Diversity and Distribution of Fin-fish and Prawn Species Assemblages of the Catch in Five Perennial Reservoirs in Northern Province, Sri Lanka

Present study was carried out to assess the abundance and diversity of fin-fish and prawn species in the selected five reservoirs in the Northern Province, Sri Lanka. Fin-fish and prawns caught by multi-mesh gillnets, and rod and hooks, were sampled from January to December 2017. Relative abundance and diversity indices were calculated and, differences and similarities among reservoirs in different months were assessed. The hierarchical clustering was used in constructing dendrogram. Of the 27 finfish species and two freshwater prawn species identified in total, the highest number of species were observed in Muthayankattu reservoir (26 species) while Puthumurippu reservoir had the lowest (14 species). Oreochromis niloticus was the dominant species in all the reservoirs. Small reservoirs (Puthumurippu, Kalmadu and Muhathankulam) showed high Shannon-Weiner diversity index and evenness, though low in dominance index, compared to large reservoirs (Vavunikulam and Muthayankattu). Selective harvest methods, number of fishers and seasonal variation of hydro-climatic factors such as rainfall, strong wind and water level significantly influenced the diversity of fish species. Increasing fishing pressure and use of destructive fishing methods during low water level were the main contributors for diversity loss in the five perennial reservoirs studied.

differences and similarities among reservoirs in different months were assessed. The hierarchical clustering was used in constructing dendrogram. Of the 27 finfish species and two freshwater prawn species identified in total, the highest number of species were observed in Muthayankattu reservoir (26 species) while Puthumurippu reservoir had the lowest (14 species). Oreochromis niloticus was the dominant species in all the reservoirs. Small reservoirs (Puthumurippu, Kalmadu and Muhathankulam) showed high Shannon-Weiner diversity index and evenness, though low in dominance index, compared to large reservoirs (Vavunikulam and Muthayankattu). Selective harvest methods, number of fishers and seasonal variation of hydro-climatic factors such as rainfall, strong wind and water level significantly influenced the diversity of fish species. Increasing fishing pressure and use of destructive fishing methods during low water level were the main contributors for diversity loss in the five perennial reservoirs studied. DOI: http://doi.org/10.4038/tar.v31i3.8400

INTRODUCTION
Reservoirs in tropics are primarily used for irrigation and/or hydroelectricity, and can cause high water level depletion since fisheries is a secondary activity (De Silva, 1996). Each reservoir is unique. Majority of reservoir fisheries are artisanal and highly seasonal in many parts of the world (Welcomme, 2001). Sri Lanka does not possess any natural lakes whereas its fish population is dominated by traditionally unexploited fish species (De Silva, 1988;Pet et al., 1999). These endemic and indigenous fish species are not completely adapted to lacustrine conditions. Therefore, native fish species alone cannot produce an adequate catch. As a result, exotic species play a major role in reservoir fisheries (De Silva, 1996).
Sri Lanka has approximately 137 species of freshwater fish fauna comprised of 90 true freshwater fishes including 53 endemics, 23 estuarine and 24 exotic species. Most of exotic fishes are well established in the island and account for an overwhelming share of fish biomass (Wildlife conservation society -Galle, 2015). Fishing methods in Sri Lankan reservoirs are mostly limited to gillnets, which are highly selective, and towing or surrounding nets are prohibited according to inland fisheries regulations (Amarasinghe and Pitcher, 1986;Amarasinghe et al., 2014). Gillnets which are passive standard gear (Hamley, 1980) are commonly used in sampling fish in reservoirs. Studies of freshwater fish species are essential for planning reservoir fisheries development and management, and especially studies on smallsized fish species which are in high abundance in most tropical reservoirs are vital. Future demand for fish can be partly met by exploiting these untapped fish resources. The objective of this study was to assess and compare the abundance and diversity of finfish and prawn species in fish catch of five reservoirs in Northern Province of Sri Lanka.

MATERIALS AND METHODS
The study was conducted in five perennial reservoirs ( Figure 1) in Northern Province of Sri Lanka, viz., Vavunikulam, Muthayankattu, Puthumurippu, Kalmadu and Muhathankulam with areas of 1275 ha, 1255 ha, 151 ha, 74 ha and 211 ha, respectively at Full Supply Level (FSL). Reservoirs were categorized into two groups for comparison according to areas at FSL. Vavunikulam and Muthayankattu, which showed FSL above 1,000 ha were considered as large reservoirs and Puthumurippu, Kalmadu and Muhathankulam, with FSL less than 1,000 ha, were grouped as small reservoirs.

Figure 1. Locations of selected reservoirs
Fish and prawns caught by multi-mesh gillnets (50.8 mm, 63.5 mm, 88.9 mm, 101.6 mm, 114.3 mm 127 mm, 152.4 mm and 177.8 mm, knot to knot) and rod and hooks were sampled from January to December 2017. These gillnets were set in almost all parts of each reservoir by local fishermen in the evenings around 4.00 p.m. to 6.00 p.m., and collected on the next day morning around 5.00 a.m. to 9.00 a.m. It was assumed that all the reservoirs used nets with same mesh size. All fish and prawn species were sorted and identified to species level according to Pethiyagoda (2006). Total number of species and individual numbers were recorded for each month.
Species diversity was assessed by using four indices namely, species richness, Shannon-Wiener diversity, Evenness and Dominance. Shannon-Weiner diversity index (H') (Shannon and Weaver, 1963) was calculated by the equation (1) where, S = the total number of species Pi = the relative cover of i th species.
Species richness was calculated by using Margalef index (d) (Margalef, 1968) using the equation ( where, S = the total number of species.
The dominance index (Harper, 1999) was measured to identify whether particular species was dominating in the specific aquatic resource. Dominance index was calculated by using equation ( where, ni = the number of individuals of species i N = the total number of individuals. The relative abundance of each species was calculated using equation (5) Relative abundance = ( ) 100 % ……………. (5) where, ai = the number of individuals caught in the i th species A = the total number of individuals One-way analysis of variance (ANOVA) was used to identify difference of diversity indices among the reservoirs and among months of the year. Tukey comparison was used to compare means (Spjotvoll and Stoline, 1973). Similarity percentages analysis (SIMPER) (Clarke, 1993) was performed to determine the percentage of similarity among reservoirs and among months of the year. The hierarchical clustering (Clarke and Warwick, 1994;Hossain et al. 2012) was calculated to produce a dendrogram. All statistical procedures were performed using Microsoft Excel 2016, Minitab 17 statistical software and Paleontological Statistics (PAST) version 3.0.

RESULTS AND DISCUSSION
A total of 43,408 individuals were quantified, comprised of 27 fin-fish species and 2 freshwater prawn species (Table 1) Present study consisted of 29 species in total compared to Amarasinghe et al. (2014) who observed 30 species in Minneriya reservoir, 21 species in Udawalwe and 18 species in Victoria reservoir, with 12 different families dominated by Cyprinidae where fishing was done by using beach seines (1, 5 and 7 mm mesh size) and multi-mesh monofilament gillnets (12.5, 16, 20, 25, 33, 37, 50, 60, 76 90 mm stretched mesh size). Chandrasoma et al. (2015) reported that species composition in the catch significantly varied after stocking with Indian major carps, where O. niloticus contributed to 80-90% before the year 2004, and then reduced in 2004 up to 64.6% in smaller reservoirs and 57.1% in larger reservoirs. Indian major carps C. catla, L. rohita and other carp species were reported contributing to 41.9% in minor and 39.5% in medium reservoirs. Their contribution was also reported, 29.1% in Senanayaka Samudra and 38% in Jayanthi wewa which are major perennial reservoirs, while M. rosenbergi contributed to only 1.6%, 0.5%, 0.4%, 1% in major, minor, Senanayaka Samudra and Jayanthy wewa reservoirs, respectively. Dematawewa et al. (2008)  At the similarity level of 22%, the relative abundance data identified two major clusters in cluster analysis ( Figure 2). Results revealed that larger reservoirs had very distant similarities (22%) with smaller reservoirs except from August to October in Muhathankulam. The reason could be that, high fishing pressure was applied by fishers during those three months through using less selective gear due to the low water level resulting from poor rainfall and high water drainage from the sluice to irrigate paddy.  Small reservoirs recorded high Shannon-Weiner diversity index and evenness compared to those recorded by large reservoirs, whereas the dominance index in large reservoirs were higher than that in smaller reservoirs. Means of Shannon-Weiner diversity index, dominance index, evenness and species richness significantly differed among reservoirs (P = 0.000) but not among months. Four diversity indices of Vavunikulam and Muhathankulam reservoirs significantly differed from all other reservoirs except species richness of Kalmadu reservoir.
Though reservoirs of Northern Province have short development history of culture based fisheries, findings revealed that exotic species dominated the catch throughout the year. Present study was purely based on catch from multi-mesh gillnets, where minimum mesh size was 88.9 mm regulated by National Aquaculture Development Authority (NAQDA), Sri Lanka. Species composition of catch in the present study would have been influenced by fishing methods and gear used for target species. Vavunikulam and Muthayanakttu Fisheries Societies had their own regulations on mesh size, where minimum size was 101.6 mm and 114.3 mm, respectively. However, such regulations on gear size did not exist in Puthumurippu, Kalmadu and Muhathankulam reservoirs. Even though mesh size limitation is there, fishers used 50.8 mm and 63.5 mm, during two to three months every year mainly targeting L. dussumieri, Ompok bimaculatus at Vavunikulam reservoir. Use of rod and hooks mainly targeting for freshwater eels (Anguilla spp.) also have been recorded in few seasons. This could be one of the reasons for low species diversity observed in the present study as most of the minor cyprinids, which are smaller in size did not get caught in this type of gear.

Figure 3. Mean value of four species diversity indices at selected reservoirs in 2017.
Use of multi mesh gillnets is often found in Sri Lankan reservoir fisheries, and it varied from season to season, due to seasonal variation of catchability of target species (Dematawewa et al., 2008). Engagement of part-time fishers in agricultural and the other activities depends on several external factors, especially the rainfall and other weather conditions (Dematawewa et al., 2008) which may also the reason for seasonal variation of fishing pressure in reservoirs. Seasonal variation of hydroclimatic factors such as rainfall and wind related factors, water level, waves, turbidity and water chemistry highly influence the productivity of reservoirs (Moses et al., 2002) and catch per unit effort in many inland reservoirs (Hontela and Stacey, 1990;Moses et al., 2002). Intense exploitation is common during low water levels whereas recovery of fish stock occurs during wet season (Dematawewa et al., 2008).
Increasing fishing pressure in Kalmadu and Muhathankulam reservoirs during low water level was observed. This is one of the main triggering factors for fisheries diversity loss. There are few technical factors, which influenced gillnet selectivity and catch size such as mesh size, hanging ratio (Fonseca et al., 2005), twine thickness (Holst et al., 2002) and colour of twine (Tweddle and Bodington, 1988). Further, presence of body projections such as teeth or spines, facilitated significant proportion of fish being entangled in gillnets (Sparre and Venema, 1998). High swimming activity of fish is another biological characteristic which can result in high probability of encounters (Rudstam et al., 1984).
The number of species or richness and the distribution of individuals among species are the two components of species diversity (Magurran, 1988). Low diversity in Vavunikulam in February could be due to high contribution of O. niloticus (87.38%) owing to highly selective harvest of tilapia. However, the high diversity observed in Muhathankulam during October could be because of high fishing pressure using mostly multi-mesh gillnets of 63.5 mm, 76.2 mm and 88.9 mm mesh. In addition, fishers tend to practice destructive water beating fishing method since water levels are low.
High dominance index involved with highly selective harvest methods and low dominance involved with less selective harvesting methods practiced by fishers. Fishers in larger reservoirs mostly practiced high selective harvest methods especially for O. niloticus. When the temporal variation of dominance status was compared, fluctuation in a narrow range was observed except in Vavunikulam during April-June where application of small mesh gillnet targeting L. dussumieri was practiced. The reason for high evenness recorded in Puthumurippu in November could be the low number of species and individuals recorded during the period. Similarly, low evenness was recorded in Vavunikulam in February because of target highly towards O. niloticus. Muthayankattu showed a high richness in January due to the use of gear of different mesh sizes together with rod and hooks, whereas Kalmadu recorded low richness in April due to low fishing intensity.
There was strong positive correlation (0.93) between Shannon-Weiner and Evenness index confirming the observation of Nair et al. (1989) who showed the similar relationship of fish species diversity in the Nair river of the Western Ghats of India. However, strong negative correlation (-0.935) was found between Shannon-Weiner and Dominance index in the present study, confirming the findings of the study in Naaf river estuary by Chowdhury et al. (2010).
Variations in biodiversity indices could be due to seasonal variations of nutrients affecting the coexistence of many fish species, atmospheric air currents, environmental conditions and seasonal fish migrations (Huh and Kitting, 1985). Ajith Kumara et al. (2009) indicated possibilities of inshore-offshore migration of species associated with the water level fluctuations, and poor breeding in lacustrine habitat. Although indigenous cyprinid species require riverine habitats for spawning, similar to the African cichlid species they exploit the lacustrine habitats of reservoirs for feeding grounds, colonizing and due to anthropogenic effects (De Silva, 1983;De Silva, 1988).
According to Keskin and Unsal (1998), the reason for showing lower species diversity in the catch in their study was due to the fishing gears which had high selectivity. In most reservoirs of Sri Lanka, fishers beat the water with wooden poles or weighted ropes to drive fish towards gillnets. The efficiency of this water beating technique is significantly higher than that of normal gillnetting, (Amarasinghe and Pitcher, 1986), and is a common method during dry season. Cichlids exhibit depth preferences according to body size, where small individuals prefer shallow littoral zones while adults prefer deeper pelagic zones (Caulton and Hill, 1973;Ribbink and Hill, 1979). Juveniles of exotic cichlids in Sri Lankan reservoirs have spatially segregated from the minor cyprinids. Therefore, when the fishing intensity targeting O. niloticus is high, it may cause low diversity in the catch (Pet et al., 1999;Ajith Kumara et al., 2008). Hence, as recommended by Amarasinghe and Pitcher (1986) and Ajith Kumara et al. (2009) introduction of an additional gillnet (< 52 mm stretched mesh) for minor cyprinids without harming juvenile cichlids is feasible in these reservoirs too.

CONCLUSION
Selected five reservoirs are valuable for fisheries and they provide favorable condition for fish abundance and diversity. High species numbers were observed in larger reservoirs, which were dominated with exotic species, while smaller reservoirs were dominated with native species, though O. niloticus was highest abundance in the catch. Smaller reservoirs have higher Shannon-Weiner diversity index and evenness than larger reservoirs, while larger reservoirs have high dominance index than smaller reservoirs. Anthropogenic factors such as fishing pressure, selective harvest methods and water effluent for irrigation act as driving forces for the abundance and diversity of fish and prawn species in the catch.