Defining and measuring the health of savanna landscapes: A north Australian perspective

Participants of the Tropical Savannas CRC have sought to develop performance benchmarks by defining healthy savannas and their attributes, and identifying indicators of health. They propose that:

A healthy savanna:
• maintains basic functions (including but not confined to nutrient cycling, water capture, provision of food and shelter for fauna) at all spatial scales;
• maintains viable populations of all native species of plants and animals at appropriate spatial and temporal scales; and,
• reliably meets the long-term needs (spiritual, aesthetic and material) of those with an interest in the savannas.
Key features of this definition of savanna health are that:

(i) attributes considered important extend beyond basic processes such as nutrient cycling to include features such as biological and landscape diversity that may make little direct contribution to production; and
(ii) attributes considered integral to health may vary across spatial scales.

In taking this broad view of application of the term health to landscapes, participants chose to avoid theoretical and semantic debate and focus instead on the primary goal: to clarify the sorts of resource management outcomes to which the TS–CRC will be seeking to contribute. While this approach requires that we embrace some diversity of views, we consider that attempts to closely circumscribe ideas at this stage in the development of the savanna health concept would be premature and potentially counter-productive.

Therefore the suite of attributes identified also varies somewhat according to prevailing land use; they are identified separately for Aboriginal, conservation, pastoral and more intensive agricultural use. However, attributes tend to converge and be less dependent on narrower sectoral goals at larger spatial scales. Proposed indicators all identify potentially measurable aspects of savanna state or process, but many require further work to simplify, verify sensitivity and cross-validate. In addition to technical validation of indicators, future research should include analysis of processes that most effectively encourage application of such benchmarks by land managers and other decision-makers.