Dire wolves are considered one of the most common and widespread large carnivores in Pleistocene America, yet relatively little is known about their evolution or extinction. To reconstruct the evolutionary history of dire wolves, we sequenced five genomes from sub-fossil bones dating from 13,000 to over 50,000 years ago. Our results indicate that though they were similar morphologically to the extant gray wolf, dire wolves were a highly divergent lineage that split from living canids ~5.7 million years ago. In contrast to numerous examples of hybridization across Canidae, there is no evidence for gene flow between dire wolves and either North American gray wolves or coyotes. This suggests that dire wolves evolved in isolation from the Pleistocene ancestors of these species. Our results also support an early New World origin of dire wolves, while the ancestors of gray wolves, coyotes, and dholes evolved in Eurasia and only colonized North America relatively recently.

2D Landmarks recorded with tpsDig from photographs taken using a Nikon DSLR with 60mm fixed focal length lens. Dental data was digitised by Ardern Hulme-Beaman; mandibular data was digitised by Carly Ameen. Each csv is in a format with specimens by row and data by column. The first column includes specimen IDs, the second column is the species and the third column is the centroid size; the remaining columns are the x and y coordinates of the individual landmarks for each specimens in the format x1, y1, x2, y2, x3 … etc. For the 2D molar shape data (Canid_M1_SIdata.csv) there are 51 landmarks and therefore an additional 102 columns. For the 2D mandibular shape data (Canid_Mand_SIdata.csv) there are 15 landmarks and therefore in the csv files there are an additional 30 columns.