Detecting an impact of predation on bird populations depends on the methods used to assess the predators
Nicoll, M. and Norris, K. (2010) Detecting an impact of predation on bird populations depends on the methods used to assess the predators. Methods in Ecology and Evolution, 1 (3). pp. 300-310. ISSN 2041-210X
Full text not archived in this repository.
To link to this article DOI: 10.1111/j.2041-210X.2010.00030.x
Summary 1. In recent decades there have been population declines of many UK bird species, which have become the focus of intense research and debate. Recently, as the populations of potential predators have increased there is concern that increased rates of predation may be contributing to the declines. In this review, we assess the methodologies behind the current published science on the impacts of predators on avian prey in the UK. 2. We identified suitable studies, classified these according to study design (experimental ⁄observational) and assessed the quantity and quality of the data upon which any variation in predation rates was inferred. We then explored whether the underlying study methodology had implications for study outcome. 3. We reviewed 32 published studies and found that typically observational studies comprehensively monitored significantly fewer predator species than experimental studies. Data for a difference in predator abundance from targeted (i.e. bespoke) census techniques were available for less than half of the 32 predator species studied. 4. The probability of a study detecting an impact on prey abundance was strongly, positively related to the quality and quantity of data upon which the gradient in predation rates was inferred. 5. The findings suggest that if a study is based on good quality abundance data for a range of predator species then it is more likely to detect an effect than if it relies on opportunistic data for a smaller number of predators. 6. We recommend that the findings from studies which use opportunistic data, for a limited number of predator species, should be treated with caution and that future studies employ bespoke census techniques to monitor predator abundance for an appropriate suite of predators.