Modelling population changes using data from different surveys: the Common Birds Census and the Breeding Bird Survey

Abstract

Capsule A method for producing and validating long-term population indices using data from the Common Birds Census and its successor, the Breeding Bird Survey, is described. Aim To investigate a means of combining site-specific records from two very different surveys into reliable population indices. Methods A generalized linear model is described for Common Birds Census (CBC) and Breeding Bird Survey (BBS) data, and used to identify species and geographical regions for which the resulting temporal trends are comparable, and hence derive long-term trends that straddle the overlap of the two surveys. Results From 1994 to 2000, when both the CBC and the BBS were in operation, no significant difference in population trends in southeastern Britain between the two surveys was detected for the vast majority of the 73 species considered. CBC data are limited outside this region, but an analysis of BBS data over the same period showed that the trends outside this region were significantly different for around half of the species considered. Conclusion Although the predominant means of gathering data on terrestrial breeding birds has changed since the 1960s, a joint analysis of the combined data from the surveys can be used to produce annual indices of abundance for most species with sufficient data for either the whole or at least a proportion of Britain.