The Demography of Farmland
Bird Population Changes

It was the BTO's Common Birds Census (CBC) that first alerted the conservation community to the recent, severe population declines among our farmland birds. The BTO has just completed a MAFF-funded (Contract BD0906) three-year integrated population monitoring study in which we took CBC analysis further, combining the data with information from other BTO schemes such as the Nest Record Scheme, the Ringing Scheme and the New Breeding Atlas to investigate the likely causes of the declines, concentrating on seed-eating species.

We tested the idea that large numbers of species had undergone similar population trends onfarmland by using objective, quantitative methods to analyse the CBC data for those species that were sufficiently well represented on farmland CBC plots. This showed that, in fact, most species' trends have been different in the long-term, so that they do not fall readily into large groups. However, we did find several small groups, most notably one including the three most common thrushes, which tend to suggest that similar environmental changes have influenced the species in each group. We were also able to identify several short periods when the population trends of large numbers of species turned in the same direction at the same time: each period where a large number of upturns or downturns was occurring is one where an environmental change with a widespread impact is likely to have occurred.

Dead recoveries of ringed birds allowed us to investigate historical changes in annual survival and nest record data provided information on past changes in breeding performance. Concentrating on seed-eating species, we linked these data to CBC trends and asked whether the changes that have occurred in breeding performance and survival are consistent with their having caused the population trends. For Goldfinch and House Sparrow (Figure 1), variation in survival explained almost all of the changes in abundance seen, but it was weakly related to population trends for Bullfinch, Chaffinch, Greenfinch and Linnet.

However, when we examined variations in survival for a wider range of species, we found that it was much more common than not for changes in survival to be consistent with their having driven population trends: this was true for 13 of 28 species, and more species might well have been added if more ring-recovery data had been available. Our analyses of Nest Record data supported this pattern: of 12 seed-eaters, only for Linnet was there significant variation in breeding performance which is likely to have caused a major population trend. Unfortunately, we were unable to investigate the importance of post-fledging survival rates or the number of breeding attempts possible in a season: changes in these demographic rates could have been very important for farmland bird populations.

We also analysed variations in breeding performance and in species' frequency of occurrence within 10-by-10km squares (from the New Atlas of Breeding Birds) with respect to agricultural land-use. These analyses showed that agricultural land-use seems to affect farmland seed-eaters in species-specific ways. We did find some notable general patterns, however, suggesting that many species prefer and are benefitted by features of more extensive farming such as fallow land, a mixture of tillage and pasture and spring-sown cereals.

Contact: ipm.demography@bto.org

Back to Farmland Birds