In this study we explored how simple optimality processes and decisions can actually lead to rather complex temporal variability and variability among individuals at a given time. We pursued these questions in relation to coral reef fish settlement decisions and the implications of these decisions on the distribution and dynamics of these fish. In many species, larvae settle into already-formed groups with conspecifics. The size of these groups can influence both individual growth rate and survival probability. Typically, a tradeoff occurs in which as group size increases survival probability increases but growth rate decreases.

First we developed a model of optimal choice rather than optimal group size. The model showed that variation in optimal group size can arise on a single day from otherwise identical individuals, and allow initial variation in group sizes to persist. We used the model to predict that during settlement D. albisella would show no group size preference early in their settling season, but later should favor smaller groups to maximize growth. Our field data showed preference for larger group sizes throughout the settling season. We also found that some individuals moved away from their groups after settlement. Our observations suggest that long term growth and survival are uncoupled from group-size choice for settlers. We proposed instead that new settlers prefer larger groups where, as the smallest individuals, they can obtain the benefits of enhanced survival but avoid costs of competition; when they become large enough to incur the costs of competition they can then choose to relocate. Movement between and away from social groups is an important factor in D. albisella that is linked to the fitness considerations associated to social dynamics, and thus needs to be included in the study of their populations.

We then incorporated group-size choice and movement in a simulation model to distinguish between pre- and post-settlement processes as factors determining the size and structure of coral reef fish population. The results demonstrated these processes do not act in isolation of each other. Processes influencing the number of new individuals coming into a population will determine the size and number of social groups of that population. Post-settlement growth then determines how many individuals within the population reach maturity within a given year and the make-up of the social groups.

Finally, given the importance of movement associated to fitness and social dynamics of D. albisella, we conducted a study to explore a potential mechanism triggering movement between and away from social groups. Individuals exposed to high levels of cortisol were more likely to move within experimental aquaria than controls. The results suggest a potential link between interrenal activation and migratory behavior. The role of cortisol and hormones in the movement of D. albisella or other reef fishes remains to be examined.

We demonstrated how by focusing on the decision process (i.e., group-size choice) rather than the outcome of fitness trade-offs (e.g., optimal group size) we recreated variation in behavior (i.e., variable group size) typically observed in nature. By incorporating individual choice and flexibility of movement to relocate between groups in a simulation model, we demonstrated that restriction of these behaviors in experimental studies attempting to discern the role of pre- and post-settlement processes in determining population size and structure could confound their results.