Relationships between site characteristics, farming system and biodiversity on Australian mixed farms

A collaborative project between researchers, regional natural resource management organisations and landholders set out to explore three questions about the relationships between biodiversity and land use in Australia's mixed-farming landscapes: (1) the extent to which farm-scale measures of biodiversity were related to agricultural production; (2) the influence of the type and intensity of agricultural production on native biodiversity on farms; and (3) the relative influence of site and farming system on selected measures of biodiversity. Four land-use types on 47 mixed farms across nine regions, derived from several of Australia's 56 natural resource management regions, were surveyed in autumn and spring 2006 and 2007. Surveys of birds, surface invertebrates (beetles, ants, spiders), vegetation and soils were undertaken on four land classes on each farm; crop, 'rotation' (break crop/pasture phase), perennial pasture and remnant vegetation. Data were collected by participating regional staff, using a standard protocol, which were sent to a central collection point for collation, analysis and interpretation. Species richness, functional diversity and vegetation structure were assessed. This introductory paper focuses on results relating to species richness, which for most taxa was greater in remnant vegetation than other land-use classes and declined along a disturbance gradient (remnant, pasture, rotation, crop). Properties with a greater proportion of perennial pastures recorded higher species richness than properties that were dominated by crop. Properties that recorded high wheat yields (t/ha) also recorded lower species richness for spiders and birds. The presence of insectivorous birds and beneficial invertebrates (spiders, beetles and ants) in all land-use classes suggests the potential to apply integrated pest management approaches to mixed-farming systems across the country. Site and system features were found to be important determinants of biodiversity, with their relative importance varying with the scale of investigation and the taxa. At the landscape scale, bird species richness was correlated with the amount of remnant vegetation within a 5-km radius of the farm boundary, and with the condition of native vegetation on the farm. The average size of remnant vegetation patches was 5 ha or less on nearly half of the properties surveyed. At the farm scale, ant species richness was correlated with site features, while beetles were correlated with management/system features such as the presence and fertility of perennial pastures. Analyses at the functional group level will provide more detailed information on relationships between different land-use types. Further experimental work needs to be undertaken to qualify the suggested impact of land management practices on different taxa, while repeated surveys will allow for the collation of datasets over time, from which population dynamics may be determined.