Assessing long-term change in rangeland ecological health using the Western Australian rangeland monitoring system

Abstract

The rangelands or semi-arid and arid regions of Western Australia occupy about 87 percent of the land area. Pastoral grazing of managed livestock, mainly sheep and cattle, occurs over much of this area, with an increasing proportion being allocated to the state conservation estate. Rangeland monitoring began at the local scale in the 1950s and since then has been closely tied to the needs of the pastoral industry. By 1992 a regional-scale, ground-based system was in place after two decades of trialling precursor techniques. The state-wide pastoral monitoring programme, known as the Western Australian Rangeland Monitoring System (WARMS), helps to monitor the state’s natural vegetation and soil resources. Change in soil and vegetation attributes through time, in response to climatic conditions, herbivore grazing, fire and other natural and anthropogenic drivers in the rangelands is known as change in range condition or range trend. When range condition is used in an ecological context, as it is in this research, an improving trend implies an improvement in ecological integrity or ecosystem health. In contrast, a declining trend implies a reduction in integrity, otherwise known as natural resource degradation. The principal objective of this study is to produce a regional-scale, long-term quantitative assessment of range condition change in the southern rangelands of Western Australia, using WARMS transect data. Previous analyses of the WARMS database have examined selected vegetation parameters, but this study is the first to calculate a single integrated range condition index. The assessment covers an area of approximately 760,000 km2, stretching southeast from the southern Pilbara region through the Gascoyne-Murchison and Goldfields regions to the Nullarbor region on the Great Australia Bight.WARMS is designed to provide data and information for assessing regional and long-term changes in rangeland ecological condition. It consists of two principal parts: (1) numerous permanent field monitoring sites and (2) a large relational database. By the end of 2006, there were 980 WARMS sites located on 377 pastoral leases (stations) in the southern rangelands of Western Australia. Average lease size is 202,190 ha and the largest is 714,670 ha. The total area occupied by leases (pastoral plus leases converted to the conservation estate) is approximately 76,250,000 ha. WARMS sites are at an average density of 2.6 sites per lease or 1 site per 77,780 ha of pastoral rangeland. Field-recorded metrics include 11 soil surface parameters and four plant parameters (location on belt-transect, species, height and maximum canopy extent). The field data collection protocol has remained essentially unchanged since 1992 and new field data are captured at each site on a 5-year cycle. This is the most extensive quantitative, ground-based rangeland monitoring system in Australia. This assessment of range condition is based a suite of soil and vegetation indices derived from the WARMS transect field metrics. Seven basic indices have been derived and algorithmically combined into three higher-order indices, one for each of three components of ecological integrity: composition, function and structure. The three indices are then combined into an overall index of ecological health called the Shrubland Range Condition (SRC) Index. In addition, the indices have been assigned to particular time-slices based on the field acquisition date of their component metrics, allowing the calculation of change through time.The combination of the hierarchical index framework, the use of time-slices and GIS mapping techniques provided a suitable analysis platform for the elucidation of spatial and temporal change in rangeland ecological integrity or health at WARMS sites. The nature of change in the SRC Index and the landscape function, vegetation structure and vegetation composition sub-indices has enabled possible causes to be inferred. The patterns of range condition and change are complex at all landscape scales. However, based on analysis of the WARMS sites, range condition is considerably more variable, in space and time, in the northern parts of the southern rangelands compared to the southern parts, with the exception of the Nullarbor region. Through time, the Ashburton and Gascoyne regions consistently demonstrate the largest area (site clusters) of change and the greatest magnitude of change. For many areas, range trend has fluctuated markedly between improvement and decline since the mid-1990s. However, there are two large clusters of sites which show continuing decline through more than two decades. The legacy of historical degradation and ongoing poor land stewardship (principally through over-stocking) is hindering the widespread recovery in range condition, despite more than a decade of good rainfall seasons. An uncommon exception to this sad story is a group of sites located in the upper region of the Gascoyne catchment, where there has been almost continuous improvement over the same period. This work also provides empirical evidence of a fundamental difference in the behaviour of surface water-flows in different catchment types.Using the Landscape Function Factor (LFF), there is conspicuous regional differentiation of sites located in exorheic catchments from those located in endorheic-arheic catchments. In general, sites located in the coastal draining exorheic catchments exhibit greater rates of soil erosion compared to sites located in the other internally draining catchment types; the different erosional regimes are probably related to the nature of the ultimate and local base-levels associated with each catchment type. This has important implications for the long-term management of the rangelands of Western Australia.