Fifteen sites on nine second- and third-order streams in the Mount Cameroon area, with varying degrees of human disturbance, were sampled during wet and dry seasons over 21 months in 2003–2005 to estimate their potential for sustainable exploitation of ornamental fishes. In total, 35 species of fish representing 22 genera and 14 families were captured. By biotope, fish populations per 20 m averaged 528 ± 318.5 fish (8.0 ± 3.02 species) in runs, 86 ± 60.9 fish (2.4 ± 1.06 species) in riffles and 819 ± 480.0 fish (7.8 ± 3.15 species) in pools. Riffles contained significantly fewer individuals and species than either runs or pools. Shannon’s index was similar for runs and pools, averaging 1.646 ± 0.429 and 1.548 ± 0.345, respectively, but differed significantly for riffles, averaging 0.652 ± 0.473. There were no significant differences in number of individual fish, number of species or Shannon’s index among stream order or season, reflecting the stability of these ecosystems throughout the year. Regression of number of fish per 20 linear m of stream, number of species and Shannon’s index against a subjective scale related to stream degradation was significant. In particular, the polluted Limbe River sites had significantly fewer individual fish, fewer species and a lower Shannon index across all biotopes than other sites. Fishing efficiency of local gears was estimated at 27% for easily captured species and 14% for the more difficult ones, and could be used to make relative estimates of fish abundance as a means of monitoring changes in fish abundance. Although species diversity in south-western Cameroon was high, typical of tropical streams globally, total abundance was low, and therefore it is unlikely that a profitable ornamental fish trade could be sustained on the basis of fish captured exclusively from the wild.
Stream degradation, fish abundance and the potential viability of ornamental fisheries in south-western Cameroon
Brummett, R.E., Cargill, C., Lekunzé, L.M., Puddister, D. (2010)
African Journal of Aquatic Science 35(2):155-164