POPULATION DYNAMICS OF SPECIES RECOVERY: MULTISCALE DEMOGRAPHY IN RESTORED HABITATS

With habitat loss remaining one of the greatest global threats to biodiversity, habitat restoration has become an important tool for species recovery and conservation. Yet despite the conceptually-appealing lens of “if you build it, they will come” (i.e., the ‘Field of Dreams’ Hypothesis), restoration outcomes are highly variable and generally lack rigorous monitoring and evaluation. Species responses to habitat restoration can vary with a wide range of factors, including life history of the focal species, multi-scale habitat attributes and local or regional demography, which highlights the need to assess species response to habitat restoration through multiple ecological frameworks. This dissertation assessed behavioral, ecological, and demographic factors affecting restoration outcomes for a Nearctic-Neotropical migratory songbird, the Golden-winged Warbler (Vermivora chrysoptera). Over the past half-century, Golden-winged Warbler populations have declined, in part, due to the loss of early-successional breeding habitat. One strategy to address declines has been to restore breeding habitat according to established and evidence-based “best management practices”. Restoration through initiatives like the Natural Resource Conservation Service’s ‘Working Lands For Wildlife’ and ‘Regional Conservation Partnership Program’ have already created > 6,000 hectares of breeding habitat since 2012, with more hectares anticipated over the coming decade. However, the response of Golden-winged Warblers to restoration remains poorly understood. In this study, we evaluated the degree to which variation in restoration outcomes were explained by habitat and landscape features, local population trends, breeding productivity, and behavioral and ecological needs at contrasting life stages (e.g., nesting, post-fledging). From 2015-2018, we surveyed male Golden-winged Warblers and measured vegetation attributes in 672 restored habitats in the Great Lakes (n = 215) and Appalachian Mountains (n = 457) regions of the breeding range. In addition to estimating occupancy, we quantified full-season reproductive productivity and survival based on the survival of 341 nests and 258 fledglings in two regions in Pennsylvania along with previously published data. Occupancy data were analyzed in program R (package unmarked) using static and dynamic occupancy models including multi-scale habitat features as model covariates, whereas nest/fledgling survival data were analyzed in program MARK to assess how components of breeding productivity varied with habitat features within focal landscapes. Our results suggest that, although Golden-winged Warblers commonly used restored habitats, occupancy probability was related to micro-habitat attributes and landscape context. Warblers were most likely to occupy structurally complex sites that were eight years post-treatment and located in landscapes with no mixed coniferous-deciduous forest within 1 km. However, even after controlling for micro-habitat and landscape attributes, occupancy rates varied widely, demonstrating that even “if you build it…” they may not come. Local breeding output (# juveniles /pair /year), was positively associated with occupancy rates, but only between focal landscapes in the Appalachian Mountains where the species was rare and declining; occupancy in Great Lakes focal landscapes were uniformly high despite intermediate levels of breeding output. In addition to regional and landscape-level variability, we found evidence that restoration outcomes differed among life stages – a pattern consistent with a growing body of literature indicating that the needs of many forest bird species differ between nesting and post-fledging periods. For Golden-winged Warblers, nest success (the probability of fledging at least one young) did not vary with micro-habitat conditions in restored habitats, but survival rates varied across life stages in ways that scaled up to yield landscape-specific differences in productivity. For example, the threefold difference in breeding output between two Appalachian landscapes (i.e., Pocono Mountains and Pennsylvania Wilds) were driven by differential nestling/fledgling ( 11 days post-fledging) survival. The greater vulnerability of nestlings and young fledglings may stem partly from developmental processes related to shifting energetic requirements and foraging development. Fledgling energetic needs are likely intense given that they replace almost all body plumage during the post-fledging period via a rapid pre-formative molt, most pronounced from 13-17 days post-fledging. Difficulty in meeting nutritional demands might also be reflected by the fact that Golden-winged Warblers began foraging almost immediately after fledging and rapidly specialized on probe-and-gape foraging (> 7 days post-fledging). Although adult Golden-winged Warblers provided extensive parental care over the post-fledging period, parental feeding may not offset the physiological and behavioral challenges faced by young fledglings during this dangerous life stage. Overall, this research provides grounds to reject the Field of Dreams Hypothesis and thereby highlights the importance of considering multi-scale habitat and demographic factors that drive restoration outcomes. Our results also emphasize the need to anticipate how stage-specific survival and life history constraints, like those we documented during the post-fledging period, may shape population-level responses to habitat restoration. This is especially important given that species conservation plans may disregard understudied life stages, including the post-fledging period. Ultimately, our analyses provide one of the most comprehensive assessments of breeding habitat restoration for an imperiled migratory songbird, while also offering new insights into the breeding biology of Golden-winged Warblers and other passerines