USE OF A NETWORK OF LONG-TERM EXPERIMENTS FOR ANALYSIS OF SOIL CARBONDYNAMICS AND GLOBAL CHANGE - THE NORTH-AMERICAN MODEL

Soils contain a large proportion of the carbon (C) in the terrestrial biosphere, yet the role of soils as a sink or a source of net atmosphe ric C flux is uncertain. In agricultural systems, soil C is highly inf luenced by management practices and there is considerable interest in adapting management systems to promote soil C sequestration, thereby h elping to mitigate atmospheric CO2 increases. Long-term field experime ntal sites represent a unique source of information on soil C dynamics , and networks of such sites provide a key ingredient for making large -scale assessments of soil C change across ranges in climate and soil conditions and management regimes. Currently, there are collaborative efforts to develop such site networks in Australia, Europe, and North America. A network of long-term experiments in North America was estab lished to provide baseline information on the effects of management (i .e. tillage, crop rotations, fertilisation, organic amendments) on soi l organic matter. Historical data on soils, primary productivity, clim ate, and management were synthesised by scientists from the individual field sites, representing a total of 35 long-term field experiments. An additional cross-site soil sampling campaign was carried out to pro vide uniform comparisons of soil C and nitrogen (N), both within and a cross sites. Long-term field experiments are a principle component nec essary for regional assessments of soil C dynamics. We describe a gene ral methodology for combining long-term data with process-oriented sim ulation models and regional-level, spatially resolved databases. Such analyses are needed to assess past and present changes in soil C at re gional to global scales and to make projections of the potential impac ts of changes in climate, CO2, and landuse patterns on soil C in agroe cosystems.