Technical Notes: Classifying herbivore diets using hierarchical cluster analysis

Abstract

Hypotheses are frequently posed that require comparisons among diet samples. Hierarchical cluster analysis is suited to this task, but has received little attention in food habits research. By grouping samples so that similar diets are close together and dissimilar diets are spaced farther apart, hierarchical cluster analysis reveals patterns in data difficult to recognize in the matrix of similarity coefficients typical of most food habits studies. We provide an example of this type of analysis, and indicate its application to management of large herbivores.