A new type of sample unit for the efficient assessment of diverse tree communities in complex forest landscapes

We present a new and versatile sample unit suitable for rapid assessments of tropical forest in heterogeneous areas. The method uses multiple applications of small and easy-to-apply variable-area subunits, in which the area is defined by simple and objective rules. Compared with any fixed-area approach the sample unit is quick and easy to apply even in difficult terrain, and the amount of information collected varies little with stem densities. Unlike most variable-area methods difficult judgements are rare. Further, it cannot be extended to arbitrary size, but remains compact, allowing data to be linked to local-site variables. Useful data will generally result even in patchy and divided environments. Here we describe the method, discuss the nature of the resulting data and show how various stand characters can be calculated. The calculation of basic stand parameters from the sample data does not require any sophisticated analyses, and some worked examples are provided to ensure that the calculations are accessible. To demonstrate the statistical theory underlying this class of methods, and the good performance of the estimation methods, a more formal theoretical treatment is included as an appendix. The approach offers considerable promise for efficient forest assessments.