HORSE OVER BISON: DIETARY PREFERENCES OF CARNIVORES IN PLEISTOCENE YUKON

How do predator-prey dynamics respond to significant ecological perturbations? Across Pleistocene climate cycles, mammal populations experienced dramatic demographic fluctuations, sometimes leading to shifts in community structure. To examine how such changes impacted predator-prey dynamics, we examined fossils recovered from permafrost deposits in the Klondike goldfields region of Yukon Territory, Canada. We used the presence or absence of carnivore modifications on fossils to evaluate changes in prey utilization between >50 kya BP and ~11 kya BP. Specifically, we tested whether the carnivore guild, dominated by Canis lupus, Panthera spelaea, and Arctodus simus, was biased towards specific prey species, or targeted prey as a function of their relative abundances. We focused on the region’s two dominant ungulates: steppe bison (Bison priscus) and caballine horses (Equus sp.). We selected mandibles (n_bison = 84, n_horse = 62) to taphonomically standardized our data. To estimate relative population sizes of bison and horse through time, we synthesized AMS radiocarbon (¹⁴C) dated specimens using KDE_model (OxCal). Blinded to ¹⁴C results, we visually evaluated fossils for punctures, scores, and furrows consistent with carnivore modification. Tooth marks cannot differentiate predation and scavenging behaviors, though they provide useful metrics of prey utilization. We found that both bison and horses reached population maxima prior to and following the Last Glacial Maximum. However, the timing of peak abundances was offset, establishing two shifts in dominance. Horse fossils were modified by carnivores across their timeseries and regardless of bison population size. Bison fossils were modified only during periods of low horse abundance (prior to ~40 kyr BP and during the Terminal Pleistocene). Furthermore, although horse fossils showed similar rates of carnivore modification for sub-adults (~33% ) and adults (~25%), carnivore modification of bison fossils was biased towards subadults (nearly 50% of subadults vs. < 10% of adults; p < 0.01). While identification of bone-modifying agent(s) requires further study, our results indicate a strong preference for horse among late Pleistocene Yukon carnivores, with bison (subadult) utilized when horse abundances were below a threshold of accessibility.