Dalal Haouchar’s research while affiliated with Curtin University and other places

Kangaroos and wallabies of the Macropus complex include the largest extant marsupials and hopping mammals. They have traditionally been divided among the genus Macropus (with three subgenera: Macropus, Osphranter and Notamacropus) and the monotypic swamp wallaby, Wallabia bicolor. Recent retrotransposon and genome-scale phylogenetic analyses clarify the placement of Wallabia as sister to Notamacropus, with Osphranter and Macropus branching successively deeper. In view of the traditional Macropus concept being paraphyletic, we undertake to resolve the species-level phylogeny and genus-level taxonomy of the Macropus complex. For the first time, we include nuclear and mitochondrial DNA covering all extant species, and the first DNA sequences from the extinct Toolache wallaby (Notamacropus greyi), which we find groups with the black-gloved wallaby (Notamacropus irma). Morphological variation was examined using geometric morphometric methods on three-dimensional surface-scanned skulls. Wallabia skull shape fell close to Notamacropus (or Thylogale when controlling for allometry). We recommend the subgenera Macropus, Osphranter and Notamacropus be elevated to genera, alongside Wallabia, based on comparisons with other established macropodine genera for cranial disparity, ecology and molecular divergence. Our time tree estimates that all four ‘Macropus’ genera diverged close to the Miocene–Pliocene boundary (~6–5 Mya), then diversified coincident with Pliocene expansion of grasslands in Australia.

The three surviving ‘brush-tailed’ bettong species—Bettongia gaimardi (Tasmania), B. tropica (Queensland) and B. penicillata (Western Australia), are all classified as threatened or endangered. These macropodids are prolific diggers and are recognised as important ‘ecosystem engineers’ that improve soil quality and increase seed germination success. However, a combination of introduced predators, habitat loss and disease has seen populations become increasingly fragmented and census numbers decline. Robust phylogenies are vital to conservation management, but the extent of extirpation and fragmentation in brush-tailed bettongs is such that a phylogeny based upon modern samples alone may provide a misleading picture of former connectivity, genetic diversity and species boundaries. Using ancient DNA isolated from fossil bones and museum skins, we genotyped two mitochondrial DNA (mtDNA) genes: cytochrome b (266 bp) and control region (356 bp). These ancient DNA data were combined with a pre-existing modern DNA data set on the historically broadly distributed brush-tailed bettongs (~300 samples total), to investigate their phylogenetic relationships. Molecular dating estimates the most recent common ancestor of these bettongs occurred c. 2.5 Ma (million years ago), which suggests that increasing aridity likely shaped their modern-day distribution. Analyses of the concatenated mtDNA sequences of all brush-tailed bettongs generated five distinct and well-supported clades including: a highly divergent Nullarbor form (Clade I), B. tropica (Clade II), B. penicillata (Clades III and V), and B. gaimardi (Clade IV). The generated phylogeny does not reflect current taxonomy and the question remains outstanding of whether the brush-tailed bettongs consisted of several species, or a single widespread species. The use of nuclear DNA markers (single nucleotide polymorphisms and/or short tandem repeats) will be needed to better inform decisions about historical connectivity and the appropriateness of ongoing conservation measures such as translocations and captive breeding.

The extent of genetic diversity loss and former connectivity between fragmented populations are often unknown factors when studying endangered species. While genetic techniques are commonly applied in extant populations to assess temporal and spatial demographic changes, it is no substitute for directly measuring past diversity using ancient DNA (aDNA). We analysed both mitochondrial DNA (mtDNA) and nuclear microsatellite loci from 64 historical fossil and skin samples of the critically endangered Western Australian woylie (Bettongia penicillata ogilbyi), and compared them with 231 (n=152 for mtDNA) modern samples. In modern woylie populations 15 mitochondrial control region (CR) haplotypes were identified. Interestingly, mtDNA CR data from only 29 historical samples demonstrated 15 previously unknown haplotypes and detected an extinct divergent clade. Through modelling, we estimated the loss of CR mtDNA diversity to be between 46% and 91% and estimated this to have occurred in the past 2000-4000 years in association with a dramatic population decline. Additionally, we obtained near-complete 11-loci microsatellite profiles from 21 historical samples. In agreement with the mtDNA data, a number of 'new' microsatellite alleles was only detected in the historical populations despite extensive modern sampling, indicating a nuclear genetic diversity loss greater than 20%. Calculations of genetic diversity (heterozygosity and allelic rarefaction) showed that these were significantly higher in the past and that there was a high degree of gene flow across the woylie's historical range. These findings have an immediate impact on how the extant populations are managed and we recommend the implementation of an assisted migration program to prevent further loss of genetic diversity. Our study demonstrates the value of integrating aDNA data into current-day conservation strategies. This article is protected by copyright. All rights reserved.

In 1933, geologist and explorer Michael Terry collected the skull of a small macropodid captured by members of his party near Lake Mackay, western Northern Territory. In 1957, this skull was described as the sole exemplar of a distinct subspecies, Bettongia penicillata anhydra, but was later synonymized with B. lesueur and thereafter all but forgotten. We use a combination of craniodental morphology and ancient mitochondrial DNA to confirm that the Lake Mackay specimen is taxonomically distinct from all other species of Bettongia and recognize an additional specimen from a Western Australian Holocene fossil accumulation. B. anhydra is morphologically and genetically most similar to B. lesueur but differs in premolar shape, rostrum length, dentary proportions, and molar size gradient. In addition, it has a substantial mitochondrial cytochrome b pairwise distance of 9.6-12% relative to all other bettongs. The elevation of this recently extinct bettong to species status indicates that Australia’s mammal extinction record over the past 2 centuries is even worse than currently accepted. Like other bettongs, B. anhydra probably excavated much of its food and may have performed valuable ecological services that improved soil structure and water infiltration and retention, as well as playing an important role in the dispersal of seeds and mycorrhizal fungal spores. All extant species of Bettongia have experienced extensive range contractions since European colonization and some now persist only on island refugia. The near total loss of these ecosystem engineers from the Australian landscape has far-reaching ecological implications.

Highly fragmented and morphologically indistinct fossil bone is common in archaeological and paleontological deposits but unfortunately it is of little use in compiling faunal assemblages. The development of a cost-effective methodology to taxonomically identify bulk bone is therefore a key challenge. Here, an ancient DNA methodology using high-throughput sequencing is developed to survey and analyse thousands of archaeological bones from southwest Australia. Fossils were collectively ground together depending on which of fifteen stratigraphical layers they were excavated from. By generating fifteen synthetic blends of bulk bone powder, each corresponding to a chronologically distinct layer, samples could be collectively analysed in an efficient manner. A diverse range of taxa, including endemic, extirpated and hitherto unrecorded taxa, dating back to c.46,000 years BP was characterized. The method is a novel, cost-effective use for unidentifiable bone fragments and a powerful molecular tool for surveying fossils that otherwise end up on the taxonomic "scrapheap".

List of primers and conditions used for amplifying macropodoid DNA. (PDF)

GenBank accession numbers for the sequences included in the mitochondrial data matrices. (PDF)

jModelTest selections for mitochondrial and nuclear data partitions. (PDF)

Maximum likelihood bootstrap identification of the number of genes required to resolve macropodid phylogeny. (A) for Macropus monophyly and (B) for the M. (Macropus)-M. (Osphranter) grouping. Simulated gene sequences (1,000 bp) were added in increments of five. Ten independent runs were continued until sufficient sequences were added for MLBP >95%. Seven of 10 simulations reached 95% MLBP with 20 genes for Macropus and 35 genes for M. (Macropus)-M. (Osphranter). (PDF)