The peninsula effect is a pattern of diversity wherein species richness decreases along a peninsula from base to tip and is attributed to three mechanisms: historical processes, habitat gradients, and immigration-extinction equilibrium. Numerous studies have reported conflicting results involving the existence, cause, and validity of the peninsula effect in part because they did not account for effects of history or habitat on species richness patterns and because most previous research focused on organisms that actively disperse, which could confound results with behavioral habitat selection. Florida poses an excellent opportunity to study the peninsula effect because of its geological history and its unique ridges have similar histories (e.g. age, elevation, and sediment). Habitat changes down the peninsula, from a warm temperate climate in the north to a subtropical climate in the south. I studied freshwater crustaceans in isolated wetlands because crustaceans are diverse and disperse passively among these discrete habitats. My study design and statistical analyses controlled for two of the three mechanisms (habitat and history) that may generate a peninsula effect to better test for the third hypothesis (immigration-extinction equilibrium) on the Florida peninsula. Thirty-one wetlands were sampled for crustaceans monthly from November 2004 through April 2005, or until a site dried. Human disturbance was minimized by choosing isolated, ephemeral wetlands located within state reserves, parks, and forests located on four major ridges: Trail, Brooksville, Mount Dora and Lake Wales. I measured several environmental variables to assess habitat variation among sites. Limnological parameters included temperature, pH, dissolved oxygen, conductivity, chlorophyll á, pheophytin, total nitrogen, total phosphorus, and total hardness. Other habitat variables included surface area, distance to nearest water body, fish presence or absence, hydroperiod, total transmitted light and canopy openness. Crustacean species were identified to the lowest practical taxonomic level (typically species) and recorded as present or absent. A total of 53 different crustaceans were identified, including 41 cladocerans, 10 copepods, and 2 ostracods. In a multiple regression, environmental variables and sampling effort accounted for 57% of the variation in species richness. Regression of remaining variation (residuals) against latitude, which measures position along the peninsula, was not statistically significant. The same pattern was obtained when the sequence of regressions was reversed. Therefore, the peninsula effect does affect the species richness of freshwater crustaceans inhabiting ephemeral wetlands on Florida's ridges. Instead, variation in species richness was determined mainly by habitat differences, particularly the complex interaction of phosphorus levels, isolation, fish presence or absence, and hydroperiod. This study may serve as a model for more thorough analyses of mechanisms (history, habitat, and immigration-extinction) of a peninsula effect in other taxa.

The red-headed woodpecker is a primary cavity excavator found throughout the Eastern Deciduous Forest and parts of the Great Plains. Although widespread, the red-headed woodpecker is generally considered uncommon, with patchily distributed populations. Over the past 50 years, this species has experienced precipitous, range-wide declines. The red-headed woodpecker uses a variety of cover types to meet resource needs across the annual cycle, ranging from oak savannas and wetlands, to mature beech forests and urban parks. Given their apparent adaptability to such a wide range of habitats, causes of declines are perplexing. To understand and stem declines, recent studies have focused on quantifying this species’ habitat requirements. Most of these studies have focused on a single cover type, often open forests, during the breeding season. However, effective management requires comprehensive knowledge of red-headed woodpecker habitat needs in a variety of cover types across the annual cycle. My thesis seeks to address this knowledge gap. I studied red-headed woodpecker habitat selection during both the breeding and non-breeding seasons at Fort A.P. Hill (FAPH) in Caroline County, Virginia. FAPH contains a variety of cover types that red-headed woodpeckers are known to use for breeding and overwintering, including: wetlands, open forests and closed forests. In Chapter 1, I investigated winter roost-site selection of red-headed woodpeckers in a burned forest stand. My study was the first to quantify winter habitat selection for this species. My aim was to identify variables driving winter roost-site selection at two relevant spatial scales, and estimate their parameter weights using logistic regression. I found that red-headed woodpeckers preferred habitat around the roost tree that contained a higher basal area of snags and mast-producing trees. They differentially selected roost trees based on taxon and decay state. My results indicate that managers should adopt practices that promote snag generation and retention, and mature oak recruitment. In chapter 2, I focused on red-headed woodpecker breeding season habitat requirements, as many other studies have done. However, unlike other studies that investigated habitat selection within a single cover type, I found and characterized nest-sites in three distinct cover types (wetlands, closed forests, open forests). I used a comparative approach to identify cover-type specific nest-habitat thresholds at the landscape, patch and tree scales using boosted regression trees (BRT). Although models at the landscape scale inadequately discriminated between nest and available sites, models at the patch and tree scales achieved excellent discrimination ability. I found that red-headed woodpeckers are consistent in their preference for a number of habitat features at the patch (high medium/large snag density, open canopy) and tree (large diameter tree with less bark) scales, irrespective of cover type context. However, I also found cover-type specific habitat preferences at the patch scale, indicating red-headed woodpeckers are flexible in their selection of features surrounding the nest and responsive to the broader habitat context. My findings suggest that there are a number of habitat features that facilitate breeding for red-headed woodpeckers generally, and management that promotes these features will be effective in a variety of habitat contexts. They also indicate that management can be tailored to provide the most suitable habitat for each of these three commonly used cover types. Together, Chapters 1 and Chapter 2 expand our knowledge of this declining keystone species’ habitat needs across the full annual cycle. But there is still much to know. The choices of habitat selection have consequences. Determining the conditions under which habitat features function to either help or harm populations is a necessary next step. While not a part of this thesis, I am currently investigating the factors that drive nest success among cover types, linking breeding-site selection to population dynamics. I also have behavioral data that will help reveal the mechanisms that either facilitate or constrain the exploitation of food resources across cover types. Finally, the red-headed woodpecker is a facultative migrant that often shifts habitat associations to take advantage of seasonally available resources. Data from my point count surveys—conducted over two successive winter and summer seasons—allow for development of dynamic occupancy models. Modeling shifts in occupancy across seasons will show the habitat factors underlying seasonal shifts. Site-specific differences in colonization, extinction and frequency of occupancy, within seasons but across years, will further our understanding of what constitutes habitat quality for this species, in both the breeding and non-breeding seasons.