Re-evaluating models for estimating prey consumption by leopards

Biomass regression models and associated correction factors derived from feeding trials are essential to convert frequency of prey occurrence from scats into biomass and numbers of prey individuals consumed by carnivores. These dietary analyses form a substantial part of many research projects on predator–prey relationships and human–carnivore conflicts. So far, diet studies of leopard (Panthera pardus) applied the linear biomass model developed for puma (Puma concolor). Recent works, however, suggested that non-linear biomass models are more meaningful for estimating prey biomass and numbers, and presented a generalized model of biomass consumption for all tropical felids. This model accounted for partial consumption of prey, but did not include ecological factors limiting prey consumption by felids. Hence, using 35 feeding trials we developed a leopard specific regression equation by setting a consumption limit for leopard per prey. This new correction factor takes into account the proportion of inedible matter of prey and daily food intake. Besides refining the leopard specific biomass models, our study also showed the saturation in scat production with consumption of larger prey, which was described in recent studies. A reanalysis of prey consumption by leopards from published diet studies by using our new regression models suggests a significant decrease in estimated absolute numbers of prey individuals consumed. This finding and the higher share of larger prey in leopard diet may affect subsequent analyses using prey profiles, and may affect decisions made in biodiversity conservation and management, especially in the area of human–carnivore conflict.