Description:
Plant diversity is threatened in many agricultural landscapes. Our understanding of patterns of plant diversity in these landscapes is mainly based on small-scale (<1000 m2) observations of species richness. However, such observations are insufficient for detecting the spatial heterogeneity of vegetation composition. In a case-study farm on the North-West Slopes of New South Wales, Australia, we observed species richness at four scales (quadrat, patch, land use and landscape) across five land uses (grazed and ungrazed woodlands, native pastures, roadsides and crops). We applied two landscape ecological models to assess the contribution of these land uses to landscape species richness: (i) additive partitioning of diversity at multiple spatial scales, and (ii) a measure of habitat specificity – the effective number of species that a patch contributes to landscape species richness. Native pastures had less variation between patches than grazed and ungrazed woodlands, and hence were less species-rich at the landscape scale, despite having similar richness to woodlands at the quadrat and patch scale. Habitat specificity was significantly higher for ungrazed woodland patches than all other land uses. Our results showed that in this landscape, ungrazed woodland patches had a higher contribution than the grazed land uses to landscape species richness. These results have implications for the conservation management of this landscape, and highlighted the need for greater consensus on the influence of different land uses on landscape patterns of plant diversity.
Description:
AimsThe effects of pastoralism on grassy vegetation are predicted to vary across spatial scales. However, local-scale effects have generally been better considered than large-scale effects that may be pertinent for plant diversity conservation. We examined variation in plant diversity in an agricultural region by comparing patterns observed at large and local scales for different land uses. We tested whether habitat specificity (which relates to how specific different species are to particular land uses) can indicate the importance of land uses to plant diversity conservation. MethodsWe surveyed 127 quadrats of grassy vegetation across four land uses: previously cultivated and never-cultivated native pastures, grazed woodland and ungrazed woodlands (grazing refugia). We assessed (1) species density at the quadrat scale, (2) correlations among explanatory variables and species composition using ordination techniques, and (3) habitat specificity at the regional scale using the -diversity contribution index. ResultsAt the quadrat scale, differences in species density among land uses were not pronounced. In contrast, habitat specificity was highest in ungrazed, unfertilized woodlands. Land uses that were grazed, cleared of trees and fertilized were generally species dense at the quadrat scale (>40 species 400m(-2)), but were comprised of a limited suite of native and exotic grasses and forbs, with relatively homogeneous composition across the region. ConclusionsIn the region studied, grazing refugia with no tree clearing or history of fertilizer application were valuable for conserving a large number of native species, and so should be protected. Such land uses are currently rare in the region. Simple measures of native or total species density were insufficient for evaluating conservation value, as they did not distinguish habitats with large numbers of common, generalist native species from those with species that were rare or restricted in the region. Analyses of habitat specificity may potentially overcome the pitfalls of basing conservation management decisions on local-scale measures. Our results also highlighted a need to understand the processes associated with management practices that serve to reduce species diversity at landscape and regional scales.