Two-phase approaches to point and transect relascope sampling of downed logs

Point relascope sampling and transect relascope sampling were recently proposed as methods for the inventory of downed coarse woody debris. By only counting logs with a relascope device, the total length squared (with point relascope sampling) or the total length (with transect relascope sampling) of downed logs in an area can be estimated. For estimates of other variables, such as volume, additional measurements on the sampled logs are required. In this article, two-phase approaches to the methods are presented that makes use of the estimates from fast counts of logs as auxiliary data. The presented approaches serve two purposes: (i) to improve the efficiency of the methods and (ii) to avoid the bias that is likely to occur if careful checks of whether or not doubtful logs should be counted are neglected. Each two-phase design was compared with a single-phase design in terms of the standard error obtained for a given inventory cost. The two-phase designs decreased the standard errors with ca. 17-18% for points and 10-15% for lines. Including subjective judgements as additional auxiliary variables further decreased the standard errors in the line case but not in the point case. In the former case, the improvement in comparison with the single-phase design was 17-23%.