The taxonomic identification of wild animal remains in archaeological deposits is crucial to our understanding of paleoecology and past food economies. Postcranial bones (including the skeleton but not the skull) make up a large proportion of paleozoological assemblages and are particularly informative for interpreting human interactions with animals because they provide information on carcass processing and body part utilisation. Delicate craniodental elements, with their wealth of taxonomic information, are often fragmented and dental morphology can become obscured through wear. Researchers often rely on postcrania to understand taxonomic diversity and human subsistence behaviour. However, in regions with high biodiversity such as Australasia, Southeast Asia, South America and the Neotropics, differentiating animals via postcranial bone is challenging owing to the large numbers of closely related and morphologically similar species.
The quantification of bone shape presents itself as a useful solution to taxon differentiation in biodiverse deposits. These morphometric approaches have revolutionised biology, and are routinely used to describe evolutionary, phenotypic and functionally adaptive traits 1–4. Archaeologists have long recognised the utility of a quantitative approach to identifying and exploring shape variation in archeozoological bone, particularly skulls and teeth 5,6. However, such approaches have largely focused on differentiating morphologically similar domesticates such as sheep and goats 7–9, understanding the morphological implications of the domestication process 10–12 and describing economic and culturally driven changes in domesticate body size and skeletal shape 13–15. In contrast, differentiation between wild species in biodiverse deposits still mostly relies on qualitative diagnostic criteria and physical reference collections. Quantitative morphometric approaches could therefore have a transformative impact in Quaternary palaeozoological research, particularly where preservation does not permit the application of proteomic or molecular methods of identification 16–18.
Accessibility is another major hurdle for the adoption of morphometrics in paleozoological research. Geometric morphometrics is a more powerful method for differentiating closely related species but is more complicated and specialised than traditional morphometrics which is more widely accessible 7,12. Functional morphometrics demonstrate that postcranial bone shape of closely related wild taxa can substantially differ in biomechanical adaptation 19–21. In Australia, functional analyses of marsupial postcrania using linear measurements have been very successful 22–24, suggesting that a traditional morphometric approach could also produce an accessible and efficient means of taxonomically classifying marsupial postcrania in archaeological deposits.
Kangaroos of the superfamily Macropodoidea (macropods) are one of the most diverse marsupial groups in Australasia and are a particularly important group for paleozoologists because of their ecological and behavioural diversity. However, researchers are often hesitant to identify macropod postcranial bone beyond the family level owing to the morphological similarity between taxa 25. Over 40 extant macropod species inhabit Australasia, ranging from under 1kg to over 90kg, at least eight species are recently extinct and a number of species are only known paleontologically 26. These herbivores have played an important role in regional ecologies and Indigenous food economies and belief systems across Australasia. They occupy a wide range of habitats, from arid deserts to tropical rainforests, and vary in their social behaviour, feeding preferences and predator avoidance strategies. Understanding macropod diversity in a deposit could therefore provide a plethora of information on local habitat structure, paleoenvironmental change and human hunting and resource management strategies in the past.
Current paleozoological interpretations of macropod remains are symptomatic of the wider challenges of dealing with a biodiverse radiation of species. Macropod postcranial remains which cannot be taxonomically identified are often differentiated into small, medium and large size categories 27–29. Size is clearly an important discriminating characteristic between taxa, but broad size categories without additional taxonomic information have limited use in answering either paleontological or anthropological research questions. The quantification of specimen size may provide more precision for distinguishing between diverse wild species, including macropods, but requires clear and replicable protocols and the collation of comprehensive metric reference data.
Here we apply an easy-to-replicate method and R code, designed for low-cost linear measurements, to quantitatively compare macropod pes (foot and ankle) bones. We quantify pes bone size and shape to explore variation between macropod species and apply discriminant analysis to taxonomically classify unknown specimens. Pes bones were chosen because individuals have only two of each (a left and a right) making them good candidates for providing unbiased quantification of minimum number of individuals (MNI) and examining taxonomic abundance 8. Mammalian feet and ankles contain much locomotor information and a range of evolution-focused studies have successfully identified functionally adaptive variation between some macropod taxa making them ideal candidates for morphometric classification 22,23, 30–32. Pes bones are also composed of compact, dense bone and are consistently found in good condition in paleozoological deposits.
We demonstrate the usefulness of our approach by classifying unknown specimens from an important archaeological site in Western Australia. Boodie Cave on Barrow Island contains the oldest known evidence for human occupation along an arid Pleistocene coastline in Australia 33. Located 60km off the Pilbara coast, the site lies on the edge of the northwest continental shelf and was occupied by people from ca. 51,000 to 7,000 calibrated years before the present (cal BP). Formerly surrounded by sandy plains of the continental shelf, the island was isolated by rising sea levels in the mid-Holocene. Situated 20 degrees south, today the island has an arid tropical climate and is located at the transition between the temperate weather systems to the south and tropical systems to the north, receiving unreliable annual rainfall from both the temperate winter rains and tropical summer monsoon 34.
Archaeological research has demonstrated a long antiquity for the occupation of the northwest coast with deep and continuous connections between the desert and the sea 33,35. Regional patterns in faunal remains indicate a broad diet along this coastline during the terminal Pleistocene along with a shift towards marine resources as sea levels rose and coastline drew closer 28,33,36. Terrestrial animals, particularly macropods, continued to play an important role in these Indigenous maritime-desert economies. Regional cave deposits and historical accounts provide evidence that, until the late nineteenth century, the mammalian fauna of the Pilbara bioregion was much more diverse that is seen today 37,38. Understanding taxonomic diversity within late Quaternary cave deposits therefore provides an important independent line of evidence for ecological change and allows us to develop a more nuanced understanding of Indigenous adaptations to climatic and environmental changes in northwest Australia over the last 50,000 years.