The Caloprymnus campestris mitogenome (16,866 bp) is ordered with 13 protein-coding genes, two ribosomal (r)RNA genes, 21 transfer (t)RNAs, and a non-coding AT-rich control region, which follows the typical configuration for marsupials39,40. The tRNAs are arranged around the origin of the L strand (A-C-W-OL-N-Y) and intersected between the NADH2 and COX1 genes. Substitution of the anticodon GCC for tRNAASP (trnD) is also consistent with RNA-editing41.
Maximum likelihood and Bayesian analyses of our mitogenome dataset produce unanimous resolution of Macropodoidea with Potoroidae as the sister to Macropodidae (Supplementary Figures S1–S6). This pivotal higher-level grouping is consistent with other crown macropodoid phylogenies12,42–45, and warrants a new taxonomic definition46, which we coin as Macropodia, new clade, herein (Table 1; Supplementary Information). Bootstrap and BPP support is >90% for almost all constituent nodes except those uniting: (1) the extinct short-faced kangaroo, Simosthenurus occidentalis, with the Banded hare-wallaby, Lagostrophus fasciatus, as basally branching macropodids (partitioned/non-partitioned bootstrap = 58/63%; MrBayespartitioned/non-partitioned BPP = 0.54/0.56; BEASTpartitioned/non-partitioned BPP = 1/1); (2) the Quokka, Setonix brachyurus, with other macropodines (bootstrap = 48/60%; MrBayes BPP = 1/1; BEAST BPP = 0.63/0.72); (3) grey kangaroos in the genus Macropus with Osphranter rufus and brush wallabies representing the genus Notamacropus (bootstrap = 80/64%; MrBayes BPP = 0.99/1; BEAST BPP = 0.99/0.96); and (4) O. rufus with Notamacropus (bootstrap = 58/55%; MrBayes BPP = 0.81/1; BEAST BPP = 0.72/0.76). As found by previous studies5,12,27,42–47, Potoroinae comprises potoroos within the genus Potorous and is distinguished from its sister clade, which we designate Bettonginae48 to include the Rufous bettong, Aepyprymnus rufescens,as the basally branching sister to C. campestris and the species of Bettongia (Table 1). Alternative monophyly of C. campestris with either A. rufescens49,50, or the species of Potorous27,46 were tested using topological constraints in PAUP* 4.0b1051 (Supplementary Table S4), but decisively rejected (P< 0.0001***). Taxonomically, therefore, we conclude that the original classification of Gould’s Desert ‘bettong’19 as generically consistent with Bettongia is feasible, but defer any formal nomenclatural amendment pending a detailed morphological re-evaluation.
Our maximum likelihood, Bayesian and time-tree analyses of the nDNA (Supplementary Figures S7–S9) and combined mitogenome/mtDNA/nDNA datasets (Figure 2; Supplementary Figures S10–S12) yield broadly compatible topologies, with the basal divergence of potoroids and macropodids, and subsequent split between potoroines and bettongines both occurring during the early Miocene (Table 2; Supplementary Table S4). Notably, this concurs with divergence estimates derived using different dating methods and constraints12,42–46,52. Furthermore, while our S-DIVA and BBM ancestral area optimisations correlate the early Oligocene (or late Eocene based on nDNA41; Supplementary Table S4) emergence of crown macropodoids with predominantly humid wooded habitats53, the earliest radiation of potoroids, together with members of the macropodid subclades Sthenurinae and Lagostrophinae + Macropodinae, all coordinate with dispersals into shrubland and woodland-forest mosaics (Figure 2; Supplementary Figures S13 and S14). These potentially included ‘mallee-like’54 sclerophyll communities, which propagated throughout central Australia from the early to middle Miocene53.
The globally recognised55 middle to late Miocene climatic transition from equable to increasingly cool, dry conditions53,56 coincides with potoroine speciations in mesic environments throughout southern Australia27,52. These are tracked by our S-DIVA and BBM optimisations, which infer occupation of primarily wooded habitats after the early-late Miocene (Supplementary Figures S13 and S14). This is also concurrent with the incipient desertification of inland Australia57, which may have promoted genetic segregation of Potorous platyops and Gilbert’s potoroo, Potorous gilbertii, in central-southern58 and southwestern Australia, versus the Long-nosed potoroo, Potorous tridactylus, and basally branching Long-footed potoroo, Potorous longipes, in southeastern Australia52. Additionally, we show that regional subspecies distinctions within P. tridactylus were completed by the early to middle Pleistocene (Table 2; Supplementary Table S5). Curiously, though, Cyt b K2P variation (Supplementary Table S6) implies substantially less genetic difference between the Tasmanian P. tridactylus apicalis and northeastern mainland P. tridactylus tridactylus (1.93%), in comparison to the southeastern mainland P. tridactylus trisulcatus (4.21%). Indeed, these values approximate those contrasting P. tridactylus tridactylus/P. tridactylus trisulcatus with P. gilbertii (2.69/5%), P. platyops (4.1/5%), and P. longipes (5.84/5.69%), supporting inferences of cryptic taxa52, but in our opinion, only up to species-level.
Despite currently limited DNA sequence coverage for the extinct Finlayson’s59 Desert bettong, Bettongia anhydra60, we derive unequivocal support (Figure 2; Supplementary Figures S1–S12) for the monophyly of Bettongia spp. (bootstrap = >90%; BPP = 1), together with close relationships between the woodland-forest dwelling Eastern bettong, Bettongia gaimardi, Northern bettong, Bettongia tropica, and Brush-tailed bettong, Bettongia penicillata penicillata (bootstrap = >99%; BPP = 1). Only a few hundred Cyt b (or control region) nucleotides are available for the Woylie, Bettongia penicillata ogilbyi61. Nevertheless, our S-DIVA and BBM optimisations suggest seminal late Miocene divergences of Bettongia spp. in open xeromorphic habitats (Supplementary Figures S13 and S14), followed by a Pliocene to as recent as middle Pleistocene radiation of B. gaimardi + B. tropica + B. penicillata subsp. (Table 2; Supplementary Table S5) in conjunction with mesic habitat variegation (Supplementary Figures S13 and S14). We attribute this to vicariant ‘reversions’5 in eucalypt woodlands and forests62–64, which contracted and fragmented with intensifying aridification over the Pliocene–Pleistocene interval65.
Bettongia is karyotypically conservative, retaining the 2n=22 chromosomal number of most macropodoids66,67. Conversely, chromosomal fission in P. longipes has produced 2n=24, while fusions (and inversions) in P. tridactylus and P. gilbertii manifest unusual reductions to 2n=12♀, 13♂68. Aepyprymnus rufescens, on the other hand, exhibits a unique karyotypic increase to 2n=32, which is the highest for any marsupial68, and presumably reflects its independent evolution since the middle or early-late Miocene (Table 2; Supplementary Table S5). Although the chromosomal arrangement of C. campestris is unknown, our robustly supported (bootstrap = >97%; BPP = 1) later-middle to early-late Miocene split from Bettongia (Table 2; Supplementary Table S5) suggests a similarly protracted ancestry, yet with genetic differentiation that approaches intrageneric levels within Bettongia spp. (Cyt b K2P variation being as little as 6.91% compared to B. penicillata: Supplementary Table S6). Significantly, our S-DIVA and BBM optimisations correlate the C. campestris-Bettongia divergence with an invasion of open-xeric environments (Supplementary Figures S13 and S14), perhaps incorporating arid chenopod shrublands and gallery woodlands that spread across central Australia from the late Miocene53,54,57. The coeval radiation of macropodines is otherwise linked to disparate woodland-forest and shrubland settings (Supplementary Figures S13 and S14), with speciation among dorcopsins and dendrolagins in humid forests coincident with uplift of the New Guinean landmass3,69,70, and macropodins in schlerophyllous woodlands, chenopodiaceous shrublands, and xeric grasslands (Supplementary Figures S13 and S14), which proliferated after the late Miocene–Pliocene53 coincident with the prevalence of grazing kangaroos9.