Vegetation management, grazing and invertebrate assemblages in the Traprock region

[Summary]: The modification of natural woodlands through tree removal or clearing and a history of livestock grazing are likely to be important factors affecting biodiversity. These variegated landscapes are found in many eastern states where both livestock grazing and tree removal represent significant anthropogenic disturbances on natural woodlands. An understanding of the effects of vegetation management and sheep grazing on a number of components of biodiversity is important to ensure more sustainable management of land units in any region. The broad aim of this research project was to determine if vegetation management and grazing has an effect on arthropod assemblages in the Traprock wool-producing region of southern Queensland. Eighteen study sites were established across the study area according to vegetation type (ironbark/gum woodland and box woodland), and density of mature trees (low [20 trees/ha]), in areas where there was an absence of woody regrowth in the understorey. Sample sites were established randomly within patches and three ‘treatments’ were set up using 6m x 6m exclosures: complete grazing exclosure (2m fully meshed fencing to exclude all vertebrate herbivores); partial exclosure (1m three wire fencing to exclude sheep but allow access for native herbivores such as wallabies and kangaroos) and unexclosed (unfenced plot marked with corner pegs). Three to four pitfall traps (9 cm deep 11cm diameter, plastic food containers) were placed in each exclosure over a period of 3 days in April 2005 and opened at the same time for a period of 6 days. All arthropods were identified to order level, except for Formicidae, which were initially taken no further than family level and collembola and myriapods, which were only identified to the level of class. One-way analysis of variance (ANOVA) was used to test for differences in arthropod abundance and taxon richness. Two diversity indices were calculated for the order level data: Simpson and Margalef. Patterns in arthropod abundance were examined using multidimensional scaling (MDS) and analysis of similarity (ANOSIM). Tree density was found to have a significant impact upon the abundance, richness and diversity of arthropods collected in pitfall traps in the Traprock region of southern Queensland. Arthropod abundance was found to be significantly greater in low tree density sites than in medium or high tree density sites. Arthropod richness (at the taxonomic level of order) and diversity were found to be significantly greater in medium and high tree density sites compared to low tree density areas, although no significant difference was found for abundance, richness or diversity between medium and high tree density sites. Vegetation type and landscape position was found to have little effect upon arthropod abundance, richness and diversity, although this may have been be a function of the coarse taxonomic resolution employed rather than a lack of assemblage distinction between the sites. These and other findings are discussed in relation to ecological theory, management recommendations and the practicalities of using arthropods as barometers indicators of system change and condition