The relationship between agricultural intensification and a decline in farmland bird populations is well documented in Europe, but the results are mostly based on data from the western part of the continent. In the former socialist eastern and central European countries, political changes around 1990 resulted in a steep decline in the intensity of agriculture. Therefore, one would expect populations of farmland birds to have recovered under these conditions of lower agricultural intensity. We explored population trends of 19 farmland bird species in the Czech Republic between 1982 and 2003 using data from a large‐scale monitoring scheme, and, additionally, we looked for relationships between such population changes and a number of variables describing the temporal development of Czech agriculture. Most farmland species declined during the focal period, and this decline was steepest in farmland specialists (Northern Lapwing Vanellus vanellus, Skylark Alauda arvensis, Linnet Carduelis cannabina and Yellowhammer Emberiza citrinella). Although the intensity of agriculture was lower after than before 1990, the overall decline continued in most farmland bird species, albeit at a slower rate. The correlations between agricultural intensity and farmland bird decline showed opposite patterns to that found in other European studies, because bird populations were highest in years with the most intensive agriculture. We speculate that this pattern could have resulted from the impact of different driving forces causing farmland bird decline in different periods. The high intensity of agriculture could have caused the decline of the originally abundant populations before 1990. After 1990, the decreasing area of arable land could be the most important factor resulting in the continued decline of farmland bird populations. Our results demonstrate that the drivers of farmland bird population changes could differ across Europe, and thus investigations into the effect of farmland management in different parts of the continent are urgently required.

Land-use practices have changed markedly in the Czech Republic, and in other central and eastern European countries, during the 20th century (Lipský 1995, Sklenička 2002). In the last few decades in particular, changes in land-use have been accompanied by changes in the extent of all major landscape types. The amount of arable land has declined as fields in montane and submontane areas were abandoned after the Second World War, when three million people were translocated from areas near state borders (Sonka 2002). As a result, forests started to regenerate on these areas (Pokorný et al. 2005). At the same time, cattle grazing was prohibited within the forested areas and the rotation period increased (Konvička et al. 2004). Forest invasion was enhanced during the transformation of the Czech economy starting in 1989; subsidies in agriculture dropped to a minimum thus restricting arable land to highly productive areas (Sklenička 2002). All of these processes have contributed to the maturation of Czech forests during the last decades (Anon. 2005). Although such large environmental changes would be expected to have an effect on bird populations, large-scale studies of population changes in birds in central and eastern European countries are rather scarce. The explanatory power of existing small-scale studies (Wesołowski & Tomiałojc 1997, Stříteský & Krist 2004) is limited, and does not explain population changes of birds at larger spatial scales (Holmes & Sherry 1988). Moreover, results at regional and local

Recent studies show diVerences in population trends between groups of species occupying diVerent habitats. In Czech birds, as well as in many other European countries, populations of forest species have increased, whereas populations of farmland species have declined. The aim of our study was to test whether population trends of particular species were related to Wner bird-habitat associations within farmland and forest birds. We assessed bird-habitat associations using canonical correspondence analysis based on data from a 400 km long transect across the Czech Republic. We calculated population trends of 62 bird species using log-linear models based on data from a large-scale annual monitoring scheme, which covers the time series from 1982 to 2005. Within forest birds, species with a closer association with lowland broad-leaved forest have had more positive population trends, whereas species with a closer association with montane and coniferous forest revealed more negative population trends. We attribute these opposite trends to the gradual replacement of coniferous forests by deciduous ones, which took place in the Czech Republic during recent decades. Our analyses revealed a hump-shaped relationship within farmland birds, species most closely associated with farmland habitat revealing the most negative trends, whereas species with intermediate association to farmland habitat showed the most positive population trends. Such a pattern can be explained by the abandonment of 3308 Biodivers Conserv (2008) 17:3307-3319 1 C previously cultivated areas followed by the spread of unmanaged meadows and scrubland. Changes in quantity or quality of preferred habitats may thus represent major drivers of observed bird population changes.

No abstract

A species’ susceptibility to environmental change might be predicted by its ecological and life‐history traits. However, the effects of such traits on long‐term bird population trends have not yet been assessed using a comprehensive set of explanatory variables. Moreover, the extent to which phylogeny affects patterns in the interspecific variability of population changes is unclear. Our study focuses on the interspecific variability in long‐term population trends and annual population fluctuations of 68 passerine species in the Czech Republic, assessing the effects of eight life‐history and five ecological traits. Ordination of life‐history traits of 68 species revealed a life‐history gradient, from ‘r‐selected’ (e.g. small body mass, short lifespan, high fecundity, large clutch size) to ‘K‐selected’ species. r‐selected species had more negative population trends than K‐selected species, and seed‐eaters declined compared with insectivores. We suggest that the r‐selected species probably suffer from widespread environmental changes, and the seed‐eaters from current changes in agriculture and land use. Populations of residents fluctuated more than populations of short‐distance migrants, probably due to the effect of winter climatic variability. Variance partitioning at three taxonomic levels showed that whereas population trends, population fluctuations and habitat specialization expressed the highest variability at the species level, most life‐history traits were more variable at higher taxonomic levels. These results explain the loss of statistical power in the relationship between life histories and population trends after controlling for phylogeny. However, we argue that a lack of significance after controlling for phylogeny should not reduce the value of such results for conservation purposes.