Assessment of Non-Wood-Goods and Services by Cluster Sampling

Abstract

Clustering of samples is an appropriate tool for increasing the efficiency of field data assessments. This is one reason for forest inventories on national levels being often based on cluster sampling. However, the cluster design should be optimised for individual conditions by selecting an optimal combination of sample size and cluster design. Therefore, information on the cost and variance structure is essential. The development of a monitoring system for non-wood-goods and services (NWGS) is a main task of the EU-funded research project “Scale Dependent Monitoring of Non-Timber Forest Resources Based on Indicators Assessed in Various Data Sources”. Field surveys were carried out in test sites of five participating countries throughout Europe. The surveys were designed in order to provide sound data for studying and optimising sampling designs for cost-efficient assessment of information on NWGS. Time studies were conducted in addition. This paper presents the analysis of various design alternatives. Results were evaluated by visualising the variance structures by means of spatial statistics. The study shows that cluster sampling is an appropriate tool for the assessment of NWGS along with timber resources in the wide variety of natural conditions found in the five test sites. L-form clusters with data assessment on lines and concentric sample plots were found to be superior to other cluster designs. Distances between plots should be selected according to the specific requirements. The statistical analysis proved the efficiency of the cluster design. The number of sample plots can be optimised by the adopted methods. Spatial statistics, namely variograms provide valuable information for the optimisation of a cluster design for specific populations.