Should we account for detectability in population trends?

Abstract

Capsule Accounting for changes in detectability over time which could invalidate population trends for common and widespread breeding birds in the UK, resulted in little change in trends for the majority of species considered. Aims To examine whether detectability has a significant influence on population trends for common and widespread breeding birds in the UK. Methods Using data collected over 16 years of the UK Breeding Bird Survey (1994–2009) and focusing on a random sample of 20 common and widespread bird species, we examine the extent to which controlling for temporal change in detectability would influence national population trends. Results Population trends were significantly different for only 2 of the 20 species, Common Blackbird and Winter Wren. This difference corresponds to an apparent fall off in detectability of these species between 2005 and 2009. For the remaining species, controlling for detectability resulted in no significant difference in population trends. Conclusions The importance of detectability should be examined as part of any long-term monitoring programme. However, currently we do not find sufficient support for routinely incorporating detectability into population trends for widespread and abundant breeding birds in the UK.