Abstract
An index to assess trophic diversity from presence-absence food data is proposed. The index is computed according to the expression \(D = - \sum\limits_{i = 1}^s {1g} {\text{ }}\hat p_i \) , where the \(\hat p_i \) 's are the frequencies of occurrence of the various prey categories. The upper and lower limits of D are derived. A test of the reliability of D was carried out by comparing D and H (Shannon's information function) values obtained from a set of twenty-three food analyses from vertebrate animals. Results show that, although a significant correlation exists between D and H, only a small fraction of H-variation is explained by D-variation. D contains two kinds of information, one referred to ‘species richness’ and another relative to the degree of between-samples heterogeneity. The former is shared in common with H and this presumably explains the fairly weak correlation found between both measures.

This publication has 4 references indexed in Scilit: