Life-History Strategies, Trophic Patterns And Community Structure In The Fishes Of Lake Biandantang

Lake Biandantang, a 330hm2 shallow macrophyte-dominated lake, is situated in the middle reach of the Changjiang River, which has been separated from Lake Bao′an through a stone dyke since 1971. Life-history strategies, trophic patterns and community structure in the fishes of Lake Biandantang were studied between 2002 and May 2004 in order to better understand ecological fundamental for regulating of fish community structure and fisheries management. The main results and conclusions summarized as: 1. Scales were reliable for determining the ages of Rhodeus ocellatus, Rhodeus fangi, Rhodeus lighti, Paracheilognathus imberbis, Acheilognathus chankaensis, Acheilognathus macropterus, Micropercops swinhonis, Rhinogobius giurinus, Rhinogobius ciffordpopei and Mugilogobius myxodermus. The annuli on the scales were obvious and could be read easily. Any population of the ten species consisted of only one year-class, the maximum life-span for both sexes was observed to be one year. There was significant difference in tolal length or body weight between both sexes of A. chankaensis, P. imberbis, M. swinhonis and R. giurinus, while no significant difference was observed in the growth of both sexes for the other 6 species. The sex-ratio (males to females) of R. ocellatus, A. chankaensis and M. swinhonis was 0.64:1 (n=482), 0.33:1 (n=809), 0.57:1 (n=706), respectively; these sex-ratioes were significantly different from the theoretic value of 1:1 (Chi-square test, P < 0.01). In contrast, the sex-ratio was 0.70:1 (n=46) for R. lighti, 0.99:1 (n=931) for R. fangi, 0.95:1 (n=788) for P. imberbis, 1.00:1 (n=540) for A. Micropterus, 1.03:1 (n=248) for R. giurinus, 0.82:1 (n=100) for R. ciffordpopei, 0.7:1 (n=17) for M. myxodermus; these sex-ratioes were not significantly different from the theoretic value of 1:1. All fish reached maturity at age 1, exhibiting secondary sexual characters during spawning. In A. macropterus and A. chankaensis, spawning started in mid-February when water temperature was about 10 ℃; in other species, spawning started in April. The average individual fecundity of R. ocellatus, R. fangi, R. lighti and P. imberbis was 114eggs, 60eggs, 124eggs and 114eggs, respectively, which were very lower than that of A. marcropterus and A. chankaensis that had 1343eggs and 1183eggs, respectively. The fecundity was 419eggs for M. swinhonis, 2462eggs for R. giurinus, 677eggs for R. ciffordpopei, and 668eggs for M. myxodermus. In addition, the size of matured eggs was measured, and the relationship between fecundity and total length or body weight was established for every species. 2. Dietary analysis of 37 species belonging to herbivores, detritivores, invertivores and piscivores, was conducted by five indices [percent weight (%W), percent occurrence (%O), percent number (%N), percent index of relative importance (%IRI) and percent index of predominance (%IP)]. Using an extensive data set on the diets, we explored similarities and differences among the above mentioned dietary indices, compared importance of prey taxa rated by the different indices, examined biases in dietary importance rating among indices related to difference in prey size, and assessed advantages and disadvantages of the indices. The results from the present study, suggested that the percent weight is a valid and scientific index, and supported the opinions that compound diet index (%IRI or %IP) is redundancy, nonadditivity and loss of information. 3. Levin's standardized niche breadths (Ba) and trophic levels of 37 fish species were determined in terms of percent weight of diet items. Trophic relationships among the fishes were invetigated and a food web of the fish assemblage was established to further reveal trophic patterns. There was a considerable difference in ecological specialization among 37 species, Ba ranged from 0.000 to 0.584, with a mean of 0.205; trophic levels averaged to be 2.83, with the lowest value of 2.00 for Ctenopharyngodon idellus, Meglobrama amblycephala and A. chankaensis, and with the highest of 3.93for Culter alburnus. Using a similarity criterion of Euclidean distance < 1.2 with all 37 fish species, 9 feeding guilds were determined and one species did not cluster according to the dendrogram from cluster analysis, with detritivore guild having the most species. Analysis of the food composition in conjunction with the main habitat occupation and activity patterns, suggested that ecological segregation existed commonly among 37 species. Furthermore, size selectivity of prey by predators was an important in resource partitioning. 4. We determined the size spectrum of atyid shrimp, Macrobrachium shrimp and fish as prey fed by 8 piscivorous species, and their corresponding prey to predator size ratio. There was no significant relationship between length of every specific predator and length of atyid prey or length of Macrobrachium prey. Prey fish length was not significantly correlated with length of Culter dabry or Mastacembelus sinensis as predator, while significant relationship between prey fish length and predator length was found for C. alburnus, Culterichthys erythropterus, Siniperca chuatsi, Channa argus or Odontobutis obscurus (P < 0.01). 5. Ecomorphological correlates were sought among 33 fish species. Three food characters were selected: food size food type (vegetable versus animal) and the vertical position of food in the water column. Significant correlates between 29 of 48 morphological features and the three functional food characters were found by Pearson correlation analysis. Furthermore, significant relationships between the diet and morphology of the fishes were examined by canonical correlation analysis. Similarity in diet and in morphology was both significantly correlated with taxonomic relatedness. Nevertheless, when taxonomic relatedness was counted for, the relationship between observed similarity in diet and similarity in diet predicted from morphology was still significant. These results lead to the conclusion that species morphology influences the diet and species having similar diet tend to converge for some morphological characters. 6. Community organization in fish assemblages was studied, by using a data set on 48 morphological features relating to many aspects of niche. In order toachieve a morphologically defined niche for each species, a principal component analysis was performed. Eight principal components with eigenvalues greater than one were selected which together accounted for 88.1% of the total variance in the data. Thus the 8 independent components were suited for defining a position for each species in 8-demensional space. The average Euclidean distances between the centers of morphologically defined niches for all 33 species were calculated. The distribution of Euclidean distances between species was significant different from that of distances between all pairs of hypothetical species, which would use resources randomly. This indicated the species currently living together are not a random assortment of species and do not randomly divide their resources. And so, it is believed that evolution has produced a nonrandom assemblage of interacting species. 7. 45 Species belonging to 14 families and 36 genera were recorded. Of these species, 29 are cyprinids, and 31 are small-sized species, respectively contributing 64.4% and 68.9% of the total. In terms of number and biomass, the small-sized fish assemblage was dominated by Micropercops swinhonis, Rhinogobius giurinus, Carassius auratus, Hemiculter leucisculus, Pseudorasbora parva and Rhodeus ocellatus, with significantly different species contribution by number of individuals between the two periods considered, 1994~1995 and 2002~2003. Mandarin fish (Siniperca chuatsi) which dominated piscivorous guild had the annual average yield of 10.8 kg/hm2 in 2001~2003, which was 1.44 time greater than that in 1991~1996. This fish played an important role in the Biandantang ecosystem and fisheries.