An Investigation Into the Accuracy of Herbivore Diet Analysis

Abstract

A commonly used technique of herbivore diet analysis consists of grinding ingested material to homogenise the size of particles, extracting a sub-set of these by sieving, and then reconstructing the diet by counting the particles referable to various plant species. A set of experiments revealed that this technique is inaccurate because of variation between species in the number of particles produced per unit dry weight, in the proportion of these retained by the sieve, and in the proportion that could be identified. Thus the technique is badly flawed even without the added complication of differential digestibility, which in itself is shown to be considerable. Certain correction factors derived to cope with these effects are shown to be only partly successful.