You are here:
US LAND-COVER MONITORING AND DETECTION OF CHANGES IN SCALE AND CONTEXT OF FOREST
Citation:
WICKHAM, J. D., K. RITTERS, AND T. G. WADE. US LAND-COVER MONITORING AND DETECTION OF CHANGES IN SCALE AND CONTEXT OF FOREST. Presented at Proceedings of Forest Stat 07, Montpelier, FRANCE, November 05 - 11, 2007.
Impact/Purpose:
Healthy Communities and Ecosystems - by providing ecosystem research and methods development at multiple scales, utilizing landscape ecology and associated techniques
Description:
Disparate land-cover mapping programs, previously focused solely on mission-oriented goals, have organized themselves as the Multi-Resolution Land Characteristics (MRLC) Consortium with a unified goal of producing land-cover nationwide at routine intervals. Under MRLC, United States land-cover mapping programs are maturing into an integrated land-cover monitoring program. MRLC has produced
land-cover data for ca. 1992 and 2001, and development of data for 2006 is underway. We have used these temporal land-cover data (1992, 2001) to detect changes in the scale at which forest dominates the landscape, and to model how continuing land-cover change will affect the landscape context of remaining forest. Numerous attributes of forest condition have been proposed and implemented, but the spatial scale of forest dominance has not been among them. The spatial scale of forest dominance is an important indicator of forest condition in humid temperate climates because our biogeography textbooks (European
and American) teach us that forest is the predominant vegetation in these climates in the absence of
anthropogenic use of the land. Present day estimates of the spatial extent of forest dominance can be compared to our biogeographic reference points to guide management options. We used area-density scaling to estimate changes in the spatial scale of forest dominance for two regions in the eastern United States. Forest dominance was define using thresholds of >50% and ¡Ý60%. Spatial scales ranges from ~2 to ~5300ha. The ratio of dominant forest loss to forest loss ranged from 1.4 to 11.5 depending on the spatial scale and region. The relationship between forest loss and dominant forest loss was nonlinear,
indicating that continued forest loss has the potential to cause abrupt transitions in the scale at which
forest dominants the landscape. The relationship between forest loss and the spatial scale of dominant forest suggests that the landscape context of extant forest will change with continuing forest loss. We modeled this relationship by applying a first-order Markov chain to a landscape classification scheme. The landscape classification scheme defines (and maps) landscapes based on the dominant land cover and other minor but significant landcover types. An example landscape would be: dominated by natural vegetation but with significant inclusions of agriculture. There are 19 classes in the scheme. The Markov results are consistent with the area-density scaling results. The Markov steady state solution suggests that areas not dominated by anthropogenic use of the land will decline to 2.5% of the landscape and that 75% of the remaining forest cover will be within that landscape type.