View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Dokumenten-Publikationsserver der Humboldt-Universität zu Berlin Paleobiology, 45(1), 2019, pp. 56–69 DOI: 10.1017/pab.2018.38 Diversity patterns of nonmammalian cynodonts (Synapsida, Therapsida) and the impact of taxonomic practice and research history on diversity estimates Marcus Lukic-Walther, Neil Brocklehurst, Christian F. Kammerer, and Jörg Fröbisch Abstract.—Nonmammalian cynodonts represent a speciose and ecologically diverse group with a fossil record stretching from the late Permian until the Cretaceous. Because of their role as major components of Triassic terrestrial ecosystems and as the direct ancestors of mammals, cynodonts are an important group for understanding Mesozoic tetrapod diversity. We examine patterns of nonmammalian cynodont species richness and the quality of their fossil record. A supertree of cynodonts is constructed from recently published trees and time calibrated using a Bayesian approach. While this approach pushes the root of Cynodontia back to the earliest Guadalupian, the origins of Cynognathia and Probainognathia are close to their first appearance in the fossil record. Taxic, subsampled, and phylogenetic diversity estimates support a major cynodont radiation following the end-Permian mass extinction, but conflicting signals are observed at the end of the Triassic. The taxic diversity estimate shows high diversity in the Rhaetian and a drop across the Triassic/Jurassic boundary, while the phylogenetic diversity indicates an earlier extinction between the Norian and Rhaetian. The difference is attributed to the prevalence of taxa based solely on teeth in the Rhaetian, which are not included in the phylogenetic diversity estimate. Examining the com- pleteness of cynodont specimens through geological time does not support a decrease in preservation potential; although the median completeness score decreases in the Late Triassic, the range of values remains consistent. Instead, the poor completeness scores are attributed to a shift in sampling and taxo- nomic practices: an increased prevalence in microvertebrate sampling and the naming of fragmentary material. Marcus Lukic-Walther. Erst-Reuter-Gesellschaft, Freie Universität Berlin, Kaiserswerther Strasse 16–18, 14195 Berlin, Germany. E-mail: [email protected] Neil Brocklehurst. *Museum für Naturkunde, Leibniz-Institut für Evolutions- und Biodiversitätsforschung, Invalidenstraße 43, D-10115 Berlin, Germany. *Present address: Department of Earth Sciences, University of Oxford, South Parks Road, Oxford, UK, OX1 3AN. E-mail: [email protected] Christian F. Kammerer. North Carolina Museum of Natural Sciences, 11 West Jones Street, Raleigh, North Carolina 27601-1029, U.S.A. E-mail: [email protected] Jörg Fröbisch. Museum für Naturkunde, Leibniz-Institut für Evolutions- und Biodiversitätsforschung, Invalidenstraße 43, and Institut für Biologie, Humboldt Universität zu Berlin, Invalidenstraße 42, D-10115 Berlin, Germany. E-mail: [email protected] Accepted: 21 September 2018 First published online: 5 December 2018 Data available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.6jk8416 Introduction diversity outside the crown. Anomodontia The Triassic represents a significant period in and Therocephalia, which were ecologically the history of life, in which ecosystems recov- important synapsid clades in the Permian, sur- ered from the most massive extinction in vived the extinction and underwent moderate Earth’s history, the end-Permian event. In the subsequent diversification in the Triassic (Frö- aftermath of this event, the earliest members bisch 2008; Abdala et al. 2014; Huttenlocker of many major modern tetrapod clades et al. 2017). The primary synapsid success appear in the fossil record, including lepido- story in the Triassic, however, was from the saurs (Jones et al. 2013), crocodylomorphs Cynodontia, the group that eventually gave (Benton and Clark 1988), testudinates (Schoch rise to mammals. and Sues 2015), and mammals (Luo 2007). Cynodonts first appear in the fossil record in Although mammals are the only members of thelatePermian(Bothaetal.2007;Kammerer the synapsid lineage that survive to the present 2016). Permian cynodonts were rare compared day, Triassic synapsids include broader with coeval gorgonopsians and therocephalians, © 2018 The Paleontological Society. All rights reserved. 0094-8373/19 CYNODONT DIVERSITY AND RESEARCH HISTORY 57 and substantial diversification of the clade did calculated. This method endeavors to include not take place until the Triassic. Even before the as-yet-unsampled portions of the fossil record origin of mammals, Triassic cynodonts exhibited in the diversity estimate by incorporating a remarkably disparate range of morphologies ghost lineages (lineages not yet sampled but and ecologies, including large predators and her- inferred from the phylogeny). The method bivores (Ruta et al. 2013). These morphotypes has been used many times in analyses of terres- were lost by the end-Triassic, but further diversi- trial vertebrates (e.g., Upchurch and Barrett fication among nonmammalian cynodonts 2005; Barrett et al. 2009; Benson et al. 2011; occurred among the herbivorous Tritylodonti- Mannion et al. 2011; Ruta et al. 2011; Brockle- dae and insectivorous Trithelodontidae in hurst et al. 2013; Walther and Fröbisch 2013) the Jurassic. The tritylodontid lineage even and has been shown by simulations to outper- extended into the Early Cretaceous (Tatarinov form the TDE under a variety of sampling scen- and Maschenko 1999; Lopatin and Agadjanian arios (Lane et al. 2005; Brocklehurst 2015). 2007), surviving alongside the increasingly rich For this purpose, we generated a supertree Mesozoic radiation of mammals (Newham (Fig. 1) from nine published phylogenies cho- et al. 2014; Close et al. 2015). sen to maximize the taxonomic sample (Marti- As the ancestors of mammals, cynodonts nelli and Rougier 2007; Watabe et al. 2007;de have been extensively studied with regard to Oliveira et al. 2010; Kammerer 2016; Martinelli the details they provide on the evolution of et al. 2017a,b,c; Melo et al. 2017; Panciroli et al. the mammalian body plan (Sidor and Hopson 2017; Supplementary Data 2), containing 85 1998; Sidor 2003; Kielan-Jaworowska et al. valid species of nonmammalian cynodonts, 2004; Kemp 2005; Ruta et al. 2013). However, using the matrix representation with parsi- during the Triassic, nonmammalian cynodonts mony (MRP) method (Baum 1992; Ragan were a diverse assemblage in their own right, 1992). The MRP matrix was analyzed in TNT and study of their evolution and diversification v. 1.1 using the new technology search incorp- is important for understanding terrestrial verte- orating the drift, sectorial search, and fusion brate faunas in the aftermath of the end- algorithms. The minimum tree length was Permian mass extinction. searched for 100 times. In all, 427 most parsi- Here we examine the changes in nonmam- monious trees (mpts) were identified, 100 of malian cynodont species richness throughout which were selected at random for subsequent their history. Additionally, we examine the analyses. The selected mpts were time cali- completeness of cynodont specimens through brated using the method of Lloyd et al. (2016), time to provide details on the quality of the itself an expansion of a method put forward raw data used in diversity analyses. Recent by Hedman (2010), implemented in R v. 3.3.2 phylogenetic analyses are used to generate a (R Core Team 2016). This is a Bayesian supertree of cynodonts, which is used to calcu- approach using the ages of successive strati- late sampling-corrected diversity estimates that graphically consistent outgroup taxa relative are of greater accuracy than raw counts of to the age of the node of interest to make infer- species through time. ences about the quality of sampling; large gaps between the age of the node of interest and that of the outgroups imply a poorly sampled fossil Materials and Methods record, and therefore the age of the node of Diversity Estimates.—A taxic diversity esti- interest may be inferred to be older. Lloyd mate (TDE; no sampling correction applied) et al. (2016) designed a procedure whereby was calculated for nonmammalian cynodonts this approach could date an entire tree rather by simply counting the number of cynodont than just a specific node. Three successive out- species known from each stage. The data set groups were used to date the root node: the first assembled may be considered complete as of appearance of Therocephalia in the Wordian March 2018 (Supplementary Data 1). To (Abdala et al. 2008) and the first appearance account for sampling heterogeneity, a phylo- of Anomodontia and Dinocephalia in the genetic diversity estimate (PDE) was also Roadian (Liu et al. 2009). A hard lower bound 58 MARCUS LUKIC-WALTHER ET AL. FIGURE 1. A, Supertree of nonmammalian cynodonts—strict consensus of the 427 most parsimonious trees. B, A randomly selected example of the time-calibrated trees used to illustrate temporal branching patterns. of 324.7 Myr was placed on the root of the tree: are drawn until a fixed coverage (proportion the age of the split between synapsids and saur- of the individuals in the original sample repre- opsids identified by molecular dating (Dos Reis sented by species drawn in the subsample) is et al. 2016). While this age may seem overly attained, measured by Good’s u (Alroy 2010). conservative,