Reliability Estimates for Ivlev's Electivity Index, the Forage Ratio, and a Proposed Linear Index of Food Selection

Ivlevˈs electivity index and the forage ratio, two commonly used measures of food selection, are significantly biased when the sizes of the prey samples from the gut of the predator and the habitat are unequal. Approximate confidence-interval expressions are derived for these indices. A stochastic (Monte Carlo) model was used to validate these expressions and to explore the statistical properties of the indices. The statistical reliability of each index is shown to be a function of the absolute and relative sample sizes and the relative abundances of the prey species in the environment. A linear index of food selection is proposed which avoids most of the statistical and mathematical inadequacies of these indices. Regardless of the index used, however, inadequate habitat sampling, differential availability of prey to the predator, and differential digestion of prey may be significant sources of error in the interpretation of food selection data.