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:
Understanding how vegetation composition varies with season and interannual climate variability is important for any ecological research that uses vegetation data derived from surveys for the basis of inference. Misunderstanding this variation can influence land management and planning decisions, leading to poor implementation of biodiversity offsetting mechanisms, for example. We monitored plots (400 m2) grazed by livestock paired with adjacent ungrazed plots in derived native pastures four times a year over 2.5 years on the North-West Slopes of New South Wales. Species density in plots varied greatly with season and interannual rainfall. Highest species density was recorded in spring, though species density in summer was not significantly lower, nor was a spring–summer peak in species density evident in the 2009 drought. Surveys in spring 2008 had the highest species density, and recorded only 60–72% of the total species recorded at each site over 2.5 years. Variation in the proportion of total site diversity represented in combinations of two or three surveys was large, though the best combinations comprised surveys from spring and summer in years of above-average rainfall, either from the same spring-summer, or from different years. Compositional differences among sites were much greater than within sites, showing that differences among sites related to broad environmental gradients were not overwhelmed by seasonal and interannual variability in site composition. When grazing was excluded, there was no evidence of competitive exclusion by the dominant grasses, and no directional shift in composition. The implications of these findings for ecological research depend on the question being addressed: if capturing a large proportion of site diversity is important, then surveys must be carefully timed, or repeat surveys must be conducted. Single surveys did not effectively capture site diversity for use in biodiversity offsetting, and the timing of repeat surveys was critical.