Assessment and monitoring of vegetation condition: Moving forward
All notes have been peer-reviewed
Assessment and monitoring of vegetation condition: Moving forward. Sue V. Briggs1 and David Freudenberger2 (1NSW Department of Environment and Conservation, c/- CSIRO Sustainable Ecosystems, GPO Box 284, Canberra, ACT 2601, Australia. Email: firstname.lastname@example.org; 2CSIRO Sustainable Ecosystems, GPO Box 284, Canberra, ACT 2601, Australia).
Key words: vegetation condition assessment, vegetation condition mapping, vegetation condition monitoring.
Introduction. The goals of the ‘Mapping Vegetation Condition in the Context of Biodiversity Conservation Workshop’ were to: (i) examine how mapping vegetation condition can contribute to improved management of biodiversity; (ii) examine methods for mapping vegetation condition at different scales; (iii) identify emerging technologies for mapping vegetation condition; and (iv) provide an overview of conceptual approaches to vegetation condition.
In this paper we: (i) review key insights from the Workshop; and (ii) discuss goals and approaches for assessing and for monitoring vegetation condition.
Vegetation condition does not have a standard definition (Gibbons & Freudenberger 2006), and the term was not defined prescriptively for the Workshop. Vegetation condition usually refers to health or quality of vegetation or other ecosystem element, such as riparian areas (Gibbons & Freudenberger 2006; Keith & Gorrod 2006). People usually interpret vegetation condition through a ‘values prism’ according to their values and goals (Gibbons & Freudenberger 2006). Good condition for one purpose may be poor condition for another (Parkes & Lyon 2006).
Vegetation condition can be assessed and monitored at a range of scales from site to regional and broader, depending on one's goals. We distinguish between the two general goals of: (i) assessment of vegetation condition; and (ii) monitoring of vegetation condition. Assessment is determining how much of something exists, usually in a defined area. Spatial assessment of vegetation condition is often displayed by mapping, which provides a snapshot in time. Monitoring is determining change or trend in something, usually over time.
Assessing vegetation condition. Vegetation condition can be assessed at a range of scales from site to regional to identify priorities for incentive funding, to analyse options for management, and to stratify sites for monitoring (Parkes & Lyon 2006; Thackway & Lesslie 2006). At regional scales, vegetation condition can be predicted and mapped to provide priorities for investment. For example, condition maps show that small patches of remnant vegetation on lower slopes and valleys are in poor condition (e.g. Zerger et al. 2006), and they need considerable investment to conserve and enhance them. In contrast, large remnants on upper slopes are in relatively good condition, and are a lower priority for conservation investment.
The Workshop demonstrated that vegetation condition could be assessed and mapped at regional scales with reasonable accuracy using remote sensing and/or abiotic variables (Drielsma & Ferrier 2006; Higgins 2006; Newell et al. 2006; Thackway & Lesslie 2006; Wallace et al. 2006; Zerger et al. 2006). Spatial models or maps of vegetation condition should be related to vegetation condition in plots on the ground, which in turn should be related to benchmarks for desirable states of vegetation condition (Ayers et al. 2005; Gibbons & Freudenberger 2006; Parkes & Lyon 2006). Site assessment, spatial modelling and remote sensing are complementary tools for vegetation condition assessment.
Monitoring vegetation condition. Burgman (2005) regarded monitoring as sampling and analysis to determine consistency with a standard or deviation from a target. He argued that monitoring should include both inputs and outputs. For vegetation condition, inputs could include rainfall and grazing pressure and outputs would be the measured condition variables. Nicholls and Wallace (2003) argued that the primary purpose of monitoring is to detect change over time, focusing on long-term changes with management.
Vegetation condition is traditionally monitored at fixed plots on the ground where condition variables are measured at regular intervals to determine changes over time.
The goals of vegetation monitoring are usually to determine trends in vegetation condition with changes in management such as grazing or fire, and with weather, particularly rainfall. Vegetation condition is monitored more commonly in rangelands than other vegetation communities (e.g. Bastin & Ludwig 2006). In rangelands, the goal is often to monitor responses of vegetation condition to rainfall under different stocking rates. In the temperate areas of southwestern and southeastern Australia we might monitor change in condition in vegetation remnants after providing incentives to control grazing or reduce firewood removal.
If the goal is to monitor trend in vegetation condition over time with rainfall or other weather variables, then stratified random plots over the area of interest are required. If the goal is to monitor trend in vegetation condition over time with grazing or other management, then plots are required with different grazing levels. If the goal is to monitor interactions between rainfall and grazing, then responses to rainfall at plots in different condition from different grazing regimes should be monitored. Bastin and Ludwig (2006) showed that responses of vegetation to rainfall could be separated from responses to grazing pressure by comparing the growth of vegetation after the same rainfall event on heavily grazed and less grazed sites.
Remote sensing can be used to measure and monitor coarse changes in vegetation condition across large areas, or in patches within larger areas (Bastin & Ludwig 2006; Stone & Haywood 2006). Coarse changes in condition can include extent and frequency of large fires, recovery from fire, dieback, or vegetation vigour (greenness) following large rainfall events. Wallace et al. (2006) showed that remotely sensed data (e.g. Landsat) could detect changes in forest health across large areas of southern Western Australia. Monitoring at sites on the ground is required to detect subtle changes in vegetation condition such as species composition, seed production or recruitment with management changes such as reduced grazing or burning individual patches. Site assessment and remote sensing are both tools for monitoring vegetation condition; sometimes they are complementary.
Monitoring can be undertaken at regional scales using a nested approach. Remnants where grazing is controlled, by fencing or otherwise, can be mapped (e.g. Freudenberger & Harvey 2004). A sample of the managed remnants can then be monitored on the ground or remotely with ground truthing of a subsample. The monitored remnants need to be stratified by vegetation type, location, grazing management, etc., using sound sampling design (see Noss 1990; Margules & Austin 1991), so that the results can be extrapolated across the population of mapped remnants. This nested approach to monitoring provides regional understanding of trends in vegetation condition with time, which can be linked with a framework such as Vegetation Assets States and Transitions (VAST) (Thackway & Lesslie 2006) to provide national perspectives on trends in vegetation condition.
Ways forward. Vegetation is similar to other assets valued by society such as roads or buildings. We want to know the location and components of the asset (assessment of vegetation condition, including mapped vegetation types and condition), whether the asset is appreciating or depreciating (monitoring), why it is changing, and what to do about it to inform management and policy. Maps of vegetation condition are planning tools for priorities in regional investment to improve vegetation condition. Monitoring vegetation condition shows trends over time and relationships with management, weather or both. Coarse monitoring can be undertaken at regional scales, with finer scale monitoring at selected sites.
Programs that assess, map and monitor vegetation condition should: (i) define the objectives; (ii) design the program at spatial and temporal scales to meet the objectives; (iii) collect data rigorously at the correct scales; (iv) analyse and interpret the data; and (v) undertake management and policy actions using the results of the monitoring and/or assessment program.
Vegetation condition is not biodiversity (S. Ferrier, pers. comm., 2005). Vegetation condition needs to be combined with other attributes of biodiversity such as vegetation type, spatial configuration and pressures to produce measures of biodiversity. The three primary attributes of biodiversity are structure, function and composition (Noss 1990). Vegetation condition encompasses aspects of structure and composition at sites (e.g. presence of structural layers, life forms and taxa) and aspects of function at site and regional scales (e.g. tree health and regeneration). Vegetation condition does not usually include components of biodiversity such as abundance, population genetics, spatial patterns and connectivity, patch sizes, nutrient cycling, or decomposition.
Vegetation condition needs to be assessed through a ‘values prism’ (Gibbons & Freudenberger 2006). For example, are we investigating vegetation condition for pastoral production, for commercial forest production, or for conservation of threatened communities? Determining the values prism for vegetation condition is necessary for setting goals for assessment and monitoring programs.
The challenges in assessing, mapping and monitoring vegetation condition are to: (i) communicate to policy makers and managers the differences between mapping and monitoring vegetation condition, and between vegetation condition and biodiversity; (ii) define goals for assessing, mapping and monitoring vegetation condition; (iii) determine the costs and feasibility of mapping and monitoring vegetation condition to meet defined goals; (iv) undertake a gap analysis of R&D required to assess, map and monitor vegetation condition, and then undertake the required R & D; and (v) resource and undertake cost-effective and technically feasible programs that meet the goals.
Acknowledgements. The authors thank Andre Zerger, Phil Gibbons, Simon Jones and Peter Lyon for organising the workshop on vegetation condition, which provided excellent insights into the topic. Funding was provided by the NSW Environmental Trust, the Department of Environment and Heritage (Australian Government), CSIRO Sustainable Ecosystems, the NSW Department of Environment and Conservation and RMIT University. The paper was improved by comments from Phil Gibbons and two anonymous referees. Additional references relevant to this short format paper are available from the senior author.