Natal dispersal in vagile species such as songbirds can shape a population's range and structure. Although effective conservation practices depend on knowledge of the scale and frequency of natal dispersal, these issues remain poorly understood because of methodological gaps. In this exploratory study, we assessed whether element signatures within natal feathers might be used to identify the geographic birth site of first-year breeders. We used two related techniques, inductively coupled plasma mass spectrometry (ICP-MS) and inductively coupled plasma with optical emission spectrometry (ICP-AES), to quantify element levels in natal feather samples from 7 species at 27 sites across the eastern United States. The techniques differed in the manner in which elements were quantified and in their detection limits. Our goal was to determine whether element analyses of feathers could discriminate (1) different species within a site and (2) different sites within a species. Additionally, because spatial autocorrelation of element levels is needed for element analysis to be an effective tool in assessing natal dispersal, we also evaluated the spatial autocorrelation of ICP-AES samples at 18 sites across the eastern United States.
Both ICP-MS and ICP-AES analyses separated species within a site with fairly high accuracy, though the discriminating elements varied with site. However, within a species, natal feather locations were not identified with high accuracy on the basis of feather elements. We were not able to determine whether there is spatial correlation among individual elements or a principal component analysis (PCA) score that described the elemental makeup of a feather. A kriging model was fit to the semivariogram of PCA scores to produce a base-map of element signatures across the eastern United States. This map was ineffective at predicting feather-element values at sample sites. Whether elemental analyses can identify natal dispersal distances requires further study. We suggest that future studies evaluate elements with ICP-MS methodologies on a single, box-nesting species that is sampled more intensively at smaller geographic scale, or on species that occur in very discrete populations. Additionally, this methodology should be evaluated in concert with stable-isotope analyses of feathers and, potentially, genetic analyses.
Suivre la Dispersion chez les Oiseaux: Évaluer le Potentiel de Marqueurs Élémentaires