I. Gamete allocation and fertilization success in coral reef fish. II. Using developmental instability to address questions in community and behavioral ecology

The four chapters that constitute this dissertation address somewhat disparate topics. In Chapter 1 I develop a series of optimization models in an attempt to identify the selective regime behind the evolution of ejaculate size in pair spawning coral reef fish. Using published data I show that: (i) fertilization is egg-density dependent; (ii) selection operates to maximize long-term average fertilization rate (FR); (iii) available data is insufficient to unequivocally infere the limits on male reproductive success.;In Chapter 2 1 test a hypothetical fertilization advantage to mass spawning and, thereby, to mating aggregations in the brown surgeonfish (Acanthurus nigrofuscus), an externally fertilizing coral reef fish. I show that mass spawning make no significant contribution to the already high FRs attained by small-group spawning (mean: 98.5%). I further demonstrate that FRs saturate well within the average lifespan of sperm, with over 60% of spawned eggs fertilized within 5 s of gamete release.;In Chapter 3 1 use floral asymmetry as a measure of developmental instability (DI), to investigate within-population variation in sex allocation by the andromonoecious plant Euphorbia xanti. Contrary to theoretical predictions, I show that DI is highest at intermediate sexual expression. I also show that fruiting success is lowest at intermediate sexual expression, providing a potential explanation of the former result.;In the final chapter, I evaluate a corollary of a hypothesis advanced to explain the often observed decline in abundance with increasing distance from the center of the species' geographic range. I predict that spatial patterns in population abundance should be mirrored by the among-population variation in DI. I test this prediction using data on 38 contemporary and Pleistocene mollusc populations, collected from a single site in the Gulf of California. Levels of DI increased with decreasing relative abundance and increasing relative distance from the center of the range, and defined the upper limit to triangular bivariate distribution patterns. Overall, however, the observed pattern deviates from the expected as it shows the majority of low-abundance peripheral populations to exhibit relatively low levels of DI.