Mammal decline in southern Western Australia – perspectives from Shortridge's collections of mammals in 1904–07

Abstract

Guy Shortridge made extensive collections of mammals in southern Western Australia in the years 1904 to 1907, less than 80 years after first European settlement of the area. This paper reviews Shortridge's unpublished letters and the published papers that resulted from his collections in a reexamination of the temporal and spatial patterns of decline of these mammals. Shortridge collected a monotreme, 25 species of marsupial, 10 species of bat, 8 species of rodent, as well as rabbits and dingoes (a total of 46 species). He collected in five broad regions: mesic woodland sites in what is now the western wheatbelt; mesic coastal scrub sites around Albany; mesic forest habitat in the far south-west corner at Margaret River, Busselton, and Bunbury; semi-arid pastoral sites at Carnarvon and Gascoyne Junction; and semi-arid mining sites at Southern Cross, Kalgoorlie and Laverton. Shortridge was struck particularly by the rich fauna of the mesic woodland and the lack of species in the semi-arid interior (he obtained 20 terrestrial native species at the former and only 3-5 species at the latter). However, since that time, several of the species Shortridge collected at the mesic woodland sites have become extinct on mainland Australia (Bettongia lesueur, Lagostrophus fasciatus, and Perameles bougainville) or are now extinct (Onchogalea lunata) and the region shows one of the highest losses of mammals within Australia. Shortridge's collections emphasise that the loss of mammals in Western Australia has been a two-step process, with an initial loss of mammals in the semi-arid interior, followed by a subsequent loss from mesic woodland and coastal sites. His collections preceded the arrival of foxes Vulpes vulpes to the State and largely preceded the spread of rabbits Oryctolagus cuniculus. Shortridge speculated that disease, predation by feral cats Felis catus, competition from house mice Mus musculus, and bushfires were major factors in overall decline, and the impact of closer settlement was important in localised declines. These, and other factors commonly cited as reasons for mammal decline (grazing and land clearing), are evaluated. The factors suggested by Shortridge for overall decline are potentially interlinked: feral cats or house mice may spread disease, feral cats may be more effective as predators in the presence of abundant house mice populations and at sites where vegetation refuges are reduced by fire, and both may reach highest densities around farms and settlements. The pattern of loss to 1907 can be explained by the spread of feral cats only if their impact on native mammals is high in semi-arid pastoral and mining areas, low in mesic areas, and low in arid areas where European impact was minimal. Possible mechanisms for such a variable impact of feral cats include: i) greater density of vegetation at mesic sites providing refuge for prey species; ii) protection conferred to prey species by poison plants Gastrolobium in the understorey of mesic woodland leading to secondary poisoning of predators; and iii) Aboriginal hunting of feral cats being effective in limiting the density and impact of cats in remote areas beyond the frontier of European colonisation.