Donovan, T.M. 1994. The demography of neotropical migrant birds in habitat sources and sinks. Ph. D. Dissertation at the University of Missouri-Columbia. 180 pp.
Many animal populations exist as geographically isolated subpopulations that are linked by dispersal. Understanding the conditions in which subpopulations persist or go extinct and the factors that influence dispersal among subpopulations is a central problem in ecology and conservation biology. I used a comparative approach to determine how habitat size and shape affects the demography of three forest nesting migrants in two regions in the United States. In both regions, I studied the distribution and reproductive success of Ovenbirds (Seiurus aurocapillus), Red eyed Vireos (Vireo olivaceus), and Wood Thrushes (Hylocichla mustelina) in fragmented and contiguous habitats. In both regions, distribution of individuals was not negatively altered by fragmentation, but total nest failure was significantly higher on fragments than contiguous forests. This increase in total nest failure and a reduction in the number of host fledglings on fragments was attributed to increased nest predation and increased brood parasitism by the Brown-headed Cowbird. The reproductive success of populations of the forest nesting passerines studied was directly related to spatial pattern of their breeding habitat. Populations in fragmented landscapes functioned as sinks and populations on contiguous landscapes functioned as sources.
I used simple population growth models to estimate viability in the absence of immigration among subpopulations, and modeled how fragmentation of contiguous habitat influenced population size and structure when individuals could disperse among subpopulations. Model results indicate that the reduction in density or extinction of some bird species on small habitats may be the result of two non-exclusive processes: 1) poor reproductive success in small habitats when immigration is negligible, or 2) fragmentation of contiguous habitat, which affects the number of immigrants available to "rescue" populations on fragments from extinction. Because migratory passerines are vagile organisms and dispersal among subpopulations is likely, I emphasize the need to identify and protect large, intact source habitats throughout a species' breeding range until the spatial scale at which subpopulations interact can be determined.