Two previous papers have attempted to identify factors associated with recent population trends of UK farmland birds. Fuller et al. (1995)
were unable to identify any factors that distinguished declining from non-declining species. Siriwardena et al. (1998)
found no evidence that diet, nesting habitat, migration strategy or taxonomic grouping were associated with a species' long-term population trend. The only factor significantly associated with population declines was whether a species was broadly categorized as a farmland generalist or specialist, with specialists declining more. However, the behavioural and life history categories used in these two analyses were generally qualitative in nature. Here, we used quantitative life history and behavioural variables.
As in the analysis by Siriwardena et al. (1998)
, long-term population change in British farmland birds was associated with a measure of specialization. In this analysis, niche position, which is a quantitative measure of how atypical the habitat use is by a particular species, was found to be a strong predictor of long-term population trend. These results are consistent with agricultural intensification as the cause of population declines, as intensification is associated with a loss of habitat diversity and simplification of farming systems (Benton et al. 2003
). It is conceivable that rarity itself, rather than specialization, makes species more vulnerable to declines. However, were this the case, geographic range should be associated with decline rate, yet it was not. An obvious conservation response given these results would be to ensure that the habitats and resources needed by the ‘atypical’ species are given priority in management plans. In fact, this is already embodied in agri-environment strategy in England, where geographically targeted ‘higher tier’ schemes provide resources for the rarest species (Aebischer et al. 2000b
), while lower tier schemes deliver for the declining but still more widespread species (Vickery et al. 2004b
Niche position has been shown to be a significant predictor of abundance and distribution of British birds (Gregory & Gaston 2000
) and here the influence is extended to population trend. One potential shortcoming of the niche variables is that they were calculated from data collected in 1996 (Gregory & Gaston 2000
), which is after the main period of population decline. The niche variables do not document spatial and temporal variability in habitat and resource use, yet it is very unlikely that there is a constant ‘equilibrium niche’. Many species are territorial in the breeding season and occupy territories despotically, in order of habitat quality, so it is possible that niche breadth, for example, will narrow as population density declines (Fretwell & Lucas 1970
; O'Connor 1987
). It is, therefore, possible that any correlation with niche breadth is a consequence rather than a cause of population declines. However, niche breadth was not a significant predictor of population declines. Moreover, it is unlikely that the correlations between niche position and population change suffer from this problem, as niche breath and position are not correlated (Gregory & Gaston 2000
). Nevertheless, spatial and temporal analyses of the intraspecific variability of niche breath and position might prove revealing in understanding patterns of distribution, abundance and population trends.
The lack of relationship between niche breadth and population trend could be an artefact of the way that the variable was estimated. Gregory & Gaston's (2000)
niche breadth measures the use of habitats by species across landscapes, rather than the specific use of food or other resources at a finer scale. An analysis that incorporates finer microhabitat and resource use of individual species may provide better measures of niche breadth. Nevertheless, similar criticisms could be made of niche position, and yet this variable does correlate with population changes in farmland birds.
The relationship between population change and brain size is intriguing and this is the first time, to our knowledge, that such a relationship has been found. Relative brain size has been associated with a number of life history traits, including low annual productivity and altricial development (Bennett & Harvey 1985
). In this study, explicit tests for life history variables were non-significant, indicating that relative brain size itself is an important factor in predicting bird population trends, rather than having an indirect relationship via life history. An obvious benefit of a larger brain is through increased cognitive skills. A recent study of invasion success in birds showed that species with larger brains were more likely to establish non-native populations, that large-brained species also have higher documented rates of behavioural innovation, and that innovation rate also correlates with establishment success (Sol et al. 2005
). Sol et al.
concluded that large brains appear primarily to help birds respond to novel conditions by enhancing their cognitive skills rather than by other mechanisms. Similarly, species with larger brains may be better able to cope with the rapidly changing nature of resources on farmland, and to respond to the availability of opportunities in other habitats, such as food provided at feeders by humans (Cannon 1999
Support for the role of cognitive ability comes from the relationships we report between population change and brain component size. The relationship between decline rates and telencephalon and brain stem sizes were stronger than for total brain size. Although the functional importance of different brain components is poorly understood, the telencephalon is generally thought to be functionally equivalent to the mammalian neocortex, in that these regions appear to be important in problem solving and complex social behaviour (Reader 2003
). The significance of telencephalon size is thus consistent with an influence of cognitive ability on population change in farmland birds. In contrast, the role of the brain stem is puzzling, as this region of the brain is generally assumed primarily important for control of vital functions such as heartbeat and metabolism. As brain stem was dropped from stepwise models, we suggest that the strong relationship between population change and brain stem size might be most probably the result of the tight correlation between brain stem and telencephalon size. Clearly, further research is required on the possible influence of brain architecture on behavioural and ecological characteristics.
In summary, our results suggest that the farmland birds whose populations have suffered most under agricultural intensification are those with more specialized resource and habitat use and lesser cognitive abilities. As this is the first study of its type, to our knowledge, to examine the association between cognitive capacity and population trends, it remains to be seen whether these conclusions apply to other taxa. In particular, the influence of cognitive abilities seems unlikely to generalize beyond the ‘higher’ vertebrates. Nevertheless, our study shows that it is possible to identify characteristics that might make some species particularly vulnerable to habitat changes, and hence whether or not the same characteristics are important for all taxa, it is an approach that could be applied to other groups. While the urgent need for robust long-term monitoring is undiminished (Balmford et al. 2005
; Gregory et al. 2005
; Loh et al. 2005
), this study suggests a potential short cut that might in some cases help to identify priority cases at an earlier stage.