Theoretical work on the size-selective vulnerability to predation during the early life history stages of fishes

Scientists believed that if we understood the factors governing the high mortality rates during the larval period of fishes, we could predict recruitment levels in coming years. This thesis concentrated in the investigation of size-selective predation mortality of fish larvae. First, I investigated the general patterns of size-selective predation from experimental studies and evaluated the empirical evidence support for the theoretical models. I found that fish larvae measuring 10% the size of predators were most susceptible to predation. This pattern was constant across a variety of experimental conditions and for four different types of predators.;Second, I investigated how encounter and susceptibility to predation generate size-selective mortality of fish larvae and how it was affected by the abundance and size of the predators. I found that encounter and susceptibility were counter-acting functions. The detection of size-selective removal of individuals and thus the balance between these two models was closely related to the cohort's overall mortality. I also found that the predator characteristics were important in determining the characteristics of survivors.;Third, I investigated how individual larval characteristics may influence their survival and how survivors differed under different selection pressures. I demonstrated that the effect of the larval characteristics in determining the number, length and growth rate of survivors depended on the characteristics of the predator population. This implied that growing faster or being larger does not translate into a universal survival advantage.;Finally, I assessed how the predictions of the individual-based model Compared with changes in the length frequency distributions observed in natural populations. I demonstrated that adding predation to the model make better predictions of the changes in the length frequency distribution observed in Conception Bay for some larval fish species. I also demonstrated that this result was highly sensitive to the accuracy of the estimates of growth rates.;This thesis is an important contribution to the theoretical framework of early life history of fishes. It posed serious questions about the effectiveness of current sampling protocols, the use of the statistical analytical tools, and most importantly the approach in the investigation of size-selective mortality of fish larvae.