Craniodental Morphology And Phylogeny Of Marsupials Author Beck, Robin M. D. School of Science, Engineering and Environment University of Salford, U. K. & School of Biological, Earth & Environmental Sciences University of New South Wales, Australia & Division of Vertebrate Zoology (Mammalogy) American Museum of Natural History Author Voss, Robert S. Division of Vertebrate Zoology (Mammalogy) American Museum of Natural History Author Jansa, Sharon A. Bell Museum and Department of Ecology, Evolution, and Behavior University of Minnesota text Bulletin of the American Museum of Natural History 2022 2022-06-28 2022 457 1 353 https://bioone.org/journals/bulletin-of-the-american-museum-of-natural-history/volume-457/issue-1/0003-0090.457.1.1/Craniodental-Morphology-and-Phylogeny-of-Marsupials/10.1206/0003-0090.457.1.1.full journal article 10.1206/0003-0090.457.1.1 0003-0090 6971356 Australidelphia Szalay, 1982 CONTENTS: Dasyuromorphia , Diprotodontia , Microbiotheria , Notoryctemorphia , and Peramelemorphia . STEM AGE: 55.1 Mya (95% HPD: 54.6–56.6 Mya). CROWN AGE: 48.0 Mya (95% HPD: 44.3–50.9 Mya). UNAMBIGUOUS CRANIODENTAL SYNAPOMORPHIES: None. COMMENTS: Recent phylogenetic analyses based on molecular data (e.g., Phillips et al., 2001 ; Amrine-Madsen et al., 2003b ; Nilsson et al., 2004 ; Phillips et al., 2006 ; Beck, 2008a ; Meredith et al., 2008b , 2009 c , 2011; Nilsson et al., 2010 ; Mitchell et al., 2014 ; Gallus et al., 2015a ; Duchêne et al., 2018 ; Álvarez-Carretero et al., 2021 ) and others based on total-evidence datasets (e.g., Asher et al., 2004 ; Beck et al., 2008 a , 2014, 2016 ; Horovitz et al., 2009 ; Beck, 2012 ; Maga and Beck, 2017 ) have consistently recovered monophyly of Australidelphia, and australidelphian monophyly has also been found in most (e.g., Horovitz and Sánchez-Villagra, 2003 ; Sánchez-Villagra et al., 2007 ; Horovitz et al., 2008 , 2009 ; Lorente et al., 2016 ; Carneiro and Oliveira, 2017a ; Carneiro et al., 2018 ; Carneiro, 2019 ) but not all (e.g., Ladevèze and Muizon, 2010 ; Wilson et al., 2016 ) recent morphological analyses. This overall pattern is confirmed here: our morphological analyses did not recover australidelphian monophyly (figs. 30, 31), whereas it was strongly supported in all our molecular (figs. 27–29) and total evidence (figs. 32, 33 ) analyses. Given that Szalay (1982a) first proposed monophyly of Australidelphia based primarily on shared derived features of the tarsus (specifically the presence of a continuous lower ankle joint, and a tripartite calcaneocuboid facet; see also Szalay, 1994 ; Beck, 2012 ), the addition of postcranial characters to our craniodental matrix may ultimately result in morphological support for Australidelphia (as in Horovitz and SánchezVillagra, 2003). No craniodental feature optimizes as an unambiguous synapomorphy of Australidelphia in our dated total-evidence analysis, but three optimize as synapomorphies under Accelerated Transformation—extracranial course of mandibular nerve fully enclosed by medial outgrowths of the auditory bulla (char. 52: 0→1; ci = 0.231); posterior limb of ectotympanic in contact with, but suturally distinct from, pars canalicularis of the petrosal and/or posttympanic process of the squamosal (char. 60: 0→1; ci = 0.333); and prootic canal foramen on tympanic face of petrosal absent (char. 69: 0→1; ci = 0.083)—but all three traits show high levels of homoplasy and undergo subsequent reversals within Australidelphia (see file S 3 in the online supplement). Our failure to identify compelling craniodental synapomorphies for Australidelphia likely reflects the fact that the ancestral australidelphian probably had a relatively generalized cranium and dentition that was little different from the plesiomorphic marsupial condition (see Szalay, 1994: 346 ) and that different lineages within Australidelphia subsequently evolved very disparate apomorphies of the dentition, cranium, or both (see also comments by Archer, 1984c: 782 ). Our estimate for the first split within Australidelphia is in the early to middle Eocene. This is younger than the early or middle Paleocene † Khasia cordillerensis , a taxon that was originally described as a microbiotherian ( Marshall and Muizon, 1988 ; see also Muizon, 1991 ; Goin et al., 2006 ; Muizon et al., 2018 ; Muizon and Ladevèze, 2020 ) and hence a crown-clade australidelphian. However, as noted above, several subsequent authors have argued that † Khasia is more likely a “pediomyoid” ( Oliveira and Goin, 2006 ; Goin et al., 2016 ), a hypothesis supported by the morphological phylogenetic analysis of Carneiro et al. (2018) . † Khasia has not been included here because it is represented only by dental specimens ( Marshall and Muizon, 1988 ; Muizon, 1991 ). If † Khasia is discounted, the oldest known australidelphian is probably † Djarthia murgonensis from the Tingamarra fossil site in eastern Australia , which has been radiometrically dated as earliest Eocene (~54.6 Mya; ( Godthelp et al., 1992 ; Godthelp et al., 1999 ; Beck et al., 2008a ). We have not included † Djarthia here due to its incompleteness (the only craniodental remains are incomplete dental specimens and isolated petrosals; Godthelp et al., 1999 ; Beck et al., 2008a ), but isolated tarsals referred to this taxon exhibit characteristic australidelphian synapomorphies ( Beck et al., 2008a ). However, † Djarthia falls outside crown-clade Australidelphia in most published analyses (with the notable exception of Maga and Beck, 2017 : fig. 38, in which it is sister to Dasyuromorphia ), and its position is unresolved with respect to the crown clade in others ( Beck et al., 2008 a , 2014, 2016 ; Beck, 2012 ; Lorente et al., 2016 ; Maga and Beck, 2017 ). The next-oldest definitive australidelphian remains are isolated tarsals from the La Barda locality in Argentina , which has been radiometrically dated as middle Eocene (between ~48 and 43 Mya; Tejedor et al., 2009 ; Lorente et al., 2016 ). The La Barda australidelphian tarsals fell within Diprotodontia (members of which are otherwise known only from Australia , New Guinea , and adjacent islands) in the phylogenetic analysis of Lorente et al. (2016) , but we consider that this biogeographically anomalous relationship warrants further testing. The woodburnodontid microbiotherian † Woodburnodon casei from the Cucullaea I Allomember of the La Meseta Formation on Seymour Island, off the Antarctic Penninsula, is a more compelling candidate for the oldest definitive crown-clade australidelphian because it preserves distinctive dental features that are characteristic of microbiotherians ( Goin et al., 2007c ), and it has been recovered within Microbiotheria in several phylogenetic analyses ( Carneiro and Oliveira, 2017b ; Carneiro et al., 2018 ; Carneiro, 2019 ). The age of the Cucullaea I Allomember has proved controversial ( Crame et al., 2014 ; Kemp et al., 2014 ; Gelfo et al., 2017 , 2019 ; Goin et al., 2020 ), but it now appears to be about 40 Mya ( Douglas et al., 2014 ; Amenábar et al., 2019 ; Mörs et al., 2020 ). This postdates our estimate for the first divergence within Australidelphia (see above) and also for the split between Microbiotheria and Diprotodontia (median = 45.6 Mya; 95% HPD: 41.4–48.8 Mya). Older putative microbiotherians have been reported from the early Eocene (probably 51.4– 56.0 Mya; Clyde et al., 2014 ; Woodburne et al., 2014 a , 2014b; Krause et al., 2017 ) Las Flores Local Fauna of southern Argentina ( Goin, 2003 ; Zimicz, 2012 ; Woodburne et al., 2014a ; Goin et al., 2016 ), which is incongruent with our estimate for the time of the Microbiotheria-Diprotodontia split, but these potentially important fossils have yet to be described.