Rangel. J. 18(2) 1996, 270-91 SPATIAL DISTRIBUTION OF SHEEP, FERAL GOATS AND KANGAROOS IN WOODY RANGELAND PADDOCKS Jill Landsberg and Jacqui St01 CSIRO Division of Wildlife and Ecology, PO Box 84, Lyneham, ACT 2602 Abstract The densities and distributions of sheep, kangaroos and feral goats were assessed from extensive dung surveys following dry, moderate and green seasons in three large paddocks in the wooded rangelands of north-westem New South Wales. Densities of sheep (219nanimals/km2) were around the long-term district average. Densities of goats (24 animals/km2) were often higher than sheep. Densities of kangaroos (11 animals/km2) were usually much lower than either sheep or goats. Animal density was usually related to vegetative cover (ground cover for sheep and kangaroos, shrub and tree cover for goats), but there were also differences among paddocks. Distribution of kangaroos showed the most differentiation according to vegetation type, with densities being consistently high on a small area of alluvial grassland and very low in the paddock with no alluvial plains and the lowest levels of ground cover. The distributions of sheep and goats were correlated in the dry season and both species showed similar ranges in preferences for different vegetation types. Of the large herbivores present in these woody rangelands, kangaroos were the most selective in terms of the vegetation types they grazed, and goats were the least selective. Because their grazing activities are focussed on alluvial grasslands, kangaroos have potential to degrade this locally uncommon vegetation type. However, the densities of kangaroos in other, more widespread, vegetation types were uniformly low. Goats were frequently the most abundant large herbivores present and were also the least selective. Therefore goats probably have the greatest potential for causing widespread grazing impacts across much of these woody rangelands. Introduction Overgrazing can undoubtedly lead to degradation of rangeland pastures. Pastoralists are generally aware of this and, where possible, adjust numbers of livestock to match perceived carrying capacities of pastures under different seasonal conditions. Livestock are not the only large animals grazing rangeland pastures, however, and are not the sole contributors to overgrazing damage. Kangaroos (Macropus rufus, M. robustus, M. giganteus, M, fuliginosus) are widespread (Caughley 1987) and so too are many feral grazing animals (Wilson et al. 1992). Feral goats (Capra hircus) are particularly abundant in the woody rangelands where sheep (Ovis aries) are the main livestock reared (Southwell and Pickles 1993, Southwell et al. 1993). Sheep, kangaroos and feral goats are all herbivores of comparable body size (Cottle 1991, Strahan 1983) and manipulated field studies have shown that their diets often overlap when they are grazing in common (Dawson et al. 1975, Harrington 1979, Squires 1982, Barker 1987, Dawson 1989, Edwards et al. 1995). In addition, grazing by kangaroos has been shown to impede regrowth of pastures from which sheep have been removed (Grice and Barchia 1992, Norbury et al. 1993). Recognition that both wild and domestic herbivores may contribute to overgrazing of pastures has led to increasing interest in determining their relative contributions to the 'total grazing pressure' exerted on pastures by their combined activities (Tatnell 199 1, Woods 1993). In this paper we explore this issue in the woody rangelands of north-western New South Wales, where the main mammalian herbivores are sheep, kangaroos and feral (Insects such as termites, ants and grasshoppers also contribute to the grazing pressure on rangeland pastures but the factors controlling their populations and distribution are likely to operate at very different scales; they are therefore not considered further in this paper.) Total grazing pressure depends not only upon the density of grazing animals but also upon how their grazing activities are distributed among different vegetation types (Coughenour 199 1 ). Pastoralists generally know the numbers of sheep in paddocks, and aerial surveys provide information about regional densities of kangaroos (New South Wales National Parks and Wildlife Service 1992) and goats (Southwell et al. 1993). However, the two scales of