HESPERHYINAE, A NEW SUBFAMILY OF EARLY-MIDDLE MIOCENE PECCARIES, AND ITS IMPLICATIONS FOR PALEOBIOLOGICAL DATA BASES

The Hesperhyinae is a monophyletic group of primitive peccaries that have long been plagued by confused systematics, invalid taxa, incorrect generic assignments, and poor specimens. Even though the currently published literature recognizes only Hesperhys (and sometimes Desmathyus), previously undescribed fossils show that there are seven distinct genera (four new) and seven species (two new). Hesperhyines are now defined by a distinctive suite of characters in the choanal region of the skull. The earliest hesperhyines include a primitive new genus and species, Lucashyus coombsi, from the late Arikareean of Wyoming, along with “Cynorca” (now Marshochoerus) socialis from the late Arikareean of the John Day Formation. Floridachoerus olseni from the early Hemingfordian of the Thomas Farm local fauna in Florida is a slightly more derived hesperhyine, as is the new genus Stuckyhyus siouxensis from the late Arikareean of Wyoming. The most derived taxon is Hesperhys vagrans (from the late Hemingfordian and early Barstovian). New material demonstrates Hesperhys a very distinctive large peccary with robust, inflated cheek teeth. Hesperhys is completely different from the smaller more gracile Desmathyus pinensis (from the late Arikareean of South Dakota) and the new taxon Wrightohyus yatkolai, from the late Arikareean-early Hemingfordian of Nebraska and Wyoming. All of three of these taxa were incorrectly lumped into Hesperhys. This revision of all the new material and mistaken systematic assignments reveals a significant evolutionary radiation and a much greater taxonomic diversity of peccaries in the early to middle Miocene than the older literature had ever suggested. It also substantiates the importance of this large monophyletic clade that was completely unrecognized by earlier scientists. Such undescribed diversity in places like the Frick Collection is a serious problem for databases of Miocene mammals, like PBDB and MIOMAP, which assume that bad taxonomy in their data is not a problem.