Search

Abstract: The tension between pyrophytic hummock grasslands and fire-sensitive mulga shrubland is a major ecological theme in central Australia. On a sandplain on the edge of the Tanami Desert, grassland and shrubland are juxtaposed with abrupt boundaries that cut across a subtle environmental gradient. We compare woody plant diversity among vegetation types and explore correlates of the distribution of woody plant species. Alpha diversity in grassland and shrubland was almost identical, but plot-level species richness was greater in grassland. Gamma (landscape) diversity was scarcely greater than alpha diversity. Few long-lived plants, and not even the dominant hummock grass and shrub, were exclusively associated with one vegetation type, but most common species were aligned with elements – soil texture, fire frequency and the extent of physical soil crusting – that characterise the environmental gradient between vegetation types. Most species were resprouters and there was no tendency for obligate seeders or facultative resprouters to be associated with the less fire-prone shrubland. Our data are consistent with the suggestion that hummock grasslands and mulga shrubland are alternative stable states maintained by fire-soil feedback loops and possibly also by competitive exclusion of pyrophytic species by Acacia aneura. [Copyright &y& Elsevier]

The Waite–Nicolson rangeland management method for semi-arid chenopod shrublands predicts that smaller paddocks with medium to moderate stocking rates help to preserve the native vegetation. Vegetation cover around waterpoints in three small paddocks (<2000 ha) from Middleback Station, South Australia was studied using multivariate analysis. Data from quadrats sampled along radiating transects were tested for correlations with a number of site features and grazing history factors. Two significant associations were detected: quadrats with an abundance of Rhagodia parabolica and less palatable species such as Maireana pyramidata, and Atriplex stipitata were correlated positively with proximity to water points, paddock age and stocking rate, and negatively with paddock size. In contrast, quadrats with species such as Rhagodia ulicina and the more palatable M. sedifolia were correlated with increasing distance from the water points and paddock size, but negatively with age and stocking rates. Transect direction was not correlated with either group. Twelve of the 20 species examined, including the important forage species A. vesicaria, also were not correlated with those paddock and grazing features included here. These results suggest that the distribution of some chenopod shrub species in fenced paddocks is still possibly affected by a combination of these factors in the long term by grazing at densities of 6 ha sheep−1 and that the method, although maintaining the fodder species, may not be preserving biodiversity at these grazing levels, although further study is needed. [Copyright &y& Elsevier]

Across nearly 100 sampling locations widely distributed within the arid shrubland of Western Australia, we demonstrated a general relationship between landscape function, primary productivity and rainfall-use efficiency. Sampling locations included landscapes that had been severely ‘degraded’ by more than 100 years of grazing, mainly by sheep. There was generally less phytomass and poorer rainfall-use efficiency on dysfunctional or degraded landscapes than on functional or non-degraded landscapes. Relationships were stronger at broader spatial scales of patch-mosaics than at the scale of individual patches and are likely to be more readily interpreted over decadal rather than yearly time-scales. A-priori assessment of landscape ‘resilience’ provided few insights into the capacity of landscapes to respond to rainfall. Contrary to expectations, herb mass increased on both resilient and non-resilient landscapes as proportional areas occupied by vegetated patches declined. [Copyright &y& Elsevier]