The Kyoto Protocol introduces the possibility that changes in carbon stock on agricultural and forest land and soils may be counted against countries' commitments to reduce their greenhouse gas emissions. Including activities related to land use change and forestry in the international climate change agreement may stimulate new incentives for soil-conservation practices domestically. However, a primary criteria for their inclusion relates to the level of accuracy and transparency with which carbon stock changes can be assessed. Parties will also be concerned with the wider environmental impact of different sequestration practices, and the impact of offsets on overall emissions targets. This paper examines these issues for agricultural soils, considering recent research in North America. It is argued that incentives for carbon sequestration practices may need to be implemented independently of actual stock changes because farm-level soil monitoring would be very costly. In the USA, priority should be given to establishing incentives for cover crops and to expanding conservation tillage programs. These activities provide a range of ancillary environmental benefits. In contrast, improvements in biomass yield tend to rely on higher fertilizer inputs with their related environmental costs. Carbon accumulated through any of these activities is easily lost if the practices are discontinued, and so assessment procedures are needed that would avoid overestimating sequestration. Annual accumulation in agricultural soils could be equivalent to about 10% of Annex I carbon dioxide emissions, and therefore options for limiting sink credits from soils should be considered.
