A method for estimating long-term average annual soil loss under different land management scenarios from the Marquis and Soufriere watersheds on St. Lucia is presented. The Revised Universal Soil Loss Equation (RUSLE) was used, and a GIS was employed to generate the required input parameters. Model execution and results were also generated within the GIS. Modelling soil loss for the different land management scenarios was based upon a land capability classification and associated conservation treatments. Soil losses under current agricultural land-use patterns were analyzed and compared to potential soil losses under conservation treatments following the criteria specified in the land capability classification. The model predicted substantial declines in soil loss where conservation treatments were assigned, as compared to soil loss under current land-use patterns. It was found that predicted soil losses from the Soufriere watershed were four times that predicted for the Marquis watershed for all the land management scenarios modelled. Of the input parameters in the model, slope steepness was most highly correlated to predicted soil loss. It is anticipated that the findings of this study will be used in the development of a decision support system for agricultural and forestry land planning on St. Lucia.

No description available.