Model of optimal behaviour in fish larvae predicts that food availability determines survival, but not growth

How do growth and survival of larval fish depend on the availability of zooplankton prey? This question connects plankton ecology to fisheries science and is typically answered by the classical ‘match/mismatch’ hypothesis for explaining recruitment variability to fish populations. In field and experimental data, however, the effects of food abundance on growth are not always unequivocal. Here, we present a detailed state-dependent optimality model of larval cod Gadus morhua in a water column with environmental gradients and circadian light cycles. Larval feeding rates and predation risk both depend on ambient light, body size, swimming activity and depth position, which together describe trade-offs between growth and survival from realistic mechanisms. The model finds the optimal depth (vertical migration) and swimming activity that maximize survival probability to a given size (15 mm). These optimal behaviours in turn determine food intake, gut fullness, growth and survival over time. The model predicts that increasing food leads to increasing survival, whereas growth remains high except when food availability is very low. When food is more abundant, larvae select deeper and safer habitats and are less active to avoid predation. This can explain why prey abundance and larval fish growth are often not strongly correlated in field studies, although there is still an important effect of prey abundance on recruitment success through its effect on predation rates.