| Abstract
| - In monitoring deer damage to plantations it is important that an assessment method which is reliable and easy to use should be available. In devising such a method either practical or theoretical considerations may be emphasised. The method presented, in which the former choice has been made, gives an estimate of the proportion of trees damaged and its standard error. It also allows the forest manager to calculate the sample size that he must use in order to obtain a target precision. An empirical trial of the method as described for a range of damage configurations which vary both in degree of clustering/regularity and intensity of damage. For all but the most clustered damage patterns the method provides estimates which are, for all practical purposes, unbiased and the standard error formula and sample size determination formula perform satisfactorily. For the most clustered damage patterns the estimates are slightly biased, but in this case the accuracy of such an estimate is well represented by the calculated standard error.
|
