An experimental investigation of differential recovery of native rodent remains from Australian palaeontological and archaeological deposits

WA Museum Records and Supplements | Updated 4 years ago

ABSTRACT – Quaternary palaeontologists are often asked to identify and interpret faunal remains from archaeological excavations. It is therefore vital that both palaeontologists and archaeologists understand the limitations imposed by recovery methods, especially sieves. Fossil small mammals, particularly rodents, in Australian archaeological deposits are very significant sources of palaeoenvironmental information. Over the last half century, recovery techniques used by archaeologists in Australia have ranged from dry screening with 10 mm sieves to wet screening with nested sieves graded down to 1 mm. Nested 6 and 3 mm sieves have been a popular combination. Experimental investigations, using owl-accumulated mammal remains from two caves, show that sieves of 2 or even 3 mm (the metric equivalent of 1/8th inch) are fine enough to recover most of the dissociated complete jaw bones used to identify Australian native rodents (all Muridae). But they fail to retain first molar teeth that have become dissociated from the jaws by pre-depositional fragmentation, predator digestion or damage during excavation. A 1.63 mm sieve (the approximate metric equivalent of 1/16th inch) differentially recovers all isolated first molars of Australian rat-sized rodents, but not small mice, or some first molars of large mouse species. Our results show that differential recovery statistically significantly biases the relative abundances of rodent species retained on a sieve, although there is considerable inter-species variability. The diagonal dimension of the mesh aperture is confirmed to be at least as important as the side dimension in determining what is lost through a sieve. The demographic structure of species can also be biased by differential loss of isolated molars from the youngest individuals, but only in species whose molar dimensions precisely span the threshold of retention on a particular sieve size. Some data suggest that a greater proportion of small objects will be lost through a sieve when immersed in water, than with sprayed water wet sieving or dry screening. Because highly fragmented faunal materials are characteristic of many Australian archaeological deposits, complete recovery of isolated first molars is essential for detection and identification of rodent species to produce the comprehensive assemblages needed for valid comparisons with accumulations by single predators such as owls, and for the multivariate analyses used in palaeoenvironmental interpretations. Complete recovery of first molars of Australian rodents requires a sieve mesh aperture no larger than 1 mm square. A recent archaeological collection made with such methodology provides an opportunity to demonstrate that a more meticulous standard is worth the time and effort to pursue.