A large-scale “experiment” to examine the effects of landscape context and habitat fragmentation on mammals

Abstract

A large-scale “experiment” was undertaken in a 100 000 ha region of south-eastern Australia to examine the response of mammals to landscape context, habitat fragmentation and other factors. The investigation examined the presence and abundance of mammals in three broad categories of sites for which there were strong contrasts in the composition of the surrounding landscape: • Large contiguous areas of Eucalyptus forest, • Areas dominated by exotic plantation softwood Radiata Pine ( Pinus radiata) trees, and • Fragments of eucalypt forest surrounded by an extensive P. radiata plantation. These sites provided the basis for assessing the effects of what we term “landscape context”. Eighty-six fragments of remnant eucalypt forest of varying size, shape, isolation age and other attributes were selected by a stratified, randomization process. Forty sites were located in large contiguous areas of Eucalyptus forest and these were matched to the sites in the remnants on the basis of forest type, geology and climatic conditions. A further 40 sites were selected in areas dominated by P. radiata trees and these also were matched to sites in the remnants and those in large contiguous areas of native forest on the basis of geology and climatic conditions. Two major surveys sampled mammals in the study. Hairtubing (a technique for detecting animals from the analysis of fur collected in a small portable bait station) was utilized at all 166 sites selected in the study. Trapping and a combination of different types of hairtubing was then employed at a subset of 58 sites. Data from these surveys was used to investigate the response of mammals to landscape context, habitat fragmentation and other attributes. A sub-theme of the study was to assess the efficacy of different methods to count mammals. There were large differences in the effectiveness of the different field techniques. The best technique (best in the sense of counting most animals) varied between species, particularly in relation to body size. Trapping and smaller-sized hairtubes were superior for small mammals such as Brown Antechinus ( Antechinus stuartii) and Bush Rat ( Rattus fuscipes). Large hairtubes performed best in the detection of large mammals like the Common Wombat ( Vombatus ursinus), Common and Mountain Brushtail Possums ( Trichosurus vulpecula and Trichosurus caninus), and Swamp Wallaby ( Wallabia bicolor). W. bicolor and V. ursinus showed no response to landscape context and were detected at similar rates in the remnants, sites in large contiguous areas of native forest and sites dominated by stands of P. radiata. Trichosurus spp. were recorded significantly less often in sites dominated by P. radiata trees. Landscape context effects for R. fuscipes and A. stuartii varied depending on the field methods employed to sample mammals. However, in general, R. fuscipes and A. stuartii were recorded significantly less frequently in P. radiata sites than sites in large contiguous areas of Eucalyptus forest or fragments of remnant eucalypt forest surrounded by the softwood plantation. An important finding of our work was that although some species were extremely rare in P. radiata stands, no significant differences were identified in mammal presence and abundance between sites located in large contiguous areas of Eucalyptus forest and sites in fragments of remnant eucalypt forest surrounded by the softwood plantation. This finding suggests that either: animals from potential population sources in contiguous eucalypt forest can move through the softwood plantation and colonise the remnants, or populations residing in the fragments of remnant eucalypt forest are large enough to resist local extinction. Softwood plantations are presently being expanded in south-eastern Australia, particularly on semi-cleared farmland that supports remnant fragments of native Eucalyptus forest and woodland. Our findings indicate that remnant native forest within plantations of exotic P. radiata trees are occupied by several species of native mammals even when these fragments are surrounded by extensive, largely unsuitable plantation forests that have been established for many decades. These fragments should not be cleared during efforts to expand the softwood plantation estate. Large remnants and those with particular habitat features such as a dense cover of vegetation should have priority for exemption from clearing.