The morphology of the mandibular coronoid process does not indicate that Canis lupus chanco is the progenitor to dogs

The domestication of wolves is currently under debate. Where, when and from which wolf sub-species dogs originated are being investigated both by osteoarchaeologists and geneticists. While DNA research is rapidly becoming more active and popular, morphological methods have been the gold standard in the past. But even today morphological details are routinely employed to discern archaeological wolves from dogs. One such morphological similarity between Canis lupus chanco and dogs was published in 1977 by Olsen and Olsen. This concerns the “turned back” anatomy of the dorsal part of the vertical ramus of the mandible that was claimed to be specific to domestic dogs and Chinese wolves C. lupus chanco, and “absent from other canids”. Based on this characteristic, C. lupus chanco was said to be the progenitor of Asian and American dogs, and this specific morphology has been continuously used as an argument to assign archaeological specimens, including non-Asian and non-American, to the dog clade. We challenged this statement by examining 384 dog skulls of 72 breeds and 60 skulls of four wolf sub-species. Only 20 % of dog mandibles and 80 % of C. lupus chanco showed the specific anatomy. In addition, 12 % of Canis lupus pallipes mandibles showed the “turned back” morphology. It can be concluded that the shape of the coronoid process of the mandible cannot be used as a morphological trait to determine whether a specimen belongs to a dog or as an argument in favour of chanco as the progenitor to dogs.

The other group claims that dogs originated before the LGM, as early as in the Aurignacian and Gravettian and thus 35,000 years ago (Bocherens et al. 2014;Germonpré et al. 2009Germonpré et al. , 2012Ovodov et al. 2011;Sablin and Khlopachev 2002). Although genetic analysis has not found any relationship between these old archaeological canine specimens (Thalmann et al. 2013) purported to be domesticated wolves and modern dogs, these researchers suggest that these animals were, however, domesticated, but did not produce surviving offspring (aborted domestication waves) (Germonpré et al. 2012;Skoglund et al. 2011). The arguments to place these pre-LGM specimens in the dog clade are based on morphology alone and mainly on wider and shorter snouts. Drake et al. (2015) have, however, demonstrated that this criterion (shorter and wider snouts) is not useful in distinguishing dogs from wolves and also identified some of the so-called pre-LGM dog fossils as wolves.
Many morphological differences have been described between wolves and dogs in the literature since the eighteenth century (Clutton-Brock 1962;Degerbøl 1961;Nehring 1888;Stockhaus 1965;Studer 1901;Wolfgram 1894). Three morphological methods were used to examine morphological differences: • The ''obvious'' visual difference in appearance (morphology, sensu stricto) (Olsen and Olsen 1977). • The difference in size (morphometry) (Benecke 1987(Benecke , 1994; Boudadi-Maligne and Escarguel 2014; Napierala and Uerpmann 2012). The most frequently reported morphological and morphometric differences used to distinguish dogs from wolves are smaller stature and thus smaller anatomical parts (e.g., skull, teeth such as carnassials, etc.), shorter and wider snouts, tooth crowding, larger orbital angles and a ''turned back'' morphology of the dorsal side (apex) of the vertical ramus of the mandible (coronoid process) (Olsen and Olsen 1977, Fig. 1 and 2, p. 534-535). The latter morphological difference is based purely on difference in shape. This distinctive morphological characteristic was described in 1977 by Olsen and Olsen (Olsen and Olsen 1977). The authors state that the origin of Asian and American (New World) dogs must have originated in the Far East and proposed the Tibetan wolf (Chinese wolf, Asian wolf, Canis lupus chanco) as the dog's ancestor (Olsen and Olsen 1977, 534). This opinion was based on the specific ''turned back'' morphology of the coronoid process of the mandible, claimed to be ''specific to domestic dogs'' and Chinese wolves, and to be ''absent from other canids'' (Olsen and Olsen 1977, 534). Based on this assumption, C. lupus chanco was said to be ''the progenitor of dogs'', and this specific morphological trait is still used in recent publications to assign archaeological specimens to the dog clade (e.g., Ovodov et al. 2011).
We tested the statement of Olsen and Olsen (1977) by examining 384 dog mandibles of many breeds, of which six breeds are Asian or American, and 60 wolf mandibles of four sub-species. Our aim is to examine whether this ''turned back'' morphology is indeed present in ''all'' dogs and only in C. lupus chanco as hypothesized.

Materials and methods
All examined mandibles are from reputable museum collections. These had been collected in historical and recent periods and were professionally prepared. All are intact and from adult animals. In total 444 dog and wolf skulls were examined (888 mandibles) including 384 dog skulls and 60 wolf skulls. For the wolves (Table 1), 37 are from the collection of The George S. Wise Faculty of Life Sciences, Department of Zoology at Tel-Aviv University, Israel (ZMTAU). Thirty-two of these were Canis lupus pallipes and five Canis lupus arabs. Seven skulls were examined from the collection of the Natural History Museum in London, Great Britain (BMNH): six C. lupus arabs, and one C. lupus pallipes. Eleven skulls are from the collection of the Natural History Museum Bern, Switzerland (NMBE), all from Eurasian wolves (Canis lupus lupus) from Central Europe or Russia. Five specimens of C. lupus chanco from the collection of the Department of Vertebrate Zoology, Smithsonian Institution at the National Museum of Natural History, Washington DC, USA (USNM), were also examined.
We also examined 123 dog skulls from the collection of the anatomy department of the school for Veterinary Medicine, Ghent University, Belgium, and 261 skulls from the collection of The Museum of Natural History, Bern, Switzerland (total 384) ( Table 2). The skulls belong to 72 different breeds, of which six breeds and 33 skulls are Asian or American. These are Alaskan malamute (5), Canadian Eskimo dog (4), Chow-Chow (16), Shar Pei (1), Tibetan Mastiff (6) and Tibetan Terrier (1).
Each mandible was digitally photographed from a distance of 40-50 cm with a digital Nikon D 700 camera with a 50 mm lens. The photographs were imported in the OsiriX Imaging Software program. A straight vertical line was then drawn confluent with the straight part of the ventral caudal border of the mandible. The mandibles were divided in two categories based on the morphology of the coronoid process and by drawing a straight line (green on the figures) coinciding with the caudal border. For Category 1, the mandible has a perfect vertical straight caudal border ( Fig. 1) or the uppermost part of the apex points minimally in the caudal direction, while the caudal border is straight (Fig. 2). In this category, the straight green line follows the caudal bony border of the vertical ramus and the dorsal aspect of the mandible does not cross the green line or transects only a very small part at the tip. For Category 2, the caudal border is concave over its entire length and has the form of a dolphin fin (Fig. 3). Here, the

Discussion
Three main claims are made in Olsen and Olsen's article (1977). The first is that the Chinese wolf is progenitor to Asian and New World dogs. When Olsen and Olsen's   (Pocock 1946), should also be placed in the same historical perspective as the 1970s were a period of a Sino-American rapprochement (Oksenberg 1982).
Recent genetic analysis has confirmed that only wolves are progenitors to dogs, contradicting older theories about different geographic domestication waves (Duleba et al. 2015;Horard-Herbin et al. 2014;Larson et al. 2012;Thalmann et al. 2013) and has revealed that New World dogs did not originate locally but invaded the continent together with early migration waves of Homo sapiens (Leonard et al. 2002;Savolainen et al. 2002). The original article shows drawings of 13 mandibles of which only ten have sufficient intact anatomy to make interpretation possible (according to personal re-examination of the published drawings by LJ). Of these, six belong to dogs, one to C. lupus chanco and three to species other than Canis lupus. All dogs and all C. lupus chanco specimens show the ''turned back'' anatomy. It is not reported if more than these seven mandibles were examined. If not, it is difficult to understand why such a general statement was published. C. lupus chanco skulls are very difficult to find in zoological and natural history collections. This may explain why only one was reported in the article. We found only eleven skulls in many worldwide collections. Of these only five had intact mandibular anatomy, of which one (20 %) had a straight caudal mandibular ramus, contradicting Olsen and Olsen's (1977) original statement.
The second assertion is that the ''tuned back'' morphology is absent from other canids. This statement is unsupportable as we have demonstrated the presence of the ''turned back'' morphology in C. lupus pallipes mandibles. Studer (1901) early on reported that from all examined wolf skulls pallipes and chanco were the most anatomically similar. This may explain why these two wolf subspecies share this ''turned back'' morphology, unseen in the two other wolf sub-species we examined.
The third statement is that ''dogs have the turned back morphology''. At one point in the article this statement is made in general: ''all dogs'' have the turned back morphology (Olsen and Olsen 1977, 534, last paragraph), while in another location it refers to ''New World and Asian dogs'' (Olsen and Olsen 1977, 533, fifth paragraph), while the title of the article refers only to New World dogs. The ''turned back'' morphology is present in the six dog mandible drawings in the article, but the same pattern was not observed in the large group of dog mandibles we examined, not in general and not in Asian or New World dogs. Indeed, only a minority of dogs (20 %) have ''turned back'' morphology. In addition there are no differences in occurrence between Asian and/or New World dogs nor in the total group of dogs (18 % in these breeds vs. 20 % in total).

Conclusion
The statement that all dogs have a specific ''turned back'' morphology of the mandibular coronoid process, and that they share this specific morphology with only one wolf sub-species (C. lupus chanco), is untenable. This morphological trait cannot therefore be used as an argument to claim that archaeological remains belong to dogs, nor to argue that C. lupus chanco is the progenitor of dogs.