Area-based indicators can be considered as abundance-based measures at the ecosystem level, corresponding to population measures at the species level. They typically express the area (km2) at a particular time of a defined ecosystem. The area may be expressed as a fraction of some reference state, such as the supposed original (potential) area. A well-known example is the Global Forest Resources Assessment (FAO 2001). Others are assessments of coral reefs (Wilkinson 2000) and mangroves (FAO 2003). Biodiversity, at all levels, needs an area in which to exist. When half the habitat of a given species is lost, the abundance of that species is roughly halved. This makes area an easily understood and easily measured indicator.
Fragmentation indicators are theoretically based on concepts of island biogeography, which predict that as fragments are isolated from a larger mass, they will lose species. They vary from the simple (e.g., mean fragment size, perimeter length to bounded area ratio) to the sophisticated (fractal-based indices). Road network density has been used as a proxy for fragmentation.
Area-based indicators are built on readily observable structural features, such as the cover by trees or coral reefs, and thus are typically at the highest level of ecosystem classification, the biome. In principle they are straightforward and uncontroversial, but in practice it is hard to ensure uniformity of application of the definition. As a result, the variation between different sources of information (compare FAO 2001 to Achard et al. 2002, for instance) usually is much higher than the temporal trend, at least in the short term. However, the relative temporal trend usually is consistent in direction, if not magnitude. Only a few time series of trends in biome extent exist at the global level (Jenkins et al. 2003).
It is not clear what the equivalent boundaries for oceanic ecosystems are because their edges would be expected to be somewhat more variable in time. The large marine ecosystems (Sherman et al. 1990) have fixed spatial definitions and therefore are not useful as indicators by themselves (although the fraction degraded within them would be useful). The marine biogeochemical provinces (Longhurst 1998) are mapped largely using remote sensing of sea surface temperature and chlorophyll content and are highly dynamic over a period of weeks, making them too variable to be useful in the short term. Marine ecosystems therefore may lend themselves better to indicators at the species or community (e.g., seagrass, coral reefs) level.
Was this article helpful?