Didelphimorphia Gill, 1872 CONTENTS: Didelphidae (see table 2 for included terminal taxa). 29 The phylogenetic analysis of Wilson et al. (2016) is a notable exception. STEM AGE: 56.2 Mya (95% HPD: 54.7–58.6 Mya). CROWN AGE: 27.1 Mya (95% HPD: 23.7–31.1 Mya). UNAMBIGUOUS CRANIODENTAL SYNAPOMORPHIES: Premolariform P2 and premolariform P3 subequal in height (char. 119: 21; ci = 0.118); premolariform P3 with well-developed anterior and posterior cutting edges (char. 124: 1→0; ci = 0.667); M4 erupts before P3 (char. 130: 2→0; ci = 0.089); and p2 distinctly taller than p3 (char. 156: 2→0; ci = 0.118). COMMENTS: As noted by numerous authors (e.g., Aplin and Archer, 1987; Goin, 2003; Forasiepi et al., 2009; Voss and Jansa, 2021), the name “ Didelphimorphia ” has long been used for a nonmonophyletic assemblage of dentally plesiomorphic metatherians that included both marsupials and stem metatherians. By contrast, we follow the stem-based phylogenetic definition of Didelphimorphia proposed by Beck and Taglioretti (2020), namely, the most inclusive clade containing Didelphis marsupialis, but not Caenolestes fuliginosus or Phalanger orientalis. The family Didelphidae (sensu Voss and Jansa, 2009) comprises the entire didelphimorphian crown clade. All four of our fossil didelphimorphian terminals († Hesperocynus, † Sparassocynus, † Thylatheridium, and † Thylophorops) fall within the crown clade, so craniodental synapomorphies for Didelphimorphia are also synapomorphies of Didelphidae in our applications of these names. All the craniodental features that optimize as unambiguous synapomorphies of Didelphimorphia / Didelphidae show high levels of homoplasy, as indicated by their low consistency indices (see above). One potential craniodental synapomorphy of Didelphimorphia / Didelphidae that shows less homoplasy is precocious fusion of the suture between the interparietal and supraoccipital (char. 31, state 1). This feature was first noted in this context by Voss and Jansa (2009), who stated that it is unique to didelphids within Marsupialia, but we also observed it in dactylopsiline petaurids (see char. 31). Nevertheless, interparietal-supraoccipital fusion optimizes as a synapomorphy of Didelphimorphia / Didelphidae only under Accelerated Transformation. The reason is that it could not be scored in key nondidelphimorphian taxa that would help unambiguously resolve the branch along which this feature arose, either because suitably preserved material of an appropriate ontogenetic age is unavailable (e.g., for paucituberculatans, † Yalkaparidon, † Mimoperadectes, and our outgroup terminals), or because this character is inapplicable (as is the case for † Herpetotherium, in which a distinct interparietal appears to be absent; see char. 30). In agreement with Goin (1991, 1995), our dated total-evidence analysis (fig. 33) indicates that the didelphimorphian crown clade is a relatively young radiation, with its earliest divergence estimated as late Oligocene or early Miocene. This estimate is also broadly congruent with the molecular node-dating analysis of Jansa et al. (2014) and with the total-evidence tip-andnode dating analysis of Beck and Taglioretti (2020); however, it is considerably younger than estimated divergence dates suggested by several other molecular studies, which have dated the last common ancestor of living didelphimorphians to the early Oligocene or Eocene (Steiner et al., 2005; Meredith et al., 2008b; 2011; Mitchell et al., 2014; Vilela et al., 2015; Álvarez-Carretero et al., 2021; see table 6). Our estimated crown age is likewise broadly congruent with the relevant fossil record (reviewed by Beck and Taglioretti, 2020; Castro et al., 2021), in which the oldest known (probable crown-clade) didelphids are from the early Miocene (Colhuehuapian SALMA, ~21.0–20.1 Mya; Goin et al., 2007a; Dunn et al., 2012; Castro et al., 2021). 30 The comparatively long branch between the divergence of Didelphimorphia from other marsupials and the first diversification of lineages ancestral to living didelphids suggests a long history (presumably largely or entirely South American; Jansa et al., 2014; Voss and Jansa, 2021) of stem didelphimorphians; because none have been reported to date—in effect, a protracted ghost lineage. However, given the lack of dental synapomorphies characterizing Didelphimorphia (Voss and Jansa, 2009), confidently distinguishing stem didelphimorphians from other dentally plesiomorphic marsupialiforms is likely to be difficult in the absence of well-preserved cranial specimens. Indeed, it seems likely that stem didelphimorphians are represented but remain unidentified among the rich dental record of Palaeogene marsupialiforms from South America (Goin et al., 2016). Postcranial material may prove useful in future attempts to identify stem didelphimorphians, because putative synapomorphies have been identified in the tarsus (Szalay, 1982a; 1994; Szalay and Sargis, 2001; Flores, 2009).

Published as part of Beck, Robin M. D., Voss, Robert S. & Jansa, Sharon A., 2022, Craniodental Morphology And Phylogeny Of Marsupials, pp. 1-353 in Bulletin of the American Museum of Natural History 2022 (457) on pages 198-199, DOI: 10.1206/0003-0090.457.1.1, http://zenodo.org/record/6971356