Open Access

Madeleine O. Rankin, William S. Semple, B. W. Murphy, Terry B. Koen

• The assumption that ‘soils’ and ‘vegetation’ are closely associated was tested by describing soils and vegetation along a Travelling Stock Reserve west of Grenfell, New South Wales (lat 33° 55’S, long 147° 45’E). The transect was selected on the basis of (a) minimising the effects of non-soil factors (human interference, climate and relief) on vegetation and (b) the presence of various soil and vegetation types as indicated by previous mapping. ‘Soils’ were considered at three levels: soil landscapes (a broad mapping unit widely used in central western NSW), soil types (according to a range of classifications) and soil properties (depth, pH, etc.). ‘Vegetation’ was considered in three ways: vegetation type (in various classifications), density/floristic indices (density of woody species, abundance of native species, etc.) and presence/absence of individual species. Sites along the transect were grouped according to soil landscapes or soil types and compared to vegetation types or indices recorded at the sites. Various measures indicated low associations between vegetation types and soil landscapes or soil types. Except for infrequent occurrences of a soil type or landscape, any one soil type or landscape was commonly associated with a number of vegetation types and any one vegetation type was associated with a number of soil landscapes or soil types. However, significant associations between some vegetation indices, mainly density or numbers of woody species, and some soil landscapes and soil types were evident. Although many species were relatively ubiquitous, some groups of species that were restricted to one or two soil types were identified. Canonical Correspondence Analysis provided some suggestions as to which properties (e.g. texture) of these soils were associated with the presence of particular species.