Gillnet selectivity of three flying fish, Cheilopogon nigricans (Bennett, 1846), Cypselurus poecilopterus (Valenciennes, 1846) and Cheilopogon suttoni
M.D.S.T. DE CROOS
On the northwestern coast of Sri Lanka, off Kandakuliya, there is a seasonal fishery for flying fish resources from October to April every year. In this fishery, major species caught were Cheilopogon nigricans, Cypselurus poecilopterus and Cheilopogon suttoni. Gillnet selectivity of these three flying fish species was determined using Baranov-Holt method. The optimal lengths (Lopt) and probabilities of capture in gillnets of 3.4 cm and 4.5 cm mesh sizes were determined for the three species separately. For this analysis, sampling was conducted from October 2002 to April 2003. Estimated values of selection range for Cheilopogon nigricans, Cypselurus poecilopterus, Cheilopogon suttoni for 3.4 cm mesh size were 18.9-26.8 cm, 16.4-24.3 cm, 18.9-29.7 cm, and for 4.5 cm mesh size were 26.3-33.1 cm, 23.0-30.9 cm, 26.7-37.5 cm respectively. Estimated Lopt for 3.4 cm and 4.5 cm mesh sizes respectively were 22.9 cm and 30.3 cm for Cheilopogon nigricans, 20.3 cm and 26.9 cm for Cypselurus poecilopterus and 24.3 cm and 32.1 cm for Cheilopogon suttoni. As these fishes are migratory, when the mesh-wise length frequency data were superimposed on the gillnet selection curves, it was evident that gillnet mesh sizes were effective for catching only a limited range of length classes of all three species studied. In 3.4 cm mesh gillnets, size classes smaller than the optimal length were virtually absent. Also in gillnet catches of 4.5 cm mesh size, size classes bigger than the optimal length were not caught. These indicate that only intermediate size ranges were available in the fishing grounds of all three species. As such, through experience, fishers have identified 3.4 cm and 4.5 cm mesh gillnets as effective fishing gear for catching these species. Hence, strict management options are not an immediately required for managing the flying fish fishery off the northwestern coast of Sri Lanka. As the catch samples of flying fish species do not represent actual size composition of populations, estimation of growth and mortality parameters using length-based stock assessment methodologies is problematic. As such, for determination of optimal fishing strategies using dynamic pool models, independent estimates of demographic parameters should be used.