Skip to main content
. 2020 Sep 1;11:4048. doi: 10.1038/s41467-020-17893-z

Fig. 2. Model of mastodon extirpation and expansion in response to glacial cycles.

Fig. 2

a Global stack of benthic foraminifera δ18O for the last 1 million years, which tracks changes in deep-water temperature and global ice volume. The y-axis has been inverted so that periods of low ice buildup (and higher temperatures—red) are at the top of the graph, and periods of greater ice buildup (and lower temperatures—blue) are at the bottom. Marine Isotope Stage (MIS) extents are indicated with black bars above (interglacials) or below (glacials) the δ18O record. δ18O values and MIS terminations can be found in Lisiecki and Raymo7. One full glacial cycle is represented, showing the change from glacial (b) to interglacial (c) conditions, followed by a fall back into another glaciation (d). North American ice-sheet cover at each stage (c, d) is approximated from recorded δ18O to similar conditions during the transition out of the last glaciation8, or from published simulations where available (b)65. The ecological implications of these transitions are summarised in (e, f), with mastodons being able to occupy most of eastern Beringia and Canada during interglacials (e), but progressively extirpated from these regions as conditions descend into the next glacial period (f). Populations would either need to retract to unglaciated regions south of the ice sheets or north to temporarily unglaciated refugia which would be unlikely to support mastodon populations throughout long glaciations.