UC Santa Barbara

Community ecology seeks to understand the assembly processes that structure the abundance, richness, and prevalence of species in the community. However, some major difficulties are that the effect of an assembly process can change between scales and that natural scales can be difficult to identify. The first two chapters of this dissertation aimed to understand the role of assembly processes across biologically defined scales. All empirical data used in this dissertation were previously collected on Palmyra Atoll (McLaughlin et al., 2023; McLaughlin, 2018), a national wildlife refuge located within the Remote Pacific Islands Marine National Monument. Chapter 1 examined the scale-dependent factors that influenced free-living arthropod species abundances across terrestrial spatial scales. We found evidence that of the island- or forest- specific covariates included in the model, only soil cation exchange capacity had population-level effects on arthropod abundances. However, we found species-specific and higher-taxon level (“order”) responses to island size, nutrient input, and canopy type. We also explored how species residual associations changed with scale as a possible indicator of biotic interactions with narrowing scale. Chapter 2 focused on the effect of host traits on parasite component and infracommunities in marine sandflat fish species. Parasite species occurrence probabilities varied between and within host species, with parasite occurrences responding to host species generality, density, and host individual weight. Another major difficulty in ecological studies is sampling error when observing species communities. Few species distribution studies account for imperfect detection, but assuming that observations perfectly capture a community can lead to biased results and inferences. In the final chapter, we assessed a method to predict true communities based on estimated false-negative probabilities in 1000 simulated datasets. Based on this analysis, predicted communities were more accurate and shared more mutual information with the true community than the observed community. We then tested these methods in a case study using empirical data from Palmyra Atoll’s Hymenoptera community.