Partitioning of fish and diet selection as methods for the reduction of cannibalism in Paralichthys dentatus larviculture

Partitioning of larvae and diet selection were examined as strategies to re duce cannibalism in summer flounder Paralichthys denatatus. Methods of part itioning included physical grading by size, baffle interference, and a cont rol. The baffle consisted of a barrier with cells 14-mm square and walls 9- mm high. The barrier was positioned horizontally at a depth where it would divide the tank into two equal volumes of water. Factors which may characte rize cannibalism were quantified, including: total length, developmental st age, pigmentation, wet weight, and biomass. Engineered to reduce cannibalis m, the baffle instead contributed to a decrease in survival for at least on e experiment (P < 0.05). Grading by size did not significantly increase sur vival over that of the untreated control (P > 0.05), Mean developmental sta ge of P. dentatus in the graded treatment was significantly less than fish in either the control or baffle treatments (P < 0.05). A factorial design o f partition strategy and weaning diet did not demonstrate any difference in survival or measured character (P > 0.05). Finally, both inert feeds and l ive Artemia sp. nauplii were examined for their contribution to the occurre nce of cannibalism. Predicted differences between live and artificial diets were not significant (P > 0.05), During experiments, cannibalism was princ ipally observed in late larval and early juvenile stages and was not restri cted to post-metamorphic juveniles feeding on premetamorphic larvae. Method s normally employed to reduce cannibalism (partitioning of larvae and diet selection) do not appear to be effective for P. dentatus.