The Golden Bandicoot, Isoodon auratus Ramsay 1887, in western New South Wales during European times M. Ellis1, P. Wilson1 and S. Hamilton1 'RZS Mammal Section, P.O. Box 20, Mosman 2088 In September 1988,while investigatingthe occurrence of the nests of sticknest rats (Leporillus sp.) in the eastern escarpment of the Gap Range at Mootwingee National Park (31°3'S 142"36'E), a small deposit of owl pellets was found. The pellets were on the floor of a small cave (approximately one metre cubed) and had been partly compacted into the floor debris by goats Capris hircus. The more intact pellets had a black coating, indicating that they were deposited by a species of masked owl, Tyto sp. (Smith and Cole 1989). A number of loose bones, mainly skulls and dentaries, were also located in the floor debris of the cave. One partially intact pellet contained the skull of a polyprotodont marsupial. Upon extraction it was determined to be a virtually complete juvenile Isoodon skull. Based on the relative size of the bullae to skull length (345 mm3 to approximately 46.5 mm), the angle and width of the ramus, and the long, f i e angular process, the skull most closely agrees with the Golden Bandicoot I. auratus as reviewed by Lyne and Mort (1981). The specimen has been lodged with the Australian Museum. Other bandicoot dentaries found in the debris have been tentatively assigned to the Western Barred Bandicoot, Perameles bougainde Quoy and Gaimard 1824. The Golden Bandicoot has been recorded from owl pellets in the Flinders Ranges, South Australia, regarded as being of 18th to 19th century origin (Smith 1977). Assuming that the material from Mootwingee is of the same age because of its similar state of preservation, this is the first definitive record of the Golden Bandicoot in New South Wales in historical times. The documented decline of the species shows that it rapidly became extinct across much of its range. It was last recorded in the northwest of South Australia in 1933 (Copley et a/. 1989), the Top End of the Northem Territory in 1912 (McKenzie 1983), and in the centre of the Northern Territory in 1952 (Burbidge et al. 1988). ln light of this rapid decline it is not unreasonable that the species occurred in western New South Wales at the time of European penetration of the region (1830-1860), but rapidly became extinct. Figure 1 shows the distribution of the Golden Bandicoot and Southern Brown Bandicoot I. obesulus from a number of sources. Both owl pellet sites are closer to known localities for I. obesulus than for I. auratus. Krefft (1866) found what he called I. obesulus to be common at the junction of the Murray and Darling rivers, but it soon became extinct. Unfortunately, no specimens from this area were used by Lyne and Mort (1981) in their review of skull morphology in Isoodon. Confusion over the identity of the Isoodon from the Murray Darling junction has continued until the present, with the animals being assigned to I. obesulus, 1. auratus or just lsoodon sp. in Seebeck et al. (1990). The terrain between the Mootwingee site and the junction of the Murray and Darling Rivers is similar to the areas where I. auratus has been collected, rather than to Figure 1. A map of Australia showing the distribution of the Golden and Southern Brown Bandicoots. tight stippling shows the Southern Brown Bandicoot distribution and coarse stippling shows the Golden Bandicoot distribution according to Strahan (1983). Other records for the Southern Brown Bandicoot are shown as solid squares (Lyne and Mort 1981) and open squares (Morton and Baynes 1985). Other records for the Golden Bandi- coot are shown as solid diamonds (Lyne and Mort 1981), open diamonds (Morton and Baynes 1985) and solid triangles (Burbidge et a/. 1988). Krefft's (1866) record is shown as a solid star. The subfossil record for the Flinders Ranges is shown as an open star, and for this study as an open circle. 36 Australian Zoologist, Vo1. 27(1 G2) June 1991