A MARKOV MODEL PROJECTION OF SOIL ORGANIC-CARBON STORES FOLLOWING LAND-USE CHANGES

Soils are major sinks of carbon, and land use can affect the magnitude s of soil organic carbon stores and the net flux of carbon between the land and atmosphere. Hence, it is of some interest to have a method f or examining the future consequences of changes in the patterns of lan d use for soil organic carbon stores, and to allow experiments to be c arried out to assess the likely effects of various policy options. We illustrate the use of a Markov model to project future areas of land u se from land cover transition matrices for England, Wales and Scotland , 1984-1990, and by the application of vectors of soil organic carbon stores for each land use types to the changes in areas to obtain proje cted changes in the soil carbon stores. In England and Wales, much dep ends on whether or not urban land is assumed to store soil carbon. For example, during 1984-1990, there was an overall decrease in potential organic carbon store in England and Wales of 32.64 MtC assuming that urban land stores no soil carbon, but that overall decrease is reduced by 73% if urban land is assumed to store 26.25 x 10(3) tC km(-2). For England and Wales, the limiting probabilities show 37.9% of the land as urban and 15.3% as arable. There would be a decrease in the overall potential soil carbon storage capacity of 610 MtC or 239 MtC, dependi ng on whether or not urban land is assumed to store soil carbon. For S cotland, the limiting probabilities show 5.31% of the land as lowland heath and 16.9% as coniferous forest. There would be a decrease in the overall potential soil carbon storage capacity of 9414 MtC if urban l and is assumed to store no carbon, and 9668 MtC if it is assumed to st ore carbon. By changing entries in the land cover transition matrices, the consequences of different policy options can be examined. (C) 199 5 Academic Press Limited