Austral Ecology (2003) 28, 526–538 Impacts of grazing and burning on spider assemblages in dry eucalypt forests of north-eastern New South Wales, Australia REBECCA HARRIS,* ALAN YORK † AND ANDREW J. BEATTIE Key Centre for Biodiversity and Bioresources, Department of Biological Sciences, Macquarie University, Sydney, New South Wales, Australia Abstract In the dry eucalypt forests of north-eastern New South Wales, Australia, cattle grazing occurs at low intensities and is accompanied by frequent low-intensity burning. This study investigated the combined effects of this management practice on the ground-dwelling and arboreal (low vegetation) spider assemblages. Spiders were sampled at 49 sites representing a range of grazing intensities, using pitfall trapping, litter extraction and sweep sampling. A total of 237 spider morphospecies from 37 families were collected using this composite sampling strategy. The abundance, richness, composition and structure of spider assemblages in grazed and ungrazed forest sites were compared and related to a range of environmental variables. Spider assemblages responded to a range of environmental factors at the landscape, habitat and microhabitat scales. Forest type, spatial relationships and habitat variability at the site scale were more important in determining spider assemblages than localized low-intensity grazing and burning. However, it is possible that a threshold intensity of grazing may exist, above which spiders respond to grazing and burning. Although low-intensity grazing and burning may not affect spider assemblages below a threshold stocking rate, that stocking rate has yet to be established. Key words: assemblage structure, grazing, pitfall trapping, spiders, sweep sampling. INTRODUCTION Cattle grazing is widespread throughout the public forests of Australia, and in many forests forms an important part of the forest management strategy. Cattle grazing is considered to have significant benefits through the reduction of fuel loads and provision of grazing lease revenue (SFNSW 1995a). In some regions, the holders of grazing leases may also be provided with a burning permit to promote grass growth and improve the quality of fodder for cattle. Grazing in these regions is therefore linked to frequent low-intensity (< 500 kW m –1 ) burning regimes (Moore & Floyd 1994; SFNSW 1995b). Much of the research on cattle grazing in Australia has concentrated on the impacts of heavy grazing in agricultural lands or on low to moderate grazing in semiarid grasslands (King & Hutchinson 1983; Scougall et al. 1993; Fensham et al. 1999). Where research has been carried out on low-intensity grazing in forests, it has focused on the response of vascular plants and vertebrates. These studies have shown that cattle grazing affects habitat structure in Australian woodlands and forests through a reduction in under- storey vegetation (Williams 1990; Pettit et al. 1995; Henderson & Keith 2002), leading to changes in vertebrate species richness and composition (Smith et al. 1994) and modifications to the litter and ground microclimate (Friedel & James 1995; Yates et al. 2000). All of these effects may be exacerbated by associated changes to the fire regime, as frequent burning has also been shown to cause a reduction in vegetation diversity and structure (Gill 1981; Fox & Fox 1986; Moore & Floyd 1994) and to affect the physical and chemical properties of soil and litter (Springett 1976; Majer 1984; York 1996, 1999a). The impact of grazing and burning on forest- dwelling invertebrate communities remains largely unknown. Studies in heavily grazed grasslands and woodlands have linked reduction in vegetation and litter cover to changes in the structure and composition of invertebrate communities (Abensperg-Traun et al. 1996; Bromham et al. 1999). These effects are known to increase with grazing pressure (King & Hutchinson 1980, 1983), but the impact of low-intensity grazing on invertebrates is less well understood. Spiders have been suggested as a potential priority group for the assessment of ecological disturbance because they are diverse, abundant and sensitive to relatively small changes in habitat structure (reviews by Turnbull 1973; Uetz 1991; Wise 1993). Several studies have shown that the abundance, species richness and *Corresponding author. Present address: Centre for Bio- diversity and Conservation Research, The Australian Museum, 6 College Street, Sydney, New South Wales 2010, Australia (Email: rebeccah@austmus.gov.au). † Present address: Forest Science Centre, University of Melbourne, Natural Resources and Environment, Creswick, Victoria, Australia. Accepted for publication April 2003.