USING GIS TO EVALUATE THE CROWN-LINE INTERSECT SAMPLING METHOD IN FOREST SURVEY

The utility of the crown-line intersect sampling (Crown-LIS) method for forest surveying was reviewed and extended. A tropical forest model based on field data from the dry tropical forest of Madagascar was constructed and manipulated within a Geographical Information System (GIS). The accuracies of the Crown-LIS method and of five other methods for estimating basal area. stem density, species diversity and species density distribution were evaluated, using the forest model. The results show no significant statistical difference between the estimates obtained by Crown-LIS, which is a line-based sampling method, and those obtained by angle-gauge sampling methods (relascope and strand) or by area-based sampling methods (strip-line and circular plots). The distance method (point-centered quadrats) was found to overestimate the basal area and stem density, but proved accurate in assessing species distribution and species diversity. The practicality of the Crown-LIS method in the field, combined with the measurement accuracies it can achieve, conclusively makes it a good alternative sampling method in forest survey.