Quantification of land-use dynamics: an illustration from Costa Rica.

In many cases, studies dealing with land degradation require the quantification of land-use dynamics. Although research has been carried out to describe land-use dynamics and its driving forces, very little has been done on the recognition of indicators for the quantification of land-use dynamics. This article presents and discusses three different indicators recognised during a Costa Rican case-study: (1) a single-time analysis of spatial patterns; (2) standard Markov chains with a soil-type modifier; (3) geo-referenced Markov chains with indices for size, shape and land cover in neighbouring polygons. The first indicator is applied to a study area of 2942 km2 in the northeast of Costa Rica for which 1992 aerial photographs are available. Spatial patterns of land cover can only be related to land cover modifications when good insight into the colonization history exists. For the study area, clear land-use sequences have been found. Both standard and geo-referenced Markov chains have been calculated for a smaller pilot area of 151 km2 for which aerial photographs taken on five different dates area available. Clear differences in probabilities are found for soil type and polygon size. For the shape and boundary index, no clear relations were found in the pilot area. The quantification of land-use dynamics using these kinds of indicators is necessary for the comparison of regions and for land degradation modelling