Composition of grazed and cleared temperate grassy woodlands in eastern Australia: patterns in space and inferences in time

Abstract

A regional vegetation survey of the temperate grassy woodlands (temperate savanna) in Australia was designed to assess the effects of clearing and grazing on the composition of vegetation remnants and the adjacent pasture matrix. Vegetation was sampled across a range of habitats using 77 0.1024-ha quadrats; the relative abundance of species was recorded. Classification analysis clustered the sites into three main groups that corresponded to intensity of grazing/clearing followed by groups based on underlying lithology (basalt, metasediment, granites). Using Canonical Correspondence Analysis, exogenous disturbance and environmental variables were related to the relative abundance of species; grazing intensity had the highest eigenvalue (0.27) followed by tree canopy cover (0.25), lithology (0.18), altitude (0.17) and slope (0.10). Based on two-dimensional ordination scores, six species response groups were defined relating to intensity of pastoralism and nutrient status of the landscape. Abundance and dominance of native shrubs, sub-shrubs, twiners and geophytes were strongly associated with areas of less-intense pastoralism on low-nutrient soils. The strongest effects on species richness were grazing followed by canopy cover. Continuously grazed sites had lower native species richness across all growth forms except native grasses. There was no indication that intermediate grazing intensities enhanced forb richness as a result of competitive release. Species richness for all native plants was lowest where trees were absent especially under grazed conditions. Canopy cover in ungrazed sites appeared to promote the co-existence of shrubs with the herbaceous layer. Predicted declines in forb richness in treeless, ungrazed, sites were not detected. The lack of a disturbance-mediated enhancement of the herbaceous layer was attributed to habitat heterogeneity at 0.1 ha sampling scale.