Muknalia minima from the Yucatán of Mexico is synonymous with the collared peccary, Pecari tajacu (Artiodactyla: Tayassuidae)

Ongoing investigation of peccary remains from fossiliferous deposits in the Yucatán resulted in reexamination of previously identified tayassuid fossils from the region. This included the recently described new genus and species of peccary, Muknalia minima, which is based on a dentary from Muknal Cave near Tulum, Quintana Roo, Mexico. Diagnostic characters of this taxon include a concave notch along the caudal edge of the ascending ramus and a ventrally directed angular process. Our assessment of the holotype indicates that these characteristics are not a reflection of the original morphology, but are instead the result of breakage and polishing of the posterior aspect of the dentary. Measurements and intact morphological features indicate the Muknal Cave specimen belongs to the extant collared peccary, Pecari tajacu.

Our team's Yucatán research has focused primarily on Hoyo Negro, a natural pit inside the submerged Sac Actun cave system, Quintana Roo, Mexico (Chatters et al., 2014;McDonald et al., 2017;Schubert et al., 2019). As Hoyo Negro contains fossil peccary skeletons, we were curious whether or not this pit may also record the recently described new genus and species (Muknalia minima) reported to occur in the same region (Stinnesbeck et al., 2017c). Muknalia minima was described as a new taxon based on unique morphological features of the dentary and its overall small size. Here we re-evaluate the holotype dentary and present a comparison of the specimen to other Quaternary peccaries. We utilize these data to comment on the taxonomic status of M. minima.

Extant Peccary Species
As part of our analysis, we compare the Muknalia minima holotype to extant peccaries. Because this group has a complex taxonomic and systematic history in the literature, we provide a brief review here. There are three living tayassuid clades, commonly known as the collared peccary, white-lipped peccary, and Chacoan peccary. The genus Tayassu has been used for both the white-lipped peccary (T. pecari) and phenotypically similar collared peccary (T. tajacu) (e.g., Kurtén and Anderson, 1980;Mayer andWetzel, 1986, 1987;Sowls, 1997). However, Woodburne (1968) separated the collared and white-lipped peccaries into distinct genera based on comparative anatomy, referring to the collared peccary as Dicotyles tajacu. Early genetic research recognized three extant peccary species, but demonstrated the Chacoan peccary (Catagonus wagneri) and white-lipped peccary (T. pecari) are more closely related to each other than either is to the collared peccary (Theimer and Keim, 1998). Based on this separation from T. pecari, and taxonomic priority, Theimer and Keim (1998) referred to the collared peccary as Pecari tajacu, and this combination has been followed by many researchers (e.g., Gongora and Moran, 2005;Grubb, 2005;Woodburne et al., 2006;Hulbert et al., 2009;Taber et al., 2011;Fariña et al., 2013;Reyna-Hurtado et al., 2018), as well as current online databases, such as the Tree of Life Web Project (2006), Animal Diversity Web (Ingmarsson, 1999), and IUCN Red List of Threatened Species (Gongora et al., 2011). Others have continued to use Tayassu for both the white-lipped and collared peccaries (e.g., Eisenberg and Redford, 1999;Gasparini et al., 2014), or agreed with the generic split but continued to use Dicotyles instead of Pecari for the collared peccary (e.g., Harris and Li-Ping, 2007;Prothero, 2009). Ramírez-Pulido et al. (2014) recently argued for a return to Dicotyles based on nomenclatural priority. To further complicate the matter, some genetic and morphological research suggests splitting the collared peccary into two or three species with the specific epithets tajacu, angulatus, and crassus (e.g., Gongora et al., 2006Gongora et al., , 2011Gongora et al., , 2018Groves and Grubb, 2011;Ramírez-Pulido et al., 2014), and the current Mammal Diversity Database (2020) follows this method. For the purposes of this paper, we follow the usage of P. tajacu in the broader sense (sensu lato), as the sole member of a monotypic genus with a high degree of intraspecific variation (see Reyna-Hurtado et al., 2018).

Methods and Materials
The holotype of Muknalia minima was assessed by BWS and JCC on November 28, 2018 at the Instituto de la Prehistoria de América lab located in Xplor Park, Playa del Carmen, Q.R., Mexico. This specimen (collection number PQR2011-PALMAS-V-1) had been altered since the original photographs in Stinnesbeck et al. (2017c), and was missing the p3 (which was used for attempting a radiocarbon date). A 10X magnification hand lens was used to examine the surface morphology of the dentary. Some pictures herein are from Stinnesbeck et al. (2017c, figures 2, 4 and 5) and are used with permission from Elsevier.
Comparative material was examined for both Pecari tajacu (including both P. tajacu yucatanensis and P. tajacu nanus) and Tayassu pecari from the USNM (Smithsonian Institution's National Museum of Natural History, Washington, DC). The USNM collection was selected so qualitative and quantitative comparisons could be made to a geographically similar sample (Supplemental Tables 1 and 2). Specimens of both taxa from the USNM were photographed and Pecari tajacu specimens were measured for standardization and scaling with calipers, followed by additional measurements using ImageJ software (by JXS). All measurements are based on standards in von den Driesch (1976) and the list of measurements reported by Stinnesbeck et al. (2017c). Measurements of the teeth included standard length and width measurements of the lower premolars (p2 -p4) and molars (m1 -m3), while dentary measurements utilized here are as follows: 1) dentary L -from anterior margin of canine (at alveolus) to posterior margin of mandibular condyle, 2) dentary H -from dorsal margin of coronoid process to ventral margin of masseteric fossa along mandibular angle, 3) dentary depth at m1 -from dorsal margin of m1 alveolus to ventral margin of horizontal ramus, 4) length of diastema -from posterior margin of canine (at alveolus) to anterior margin of p2 (at alveolus), 5) length of tooth row -from anterior margin of canine (at alveolus) to posterior margin of m3, and 6) length of cheek tooth row -from anterior margin of p2 to posterior margin of m3. For details on measured specimens see Supplemental Tables 1 and 2.

Results and Discussion
Our re-examination of the Muknalia minima holotype (PQR2011-PALMAS-V-1) and comparison to other peccaries resulted in new interpretations of the taxonomic assignment for this fossil. The original diagnosis of the genus and species was based on characters of the dentary that were divided into seven categories, five morphological and two size-related (Stinnesbeck et al., 2017c). The morphological characters are: 1) "Deep concavity formed by the caudoventral edge of the condylar process and the caudal edge of the mandibular ramus, with an aperture angle of 110 degrees", 2) "Ventrally pronounced angular process, reaching almost one third of the maximum mandibular height", 3) "Angular process is laterally and ventrally convex, while the medial surface of the mandible is concave", 4) "Condylar process forms the caudal-most extremity of the mandible", 5) "Outline of the coronoid process is trapezoidal in lateral view, with a horizontal dorsal margin." In contrast to the original report (Stinnesbeck et al., 2017c), we interpret the diagnostic morphological features as the result of breakage and polishing, not intact morphology (Figure 1). The angle of the ramus has been removed, leaving a ~1 10 degree angle gap with one edge that is nearly parallel to the long axis of the specimen ( Figure 1B). The posterior edge of the remaining portion of the ramus is nearly vertical ( Figure 1B) and exhibits readily discernible fractures ( Figure 1E). Along the vertex of the notch angle and back to the posterior edge of the remaining condyloid process ( Figure 1D and G), the dentary is rounded and polished. As a result, the remaining surfaces superficially appear to be intact cortical bone. However, under examination with a hand lens, these surfaces include trabecular bone that exhibits a polished sheen (Figure 1G). Such a sheen can occur on bones that are repeatedly rubbed by fibrous material such as cloth or leather, or handled by people (e.g., d 'Errico, 1993;LeMoine, 1997;Backwell and d'Errico, 2004;Rosell et al., 2011;Rots, 2015).
In addition to the polishing, distinct striation marks occur on the ramus. These marks are illustrated by Stinnesbeck et al. (2017c, figs. 2-3) but are not interpreted in that paper. In a follow-up article about the site, 112 cut and scratch marks are noted on the dentary and these are attributed to human modification (Stinnesbeck et al., 2018b). A series of these striations are roughly perpendicular to the polished edge of the vertex of the rightangled notch in the ramus (Figure 1C), further supporting anthropogenic attribution. The concentration of striae in this area may be a remnant of the action taken to create either a notch at the midpoint of the caudal edge of the ramus, or a perforation that subsequently wore through.
In response to characters 1-5 of Stinnesbeck et al. (2017a), the deep concavity in the caudoventral aspect of the mandible (character 1), appearance of the "angular process" in this specimen (characters 2 and 3), and fact that the condylar process is the caudal-most extent of the dentary (character 4), are all a direct consequence of breakage and human modification. Further, while the expanded masseteric fossa occupies the angle of the mandible in peccaries, the preserved portion of the Muknal specimen does not represent a true angular process with a distinctive shape (character 3), but rather is an incomplete portion of a larger structure. The outline of the coronoid process is variable in the sample of Pecari tajacu we have studied from Mexico, and in most studied specimens from the Yucatán (USNM 36968, 108281, 108514, 108515, 108517, 108518) the process has a trapezoidal shape in lateral view and flat (horizontal) dorsal margin (character 5).
The other primary diagnosing characters for Muknalia minima (Stinnesbeck et al. 2017c) relate to the relatively small size of the peccary: "6) Diastema between canine and p2 having approximately one eighth of the mandibular length. In all other Tayassuidae this diastema is one third to one half shorter" and "7) Small body size (approximately 17-24 kg body mass); mandible at least one fourth to one half smaller than mandibles of adult Pleistocene (Mylohyus and Platygonus) and extant Tayassuidae." Woodburne (1968) reported measurements of a large sample (N = 71) of Pecari tajacu (as Dicotyles tajacu) from the American Museum of Natural History, and we have examined smaller samples of subspecies that inhabit the Yucatán region today, P. tajacu yucatanensis and P. tajacu nanus (Tables 1 and 2). Based on these data, the Muknal dentary diastema (lower canine to p2) of 21 mm reported by Stinnesbeck et al. (2017c) fits into the size range (17.60-38.40 mm) of P. tajacu reported by Woodburne (1968) and is also very similar in size to multiple specimens in our extant sample from the region ( Table 1). The size of the diastema in all of these samples is considerably smaller   than the next-largest extant species, Tayassu pecari, which has a diastema that ranges from 27.35-41.80 mm (N = 41) (Woodburne, 1968). It should also be noted that the Muknal dentary is incomplete in the region of the canine alveolus, and therefore the published length for the diastema is an estimate. The seventh and final diagnostic feature included small body size (listed as weight) and a mandible that is smaller than any living peccary species. In their conclusions, the authors note the "most striking feature of the animal is its small body mass of 17-24 kg compared to Mylohyus and Platygonus" (Stinnesbeck et al., 2017c). The weight approximation is not supported by any calculations in the paper and appears to rely on the interpretation that it is smaller than P. tajacu based on the dentary. Our sample of the pygmy collared peccary from Cozumel Island, P. tajacu nanus, is consistently smaller than the Muknal dentary in nearly all measurements (Tables 1 and 2). That sample is also smaller than values reported for P. tajacu reported by Woodburne (1968), with only a few exceptions. These samples also suggest the length of the diastema is not closely related to overall body size in these peccaries; while most measurements of P. tajacu nanus are smaller than other samples of P. tajacu, the size range of the diastema broadly overlaps with other samples (Tables 1 and  2, Supplemental Table 1).
In overall morphology and measurements (Figures  1-2; Tables 1-2), the Muknal dentary is consistent with variation observed in Pecari tajacu, and within the range of studied specimens of P. tajacu yucatanensis from the Yucatán. According to Woodburne (1968, p. 29), the dentary of P. tajacu can be distinguished from T. pecari based on morphological characters. Those that are visible in the fractured, modified, and weathered Muknal specimen and are characteristic of P. tajacu include an ascending ramus that rises well behind m3, and a mandibular angle that is slightly inflected (Woodburne, 1968). In our observation of specimens, we also noted that the condyloid process is in-line with or just slightly above the cheek tooth row in P. tajacu and the Muknal dentary, while it is substantially elevated relative to the tooth row in T. pecari. In addition to these characters separating Tayassu and Pecari, our observation of modern comparative material exemplifies a high degree of variation in the morphology of P. tajacu dentaries, and further supports the identification of the fossil as this species (Figure 2).
Teeth from the Muknal dentary are thoroughly illustrated in Stinnesbeck et al. (2017c) and measurements are given. However, these measurements and the dental morphology are not discussed. All dental measurements are consistent with the observed range of P. tajacu listed in Woodburne (1968, p. 44), except for the m1 length and m2 width ( Table 2). The m1 length is 0.15 mm smaller than reported for P. tajacu by Woodburne (1968), only 0.01 mm smaller than studied specimens of P. tajacu yucatanensis, and larger than all studied specimens of P. tajacu nanus. The m2 width is reported as 9 mm, notably narrower than the recorded size range in Woodburne (1968) and P. tajacu yucatanensis, but again larger than all studied specimens of P. tajacu nanus ( Table 2). While we did not re-measure the teeth in our examination of the specimen, close inspection of fig 5 in Stinnesbeck et al. (2017c) indicates that the dimension they report may be an error. When the provided scale bar in this figure is rotated, the width of the m2 crown appears to be ~1 0 mm, and other dental measurements increase as well (see Figure 1A). It is not known at this time whether or not these measurements in Stinnesbeck et al. (2017c) are correct or if the scale bar is too small to accurately assess dental dimensions. Premolar and molar morphology of the Muknal peccary is also consistent with P. tajacu, and unlike other extant and extinct species from the Quaternary (e.g., Platygonus and Mylohyus). The flat-headed peccary (Platygonus) and Chacoan peccary (Catagonus wagneri) have more-hypsodont cheek-teeth with distinct p4 morphology, and the long-nosed peccary (Mylohyus) has a p4 and molars with cusps and cuspules that are swollen and rounded (Kurtén and Anderson, 1980;Mayer and Wetzel 1986;Hulbert et al., 2009). The white-lipped peccary (Tayassu pecari) has brachydont and bunodont teeth quite similar to P. tajacu, but according to Woodburne (1968, p. 29) the p4 and m1 crown heights are the same in the former, while in the latter the p4 has a higher-crown than the m1. All specimens of P. tajacu that we examined from Mexico and specimens of T. pecari from Mexico and Guatemala consistently agree with the assessment of Woodburne (1968), with P. tajacu bearing a higher crowned p4 compared to the m1. As demonstrated in Figures 1 and 2, the Muknal specimen is brachydont with a p4 crown that is clearly higher than the m1 crown. Thus, all evidence here supports identification of the Muknal peccary as P. tajacu.
Although P. tajacu currently has a large range that extends from Argentina to the southwestern United States (Sowls, 1997, Reyna-Hurtado et al., 2018, it is rare in the fossil record of Central and North America. In fact, the lack of records led researchers to infer a northern dispersal from South or Central America to northern Mexico and the southwestern U.S. in the Holocene (e.g., Woodburne, 1969;Kurtén and Anderson, 1980). Woodburne (1969) reported a Pleistocene specimen from Guatemala, and Mexico has Pleistocene occurrences from the Yucatán (Arroyo- Cabrales and Polaco, 2003;Ferrusquia-Villafranca et al., 2010). Other than these records, pre-Holocene collared peccaries appear to be lacking from Central America and Mexico. From the contiguous United States, Pleistoceneage collared peccaries have only been reported from Florida, where the genus (Pecari sp.) is described from three localities (Hulbert et al. 2009). This sparse record for Central and North America makes new occurrences from this region significant for understanding the evolution, biogeography, and dispersal of this group. Additional study of peccaries from other sites in the region (e.g. Hoyo Negro) is beyond the scope of the current paper, but additional skeletons from underwater caves have the potential to further clarify the taxonomic character of fossil peccaries from the region and yield insights into the origin of peccaries living there today.

Summary
Underwater caves of the Yucatán continue to provide exciting discoveries of new species and range extensions that date to the Pleistocene and early Holocene. However, our assessment of the new peccary, Muknalia minima (Stinnesbeck et al., 2017c) from Muknal Cave indicates that this taxonomic assignment is invalid. The shape of the holotype dentary that was interpreted as having unique morphologies is the result of breakage and human modification. The interpretation of "M. minima" being distinct because of its small size is also incorrect because dimensions fit within the observed range of Pecari tajacu, the smallest Quaternary peccary. In addition, intact morphology of the dentary and teeth are in-line with P. tajacu and distinct from other extant and extinct species. Thus, the Muknal dentary is reassigned here to P. tajacu, the collared peccary.

Additional Files
The additional files for this article can be found as follows: