GSA Annual Meeting in Seattle, Washington, USA - 2017

Paper No. 10-14
Presentation Time: 11:45 AM

MECHANISTIC MODELING OF ENVIRONMENTAL DRIVERS OF WOOLLY MAMMOTH CARRYING CAPACITY AND EXTINCTION ON ST. PAUL ISLAND, AK


WANG, Yue1, PORTER, Warren2, MATHEWSON, Paul2, MILLER, Paul3, GRAHAM, Russell W.4 and WILLIAMS, John W.5, (1)Department of Geography, University of Wisconsin, 550 North Park St, Madison, WI 53706, (2)Department of Zoology, University of Wisconsin, Madison, WI 53706, (3)Physical Geography and Ecosystem Science, Lund University, Lund, SE-221 00, Sweden, (4)Geosciences, The Pennsylvania State University, 116 Deike, University Park, PA 16802, (5)Department of Geography, University of Wisconsin-Madison, 550 N Park St, Madison, WI 53706, jww@geography.wisc.edu

The Holocene population of woolly mammoths (Mammuthus primigenius) on St. Paul Island, AK, offers a natural system for studying the environmental drivers of species persistence and extirpation. St. Paul, part of the Pribilof Islands, was isolated from the Bering Land Bridge between 14,700 and 13,500 years ago as sea levels rose with the last deglaciation. There is no evidence for human presence on St. Paul until the late 18th century. Recently, Graham et al. (2016) placed the timing of extinction timing for St. Paul mammoths to 5,600 +/-100 years ago, based on five separate indicators. Graham et al. also presented evidence from limnological paleoclimatic proxies that the mammoth disappearance coincided with declining availability of freshwater and hypothesized a causal relationship. Here, we assess the carrying capacity of St. Paul Island for woolly mammoths and assess the sensitivity of mammoths to changes in climate, forage availability, and freshwater availability, using the mechanistic models Niche Mapper and LPJ-GUESS. Niche Mapper is a first-principles bioenergetics model for animals that simulates an animal's energetic balance given its ambient environment and morphological, physiological, and behavioral traits. LPJ-GUESS is a dynamic vegetation model used here to simulate net primary productivity on St. Paul Island during the changing climates of the last deglaciation. Paleoclimatic driver variables are from alkenone-based reconstructions from the Gulf of Alaska and the CCSM3 SynTrace experiments. Sensitivity analyses indicate that temperature is the most important predictive climatic variable, while fur length and density are important controls on mammoth energetic balance. Mammoths could have fasted for only 2-3 months, indicating a need and ability to access snow-buried forage. The carrying capacity of St. Paul may have been several hundred individuals at the time of isolation from the mainland. Perhaps only 2-3 individuals could have been sustained by the freshwater surplus in crater lakes, making this population highly dependent on coastal freshwater sources. The simulations are consistent with the available proxy data and reinforce the vulnerability of island megaherbivore populations to freshwater limitation and extinction.