Migratory connectivity and wintering habitat structure of loggerhead shrikes : inferences from stable hydrogen isotope and microsatellite DNA analyses

Abstract

The linking of breeding and nonbreeding grounds of migratory birds is of great conservation and theoretical importance. In theory, connecting these geographically disparate areas allows for a more complete understanding of annual events, and a first step into identifying where in the annual cycle limitations to fitness may be occurring. The Loggerhead Shrike (Lanius ludovicianus) is a Species at Risk in Canada, and its decline is attributed to habitat loss on both the breeding and wintering grounds. In the fall, Loggerhead Shrikes from breeding areas of prairie Canada (L. l. excubitorides), the focus of this study, move south along the Gulf States of the United States and Mexico to wintering areas that are already occupied year-round by resident shrikes. However, the habitat structure and variability for both migrants and resident Loggerhead Shrikes have not been studied well on the wintering grounds. Thus, it has been difficult to adequately evaluate the relative contribution of changes in wintering areas to the decline of Loggerhead Shrike populations. To identify the wintering grounds of Loggerhead Shrikes that breed in prairie Canada, I used stable hydrogen isotope (&#948;D) analysis of feathers and claws and microsatellite DNA from feathers. In North America, &#948;D measurements from feathers provide information on latitude of origin, while DNA can be related to an affiliation to a breeding population. Since several authors suggested that Loggerhead Shrike completed their preformative and formative molt on the breeding grounds prior to fall migrations, I reasoned that &#948;D analysis of tertial feathers sampled on the wintering grounds would correspond to &#948;D in precipitation from the breeding grounds. Similarly, I used DNA microsatellite markers and Bayesian clustering analysis to detect patterns of population genetic structure within the range of Western Loggerhead Shrikes (L. l. excubitorides) in Canada and consequently use these to infer breeding origin of shrikes sampled in Mexico. I based the categorization of resident and migrant Loggerhead Shrikes sampled in Mexico on &#948;D analysis of feathers. Based on evaluation of observed and expected &#948;D values of feathers, I determined that migrant shrikes used northeastern (63.8%) and south-central (73.7%) Mexico to winter. Microsatellite DNA and assignment tests, suggested that wintering migrant shrikes occupied north-central (18.6%) and northeastern (20.3%) Mexico. Differential habitat occupancy analyses, suggested that, in northeastern Mexico, wintering sites occupied by Loggerhead Shrike sites were structurally different from random unoccupied sites (MRPP, T = -8.04, P < 0.001, n = 354). An important difference was that, on average, occupied habitat contained shorter tall shrubs and huisache and fewer tall shrubs, mesquite and huisache. Similarly, residents shrikes occupied structurally different habitats (MRPP, T = -2.95, P = 0.01, n = 146) that had less percent cover of bare ground than those sites occupied by migrants. Based on these habitat results, I surmise that habitat availability may be a limiting factor for both resident and migratory shrike populations in northeastern Mexico.