The diverse living Australian lizard fauna contrasts greatly with their limited Oligo-Miocene fossil record. New Oligo-Miocene fossil vertebrates from the Namba Formation (south of Lake Frome, South Australia) were uncovered from multiple expeditions from 2007–2018. Abundant disarticulated material of small vertebrates was concentrated in shallow lenses along the palaeo-lake edges, now exposed on the western shore. The fossiliferous lens occurring within the Namba Formation, also known from Billeroo Creek 2 km northeast of Lake Pinpa, includes abundant aquatic (such as fish, platypus Obdurodon, and waterfowl) and diverse terrestrial (such as possums, dasyuromorphs, and scincids) vertebrates and is hereafter recognised as the Fish Lens. The stratigraphic provenance of these deposits in relation to prior finds in the area is also established. A new egerniine scincid taxon Proegernia mikebulli sp. nov. described herein, is based on a near-complete reconstructed mandible, maxilla, premaxilla, and pterygoid. Postcranial scincid elements were also recovered with this material, but could not yet be confidently associated with P. mikebulli. This new taxon is recovered as the sister species to P. palankarinnensis, in a tip-dated total-evidence phylogenetic analysis, where both are recovered as stem Australian egerniines. These taxa also help pinpoint the timing of the arrival of scincids to Australia, with egerniines the first radiation to reach the continent.

In order to better understand the timing of the Australian colonisation by the Egerniinae, both molecular and morphological data (including fossils) are required to generate tip-dated phylogenies. Undated parsimony and tip-dated Bayesian analyses infer, respectively, the phylogeny with the least homoplasy, and the most probable dated phylogeny.

Morphological characters

                Morphological characters used in the following analyses consisted of 102 discrete and 48 continuous traits, forming an expanded matrix from Thorn et al. (Thorn, Hutchinson et al. 2019). The expanded character list is included. Continuous characters, derived from the measurements of the individual bones or teeth from the dentaries and maxillae, were taken from either Micro-CT scan data in Avizo Lite (v. 9.0) or SEM at Flinders Microscopy, to the nearest micrometre, or with digital callipers to the nearest ten micrometres. All measurements were converted to ratios of either dentary or maxilla length to standardise for size. Continuous character states were linearly scaled to values spanning 0–2 to replicate the mean number of discrete character states (three), for analyses in both TNT (Goloboff and Catalano 2016) and BEAST 1.8.3 (Drummond, Suchard et al. 2012), so that they do not have a disproportionate weight.

Molecular partitions

                Molecular data sourced from Tonini et al. (Tonini, Beard et al. 2016) and Gardner et al. (Gardner, Hugall et al. 2008) were analysed using Partition Finder 2 (Lanfear, Frandsen et al. 2016) to find optimal partitions and substitution models [15]. The same six molecular (gene) partitions, 12s (412 base pairs [bps]), 16s (681 bps), ND4 (693 bps), BDNF (699 bps), CMOS (835 bps) and B-fibrinogen (1051 alignable bps) and substitution models chosen in that study (Thorn, Hutchinson et al. 2019) are used again here.

Maximum Parsimony

                The parsimony analyses for the combined discrete morphological, continuous morphological, and molecular data were performed using TNT v.1.5 (Goloboff and Catalano 2016). Eutropis multifasciata was set as the most distant outgroup following the phylogenetic interpretations of Gardner et al. (Gardner, Hugall et al. 2008) and Thorn et al. (Thorn, Hutchinson et al. 2019). The most parsimonious tree (MPT) for the combined data was found using 1000 replicates of tree-bisection-reconnection (TBR) with up to 1000000 trees held.

To assess clade support, 200 partitioned bootstrap replicates (with discrete characters, continuous characters, and each gene locus treated as a separate resampling partition), were performed using TNT, using new search methods (XMULT) with 1000 replicates and 1000000 trees held. The MPT and bootstrap trees from TNT were exported in nexus format, and continuous and discrete characters were traced (in Mesquite; Maddison and Maddison 2017). The executable files for finding the Most Parsimonious tree, and for performing 200 reps of Partitioned Bootstrap resamples can be found in the SI data files Namba_Egerniines_Topology.tnt (MPT file) and Namba_Egerniines_PartitionedBootstrap.tnt.

Bayesian analysis

                The discrete and continuous morphological data, and molecular data were simultaneously analysed in BEAST v1.8.4 using tip-dated Bayesian approaches (Drummond, Suchard et al. 2012). Eutropis multifasciata was again set as the furthest outgroup. Polymorphic discrete morphological data were treated exactly as coded rather than as unknown, i.e. if coded as states (0,1) it was treated as 0 or 1, but not 2. The discrete character set was analysed using the Mkv-model with correction for non-sampling of constant characters (Lewis 2001, Alekseyenko, Lee et al. 2008). Despite recent disputes over the effectiveness of this model (Goloboff, Pittman et al. 2018), it is well-tested (Wright and Hillis 2014, O'Reilly, Puttick et al. 2016) and is still widely accepted and applied to morphological data (Harmon 2019). Continuous characters, transformed to span values between 0 and 2, were analysed with the Brownian motion model. Bayes factors were used to test the need to accommodate among-character rate variability for both discrete and continuous morphological characters (i.e. gamma parameter).

                The stratigraphic data used for tip-dating analyses were derived from fossil taxa and their associated stratigraphy noted in Table 1. No node age constraints were imposed in this analysis, all dates are retrieved from the morphological and stratigraphic age ranges from the noted fossil taxa (tips). The most appropriate available model in BEAST v.1.8.4, birth-death serial sampling (Stadler 2010), was applied. An uncorrelated relaxed clock  (Drummond, Ho et al. 2006) was separately applied to the molecular and morphological data.

                Each Bayesian analysis was run for 100,000,000 generations with a burn-in of 20%. The analysis was conducted four times to confirm stationarity. The post-burnin samples of all four runs were examined in Tracer 1.7.1 (Rambaut, Drummond et al. 2018) to ensure convergence was achieved. All four runs were combined in LogCombiner, and the consensus tree produced by TreeAnnotator (Drummond, Suchard et al. 2012). The executable .xml file for BEAST, all output log files, and the final consensus tree file (.tree) are available as supplementary information.

Alekseyenko, A. V., C. J. Lee and M. A. Suchard (2008). "Wagner and Dollo: a stochastic duet by composing two parsimonious solos." Systematic Biology 57(5): 772–784.

Drummond, A. J., S. Y. W. Ho, M. J. Phillips and A. Rambaut (2006). "Relaxed phylogenetics and dating with confidence." PLOS Biology 4(5): e88.

Drummond, A. J., M. A. Suchard, D. Xie and A. Rambaut (2012). "Bayesian phylogenetics with BEAUti and the BEAST 1.7." Molecular Biology and Evolution 29(8): 1969–1973.

Gardner, M. G., A. F. Hugall, S. C. Donnellan, M. N. Hutchinson and R. Foster (2008). "Molecular systematics of social skinks: phylogeny and taxonomy of the Egernia group (Reptilia: Scincidae)." Zoological Journal of the Linnean Society 154(4): 781–794.

Goloboff, P. A. and S. A. Catalano (2016). "TNT version 1.5, including a full implementation of phylogenetic morphometrics." Cladistics 32(3): 221–238.

Goloboff, P. A., M. Pittman, D. Pol and X. Xu (2018). "Morphological data sets fit a common mechanism much more poorly than DNA sequences and call into question the Mkv model." Systematic Biology 68(3): 494–504.

Harmon, L. J. (2019). Phylogenetic comparative methods. Published Online, Luke Harmon.

Lanfear, R., P. B. Frandsen, A. M. Wright, T. Senfeld and B. Calcott (2016). "PartitionFinder 2: new methods for selecting partitioned models of evolution for molecular and morphological phylogenetic analyses." Molecular Biology and Evolution 34(3): 772–773.

Lewis, P. O. (2001). "A likelihood approach to estimating phylogeny from discrete morphological character data." Systematic Biology 50(6): 913–925.

Maddison, W. and D. Maddison (2017). Mesquite: a modular system for evolutionary analysis. Version 3.2.

O'Reilly, J. E., M. N. Puttick, L. Parry, A. R. Tanner, J. E. Tarver, J. Fleming, D. Pisani and P. C. Donoghue (2016). "Bayesian methods outperform parsimony but at the expense of precision in the estimation of phylogeny from discrete morphological data." Biology Letters 12(4): 20160081.

Rambaut, A., A. J. Drummond, D. Xie, G. Baele and M. A. Suchard (2018). "Posterior summarisation in Bayesian phylogenetics using Tracer 1.7." Systematic Biology 67(5): 901–904.

Stadler, T. (2010). "Sampling-through-time in birth–death trees." Journal of Theoretical Biology 267(3): 396–404.

Thorn, K. M., M. N. Hutchinson, M. Archer and M. S. Y. Lee (2019). "A new scincid lizard from the Miocene of Northern Australia, and the evolutionary history of social skinks (Scincidae: Egerniinae)." Journal of Vertebrate Paleontology 39(1): e1577873.

Tonini, J. F. R., K. H. Beard, R. B. Ferreira, W. Jetz and R. A. Pyron (2016). "Fully-sampled phylogenies of squamates reveal evolutionary patterns in threat status." Biological Conservation 204, Part A: 23–31.

Wright, A. M. and D. M. Hillis (2014). "Bayesian analysis using a simple likelihood model outperforms parsimony for estimation of phylogeny from discrete morphological data." PLoS One 9(10): e109210.

See README_NambaEgerniinesRunFiles for detailed information.

Phylo files are named accordingly:

Run file: Project_numberofcharacters_numberoftaxa_numberofruns.file

Consensus trees: Project_numberofcharacters_numberoftaxa_CONSENSUS.tree

Log files: Project_numberofcharacters_numberoftaxa_clockmodels_calibrations.log

The file 'Namba_matrices_for_Mesquite.txt' can be opened in the freely available Mesquite software (Maddison and Maddison, 2017), and allows the morphological characters to be traced across the MP tree.

The Excel spreadsheet containing the continuous data (Namba_continuous_data.xlsx) includes the xml codes required to convert the morphological data set for import into xml file for analyses in BEAST.

Maddison, W. and D. Maddison (2017). Mesquite: a modular system for evolutionary analysis. Version 3.2.