Special Issue

Nature Conservation 2009 — 1. 9. 2009 — Special Issue

The impact of global climate change on trends in bird numbers in the Czech Republic

authors: Jiří Reif, Petr Voříšek, Zdeněk Vermouzek, Karel Šťastný, Michaela Koschová, Vladimír Bejček

The climate has pervasive effects on the distribution of organisms over the Earth (Lomolino et al.2005) and it has recently been found that it also substantially affects changes in the bird numbers over time (Crick 2004, Drent 2006, La Sorte & Thompson 2007).

Mostly, indirect impacts are involved, where the climate changes adversely affect breeding performance locally due to mismatches between the timing of breeding and peaks of major food resources or the phenology of vegetation growth, e.g.onset of growth of leaves by vegetation. Birds respond to current temperature changes much more slowly than insects, which are the key component in their diet during the nesting period, and plants where they locate their nests. Both are manifested in a reduction in the breeding success of the relevant species (Visser et al.1998, Sanz et al.2003, Both et al.2006, Martin 2007). In addition, the arrival dates of many bird species advance at their nesting sites, particularly for migrants wintering at temperate latitudes (Ahola et al.2004, Jonzén et al.2006, 2007): therefore many species stay longer at their breeding grounds (Thorup et al.2007). As a consequence of the early arrival, which is not followed by a shift in the nesting period, the period during which the individual birds intensively defend their territories is prolonged and thus they expend more energy, which is then lacking during the parental care. In some cases, a direct response by birds to climatic extreme events has been observed – the heat waves in France in 2003 caused greater mortality rates of a lot of common species with lower tolerance to climate fluctuations, followed by a reduction in their population size (Julliard et al.2004).

At both the regional and global scale, there are shifts in the distribution of main habitats, resulting in the movement of the distribution range limits in many bird species (Thomas & Lennon 1999, Böhning-Gaese & Lemoine 2004, Brommer 2004). Even faster changes in the bird distribution are projected for the future, particularly in the northern hemisphere temperate zone and in the Arctic (Jetz et al.2007, Huntley et al.2008). In contrast to a lot of studies dealing with the continental distribution of birds, changes in the bird numbers and distribution at the level of individual countries has received much less attention (Böhning-Gaese & Lemoine l..c, Lemoine et al.2007). When we realize that nature conservation instruments including legislative tools creating conditions and even international ones are mostly passed through national governments and parliaments (Watzold & Schwerdtner 2005), the fact is highly alarming. In order to provide environmental policies in the individual countries with the appropriate data and evidences, it is necessary to begin to exa­mine the climate change impacts on the bird po­pulations breeding within them.

In the trends in the climate in the Czech Republic, there has been a slight increase in average annual temperatures over the last few decades (Solomon et al.2007, Pretel 2009). Birds are affected most by the temperatures in spring, when almost all species in the Czech Republic breed and the current state of the weather determines their reproductive success, and in winter, when the resident (se­dentary) bird populations are limited by low temperatures (Newton 1998). No clear trend has been observed in winter temperatures in 1975–2007, so they will not be further considered here. However, the spring temperatures have significantly increased during this period. Thus, we will seek for possible manifestations of this increase in bird breeding population sizes in the Czech Republic.

In order to find how climate changes are reflected in changes in bird numbers, it is necessary to more exactly understand the individual species climatic requirements. It is assumed that species with different climatic requirements should react differently to long-term changes in the climate, which should be manifested in changes in their abundance (Jiguet et al. 2006, 2007, Lemoine et al.l.c., Huntley et al.l.c.). The climatic requirements of individual bird species can be determined even when we do not have access to large-scale climatic data. We know that the average temperature decreases with increasing latitude and thus it can be assumed that species occupying the more northern parts of the study area will prefer a colder climate. On the other hand, species that nest in the more southern areas will prefer a warmer climate. We can thus use the geographic characteristics of the distribution range to reflect the climatic requirements of the particular bird species. Useful geographic characteristics of the distribution range are generally considered to be the latitude of the distribution range’s centre and the latitude of the latitudinal midpoint, determined from the difference between the latitude of the northernmost and southernmost points of the distribution range.

Formulation of the hypotheses which will be tested by the study will be based on the assumption that the same mechanisms could be in place in Central Europe as those that cause bird distribution shifts on a continental scale, except that the consequences of the former will be less obvious. As a result of the average temperature increase, the distribution ranges in the individual species will be shifted to the north and will follow their climatic optima (Huntley et al.l.c.). However, the climate change impacts should vary according to which parts of Europe the individual bird species occupy. The population numbers in species with centres of distribution in South Europe and occurring in the Czech Republic should increase, while species with centres of distribution in North Europe should decrease in numbers. Thus, we will test the following statements:

(i) the abundance of species that have a broad area of distribution throughout Europe has not been changing in the Czech Republic in the long term;

(ii) the abundance of species with southern-centred distributions has been increasing in the Czech Republic in the long term;

(iii) the abundance of species with northern-centred distributions has been decreasing in the Czech Republic in the long term;

(iv) the abundance of species that are limited to the central part of the European continent has not been changing in the Czech Republic in the long term.

Materials and methods

Quantitative data on trends in bird numbers in the Czech Republic have been gathered within the Breeding Bird Monitoring Programme in the Czech Republic (BBMP). The field BBMP methodology is based on long-term monito­ring of nesting bird populations by skilled volunteers, members of the Czech Society for Ornithology, and has been published yet (e.g., Janda & Šťastný 1984, Reif et al.2006).

Although BBMP has been a on-going process and has been continuing to collect new data, we limited the time series used to the period 1982–2006, for which the corresponding climatic data were available. From the raw data from the field co-workers, obtained for a total of 335 sites, we employed log-linear models in the TRIM 3.51 program to calculate two characteristics of changes in the individual species numbers – the abundance index and the trend in the abundance (Pannekoek & Van Strien 2001). The abundance index gives the percentage by which the particular species numbers changed in the given year compared to the first year of the time series, which is always set at 100 %. The population trend summarizes the trend in the particular species numbers over the whole study period in a single number. This can be described as the average of annual trends (mean relative population change) over the whole study period (i.e. 1982–2006 in our case). Because the abundance of birds is calculated on a logarithmic scale, a trend in the abundance equal to one means that the abundance of the species did not change over the given period. Trends greater than 1 correspond to species on the increase and will be called positive; trends less than 1 are characteristic for declining species and will be called as negative. For example, a trend with a value of 1.02 means that the species numbers over the given period increased on an average by 2 % annually.

All 152 species, for which we analyzed trends in their numbers, were classified according to their population trend values with 95 % confidence intervals into several groups. We applied the following categories of trends, based on the criteria introduced by Gregory et al.(2007) and used in the Czech literature, e.g.by Reif et al.(2006). Strongly increasing– lower limit of the confidence interval is greater than or equal to 1.05; slightly increasing- lower limit of the confidence interval is greater than 1.00 and less than 1.05; stable– the confidence interval includes 1.00 and simultaneously its lower limit is greater than or equal to 0.95 and the upper limit is less than or equal to 1.05; slightly decreasing– the upper limit of the confidence interval is less than 1.00 and greater than 0.95; strongly decreasing– the upper limit of the confidence interval is less than or equal to 0.95; uncertain– the confidence interval includes 1.00 and simultaneously its lower limit is less than 0.95 and upper limit is greater than 1.05. We excluded from further analyses 43 species with uncertain trends in numbers and other six species that were found at less than 30 sites and whose trends could thus be substantially affected by local inter-annual fluctuations in abundance without any relationship to the overall abundance in the Czech Republic (Fox 2004). We used a set consisting of 103 bird species inhabiting the Czech Republic as breeders for the subsequent analysis (Reif et al.2008a).

The climatic requirements of the individual bird species of birds described using geographic variables characterizing their distribution in Europe. For this purpose, we used maps of the breeding bird distribution in Europe published in the EBCC Atlas (Hagemeijer & Blair 1997). We calculated the latitudinal midpoint of distribution range for each species, i.e.the difference between the latitudes of the northernmost and southernmost points of the distribution range divided by two and subtracted from the northernmost point (Lemoine et al.2007). Then we classified the individual species into four latitudinal groups, differing in the individual species’ areas of distribution. They included: northern species, southern species, central species and widespread, i.e.widely di­stributed species. Classification into groups was performed according to the methodo­logy which has been published (Reif et al.2008a): northern species(n = 28) – their ranges cover less than 30 % of the southern region (e.g., the Fieldfare Turdus pila­ris); southern species(n = 19) – their ranges cover less than 30 % of the northern area (e.g.the Nightingale Luscinia megarhynchos); central species(n = 9) – their ranges cover less than 30 % of the areas of the southern and nort­hern regions simulta­neously (e.g., the Marsh Tit Poecile palustris); widespread species(n = 47) – their ranges cover more than 30 % of the area of each region (e.g.,the House Sparrow Passer domesticus). Although such species sorting is arbitrary to some extent, and indeed 30 % has no biological meaning, we argue that it mirrors the real latitudinal preferences of particular species and thus, can be used in further analysis.

We related species’ population trends to the latitudinal midpoints of their distributions by linear regression. We compared population trends between range-defined species’ groups using one-way ANOVA (analysis of variance). In the next step, we calculated the geometric mean of indices for each range-defined species’ group” it was carried out for each year within the study period (Gregory et al.2005). We correlated the geometric means with mean annual spring temperatures. Reif et al.(2008a) gives the basic data used for all the analyses in this study.

Results

The latitudinal midpoint explained a relatively small, but significant part of the varia­bility in species population trends. As predi­cted, the relationship was negative illustrating that Czech populations of in Europe more southerly distributed species increased, while populations of more northerly distributed showed a general decline (Fig. 1).

The annual changes in the geometric means of species’ indices differed between the particular species’ groups defined above by their European distributions (Fig. 2). Whereas northern species revealed a marked decline, the population size increase of southern species was of lower magnitude and therefore, less obvious. It is remarkable, that the population changes of central species were similar to those of southern species; simultaneously, the abundance of the widespread species decreased similarly to that for northern species. A comparison of the population trends between the breeding range-defined species’ groups showed signi­ficant differences between groups. The trends in northern species were more negative than the trends in the southern species (Fig. 3)

The individual groups, defined by the latitudinal ranges of the individual species, also differed in the relationship between their population indices and annual changes in the average spring temperature. While these relationships were insignificant in southern species, central and widespread species, northern species showed a significant negative correlation.

Discussion

We found that changes in abundance of 103 bird species from 1982 to 2006 in the Czech Republic were negatively related to latitudinal midpoints of the European breeding distribution of these species. The more northerly the breeding range, the greater the decrease in the numbers was.

These results confirm our first prediction stating that as a consequence of the increase in spring temperatures, species that occupy the colder parts of Europe would decrease in numbers in the Czech Republic while species that are distributed mainly in the warmer parts of the continent should occur in the Czech Republic in greater numbers. We can consider our findings to be a clear, although indirect evidence of the global climate change effects on the long-term population trends in birds in the Czech Republic.

German ornithologists have published similar results; on the basis of breeding bird distribution mapping at Lake Constance, they showed that species with more northern latitudinal midpoints have declined (Lemoine et al. l.c.). This supports the view of Central Europe as a window through which range dynamics are visible via species’ population processes which is actually occurring at much larger spatial scale than the limited area of our region, but is also manifested in changes in the local abundances of the individual species, which we can record at this level. The projections of bird species’ distribution, modelled on the basis of breeding bird distribution mapping of the individual species and maps of the climate in the coming decades (Huntley et al. 2007), are thus well supported by current empirical data.

The individual groups of species, defined by their European breeding ranges, differed not only in their population trends, but also in the intensity of population changes. Northern species showed the steepest population decline, but southern species revealed rather a slow population increase. The pattern could be caused by latitudes of the Czech Republic, being closer in proximity to the range margins of northern species compared with that of southern species (Reif et al., unpubl. data). The above-described mechanisms could have impacts that are stronger near range edges and lower in core areas of species’ di­stribution ranges (Gaston 2003). In addition, northern had the most negative population trends and, at the same time, they were the only group with a significant (and negative) correlation with the temperature. Thus, we can speculate that the temperature has the greatest effect on just these species.

Although we studied only changes in bird numbers, it is highly probable that a similar effect of climate changes would also be found for other groups of organisms (Plesník 2009). For example, Konvička et al.(2003) found that the distribution of some butterfly species in the Czech Republic is also shifting towards higher altitudes, causing that species that are limited to higher areas have been facing a high risk of extinction from nature in this country. Thus, the climate change impact on the biosphere can be relatively serious and could possibly lead to species richness decline. In the Czech Republic, the total species richness could even increase as a consequence of climate warming, because warmer part of the Earth generally harbor more species (Hawkins et al.2003), but there is a risk of irreplaceable and, within the near future, irreversible loss of those species whose climatic requirements will not be met by he altered conditions. Therefore, efforts to save the above species are the important task for nature conservation.

In addition to climate changes, the abundance of birds in the Czech Republic is also influenced by other factors, such as changes in the landscape (Reif et al.2008b, 2008c). Changes in the numbers of some species, explained here through climate changes, could just as easily have been caused by land use changes and the connection with climate warming could be accidental – for example, the increase in the Gadwall (Anas strepera) numbers is most probably caused by changes in fishpond management in the Czech Republic (Musil et al.2001). In addition, the habitat requirements of the individual species can also effect how sensitively they respond to climate changes. Food generalists, such as the Red-backed Shrike (Lanius collurio), will probably be far less affected by a shifting the peak in available food (Hušek & Adamík 2008) than specialists, e.g.the Pied Flycatcher (Ficedula hypoleuca, Both et al.2006). Further, we can assume that species inhabiting more habitat types will, in general, be more resistant to the climate change effects. The assumption was partly confirmed by Jiguet et al.(2007) who, however, at the same time stress, that specific habitat requirements make the species vulnerable, particularly to anthropogenic changes in the land cover (Archaux 2007).

Implications for nature conservation

The following steps can be proposed for better coping with the consequences of climate changes, some of which correspond directly to the recommendations published by the Council of Europe (Huntley 2007, Usher 2007):

• for each species, determine stipulate a degree of risk (vulnerability) from climate change in the context of other factors (drivers);
• for the most threatened species (at the pan-European level), develop exact plans for minimization of the climate change effects and consequently to allocate the relevant finances, or revise the relevant state policies (e.g., subsidy policy in agriculture and rural development);
• develop and agree indicators for assessing the climate change effects on biodiversity and publish them regularly for use at all the decision-making levels;
• update, based on the outputs of extended analyses, the specially protected species lists, of course, taking into account further criteria for listing the species as specially protected.

Summary

Numerous studies have shown that climate changes associated with increasing global temperature affect bird species. For example, some long-distance migrants are not able to respond adequately to rapid advances in spring phenology, e.g. by adapting their breeding season to the earlier onset of spring. Therefore, their populations have been declining due to lower nesting success (lower breeding performance). Moreover, it was found that many species in the Northern Hemisphere have shifted their northern breeding range boundaries further to the north. However, there are surprisingly only few studies focusing on bird population trends in relation to climate change at the scale of individual countries, which make decisions on most steps related to environmental policy. We hypothesized that bird species with different European latitudinal breeding distributions would have different long-term population trends in the Czech Republic, as a result of range dynamics caused by increasing spring temperature. In accordance with this prediction, the results of an analysis based on large-scale monitoring data (from 1982 to 2006) showed that species with more northern latitudinal distributions had more negative population trends in the Czech Republic, particularly compared to southern species.

Acknowledgements

We would like to thank the hundreds of members the Czech Society for Ornitho­logy for their generous voluntary field work. We are grateful to B. Huntley for valuable comments on drafts of the manuscript. The study was supported by the Agency for Nature Conservation and Landscape Protection of the Czech Republic, the Grant Agency of the Academy of Sciences of the Czech Republic (KJB601110919) and the Czech Science Foundation (206/97/0771 and 206/04/1254).

J. Reif is at the Institute for Environmental Studies, Faculty of Science of Charles University Prague;

P. Voříšek coordinates the Pan-European Common Bird Monitoring (PECBM);

Z. Vermouzek is at the ORNIS Muzeum Komenského Přerov;

K. Šťastný and V. Bejček are professors of the Department of Ecology, Faculty of Environmental Sciences of the Czech University of the Life Sciences Prague;

M. Koschová is a student at the Department of Ecology, Faculty of Science of Charles University Prague.