Using a multi-isotope approach to understand waterfowl movement in southern Africa. (20th November 2019)

• Record Type: Journal Article
• Authors: Mutumi, Gregory L; Cumming, Graeme S; Sullivan, S Mažeika P; Caron, Alexandre; Cáceres, Carlos
• Is Part Of: Condor. Volume 121:Number 4(2019)
• Publication Date: 2019-11-20
• Subjects: linear discriminant function analysis -- molt location -- southern Africa -- stable isotopes -- waterfowl movement -- wetlands; analyse discriminante linéaire -- déplacement de la sauvagine -- isotopes stables -- milieux humides -- site de mue -- sud de l'Afrique; Birds -- Periodicals; Birds -- California -- Periodicals; Ornithology -- Periodicals; Oiseaux -- Périodiques; Oiseaux -- Pacifique, Côte du -- Périodiques; Birds; Ornithology; California; Periodicals; 598
• DOI: 10.1093/condor/duz049
• Languages: English
• ISSNs: 0010-5422

Abstract

Abstract: Many far-ranging species depend heavily on relatively small or temporary resources within a heterogeneous landscape. For waterfowl, most species rely on deep, permanent waterbodies as refugia from predators during annual flightless molt periods when synchronous loss and regrowth of the flight feathers occurs. The movements of ducks to and from molt sites are, however, poorly documented for most Afrotropical species and the dependencies of Afrotropical ducks on key sites are unclear, yet this information is integral to conservation and management efforts. We asked whether stable isotopes of wing feathers could be used to determine the molting origins of Afrotropical ducks in southern Africa. We analyzed isotope ratios of carbon, nitrogen, oxygen, and hydrogen in feathers from 4 different species across 5 different sites (wetlands, ponds, lakes) in South Africa, Zimbabwe, Mozambique, and Botswana. We observed differences among sites for all isotopes ( P < 0.05), especially δ 13 C and δ 15 N. Based on these differences, we conducted linear discriminant function analysis (LDA) to assess the utility of these isotopes to assign birds to molt locations. We obtained a global classification accuracy = 0.59, although accuracies differed among sites. Our results demonstrate the potential of a multi-isotope approach to discriminate among specific molt locations and to provide an initial estimate of molt site. Rigorous documentation of molt site from wing feathers is plausible, but will require large sample sizes, extensive spatial coverage, and careful calibration.