Diversity of fishes in Beles and Gilgel beles Rivers, abay basin, Ethiopia

Species diversity and abundance refl ect the quantity and quality of the available habitat. The decline in abundance of freshwater fi sh in the world has been of concern for over one hundred years. Since the twentieth century, many fi sh species have suffered continuing declines in abundance and distribution, some at alarming levels. This includes many of the smaller species as well as all of the species targeted by inland commercial and recreational fi sheries. These declines in abundance are commonly attributed to factors such as general habitat degradation [1], modifi ed patterns of stream fl ow [2], interrupted migratory pathways [3], reduced water quality and pollution [4], introduction of alien fi sh and diseases [5], illegal fi shing and commercial over fi shing [6] and altered biotic interactions [7].


Introduction
Species diversity and abundance refl ect the quantity and quality of the available habitat. The decline in abundance of freshwater fi sh in the world has been of concern for over one hundred years. Since the twentieth century, many fi sh species have suffered continuing declines in abundance and distribution, some at alarming levels. This includes many of the smaller species as well as all of the species targeted by inland commercial and recreational fi sheries. These declines in abundance are commonly attributed to factors such as general habitat degradation [1], modifi ed patterns of stream fl ow [2], interrupted migratory pathways [3], reduced water quality and pollution [4], introduction of alien fi sh and diseases [5], illegal fi shing and commercial over fi shing [6] and altered biotic interactions [7].
As in many parts of the world, population growth, agricultural development and industrialization contribute to the loss of species diversity of freshwater fi shes in Ethiopia [8]. Wide spread deforestation, degradation of the pristine environment, and other human induced factors might have left many Ethiopian streams, specially the northern ones, devoid of fi sh but the apparently resilient cyprinids [8]. As in many parts of the Ethiopia, human activities degrade fi sh habitat in numerous ways in study area. Wild fi re, logging, impoundment, canalisation and agricultural activities are some of the major activities that degrade fi sh habitat. At present, we have no evidence of species extinction from Ethiopian freshwaters (Harrison and Stiassny, undated cited in Abebe Getahun and Stiassny, 1998) resulting from degradation of environment. One of the main reasons is a lack of defi nitive information on diversity of freshwater fi shes and inconclusive data on the freshwater fi sh species.

Materials and methods
Site selection: -A reconnaissance survey was conducted together with the research advisor to fi x sampling sites. The survey was conducted in four sub areas along the Beles and Gilgel Beles Rivers. Two sampling sites were selected from each river taking into consideration the velocity of water, habitat type, altitude, depth of water, vicinity to road and substrate type Table 1, Figure 1. lines were used in areas where gill nets were not suitable. Immediately after retrieval, fi shes were removed and total length and total weight of each specimen were measured. Total length was measured to the nearest 0.1 cm.
Laboratory studies: -The specimens were soaked in tap water for a week to wash the formalin from the specimens. Then, they were transferred to 75 % ethanol. Identifi cation was made to species level by comparing the sample characters with taxonomic keys found in the literature and specimens deposited at the Fisheries Laboratory, Department of Biology, AAU, and also at Bahir Dar fi sheries and other aquatic life Research Center, and at National Fisheries and Other Living Aquatic Resources Research Centre, Sebeta. Keys found in Shibru Tedla [9], Boulenger [10][11][12][13], Lévêque et al. [14,15], Eschemeyer [16], Nagelkerke [17], Bishai and Khalil [18] and Golubtsov, et al. [19], were used for identifi cation. Meristic and morphometric characters were assessed for comparison purpose. Where, pi is the proportion of individuals found in the ith species. Shannon's diversity index (H'´) was used to indicate diversity at different sampling sites and/or rivers. A high value indicates high species diversity. Signifi cance of differences in species diversity between sampling sites and/or rivers was tested using T-test.

Fish species composition of beles and gilgel beles rivers
A total of 23 fi sh species were recorded during the present study from Beles and Gilgel Beles Rivers (Table 2). These fi shes were represented by a single class Actinopterygii (rayfi nned fi shes), seven families and fi ve orders ( Table 2). The Cyprinidae, Bagridae and Characidae were the best-represented families with respect to numbers of species; with 11, 3 and 3 species, respectively ( Table 2). Labeo and Labeobarbus were the best-represented genera with numbers of species; with fi ve and four species, respectively ( Table 2). The freshwater fi sh fauna of Beles and Gilgel Beles Rivers contain a mixture of Nilo Sudanic (e.g. B. docmak, B. bajad, H. forskhalii, L. forskalii, M. kannume, S. serratus and S. schall ), highland East African (e.g. L. intermedius, L. nedgia, C. gariepinus and O. niloticus) and Endemic forms ( e.g. V. beso).

Species diversity
A higher number of species were recorded in the Beles River (22 species) than in the Gilgel Beles River (5 species) in  Gilgel Beles River. L. forskalii is recorded in the present study from Gilgel Beles River but not JERBE [20]. However, Garra sp.
was recorded by JERBE [20], from Gilgel Beles River but not in the present study.
Differences seen in the species composition between the present study and that of JERBE [20,21], might be due to differences in the sampling effi ciency, habitats and seasons.   Gilgel Beles River. There was signifi cant variation in both mean number of fi sh species and diversity index between Beles and Gilgel Beles Rivers (P < 0.05) ( Table 5).
of substrate, river depth and width might contribute to high species diversity in Beles River than in Gilgel Beles.

Species diversity during wet and dry seasons
A higher number of species was recorded in the dry than in the wet season in total catch (22 versus 18 species) ( Table 5) There was no pronounced disparity in the species composition between wet and dry seasons in Gilgel Beles River (Table 6).
However, there was marked difference in species composition during wet and dry seasons in Beles River ( Table 6).
The value of Shannon's diversity index was 0.81 and 0.66 in dry and wet seasons in Gilgel Beles River, respectively ( Table 6).
The index was also higher in the dry (H´=2.48) than wet season (H´ = 2.21) in Beles River (Table 6). Shannon's diversity index (H`) indicated that species diversity was higher in dry than wet season in each river. The species diversity was also higher in the dry (H´=2.29) than wet season (H´ = 1.99) for the total catch ( Table 6). The highest species diversity was obtained in Beles River during dry season (H´ = 2.48) while the lowest in Gilgel Beles River during wet season (H´= 0.66) ( Table 6).   carried out in other areas by Nikolsky [26], Sydenham [27] and Golubtsov and Mina [28]. The increase in species number from up stream sites to down stream sites was associated with change in catchment area, canopy closure, substrate type, distance from source, depth and width of rivers [29]. These variables refl ect longitudinal gradient in the study area. Width of river was the most important variable that coincided with increase in species number from Mh sites to BAB sites. A total of 21 fi sh species found in Beles River at BAB sampling site with its mean river width of 56.5 + 2.12 m while the lowest number of species (4) in Gilgel Beles River at Mh sampling site with its mean river width of 31 + 2 m. This result is consistent with the studies carried out in other areas. In tropical area as Angermerier and Karr (1983) in Panama, EDDS [30] in India and Toham and Teugels [29] in Cameroon found respectively a signifi cant relation ship between species number and width of the river, and species number and increasing gradient of depth.
A total of 12 fi sh species found in Beles River at BB sampling site with its mean river depth of 6.3 + 0.5 m while the lowest number of species (4) in Gilgel Beles River at Mh sampling site with its mean river width of rive 2.42 + 0.28m. A similar result using depth gradient [31], stream order or river position in the gradient [32][33][34][35][36], have also been reported for temperate rivers. In addition, canopy closure and diversity of substrate type (sand, gravel and large rocks) were also most probable environmental gradient explaining the spatial distribution of species in the sampling sites. Thus, the presence of year round dense vegetation and higher catchment area, diversity The number of fi sh species ranged from 3 to 14 with a mean ± SD of 6.57

Conclusions
Beles River is richer than Gilgel Beles River, in terms of both total numbers of fi sh caught and numbers of species recorded.
A higher number of species were recorded in the Beles River (22 species) than in the Gilgel Beles River (5 species) in the present study. The increase in number of fi sh species from lower to upper reaches of study area coincide with decline in elevation.
The Cyprinidae, Bagridae and Characidae were the bestrepresented families with respect to numbers of species; with 11, 3 and 3 species, respectively. The Labeo and Labeobarbus were the best-represented genera with numbers of species.
L. nedgia and L. degeni considered belonging to a single lip morphotype (Labeobarbus nedgia) endemic to Lake Tana by Nagelkerke and Sibbing (1996) are found in Beles River.
The species diversity was also higher in the Beles River (H´= 2.42) than in the Gilgel Beles River (H´ = 0.88) for total catch. A higher number of species was recorded in dry than wet season in total catch (22 versus 18 species). The species diversity was also higher in the dry (H´=2.29) than wet season (H´ = 1.99) for the total catch.

Recommendation
In order to have a better knowledge of the fi sh populations