Systematic Sampling in Timber Surveys

The problem of estimation of error variance of a systematic sample has been reviewed. Two alternative designs have been suggested. One envisages modification of the stratified random sampling design such that the sample becomes partly systematic and partly random. The other design is a modification of a latin square restricted sampling design which again is partly systematic and partly restricted random. The designs need to be tested before application on different areas. Further studies are desirable in this direction.