Microscopic Analysis of Herbivore Diets - a Problem and a Solution

Abstract

Three microscopic techniques are commonly used to score the relative abundance of types of plant epidermis in the diet of herbivores: 1 frequency of occurrence; 2 the number; and 3 the area covered by different fragment types. A survey of the literature indicates that the relative area of fragments, determined by point quadrat analysis, is the most precise means of diet analysis. However, none of these techniques can identify every plant fragment. I describe an experiment that shows that the proportion of the unidentifiable component varies between four plant species, indicating that diet descriptions based only on identifiable tissue are unreliable. By knowing the proportion of identifiable tissue for each species, a conversion equation may be calculated that significantly improves point frequencies when compared with known dry weights. The techniques of point quadrat analysis and converted point frequencies may provide accurate estimates of relative abundance of other plant species in samples of herbivore diets.