CARRY-OVER EFFECTS FROM THE NON-BREEEDING SEASON INFLUENCE SPRING ARRIVAL DATES, REPRODUCTIVE SUCCESS, AND SURVIVAL IN AN ENDANGERED MIGRATORY BIRD, THE KIRTLAND'S WARBLER (SETOPHAGA KIRTLANDII)

Abstract

I used a unique model system to evaluate potential sources of population limitation in migratory birds that travel seasonally between temperate breeding grounds and tropical non-breeding grounds. The Kirtland's warbler (<italic>Setophaga kirtlandii</italic>) is an endangered species that is well-studied during its breeding season in northern Michigan, but how winter events may carry over to affect demographic processes remains unknown. Stable-isotope analyses of birds' tissues collected upon spring arrival revealed inconsistent yearly effects of &#948;<super>13</super>C, &#948;<super>15</super>N, and &#948;D signatures on arrival schedules. Males departing from more mesic winter habitats (depleted &#948;<super>13</super>C), consuming more insects (enriched &#948;<super>15</super>N), and wintering further south (enriched &#948;D) arrived earlier to temperate breeding grounds. However, these patterns were not significant across all years. Differences in &#948;<super>13</super>C and &#948;<super>15</super>N signatures between age classes were suggestive of age-related dominance relationships in winter. Winter habitat and diet differences among males within years did not seem to have strong carry-over effects, but mean spring arrival dates were delayed following drier winters, suggesting that carry-over effects in this species may be driven more by interannual variation in rainfall. There was a strong male age by rainfall interaction, with first-time breeders responding more strongly to changes in rainfall than experienced adults. Regardless of age, drier winters and delayed arrival and nest initiation were significantly associated with fewer offspring fledged in the subsequent breeding season. Analysis of our 5-year mark-recapture dataset showed that apparent survival is also positively associated with winter rainfall in the Bahamas. Kirtland's warbler survival probabilities are relatively high within the summer and winter stationary periods, but are lower during migration. Combining my survival and productivity data into a simple projection model revealed that the population growth rate of this species may become negative if Bahamas March rainfall drops more than 18% from its current mean. Climate change models predict continued drying trends in the Caribbean, which could have negative consequences on the population dynamics of the Kirtland's warbler. These results emphasize the importance of understanding the effects of climatic variation on demographic rates, and underscore the need for continued research on the ecology of migratory animals throughout the annual cycle.