Bettongia gouldii, Waterhouse, 1846

Bettongia gouldii

Only a few specimens were found that could represent individuals of Bettongia gouldii . As the holotype is a juvenile specimen with DP2, DP3, M1 and an M2 that is still in crypt, these specimens, which had worn dentition, could not be reliably compared. When performing the PCA and LDA analysis, it was instead determined that our identified specimens of Bettongia gouldii were representatives of other taxa. Only one specimen was not reliably identified as being from other taxa, that being specimen SAMA M8554, a subfossil skull of a subadult specimen from Cleve, South Australia. When examined in a confusion matrix, SAMA M8554 was identified as B. gaimardi , despite this species not being known from this area. Additionally, when allometry was removed from the PCA and LDA analyses, SAMA M8554 grouped with Bettongia penicillata . While SAMA M8554 does display characters that are different from B. penicillata , such as smaller and less inflated bulla, the authors cannot confidently say that SAMA M8554 is a representative of B. gouldii , or that B. gouldii is a valid taxon. For this reason, the only way to resolve this issue is with a genetic evaluation of the type specimen against other Bettongia . As B. gouldii was never formally synonymised and our investigation of this taxon is inconclusive, we believe that the taxon can be formally classified as nomen dubium.

Phylogenetic analysis

The results of our phylogenetic analysis were inconclusive, with all taxa within Bettongia forming a polytomy. This resulting matrix was also unable to separate Macropodidae from Potoroidae indicating that more research is needed into the taxonomy and evolutionary history of Potoroids. One key cranial character that is used for differentiating Potoroidae and Macropodidae is that in the former the squamosal bone contacts frontal bone, unlike in macropodids where a parietal and alisphenoid contact is present ( Seebeck and Rose 1989). However, in this investigation it was determined that in some Bettongia ( B. haoucharae , B. o. ogilbyi , B. o. sylva, and B. o. odontoploica) the inverse is true. Additionally, fossil species of Potoroids are known to be problematic as the taxonomy is currently unresolved. Initially the fossil taxa were included in this investigation, particularly in the phylogenetic tree, but were removed as we were unable to construct a conclusive phylogeny. Previous investigations have also noted that the Potoroids need work ( Flannery & Archer 1987; Cooke et al. 2015; Travouillon et al. 2016; Travouillon et al. 2022), with the taxonomy of several taxa unresolved within the family. A full taxonomic revision of Potoroidae is considered out of scope for this investigation.

Conservation implications

As a result of the taxonomic revision presented in this investigation, Bettongia penicillata is considered extinct. Additionally, the taxonomic changes will have a myriad of effects on the current conservation practices for Bettongia ogilbyi , and its subspecies. As mentioned previously, Bettongia ogilbyi ogilbyi is critically endangered, with a single wild population from Tutanning Nature reserve remaining. Based on museum specimens, the decline of B. o. ogilbyi likely began in the 1960s, with fewer specimens recorded in the WAM. The exact cause of the decline is still unknown. This also indicates that the current populations of Woylies at Perup, Kingston, and Dryandra are Bettongia ogilbyi sylvatica . Likely the Dryandra population of B. o. sylvatica was introduced during an attempt to increase gene flow between the populations ( Pacioni et al. 2013). Bettongia ogilbyi sylvatica is critically endangered but recovering due to current conservation efforts. Based on available locality data of each taxon (Supp. 4), B. o. ogilbyi occurred in open woodland to scrub habitats, while B. o. sylvatica occurred in denser woodland and forest habitats such as the Jarrah Forest (pers. obs. Newman-Martin & Travouillon). Referring to museum material, B. o. ogilbyi could be found in western Australia, all the way from the southwest to the Roe Plains. Bettongia ogilbyi ogilbyi could also be found far north into Shark Bay (Fig. 36). Museums specimens of B. o. sylvatica have been found to not occur past Esperance to the east or Yanchep to the north (Fig. 36). Both species were found in sympatry on the Swan Coastal Plain and into the Perth Hills.

FIGURE 36. Distribution of the Bettongia penicillata complex based on the results of this taxonomic review. Distribution map was created using known modern localities as well as verifiable historic and subfossil specimens. Note that temporal changes are not shown, and the distribution of taxa would have changed through time with changes to the environment.

With the advent of two subspecies within B. ogilbyi , another factor to consider is outbreeding depression. Outbreeding depression occurs when two populations of a taxa are introduced in an attempt to maximise genetic diversity. When the populations reproduce, the alleles from the introduced population may not aid the offspring in survival, and instead lead to a loss of fitness ( Frankham et al. 2011). According to Frankham et al. (2011) the probability of outbreeding depression occurring between two populations increases when the populations have at least one of the following characteristics: are distinct species, have fixed chromosomal differences, have exchanged no genes in the last 500 years, or inhabit different environments. As stated previously, the B. ogilbyi subspecies are found in different habitats from one another, with sympatry only occurring on the Swan Coastal Plain and Perth hills. Previously, outbreeding depression has been suggested to be unlikely for B. ogilbyi , due to the species belonging to a single wide range and adapted to many habitats ( Pacioni et al. 2018). However, our investigation indicates that B. ogilbyi was not adapted for many habitats and did not occur over a wide range (Fig. 36) as has previously been suggested. For this reason, the authors believe that outbreeding depression cannot be ruled out and caution should be taken when translocating and mixing populations of B. ogilbyi . Whilst it is likely that Bettongia ogilbyi ogilbyi is functionally extinct it is possible that during translocation, some hybrid individuals with Bettongia ogilbyi sylvatica were produced.

Previous translocations within the indigenous southwest populations (Perup, Kingston, Dryandra woodland and Tutanning nature reserve) have occurred ( Pacioni et al. 2013). However, the effects of hybridising the two Bettongia ogilbyi subspecies are yet unknown. It is also possible that an outbreeding depression was the cause of the original population crash of B. ogilbyi in the early 2000s, however, without more material this is just speculation. Additionally, current reintroduction projects across Australia must be labelled as introductions, where B. ogilbyi is being used as an ecological replacement for locally extinct Bettongia taxa. The findings presented here could also help to explain why some reintroduction attempts have failed while other have succeeded. For instance, in this investigation we identified B. ogilbyi specimens from St. Peter Island (NMV C32743 and NMV C32742), which has seen a successful reintroduction ( Start et al. 1995). However, B. ogilbyi released on St. Francis Island failed to establish ( Start et al. 1995). The subspecies B. o. francisca was identified from this island. This indicates that B. o. francisca was specifically adapted to the conditions of St. Francis Island.