C.E., 1948. A mathematical theory of communication. Bell system Technical Journal 27: 379-423. Schmutz, S., M. Kaufmann, B. Vogel, M. Jungwith & S. Muhar, 2000. A multi-level concept for fish-based, river-type-specific assessment of ecological integrity. Hydrobiologia 422: 279–289. Silva, E.I.L & R.W Davies, 1986. Movements of some indigenous riverine fish in Sri Lanka. Hydrobiologia 137: 265-270. Smith, H.M., 1945. The Freshwater Fish of Siam or Thailand. United States Government Printing office, Smithsonion Institution . Bull. no. 188. Washington. 622 p. Sodhi, N.S., L.P. Koh, B.W. Brook & P.K.L. Ng, 2004. Southeast Asian biodiversity: an impending disaster. Trends in Ecology and Evolution 19: 654-660. Sullivan, S.M. P., M.C. Watzin, & W.C. Hession, 2006. Influence of geomorphic condition on stream fish communities in Vermont, USA. Freshwater Biology 51: 1811-1826. Takeuchi, K., A.W. Jayawarderna, Y. Takahasi & B. Machbub, 2005. Catalogue of rivers for Southeast Asia and the Pacific Vol. I and II (CD-rom). UNESCO, Jakarta. Taki, Y., 1974. Fishes of the Lao Mekong Basin. United States Agency for International Development Mission to Laos Agriculture Division. 232 p. 170    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Taki, Y., 1978. An analytical study of the fish fauna of the Mekong Basin as a biological production system in nature. Research Institute for Evolutionary Biology, Tokyo. Ter Braak, C.J.F., 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis Ecology 67: 1167-1179. Therneau, T. M., & B. Atkinson, 2010. The rpart package. http://www.cran.r-project. org/doc/packages/rpart.pdf. Tomanova, S., P.A.Tedesco, M.Campero, P.A.van Damme, N. Moya & T. Oberdorff, 2007. Longitudinal and altitudinal changes of macro-invertebrate functional feeding groups in neotropical streams: a test of the river continuum concept. Fundamental and Applied Limnology / Archiv für Hydrobiologie 170: 233–241. Tongnunui, S. & F.W.H. Beamish, 2009. Habitat and relative abundance of fishes in small rivers in eastern Thailand. Environ Biol Fish. 85: 209–220. Vidthayanon, C., Karnasuta, J. & Nabhitabhata, J., 1997. Diversity of Freshwater Fish in Thailand. Office of Environmental Policy and Planning, Bangkok, 102 pp. Wanner, G.A., M. A. Pegg, S. Schainost, R. A. Klumb & D. A. Shuman, 2011. River geomorphology and fish barriers affect on spatial and temporal patterns of fish assemblages in the Niobrara River, Nebraska. Nebraska Game and Parks Commission, Alliance, Nebraska. Ward, J.H., 1963. Hierarchical Grouping to optimize an objective function. Journal of American Statistical Association 58: 236-244. Ward-Campbell B.M.S, F.W.H. Beamish & C. Kongchaiya, 2005. Morphological characteristics in relation to diet in five co-existing Thai fish species. Journal of Fish Biology 67: 1266–1279. Welcomme, R.L., K.O. Winemiller & I.G. Cowx, 2006. Fish environmental guilds as a tool for assessment of ecological condition of rivers. River Research & Applications 22: 377-396. Wright, J. P. & A.S. Flecker, 2004. Deforesting the river-scape: the effects of wood on fish diversity in a Venezuelan piedmont stream. Biological Conservation 120: 439–447. 171    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Yamazaki, Y., S. Haramato, & T. Fukasawa, 2006. Habitat uses of freshwater fishes on the scale of reach system provided in small streams. Environmental Biology of Fishes 75: 333–341. 172    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand P5 Fish communities in highland tropical streams connected to a reservoir. Suvarnaraksha, A., S. Lek, S. Lek-Ang and T. Jutagate (In preparation) 173    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand FISH COMMUNITIES IN HIGHLAND TROPICAL STREAMS CONNECTED TO A RESERVOIR D Apinun SUVARNARAKSHA1,2,3, , Sovan LEK2,4, Sithan LEK-ANG2,4, Tuantong JUTAGATE1 1. Faculty of Agriculture, Ubon Ratchathnai University, Warin Chamrab, Ubon Ratchathani, Thailand 34190 2. University of Toulouse III, Laboratoire Dynamique de la Biodiversité, UMR 5172, CNRS – UPS, 118 route de Narbonne, 31062 Toulouse Cedex 4, France 3. Faculty of Fisheries Technology and Aquatic Resources, Maejo University, Sansai, Chiangmai, Thailand 50290 4. University of Battambang, Cambodia Running title: Fish communities in a highland tropical reservoir D  Corresponding  author:  Faculty  of  Fisheries  Technology  and  Aquatic   Resources,  Maejo  University,  Sansai,  Chiangmai,  Thailand  50290  E-­‐mail:   apinun@mju.ac.th   174    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   ABSTRACT Fish communities in the high altitude streams connected to the Mae-ngad reservoir (412 to 425 m ASL), Thailand, were investigated both in the streams (10 stations) and the reservoir (2 stations). The study was carried out from October 2002 to September 2003 with a total of 144 surveys. Fish belonging to 66 species and 21 families were captured and almost one-half (48%) of the species caught were insectivores. The dominant family was the Cyprinidae (23 species) followed by Balitoridae and Cobitidae (each containing 7 species), which both exclusively inhabit the strong current stream. A self-organizing map (SOM) was used to cluster the fish community, according to the similarities in fish composition in each survey. Three fish communities were obtained, namely reservoir-, stream- and intermediatecommunities. The reservoir communities were characterized by “lentic-adapted” fish such as Labiobarbus lineatus, (Sauvage, 1878) and Puntioplites proctozysron (Bleeker, 1865), whereas rheophilic species, such as Rasbora paviana Tirant, 1885 and Channa gachua (Hamilton, 1822), were dominant in the stream community. The intermediate community, which contained a mixture of species from both the other communities, was found during the rainy season. A classification and regression tree was used to examine the contribution of environmental variables to the composition of the communities and to build predictive models. Six variables were selected as predictors, of which water depth was the major parameter to predict community types, followed by water chemistry. The overall percentage of successful prediction by the model was 66.0 %: the model was 100% accurate for the prediction of the reservoir community but very low for the stream community (40%). Keywords: lentic-adapted species, rheophilic species, Self organizing map, environmental variables, Thailand 175    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   INTRODUCTION River impoundment changes the water body from “rivers” to “reservoirs”, affecting not only the hydrology but also the physical, chemical, and biological characteristics. The result of these environmental alterations is a progressive decrease in the number of individuals and species of the native flora and fauna (Barrella and Petrere 2003). Impoundment also has an immediate impact on fish assemblages (Gao et al. 2010) and long- term changes on fish communities (Taylor et al. 2001). The impact of downstream damming on fish communities has been well documented in past decades and including recent studies of an upstream area, especially in connected tributaries. However there are still relatively few studies (e.g. Falke and Gido 2006; Penczak et al. 2009). The case is different if the damming is in the upper river course in a mountainous area connected to the first- and second- order streams. These have comparatively low fish species richness (Oberdoff et al. 1995; Welcomme et al. 2006), and most of the resident fish have the specific characteristics of living in a strong current with turbulent water flow and rocky substrates, i.e. rheophilic species (Casatti and Castro 2006). Alterations in the river discharge patterns also affect the structure of the stream fish assemblages in the upper river course (Poff and Allan 1995). Fish assemblage structure varies with increasing distance from a reservoir and the abundance of reservoir fish in the upstream reaches declines with the distance from a reservoir (Falke and Gido 2006). Meanwhile, fish species that had successfully colonized the reservoir after impoundment could expand into the inflowing river (Hladík et al. 2008). For example, piscivorous fish can migrate into nearby streams and predate on the stream residents (Martinez et al. 1994, Matthews et al. 1994), while omnivorous fish could also move into the stream and compete for food sources with stream residents or alter the ecosystem in these streams (Gido and Matthews 2000). Changes in fish communities in the reservoir, therefore, would be expected to be the communities of the species that could adapt to both lotic and lentic habitats and those, which migrate and inhabit exclusively the streams (McCartney 2009). Moreover, the composition of fish migrating from the reservoir into the inflowing river can be reflected in the changes in the fish assemblage in the inflowing river (Hladík et al. 2008). Variability in fish abundance and community structure is also 176    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   governed by environmental factors along river courses (Lasne et al. 2007), which is also the case when the river is dammed and there is a change in environmental factors (Barrella and Petrere 2003). Therefore, to evaluate that if there were any differences in the fish community structures induced by damming the upper reach of a tropical region, the objective of this study was to examine the fish community patterns in the streams that connected to a reservoir as well as the contribution of environmental variables to the assemblages. MATERIALS AND METHODS Study area and its characteristics The Mae-ngad reservoir is located in Chiangmai province, in northern Thailand (19º15.18 N, 099º 03.35 E to 19º 15.25 N, 099º 17.43 E, Fig. 1). It is multi-purpose and encompasses fisheries as an asset. Its elevation ranges from 412 to 425 m ASL with a catchment area of 1,309 km2, a water surface of 16 km2 and it can store up to 265 million m3 of water. It was dammed across the Mae-ngad stream, one of the first order stream tributaries of the Ping River Basin. The maximum depth of the reservoir area is 30 m with a mixed clay and silt bottom. Meanwhile, the depth of the tributary streams, connected to the reservoir, ranges between 0.25-2.0 m and there are various bottom types (i.e. rock, gravel, sand, silt and mud) along the stream gradient. There is an area of forest cover around the reservoir without any agricultural activities or villages in the vicinity. Field sampling Fish were sampled monthly from October 2002 to September 2003 from 10 stations in the tributaries and 2 stations in the reservoir (Fig. 1). The sampling of the tributaries was done by using electro-fishing with a gasoline-powered electro shocker (Honda EM 650, DC 220 V 550BA 450VA, 1.5–2 A, 50 Hz). Electric shocking at each tributary sampling station was carried out for 45 to 60 minutes and the area covered was about 100 m2, in various microhabitats, according to the bottom types. Sampling of the reservoir were done by gillnetting (mesh size 20, 40, 70, and 100 mm stretched mesh and each net’s dimension was 25 x 1.2 m2) from 0600 pm to 0600 am at the two sampling stations. Samples were identified in the field, sacrificed in a lethal 177    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   Figure 1. Map of the location of the sampling stations. Stream stations: S = Huay Mesoon, K = Huay Mekhod, P = Huay Mepang, T = Huay Tontong, Y= Huay Tonyang, M= Huay Mekua, H = Huay Phakub, J = Huay Mejog, W = Huay Panwa, C = Huay Chompoo and Reservoir stations: L = Lower part of the reservoir and U = Upper part of the reservoir solution of anesthetic, counted, weighed (to the nearest 0.1 g), measured (to the nearest 0.1 mm), and fixed in 10% formalin. Unidentified samples were taxonomically classified in the Laboratory of Ichthyology at Maejo University and specimens were kept in the Maejo Aquatic Resources Natural Museum (MARNM), Chiangmai, Thailand. Environmental variables were recorded at each sampling station, consisting of 9 physicochemical and 2 geo-morphological variables. Alkalinity (mg l-1), hardness (mg l-1 as CaCO3), Total Dissolved Solid (mg l-1 FRQGXFWLYLW\ ȝ6FP-1), dissolved oxygen (DO: mg l-1), and pH were measured in-situ using a YSI Model 85 instrument. The ammonia (mg l-1), nitrite (mg l-1) and total phosphorus (mg l-1) were measured in the laboratories according to the standard methods of APHA (1991). Stream width 178    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   and depth were measured every 25 m along 100 m stream reach and then averaged to be a single estimator. In the reservoir, depth was measured at the sampling station, where gillnets were set. Meanwhile, width was estimated at 100 m according to the length of the series of gillnets. Data analyses and modeling procedures The contribution of individual species was presented as the percentage of the relative abundance (%RA) and the percentage of occurrence frequency (%OF). A data matrix was constructed with each row comprised of 66 species and 11 environmental variables of 144 surveys (i.e. the combination of station * month, e.g. L09 is the survey of station L in September). To present the fish assemblages, a self-organizing map (SOM), which is an unsupervised algorithm of an artificial neural network model (Kohonen 2001), was used. The advantage is that this method can be used to analyze complex data sets and for the analysis of non-linear relationships (Kohonen 2001), and to obtain a two-dimensional map for easy interpretation. The SOM has proved to be an effective and powerful tool for describing species distributions and assemblages (Suryanarayana et al. 2008) The principle of SOM analysis is to classify the sample vectors (SVs), described by a set of descriptors on the map according to the similarities between the descriptors (i.e. fish species). Two SVs that are similar (from the descriptor point of view) are classified in the same or neighboring cells, whereas two different SVs are classified in separated cells that could be distant from each other (Tudesque et al. 2008). The processing elements in the network, called neurons, are arranged in a layered structure. The first layer, called the input layer, connects with the input variables. In our case, this was comprised of 66 neurons connected to the corresponding 144 surveys (i.e. 144 SVs). Then the second layer is the output layer that connects to the output variables. The output layer was made up of 56 output units in the hexagonal lattice (i.e. 8 x 7 cells), which provided the best results with which to classify community structures. The learning process of the SOM was carried out by using the SOM Toolbox (Vesanto et al. 1999) and the similarities between each cluster by mean of an analysis of similarities (ANOSIM) by analyzing the occurrence probability (OP) of individual species, which was obtained from the weighted vectors 179    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   of the trained SOM (Kohonen 2001). The analysis of similarity (ANOSIM), a nonparametric test of significant difference between two or more groups, based on any distance measure (Clarke 1993), was used to assess significant differences between communities. The predictive power and contribution of each environmental parameter to the patterns of fish assemblages was carried out by using the Classification and Regression Tree (CART: dé Ath and Fabricius 2000). CART explains variations in a single response variable using one or more predictor variables. To make a tree, the entire data set is referred to as the root node of the tree. This root node is partitioned into subsets of data that then comprise subsequent nodes. If a node is not subject to further partitioning, that node is called a terminal node (Anderson et al. 2000). The process is repeated until the tree can no longer be grown based on a set of stopping rules and cross-validation of the model. The graphics and statistical analyses were carried out with version 2.7.0.0 of the R-Program (R Development Core Team 2009). RESULTS Species composition and community assemblages A total of 11,763 individuals from 66 species and 21 families (Table 1) were sampled. The dominant families were Cyprinidae (34.9 %), Balitoridae and Cobitidae (10.6 %) and Bagridae (6.1%). In terms of the trophic guilds, they were dominated by invertivores (47.5 %), followed by carnivores (31.8 %) and herbivores (20.7 %). The first three species that had highest percentage of relative abundance (%RA) were Henicorhynchus siamensis (Sauvage, 1881), Mystacoleucus marginatus (Valenciennes, 1842) and Puntioplites proctozysron (Bleeker, 1865) (Fig. 2). The highest percentages of occurrence frequency (%OF) were shown by M. marginatus, Oxyeleotris marmorata (Bleeker, 1852), and Hampala macrolepidota Kuhl & Van Hasselt, 1823 (Fig. 2). According to the nature of the surveys contained in each cluster (Fig. 3), the clusters can be designated into reservoir-, stream-, and intermediate- communities, in which there were highly significant variations in the community structures (i.e. occurrence probability (OP) of individual species) among communities (ANOSIM, R=0.757, P<0.001). The reservoir community (RC) was characterized by the surveys 180    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand a   Total length (mm) Family Scientific name Abbrev. Habitat Guilds Mean ±SD Notopteridae Cyprinidae 181 Balitoridae a     Notopterus notopterus (Pallas, 1769) Barilius koratensis (Smith, 1931) Barilius pulchellus (Smith, 1931) Danio albolineatus (Blyth, 1860) Esomus metallicus Ahl, 1923 Rasbora paviana Tirant, 1885 Barbonymus gonionotus (Bleeker, 1850) Barbonymus altus (Günther, 1868) Cirrhinus cirrhosus (Bloch, 1795) Cyclocheilichthys armatus (Valenciennes, 1842) Cyprinus carpio Linnaeus, 1758 Discherodontus schroederi (Smith, 1945) Garra cambodgiensis (Tirant, 1883) Hampala macrolepidota Kuhl & Van Hasselt, 1823 Henicorhynchus siamensis (Sauvage, 1881) Labeo chrysophekadion (Bleeker, 1850) Labeo rohita (Hamilton, 1822) Labiobarbus lineatus (Sauvage, 1878) Mystacoleucus marginatus (Valenciennes, 1842) Neolissochilus stracheyi (Day, 1871) Puntioplites proctozysron (Bleeker, 1865) Puntius brevis (Bleeker, 1850) Puntius stoliczkanus (Day, 1871) Puntius orphoides (Valenciennes, 1842) Homaloptera smithi Hora, 1932 Homaloptera zollingeri Bleeker, 1853 Nemacheilus binotatus Smith, 1933 Schistura breviceps (Smith, 1945) Schistura obeini Kottelat, 1998 Schistura sexcauda (Fowler, 1937) Nnot Bkor Bpul Dalb Emet Rpav Bgon Balt Ccir Carm Ccar Dshc Gcam Hmac Hsia Lchr Lroh Llin Mmar Nstr Ppro Pbre Psto Porp Hmit Hzol Nbin Sbre Sobe Ssex 181 IC SC SC SC SC SC IC IC RC IC RC SC SC IC IC RC RC IC IC SC IC SC SC IC SC SC SC SC SC SC CAR INV INV INV INV INV HER HER HER INV HER INV HER CAR HER HER HER HER INV HER HER INV INV INV INV INV INV INV INV INV 239.4±44.3 51.2±11.3 57.8±14.7 52.3±7.6 49.8±6.0 50.7±12.3 226.8±73.6 180.0 325 48.0±25.9 295.0 46.2±10.1 43.5±11.8 225.9±67.5 220.6±30.8 218.3±19.8 242 221.9±33.2 67.0±28.0 162.0 179.9 ±26.6 41.5±16.5 39.1±7.9 111.5±64.8 25.8±5.9 27.3±5.6 28.3±9.6 36.0±13.4 42.7±16.7 40.3±13.3 Range 132-395 32-86 30-105 43-69 40-61 10-90 15-580 180 325 22-132 295 26-88 11-99 9-600 10-320 175-265 242 114-275 15-152 162 10-235 10-96 25-55 12-210 19-48 15-55 15-52 15-75 21-72 21-72 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 1 Species composition of fish collected in the Mae-ngad reservoir and its tributaries between October 2002 and September 2003 a   Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   2003 Total length (mm) Family Scientific name Abbrev. Habitat Guilds Mean ±SD Cobitidae 182 Amblycipitidae Bagridae Clariidae Pangasiidae Sisoridae Loricariidae Synbranchidae Mastacembelidae Belonidae Chandidae Cichlidae a     Tuberoschistura baenzigeri (Kottelat, 1983) Acanthopsoides delphax Siebert, 1991 Acantopsis choirorhynchos (Bleeker, 1854) Acantopsis thiemmedhi Sontirat, 1999 Syncrossus beauforti (Smith, 1931) Yasuhikotakia morleti (Tirant, 1885) Lepidocephalichthys hasselti (Valenciennes, 1846) Pangio anguillaris (Vaillant, 1902) Amblyceps mucronatum Ng & Kottelat, 2000 Hemibagrus nemurus (Valenciennes, 1840) Pseudomystus siamensis (Regan, 1913) Mystus mysticetus Roberts, 1992 Mystus singaringan (Bleeker, 1846) Clarias batrachus (Linnaeus, 1758) Clarias hybrid Pangasianodon gigas Chevey, 1931 Pangasianodon hypophthalmus (Sauvage, 1878) Glyptothorax trilineatus Blyth, 1860 Pterygoplichthys disjunctivus (Weber, 1991) Monopterus albus (Zuiew, 1793) Macrognathus siamensis (Günther, 1861) Mastacembelus armatus (Lacepède, 1800) Mastacembelus cf. tinwini Britz, 2007 Xenentodon cancila (Hamilton, 1822) Parambassis siamensis (Fowler, 1937) Oreochromis hybrid Oreochromis niloticus (Linnaeus, 1758) Tbae Adel Acho Athe Sbea Ymor Lhas Pang Amuc Hnem Pssi Mmys Msin Cbat Chyb Pgig Phyp Gtri Pdis Malb Msia Marm Mtin Xcan Psia Ohyb Onil 182 SC SC SC SC SC SC SC SC SC IC IC IC IC SC RC RC RC SC RC SC SC SC SC SC IC IC IC INV INV INV INV INV INV INV INV INV CAR CAR CAR CAR CAR CAR HER HER INV HER INV INV INV INV INV INV HER HER 32.2±7.5 39.4±10.3 57.3±13.8 63.1±18.1 67.5±20.0 47.4±11.7 41.8±11.5 68.7±18.7 54.0 191.7±87.3 37.1±20.0 137±66.7 107.1±29.9 119.4±28.5 250.0 940 679.4±101.2 47.5±17.7 301.0 300.7±132.9 99.7±48.2 95.7±18.4 92.8±17.8 166.4±48.2 34.3±3.9 180.0 156.4±97.3 Range 20-34 24-66 28-154 32-95 21-90 39-71 27-75 52-89 54 38-390 15-69 135-315 55-167 95-180 250 940 305-900 27-59 301 46-888 54-150 60-145 80-124 90-222 28-45 180 30-400 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 1 (cont.) Species composition of fish collected in the Mae-ngad reservoir and its tributaries between October 2002 and September    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand a   Total length (mm) Family Scientific name Abbrev. Habitat Guilds IC IC IC SC SC IC IC IC RC CAR CAR HER INV INV INV CAR CAR CAR Mean ±SD Eleotridae Anabantidae Osphronemidae Nandidae Channidae 183 Tetraodontidae Oxyeleotris marmorata (Bleeker, 1852) Anabas testudineus (Bloch, 1795) Osphronemus goramy Lacepède, 1801 Trichogaster trichopterus (Pallas, 1770) Trichopsis vittata (Cuvier, 1831) Pristolepis fasciatus (Bleeker, 1851) Channa gachua (Hamilton, 1822) Channa striata (Bloch, 1793) Tetraodon leiurus Bleeker, 1851 Omar Ates Ogor Ttri Tvit Pfas Cgac Cstr Tlei 141.1±94.4 151.5±18.3 162.7±58.1 74.6±18.9 40.2±10.4 141.5±17.1 95.2±32.0 209.9±154.3 126.9±11.4 Range 20-450 120-190 115-380 50-92 22-55 100-187 40-178 15-590 95-148 Note: Habitats: RC=Reservoir community, IC=Intermediate community, SC=Stream community; Fish guilds: INV=Invertivorous (including; benthic invertebrate, insectivorous, molluscivorous), CAR=Carnivorous, HER=Herbivorous (including; planktivorous, omnivorous) a     183 Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand Table 1 (cont.) Species composition of fish collected in the Mae-ngad reservoir and its tributaries between October 2002 and September 2003    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   during the early and late part of the year from the reservoir stations (e.g. L01 and U10). The stream community (SC) contained the surveys from stream stations during the beginning and late part of the year (e.g. K02 and T04) and only one survey from the reservoir (U05) was included in this group. The remaining surveys were grouped together in the intermediate communities that included almost all the surveys during the mid part of the year that coincided with the rainy season. Figure 2. Percentages of relative abundance (%RA) and occurrence frequency (%OF) of the fish samples found in the overall study Figure 3. Results of the SOM model (a) Distribution of the surveys based on the SOM map of 144 surveys according to the similarity of fish composition. Each survey is represented by the abbreviated station-month names (e.g. S02 is sampling at the Huay Mesoon station in February) (b) Hierarchical clustering of sampling stations showing the three communities. 184    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   Figure 4 Community characteristics for each cluster as shown by the occurrence probability (OP) of individual species. The dotted line at 0.3 indicates the dominant species. Species abbreviations are shown in Table 1. Table 2 Mean values rSD of environmental variables in the three communities Variables RC IC SC 110.16±36.9a 116.16±64.84a 102.87±47.2b Hardness (Har: mg l as CaCO3) 84.2 ± 13.4a 93.2 ± 42.4ab 104.4 ± 31.5b Total dissolved solid (TDS: mg l-1) 130.7 ± 23.7a 149.9 ± 65.1a 121.8 ± 67.4a &RQGXFWLYLW\ &RQȝ6āFP-1) 158.6 ± 40.7a 185.2 ± 115.7a 160.0 ± 85.1a 0.2 ± 0.2a 0.5 ± 0.4b 0.3 ± 0.3ab 0.06 ± 0.02a 0.10 ± 0.03b 0.09 ± 0.05ab 0.03 ± 0.04a 0.03 ± 0.04a 0.03 ± 0.03a Dissolved oxygen (DO: mg l ) 6.5 ± 0.3a 6.9 ± 0.5a 6.8 ± 0.3a pH 6.7 ± 0.3a 6.7 ± 0.5a 7.0 ± 0.4a Depth (Dep: m) 2.2 ± 0.0a 0.9 ± 0.5b 0.8 ± 0.1b Width (Wid: m) 100.0 ± 0.0a 32.4 ± 25.9b 21.2 ± 7.8b -1 Alkalinity (Alk: mg l ) -1 -1 Ammonia (Amm: mg l ) Phosphate (Pho: mg l-1) Nitrite (Nit: mg l-1) -1 Note: The same letter above a value indicates that the values are not statistically different (Tukey H6'WHVWVĮ  The distributions of OP of each species in each community can be expressed as the community characteristics (Fig. 4) and the base line of 0.3 was arbitrarily set to show 185    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   the dominant species in each community but two species gave the highest OP of all communities i.e. M. marginatus and O. marmorata. Two other species contributed to a high OP in the SC i.e. Rasbora paviana (Tirant, 1885) and Channa gachua (Hamilton, 1822). The IC was dominated by Cyclocheilichthys armatus (Valenciennes, 1842), Barilius koratensis (Smith, 1931), Garra cambodgiensis (Tirant, 1883), Pseudomystus siamensis (Regan, 1913), Schistura sexcauda (Fowler, 1937), Acantopsis choirorhynchos (Bleeker, 1854) and Mastacembelus armatus (Lacepède, 1800). It is also worthy to note that the dominant species in the SC and IC were either invertivores or carnivores. Meanwhile, the RC was dominated by a number of species that were mostly herbivores e.g. Labiobarbus lineatus, (Sauvage, 1878), P. proctozysron, H. macrolepidota, Pristolepis fasciatus (Bleeker, 1851) and H. siamensis. Table 3 Confusing matrix showing the cross validation of the CART model using the six environmentalvariables on 3 communities (overall percentage of successful prediction is 66.0 %) % success Predicted Observed RC IC SC RC 14 0 0 100 IC 8 15 64 73.6 SC 1 17 25 39.5 Prediction of community assemblages and the contribution of environmental variables The average values of the physicochemical and geo-morphological variables, obtained from the three communities, are shown in Table 2 and they were used as predictors in the CART model to discriminate the clusters of fish communities. Based on the three communities (i.e. RC, IC, and SC), six environmental variables (i.e. water depth, ammonia, hardness, alkalinity, phosphorus and nitrite) were selected to predict the response variables, i.e. community types (Fig. 5). The major variables 186    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   corresponding to assemblages were water depth, which separated the RC from the other communities. Meanwhile, the overlaps between the IC and SC were distinguished by physicochemical variables such as hardness, ammonia, alkalinity, orthophosphate and nitrite. The overall predictive power of this model, i.e. how successfully the model could predict the assigned survey to the right community, was on average 66.0 % (Table 3). Figure 5. CART model of the fish communities in the study area by using 6 environment variables as predictors. DISCUSSION The fish communities in highland tropical streams connected to a reservoir were dominated by cyprinids. Three communities of fish were found in this study. The reservoir community was located in the reservoir with well-adapted riverine species e.g. Labiobarbus leptocheila, Henicorhynchus siamensis, and Hampala macrolepidota. The stream community was located in the uppermost part of the stream; it was dominated by Channa gachua and Rasbora paviana. The intermediate community was located between the other two communities: it contained transitional species e.g. Cyclocheilichthys armatus and Pseudomystus siamensis. There were clear 187    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   differences between fish species in the RC and the other communities, although some species were overlapping between the IC and SC. Particularly, two species were found to be well adapted to all the communities i.e. Mystacoleucus marginatus and Oxyeleotris marmorata. Water depth had the main impact on the change in the communities. Fish in tropical Southeast Asia river basins are dominated by cyprinids followed by silurids (Matin-Smith 1998; Campbell et al 2006; Nguyen and De Silva 2006) but being followed by the Balitoridae and Cobitidae, as in this study, is unique for the stream areas in the region (Kottelat 1998). Meanwhile there are the silurids and characids with fusiform bodies and expanded pectoral fins in the Neotropical regions (Casatti and Castro 2006). These groups are commonly insectivores and this was also reflected in the predominance of this trophic guild in the SC and IC since various stages of insects need to develop in highly oxygenated water (Kottelat 1998; Casatti and Castro 2006; Rolla et al. 2009). Differences in the observed communities can be provided as an assessment of the ecological status (Lasne et al. 2007). The two communities, i.e. SC and RC, which showed the most distinct differences could be described as the communities under “stream environment” and “reservoir environment”, respectively and could be related to the distinction between the rhithron and the potamon in the river course (Welcomme et al. 2006), where the hydrological regime was the major factor controlling fish community patterns (Welcomme and Halls 2005). It was the periods between the beginning and late parts of the year, which coincided with the dry season that made the difference between the RC and SC. During June to October is a rainy season in the area, which results in an increase in the water surface of the reservoir. The increase in the water surface also improves the connectivity between the reservoir and the tributaries and that increases the aquatic biodiversity (Amoros and Bornette 2002; Falke and Gido, 2006) as seen in the results obtained for the IC. The variation in the occurrence probability (OP) of individual species in each cluster indicated the preferred habitat of the species. In the SC, the members were mostly rheophilic species such as B. koratensis, G. cambodgiensis and S. sexcauda, commonly found in small to medium-sized streams in upland areas (Kottelat 1998). They were sensitive to catastrophic and habitat flows (Welcomme et al. 2006) and 188    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   required strong flow conditions to live. Meanwhile C. gachua lives in the backwaters of first order streams and R. paviana is usually found in shallow and moderately flowing streams (Kottelat 1998). In the RC, the species found were the lentic-adapted species, the so called “facultative reservoir species”: they are generally native to the lower portions of a river course (Falke and Gido, 2006). There is a general consensus that fish which are originally riverine concentrate in the reservoir environment that is most similar to a river, i.e. the tributary and littoral areas of reservoirs (see Prchalová et al. 2008). Variations and high overlaps among communities could be due to some species moving in and out of the tributaries during their life cycle (Borcherding et al. 2002). For example, the cyprinid species such as L. lineatus, P. proctozysron, H. macrolepidota and H. siamensis migrate upstream annually to spawn on shallow gravel beds at the confluence or in small rivers during short periods in rainy season, i.e. June to August (de Graaf et al. 2005). This is why these fish also showed ample OP in the IC. Meanwhile, high OP in all communities of O. marmorata and M. marginatus could be caused by movement either for feeding or spawning purposes (Kottelat 1998). The community structure in the headwater depends on abiotic- rather than biotic- factors (Schlosser 1987). Among the selected controlling variables in this study, water depth is the main environmental factor that affected the fish community patterns. An increase in species diversity along the river course from the shallow upstream areas to the deeper areas downstream was emphasized (Martin-Smith 1998). Prchalová et al. (2009) mentioned that the complexity of species composition in a reservoir, increased heading towards the tributary and peaked close to or at the tributary part of reservoir, which agreed with our results obtained for the complexity of the OP in the IC. Other selected variables in the CART to discriminate between the SC and IC were related to the major nutrients in the ecosystem i.e. phosphorus and nitrogen (i.e. in forms of nitrite in this study). Both nutrients always increase during the rainy season and are released from upstream to downstream as well as from the land to the water body and then stimulate primary productivity in the ecosystem (Allen 2001; Wondie et al. 2007). This phenomenon is eventually made more complex in the fish community in the area, at least for feeding purpose (Hoeinghaus et al. 2008). The one hundred percent predictive power for the RC indicated that the 189    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   community assemblages in that area were relatively stable, while the low predictive power for the SC (39.5 %) implied the movement of downstream species into the stream (Grossman et al. 1990). CONCLUSION Results of the study showed two distinct fish community modes that were induced by the reservoir environment (Lienesch et al. 2000), the lentic-adapted species were common in the reservoir (i.e. the RC) and they could invade the tributaries during a certain period in rainy season as shown in the IC and SC (Fig. 4). Meanwhile, the species in the SC could be found in the IC but they were not found in the reservoir area (Fig. 4). Further studies on the function of individual species in each community are recommended. Moreover, an examination of the fish larvae and juveniles in the system should be also be considered since they also move and distribute in the reservoir (Quist et al. 2004); This would also provide information on species interaction and recruitment to the reservoir system. ACKNOWLEDGEMENT A. Suvarnaraksha is grateful to the Royal Golden Jubilee Program of the Thailand Research Fund for supporting his Ph.D. study (Grant PHD/0290/2549). A research grant was also awarded by the National Research Council of Thailand (Grant MJ. 1-45-017.1). Partial support by Franco-Thai Academic Collaboration (Grant PHC 16598RJ) and the French Embassy in Thailand (Grant CNOUS: 2009/2349) made it possible for him to work at the Laboratoire Evolution et Diversité Biologique under the convention for the joint supervision of theses between Ubon Ratchathani University and Paul Sabatier University (Toulouse III). REFERENCES Allen DJ. (2001) Stream Ecology. Kluwer Academic Publications, Massachusetts. Amoros C, Bornette G (2002) Connectivity and biocomplexity in waterbodies of riverine floodplains. Freshwater Biol 47:761-776. 190    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   Anderson MC, Watts JM, Freilich JE, Yool SR, Wakefield GI, McCauley JF, Fahnestock PB (2000) Regression-tree modeling of desert tortoise habitat in the central Majove desert. Ecol Appl 10:890-900. APHA (1991) Standard Methods for the Examination of Water and Wastewater: 16th edition. American Public Health Association, Washington DC. Barrella W, Petrere M (2003) Fish community alterations due to pollution and damming in Tietê and Paranapanema Rivers (Brazil). River Res Appl 19:5976. Borcherding J, Bauerfeld M, Hinzten D, Neuman D (2002) Lateral migrations of fishes between floodplain lakes and their drainage channels at the Lower Rhine: diel and seasonal aspects. J Fish Biol 61:1154–1170. Clarke KR (1993) Non-parametric multivariate analysis of changes in community structure. Austral J Ecol 18:117-143. Casatti L, Castro RMC (2006) Testing the ecomorphological hypothesis in a headwater riffles fish assemblage of the Rio São Francisco, southeastern Brazil. Neotrop Ichthyol 4:203-214. Campbell IC, Poole C., Giesen W, Valbo-Jorgensen J (2006) Species diversity and ecology of Tonle Sap Great Lake, Cambodia. Aquat. Sci. 68:355–373. dé Ath G, Fabricius K.E (2000) Classification and regression trees: a powerful yet simple technique for ecological data analysis. Ecology 81:3178–3192. de Graaf M, Nentwich E, Sibbing FA, Osse JW (2005) Lacustrine spawning, is this a new reproductive strategy among ‘large’ African cyprinid fishes? J Fish Biol 66:1214-1236. Falke JA, & Gido KB (2006) Effects of reservoir connectivity on stream fish assemblages in the Great Plains. Can J Fish Aquat Sci 63:480-493. Gao X, Zeng Y, Wang J, Liu H (2010) Immediate impacts of the second impoundment of fish communities in the Three Georges Reservoir. Environ Biol Fish 87:163-173. Gido KB, Mathews WJ (2000) Dynamics of the offshore fish assemblage in a Southwestern reservoir (Lake Texoma, Oklahoma-Texas). Copeia 2000:917– 930. 191    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   Grossman GD, Dowd JF, Crawford M (1990) Assemblage stability in stream fish: a review. Env Manag 14:661–671. +ODGtN 0 .XEHþND - 0UNYLþND 7 ýHFK 0 'UDãWtN 9 )URX]RYi ( 0DWČQD - 0DWČQRYi9.UDWRFKYtO M, Peterka J, Prchalová M, Vašek M (2008) Czech J Anim Sci 12:537-547. Hoeinghaus DJ, Winemiller KO, Agostinho AA (2008) Hydrogeomorphology and river impoundment affect food-chain length of diverse Neotropical food webs. Oikos 117:984-995. Kohonen T (2001) Self-Organizing Maps: 3rd edition. Springer, Berlin. Kottelat M (1998) Fishes of the Nam Theun and Xe Bangfai Basins, Laos, with diagnoses of twenty-two new species (Teleostei: Cyprinidae, Balitoridae, Cobitidae, Coiidae and Odontobutidae). Ichthyol Explor Fres 9:1-128. Lasne E, Bergerot B, Lek S, Laffaille P (2007) Fish zonation and indicator species for the evaluation of the ecological status of rivers: example of the Loire basin (France). River Res Appl 23:877-890. Lienesch PW, Lutterschmidt WI, Schaefer JF (2000) Seasonal and long term changes in fish assemblage of a small stream isolated by a reservoir. Southwest Nat 45:274-288. Martinez PJ, Chart TE, Trammell MA, Wullscheger JG, Bergersen EP (1994) Fish species composition before and after construction of a main stem reservoir on the White River, Colorado. Environ Biol of Fish 40:227–239. Martin-Smith KM (1998) Relationships between fishes and habitat in rainforest streams in Sabah, Malaysia. J Fish Biol 53:458-482 Mathews WJ, Harvey BC, Power ME (1994) Spatial and temporal patterns in the fish assemblages of individual pools in a Midwestern stream (USA). Environ Biol Fish 39:381–397. McCartney M (2009) Living with dams: managing the environmental impacts. Water Pol 11:121–139. Nguyen TT, de Silva SS (2006) Freshwater finfish biodiversity and conservation: an Asian perspective. Biodivers Conserv 15:3543–3568. Oberdorff T, Guegan JF, Hugueny B (1995) Global patterns of fish species richness in rivers. Ecography 18:345–352. 192    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   Penczak T, Agostinho AA, Gomes LC, Latini JD (2009) Impacts of a reservoir on fish assemblages of small tributaries of the Corumbá River, Brazil. River Res Appl 25:1013-1024. 3UFKDORYi0.XEHþND-9DãHN0HWDO  'LVWULEXWLRQSDWWHUQVRIILVKHVLQ D canyon-shaped reservoir. J Fish Biol 73:54–78. 3UFKDORYi 0 .XEHþND - ýHFK 0 HW DO (2009) The effect of depth, distance from dam and habitat on spatial distribution of fish in an artificial reservoir. Ecol Freshw Fish 18:247–260. Poff NL, Allan JD (1995) Functional organization of stream fish assemblages in relation to hydrological variability. Ecology 76:606–627. Quist MC, Pember KR, Guy CS, Steohen, JL (2004) Variation in larval fish communities: implications for management and sampling designs in reservoir systems. Fish Manag Ecol 11:107-116. R Development Core Team. 2009. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. Roola APPR, Esteves KE, Ávila-da-Silva AO (2009) Feeding ecology of stream fish assemblage in an Atlantic forest remnant (Serra do Japi, SP, Brazil). Neotrop Ichthyol 7:65-76. Schlosser I.J. (1987) A conceptual framework for fish communities in small warmwater streams. In: Mattews WJ, Heins DC (eds.) Community and evolutionary ecology of North American stream fishes. University of Oklahoma Press, Norman, pp. 17-24. Suryanarayana I, Braibanti, A, Rao RS, Raman VA, Sudarsan D, Rao GN (2008) Neural networks in fisheries research. Fish Res 92:115–139. Taylor CA, Knouft JH, Hiland TM (2001) Consequences of stream impoundment on fish communities in a small North American drainage. Regul Riv Res Manage 17:687-698. Tudesque L., Gevrey M, Grenouillet G, Lek S (2008) Long-term changes in water physiochemistry in the Adour-Garonne hydrographic network during the last three decades. Water Res 42:732-742. Vesanto J, Himberg J, Alhoniemi E, Parviainen J (1999) SOM Toolbox for MATLAB 5. SOM Toolbox Team, Helsinki University of Technology. 193    Biology of keystone fish species and fish assemblage patterns in tropical river basin, the modeling approaches: Case study of Ping River Basin, Thailand   Welcomme RL, Halls A (2005) Dependence of tropical river fisheries on flow. In: Welcomme RL, Petr T (eds.) Proceedings of the Second International Symposium on the Management of Large Rivers for Fisheries Vol. 2, Mekong River Commission, Vientiane, pp. 267–284. Welcomme RL, Winemiller KO, Cowx IG (2006) Fish environmental guilds as a tool for assessment of ecological condition of rivers. Riv Res Appl 22:377-396. Wondie, A., Mengistu S, Vijverberg J, Dejen E (2007) Seasonal variation in primary production of a large high altitude tropical lake (Lake Tana, Ethiopia): effects of nutrient availability and water transparency. Aqua Ecol 41:195-207. 194