Résumé
Effcient biodiversity management strategies aim to allocate conservation efforts so as to maximize diversity in ecological systems. Toward this end, dening a diversity criterion is an important but challenging task, as several different indices can be used as biodiversity measures. This paper elicits and compares two criteria for biodiversity conservation based on indices stemming from different disciplines: Weitzman's index in economics and Rao's index in ecology. These indices use different approaches to combine information about measures of (1) the probability distributions of the species that are present in an ecosystem (i.e. survival probabilities) and (2) the degree of dissimilarity between these species. As an important step toward in situ conservation criteria, we add to these elements information about (3) the ecological interactions that take place between species. Considering a simple three-species ecosystem, we show that criterion choice has palpable policy implications, as it can sometimes lead to divergent management recommendations. We disentangle the roles played by elements (1), (2) and (3) in the ranking outcomes, which allows us to highlight several specificities of the two criteria. An important result is that, other things being equal, Weitzman's in situ ranking tends to favor robust species that are least concerned with extinction, while Rao's in situ ranking generally gives priority to more vulnerable species that are closer to extinction.