2025 Theses Doctoral

The impacts of within-species trait variation on community ecology: Theoretical modeling and empirical studies across Puerto Rico

Arroyo, Eva E.

Across the environmental conditions they occur, species display considerable variation in their phenotypes, intraspecific variation. In this dissertation, I explore how intraspecific trait variation (ITV) shapes species' ecological responses to environmental heterogeneity. Across three chapters, I examine the relationships between ITV, environmental gradients, and species coexistence to ask when and why variation within species matters.

In Chapter 1, we used trait and occurrence data for 33 tree species across eight forest sites in Puerto Rico to test how ITV is structured by environmental variability. Specifically, we asked (1) If within-site ITV is positively correlated with site-level temporal and spatial environmental variation, and (2) How across-site ITV relates to the breadth of species' environmental distributions on the island. Two key plant traits, leaf mass per area (LMA) and wood density, were used to understand the association between ITV and environmental variation. We examined ITV across species and sites using statistical models to assess the relationship between (1) within-site ITV and site-level spatial (topography) and temporal (rainfall) environmental variability; (2) across-site ITV and environmental variation across the species geographic range on the island. We also assessed the relationship between across-site mean trait values and island-wide environmental ranges. (1) Across all species, we did not find any significant associations between within-site ITV and site-level temporal or spatial variability. (2) Greater ITV values of LMA and wood density were associated with larger environmental ranges. We also found that species with high wood density, a trait associated with a conservative growth strategy, had narrower ranges across climatic conditions, but this pattern was not evident for LMA. Our findings emphasize the complexity of the relationships between ITV and species distributions with respect to environmental heterogeneity at different spatial and temporal scales.

Chapter 2 asks when plasticity evolves and how plastic and non-plastic types can coexist. Using a mathematical model of species distributed across two- patch landscapes, we investigate how environmental stress and landscape heterogeneity shape the evolution of plasticity. Consistent with prior theory, our results show that plasticity is favored in heterogeneous, low-stress environments. However, we also find that environmental stress gradients can select for non-plastic strategies, and unexpectedly, expand the potential for coexistence between plastic and non-plastic species.

In Chapter 3, we use wood density, growth, and spatially explicit location data for trees in 11 sites across Puerto Rico, to evaluate whether incorporating ITV strengthens signatures of environmental filtering and competition and improves trait-based demographic models. Incorporating site-specific trait variation significantly altered static trait distributions, increasing kurtosis while reducing quadrat-level trait ranges. These patterns suggest that incorporating ITV detects stronger environmental filtering but weaker niche differentiation. Using hierarchical Bayesian models applied to tree growth data, we compared niche-based and hierarchical trait formulations of competition. Heterospecific competition was best captured by a niche-differentiation framework, whereas conspecific interactions aligned more closely with a hierarchical trait structure; however, trait-mediated competitive effects were weak overall. Incorporating ITV into this Bayesian growth model did not improve model fit nor strengthen trait–competition relationships.

In this dissertation, I argue that intraspecific trait variation matters, but not always in the ways we expect. ITV does not simply track environmental heterogeneity and therefore does not uniformly predict broader ranges. The effect of ITV in community ecology, and coexistence, is trait-specific and often mediated by broader ecological dynamics, especially resource stress. This work offers a framework for understanding how variation within species contributes to range size, demographic performance, and coexistence.