Linking Isotopes and Panmixia: High Within-Colony Variation in Feather δ2H, δ13C, and δ15N across the Range of the American White Pelican

• Published in: PloS one 2016-03, Vol.11 (3), p.e0150810-e0150810
• Main Authors: Reudink, Matthew W, Kyle, Christopher J, McKellar, Ann E, Somers, Christopher M, Reudink, Robyn L F, Kyser, T Kurt, Franks, Samantha E, Nocera, Joseph J
• Format: Article
• Language: English
• Subjects: Algorithms; Animal behavior; Animal feathers; Animals; Biogeochemistry; Biology; Biology and Life Sciences; Birds - physiology; Breeding; Breeding sites; Clustering; Clusters; Colonies; Dispersal; Dispersion; Ecosystem; Forensic sciences; Forestry; Gene flow; Gene Flow - physiology; Isotopes; Isotopes - metabolism; Multivariate analysis; Natural resources; North America; Philopatry; Physical Sciences; Powell, Jerome; Range extension; Species
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0150810

Complete panmixia across the entire range of a species is a relatively rare phenomenon; however, this pattern may be found in species that have limited philopatry and frequent dispersal. American white pelicans (Pelecanus erythrorhyncos) provide a unique opportunity to examine the role of long-distance dispersal in facilitating gene flow in a species recently reported as panmictic across its broad breeding range. This species is also undergoing a range expansion, with new colonies arising hundreds of kilometers outside previous range boundaries. In this study, we use a multiple stable isotope (δ2H, δ13C, δ15N) approach to examine feather isotopic structuring at 19 pelican colonies across North America, with the goal of establishing an isotopic basemap that could be used for assigning individuals at newly established breeding sites to source colonies. Within-colony isotopic variation was extremely high, exceeding 100‰ in δ2H within some colonies (with relatively high variation also observed for δ13C and δ15N). The high degree of within-site variation greatly limited the utility of assignment-based approaches (42% cross-validation success rate; range: 0-90% success). Furthermore, clustering algorithms identified four likely isotopic clusters; however, those clusters were generally unrelated to geographic location. Taken together, the high degree of within-site isotopic variation and lack of geographically-defined isotopic clusters preclude the establishment of an isotopic basemap for American white pelicans, but may indicate that a high incidence of long-distance dispersal is facilitating gene flow, leading to genetic panmixia.