MEPS 349:43-61 (2007)  -  DOI: https://doi.org/10.3354/meps06963

ABSTRACT: Much research has been done on larval settlement cues. Rather than having simple fixed responses to constant environmental stimuli, it seems likely that settlement decisions made by individual larvae should vary depending on the individual and the conditions under which it encounters that cue. Here, we present a simple stochastic dynamic programming model that explores the conditions under which larvae may maximize their lifetime fitness by accepting lower quality habitat rather than continuing to search for superior habitat. Our model predicts that there is a relatively narrow range of parameter values over which larval selectivity among habitat types changes dramatically from 1 (larvae accept only optimal substrata) to 0 (indiscriminant settlement). This narrow range coincides with our best estimate of parameter values gleaned from empirical studies, and the model output matches data for the polychaete worm Hydroides dianthus remarkably well. The relative availability of habitats and the total time available to search for high quality habitat (i.e. the ability to delay metamorphosis) had the greatest effects on larval selectivity. In contrast, intuitive factors, including larval energetics and mortality, showed little effect on larval habitat preference, but could still alter the proportion of larvae settling in different habitats by reducing search time. Our model predicts that a given larva may behave differently depending on where it falls in the optimality decision matrix at the instant in which it locates substrata. This model provides a conceptual framework in which to conduct future studies involving variability in settlement decisions among individual larvae, and in which to consider the selective forces driving the evolution of specific larval settlement cues. Our results suggest that a combination of the maximum search period and the relative frequency and quality of optimal habitat likely exert the greatest influence on the evolution of larval selectivity in the field.

KEY WORDS: Habitat choice · Larval settlement behavior · Desperate larva hypothesis · Stochastic dynamic programming model