Assessing bias in diet methods for the Long-legged Buzzard Buteo rufinus

We analysed diet data in a Long-legged Buzzard Buteo rufinus population using four different methods (direct observations, pellets, prey remains and combined pellets and prey remains) from Cyprus, during 2005–2007. Direct observations can provide reliable diet information. However, the small sample size of monitored nests, the subjective selection of accessible nests, the differential prey size deliveries by different sexes, and the different recognition of different prey species could involve biases, therefore accounting for the low dietary breadth. Prey remains differed significantly from pellet analysis due to more birds and fewer lizards, but the broadest trophic spectrum was obtained. Pellets analysis failed to detect small-sized prey items, but provided a high number of species present in the diet. Pellets and pooled data corresponded closely with the results of direct observations. Pooling data from pellets and prey remains overcame the problems associated with these two methods and revealed a high diet diversity and a great similarity to direct observations. Our results suggest that the pooled data method will be effective for monitoring dietary trends in the breeding and non-breeding Long-legged Buzzard population.

► We assess biases associated with four diet methods of Long-legged Buzzards.
► Direct observations provide reliable diet information but prove a narrow diet breadth.
► Prey remains provide the broadest trophic spectrum but a low number of samples.
► Pellets fail to detect small-sized prey items.
► Pooled data from pellets and prey remains compensate for discrepancies between them revealing high diet diversity.