Simulation Studies of Fish Larval Survival

Abstract

The results of simulation studies of the fish larval stage within a closed ecosystem model using input parameters appropriate to a hypothetical, but largely “haddock-like” species are described. In a seasonally varying world, a fish larva feeding on rapidly growing food organisms, is likely to experience a relatively high concentration of food for only a relatively short period of time. The implication is that the probability of larval survival could be relatively high provided two main criteria are met: (1) a larva should be in the right place at the right time, to start feeding, just as the biomass of suitable food organisms is beginning to increase, and (2) a larva does not happen to be part of a larval cohort so numerous that its overall grazing power is likely to shorten the time during which the concentration food organism exceeds a critical value. To survive, it is essential that while food is abundant, a larva is able to grow large enough to be able to begin feeding on other and larger food organisms. In practice, some larvae are likely to grow well with a high probability of continued survival whilst others are likely to grow more slowly with a much lower probability of ultimate survival. Individuals that grow slowly may be eliminated by predators before they have had time to exhibit obvious signs of resource limitation, much less signs of starvation. The main conclusion however is that whether larvae die of predation or starvation, the actual number dying is likely to be largely a function of resource limitation, and not something that can be predicted simply from a knowledge of the number of larvae and of the number of predators. Simulation studies suggest that there should be an optimum period for first feeding of larvae but that this is likely to depend on the number of first feeding larvae. In practice therefore, the optimum period for first feeding (and hence spawning) is not likely to to be precisely defined and could easily be spread out over several weeks as is normally observed. The highest levels of simulated larval production were obtained using relatively small biomasses of first feeding larvae. Theoretical results suggested a biomass of first feeding larvae of around 2–5 mg N/m² for optimum larval production.