Use of stable isotope fingerprints to assign wintering origin and trace shorebird movements along the East Atlantic Flyway

Migratory connectivity can be defined as the flux of individuals or populations among areas between stages of an animal's life cycle. Many shorebird species perform long-distance migrations and while moving between breeding and wintering grounds, they depend on a network of intermediate wetlands (stopover sites) where populations of different origins extensively overlap. The difficulty to discriminate such populations represents a serious obstacle to the identification of the links between breeding or wintering areas and stopover sites, and also precludes the estimation of demographic parameters for each population. In this study, we test if linear discriminant models based on stable carbon and nitrogen isotope ratios in toenails can he used to identify populations of several shorebird species of different wintering origins overlapping at two stopover sites of the East Atlantic Flyway. In addition, we evaluate the ability of this approach to infer migratory phenological patterns of shorebirds. Linear discriminant analyses performed overall well in distinguishing the isotopic signals of birds from wintering areas (in France, Portugal, Morocco, Mauritania and Guinea-Bissau) most species, correctly classifying over 80% (n = 542) of all wintering individuals sampled at these areas. Assignment rates of shorebirds captured during spring migration were also high (96%, n=323) at the Tejo estuary, Portugal, but lower (40%, 0 = 185) at Marennes-Oleron Bay in France, and also differed among species. A large proportion of spring migrants captured in Portugal and France were assigned to Banc d' Arguin in Mauritania, the most important wintering area in the flyway. Phenological patterns derived for dunlins (Calidris alpina), common ringed plovers (Charadrius hiaticula) and grey plovers (Pluvialis squatarola) suggest that the first northward migrants started arriving at the Tejo estuary during the second half of March, with peaking numbers occurring during April or May