Dogs display a remarkable amount of phenotypic diversity in coat color, texture, and cranial and skeletal proportions, yet as a species they are distinct from wild canids (Wayne 1986a,b). Similarly, considerable variation in behavior and physiology are evident among breeds and between dogs and wild species (Hart 1995, Hare et al. 2002, Spady and Ostrander 2008). Considering the great diversity of dogs, Darwin and others postulated that domestic dogs were founded from more than one canid species (Darwin 1859, Lorenz 1954, Coppinger and Schneider 1995). However, molecular data showed conclusively that dogs originated from the gray wolf (Canis lupus) (Morey 1994, Vilà et al. 1997, Savolainen et al. 2002, Pang et al. 2009). Nonetheless, because of the morphologic divergence between dogs and gray wolves, identifying a specific ancestral gray wolf population is difficult.

The earliest proto-dogs were likely similar to wolves but subsequent evolution has rendered all breeds and dog populations as morphologically distinct from wild progenitors (Zeuner 1963, Epstein and Mason 1972, Olsen 1985, Morey 2006). Archaeological evidence suggests that the earliest dogs are from central Russia (15,000 y bP, Sablin and Khlopachev 2002), Western Europe (31,000 y bP, Germonpre et al. 2009) or the Middle East (12,000 y Bp, Nobis 1979, Olsen 1985, Davis 1987, Dayan 1994, Clutton-Brock 1995, Morey 2006). A problem with the archaeological data is that putative ancient dogs may in reality be differentiated varieties of local wolves (e.g., Leonard and Wayne 2008) or already so derived that the origin may be elsewhere. By contrast, East Asia is supported as a center of origin for dogs based on higher mitochondrial DNA (mtDNA) sequence diversity in dogs that originated there (Savolainen et al. 2002). However, mtDNA sequence is information from essentially a single gene and thus may not represent the history of the genome or the actual species tree (Wayne et al. 2006). For example, high mtDNA sequence diversity may reflect a center of regional trade rather than the location of species origin (Wayne et al. 2006). Consequently, genome-wide approaches are needed to sample a wider variety of gene histories and to provide a better consensus of evolutionary history.