Intrinsic Markers In Avian Populations: Explorations In Stable Isotopes, Contaminants, And Genetics

Abstract

This research outlines the diversity of questions that intrinsic markers have the potential to answer and demonstrates some of these marker's limitations and successes. To test the working hypothesis that feathers grown on different continents have significantly different stable isotope ratios in commonly used markers, I analyzed stable isotopes in two generations of feathers from three species of birds that breed at high latitudes and winter on different continents. As expected, significant differences in stable isotope ratios were detected between summer- and winter-grown feathers in both plover species (Pluvialis fulva and P. domininca). However, no differences were found between the two groups of winter-grown plover feathers, despite being grown on different continents. Similarly, no differences were detected in isotope values between summer- and winter-grown feathers in northern wheatears (Oenanthe oenanthe). Large variances in isotope ratios limited the percentage of feathers correctly assigned to their origins to 41%. Atmospheric transport has been identified as the source of pollutants in several arctic ecosystems and has the potential to severely impact high-latitude populations. To determine whether long-range atmospheric transport, point sources, or migratory prey were sources of contaminants in the North Pacific, birds from two trophic levels were sampled across the longitudinal transect of the Aleutian Archipelago. Carbon isotope ratios differed among islands, thereby linking birds to island food webs and ruling out contaminant transfer through migratory prey. Patterns in some PCB congeners indicated local point sources, but significant west-to-east declines in contaminant concentrations for the majority of detected organochlorines provided evidence of long-range transport. Linking individuals to source populations using intrinsic markers has only been successful at broad scales. To determine whether increased resolution among populations could be achieved by merging multiple intrinsic marker classes, a new analytical procedure was developed. Discrete and continuous markers were combined to evaluate a Bayesian method of assignment across marker classes. For three datasets, two real and one simulated, the percentage of individuals assigned to correct source populations increased with the addition of markers and marker classes. In all cases, the maximum number of individuals was correctly assigned when all marker classes were combined.