Abstracts Biological Sciences

SCALE-DEPENDENT MECHANISMS OF HABITAT SELECTION AND

by Kerri Lyn Cornell

The process of habitat selection involves discriminating among alternative habitats that differ in quality in terms of survival and reproduction. Thus, although habitat selection has important ecological and evolutionary consequences for individual organisms and species, the mechanisms that drive habitat selection decisions remain poorly understood. Further, human alteration of habitats that are critical for survival and reproduction currently poses a significant threat to wildlife populations worldwide and may interfere with habitat selection decisions. Understanding the factors that drive habitat selection and the demographic consequences of those decisions is important for understanding population dynamics and can provide critical information about species habitat requirements to conservation planners. I studied the process of habitat selection and fitness consequences for a Neararctic-Neotrpical migratory forest songbird, the black-throated blue warbler (Dendroica caerulescens) in a heterogeneous landscape in west-central Vermont. My research had three parts. First, I investigated the relative effects of habitat features at three spatial extents on four different demographic parameters including abundance, age ratio, pairing success, and annual fecundity. I used a model selection analysis framework to determine which habitat levels were most important for each demographic parameter. I found that that the distribution of warblers across the landscape in Vermont matches patterns described for this species in intact landscapes, suggesting that warblers use specific proximate cues for territory selection. However, reproductive success was negatively affected by the degree of fragmentation. These results suggest that proximate habitat cues used for territory selection may be decoupled from realized fitness in this system. Second, I conducted a conspecific playback experiment to evaluate whether conspecific attraction is important for determining abundance and occupancy patterns in high and low quality habitats by playing warbler vocalizations in previously unoccupied habitats in Vermont, USA. I used multi-season occupancy models in Program MARK to identify whether territory-level shrub density, landscape-level habitat patterns, or the attraction by conspecifics were most important for predicting territory occupancy and abundance. I found that habitat features were more important for determining both abundance and occupancy than conspecifics. Finally, I assessed relationships between multiple demographic variables to determine whether warbler abundance can be used as a surrogate indicator of habitat quality. I found that warbler abundance is significantly positively related to both reproductive success and survival for this species in Vermont indicating that count data may be sufficient for long-term population monitoring for black-throated blue warblers.