Habitat preferences of the Syrian Woodpecker Dendrocopos syriacus in urban environments: an ambiguous effect of pollution

Abstract

Capsule The occurrence of the Syrian Woodpecker Dendrocopos syriacus in an urbanized habitat is determined by the number of trees and pollutant emissions. Air pollution may weaken trees and increase the number of insects inhabiting them.

Aims To determine the habitat preferences of the Syrian Woodpecker in the highly urbanized environment of the city of Kraków.

Methods A set of 13 habitat and environmental parameters of 50 breeding territories and 50 random points were evaluated. A logistic regression was used to determine the probability of species occurrence, and Akaike's information criterion was used for model selection.

Results The number of trees, coverage of woody vegetation, total vegetation cover and level of pollutant emissions were significantly higher in Syrian Woodpecker breeding territories than in the random points. The model best explaining the probability of species occurrence consisted of four parameters: the number of trees, coverage of the total built-up area, total vegetation cover and pollutant emissions. The parameters best explaining a high probability of species occurrence were high number of trees and high pollutant emissions.

Conclusion Air pollution may weaken trees and potentially increase the number of insects inhabiting them, an important component of the Syrian Woodpecker's diet. However, the negative impact of air pollution on birds may impact on their condition. Urban habitats could, therefore, act as an ecological trap for this species.

Urban ecosystems have different compositions of bird species which change along the urban gradient (Blair 1996, Tomiałojć 1998, Clergeau et al. 2002). The main habitat parameters affecting the diversity of birdlife in towns and cities include the degree of urbanization, the types of buildings, the numbers and sizes of trees, and the extent and distribution of tree-covered areas (Melles et al. 2003, Palomino & Carrascal 2006, Sandström et al. 2006, Stagoll et al. 2012, Sushinsky et al. 2013). Factors, such as refuse management, bird feeders and drinkers, artificial nesting sites, and the maintenance of urban greenery also affect the presence and distribution of birds. Bird populations in urban areas are also limited by road traffic and the associated mortality (Forman et al. 2002, Bujoczek et al. 2011), collisions with buildings (Longcore et al. 2013), air pollution (Eeva et al. 2009, 2012, Berglund et al. 2011), vibration and noise (Nemeth et al. 2013, Proppe et al. 2013, Ríos-Chelén et al. 2013). Number and pressure of natural predators in urban areas are much lower than in natural and semi-natural habitats (Randa & Yunger 2006). However, in urbanized areas, there is ever-present disturbance and pressure on the part of humans and their pets (dogs, cats and rats) (Luniak 2004, Baker et al. 2008).

The Syrian Woodpecker Dendrocopos syriacus is the least known woodpecker species in Europe (Pasinelli 2006). The estimated global population of this species is 2.2–6.6 million individuals (IUCN 2012) and its numbers in Europe are estimated at 0.5–1.1 million breeding pairs, with the most numerous populations breeding in Romania, Bulgaria and Turkey. The Syrian Woodpecker is a relatively new species in urban environments in central and south-eastern Europe (Cramp 1985, IUCN 2012), including south-eastern Poland (Tomiałojc´ & Stawarczyk 2003). The expansion of this species began in the late 19th century, starting from the Balkans, from where the species reached Romania and Hungary in the 1930s and the former Czechoslovakia in the 1940s (Cramp 1985). The Syrian Woodpecker was first recorded in Poland in 1979 (Ciosek & Tomiałojc´ 1982). After three decades of expansion, the species' contiguous range now covers south-eastern Poland, and there are scattered breeding localities or records of non-breeding birds from practically the whole country (Tomiałojc´ & Stawarczyk 2003, Buczek 2004, Biaduń & Stachyra 2005). In Poland, this species has colonized wooded river valleys, farmland with clusters of trees and also towns and villages; its populations in urban areas can be fairly large (Luniak et al. 2001).

As woodpeckers are highly specialized birds, their occurrence is associated with a very specific set of environmental features (Roberge et al. 2008). The habitat preferences of woodland species are relatively well known: they generally relate to the type of woodland and the presence of sufficient quantities of dead and dying trees (Roberge et al. 2008). But the habitat preferences of a non-woodland species like the Syrian Woodpecker are almost completely unknown. In general, the species mainly inhabits the farming landscape and to a lesser extent, sparse woodlands in river valleys or woodland margins, avoiding closed-canopied forests (Cramp 1985). It is important to determine the specific habitat requirements of the Syrian Woodpecker to understand the reasons for the species' recent spectacular expansion in Europe and its possible further progress. Furthermore, being a primary hole-nester, the Syrian Woodpecker will affect the occurrence of secondary hole-nesters by supplying them with nesting sites (Szlivka 1981, Orchan et al. 2013), thereby potentially affecting the biodiversity of urban areas.

The aim of this work was to study the habitat preferences of Syrian Woodpeckers in the highly urbanized conditions of a city. We predict that in towns and cities this species, as a tree-dwelling and primary cavity nester, will prefer tree-covered areas and avoid densely built-up areas, which are typically devoid of large trees. We also predict that air pollution and road noise may have an adverse effect on its occurrence, as has been noted in other bird species.

METHODS

Study area

The study was carried out in Kraków (50°05′N, 19°55′E), the largest urban area in south-eastern Poland (Fig. 1). The city has a surface area of 327 km² and a human population density of 2310 persons/km² (GUS 2011). Kraków lies in a basin that extends along the valley of the River Vistula between the Kraków-Cze˛stochowa Upland and the Western Beskid Foothills (Kondracki 2000). The city is highly urbanized, and the buildings are of various types: tenements, residential areas with detached housing, high-rise blocks of flats, four-storey blocks of flats and scattered buildings on the city's outskirts.

Figure 1. Breeding territories of the Syrian Woodpecker (black dots) in the city of Kraków in 2002–2012 and random points (white dots).

These various forms of buildings are accompanied by different types of urban greenery, which usually comprises gardens (14.4% of the city area), road verges and squares (10%), allotments and orchards (4.3%), and parks and cemeteries (3%). There is also derelict land, often with trees and shrubs (4.7%), and commercial properties (14.7%) on which woody vegetation is infrequent. In the youngest parts of the city, there is arable (14.4%) and fallow land (13.3%), as well as undeveloped land, the vegetation of which consists of around 100 plant communities (Bartoszek et al. 2008). The urban greenery in Kraków consists of native and non-native trees and shrubs. A number of small rivers flow into the Vistula within the city boundaries – the Białucha, Rudawa, Dłubnia, Drwina, Długa and Wilga – along which there are larger or smaller areas of riparian or damp woodland (MIIP 2012). Air quality in Kraków is among the worst in Europe (AQE 2013). The city is surrounded by an agricultural landscape with numerous villages.

Woodpecker data

Hitherto unpublished Syrian Woodpecker records in Kraków between 1986 and 2012 were obtained from the Małopolska Ornithological Trust database, mailing lists and discussion forums on the region's birdlife and directly from bird watchers. Because, for many years, all records of the species had to be verified by the Polish Rarities Committee, every record of this woodpecker was noted by observers in detail. Most are from 2004 to 2006, when intensive field work was under way for the atlas of breeding birds in Kraków. In total, there were 225 sightings of the Syrian Woodpecker from 41 observers.

These data on Syrian Woodpeckers in Kraków were used to create a vector layer together with a database containing details of the records, that is, numbers, breeding status, age, sex, and behaviour. Observations more than 10 years old or relating to an inaccurately defined locality were rejected. Records indicative of certain or probable breeding were extracted and more than one record relating to the same locality were rejected to avoid pseudoreplication. All possible records gave a set of 50 localities where Syrian Woodpecker had certainly or probably bred between 2002 and 2012 and these were used in further analyses (Fig. 1). Since the peripheral part of the city might have low observer coverage influencing the number of records, comparative random locations were selected within the evenly covered main part of the city.

Environmental data

The environmental parameters of the breeding localities of Syrian Woodpecker and the comparative layer of 50 random points were defined on the basis of orthophotomap (GUGiK 2009), topographic map (WODGiK 2009), pollution emission map (MIIP 2012), noise emission map (Mrugała 2009) and the vegetation atlas of Kraków (UMK 2012).

The points on the comparative layer were randomly selected using Quantum GIS software (QGIS 2012) within the boundaries of the polygon delimited by the extreme Syrian Woodpecker localities in Kraków (Fig. 1). The position of the nesting hole or the centre of the sighting locality was assumed to be the centre of the territory. The exact size of a Syrian Woodpecker territory is not known, but published records suggest that it is fairly small with nest holes as little as 30 m apart (Cramp 1985). A buffer zone 140 m in radius (6.2 ha) was created for every point on the Syrian Woodpecker locality layer and the comparative layer. This buffer zone radius corresponds with the size of the core area of the species' territory and was calculated on the basis of the birds' reactions to playback stimulation conducted on sample plots (Fröhlich & Ciach 2013). Thirteen habitat and environmental parameters were defined within the buffer zones (Table 1).

Table 1. Habitat and environmental parameters measured in random points and those occupied by the Syrian Woodpecker in an urban environment (all measurements except WATER were taken in the buffer zone of radius 140 m).

Parameter	Code	Method of measurement (unit)	Data source
Length of roads	ROAD	Total length of hardened roads (m)	Orthophotomap
Length of railway tracks	RAIL	Total length of railway tracks (excluding tramlines embedded in roadways) (m)
Total length of roads and railway tracks	ROAD + RAIL
Area of low buildings	LOWBUILD	Total roof area of low buildings (m^2)
Area of tall buildings	TALLBUILD	Total roof area of tall buildings (m^2)
Built-up area	LOWBUILD + TALLBUILD
Number of trees	TREES	Total number of trees
Area of garden vegetation	GARDENVEG	Total area of garden vegetation and orchards (m^2)	Vegetation Atlas of Kraków
Area of woody vegetation	WOODVEG	Total tree cover (woodland, tree clusters, parks, cemeteries) (m^2)
Total vegetation cover	GARDENVEG + WOODVEG
Distance to water	WATER	The shortest distance between the buffer zone boundary and the boundary of a water body	Topographical Database
Road noise	NOISE	Maximum emission (class)	Map of road traffic noise emission in Kraków
Pollutant emission	POLLUTANT	Dominant level of emissions (class)	Map of pollutant emissions in Małopolska province

Linear vector layers were created to calculate road lengths (ROAD) and railway track lengths (RAIL) from the orthophotograph. These linear variables were considered as indicators of modification and use of the terrain. Every hardened road more than 2 m wide was marked on the ROAD layer. Larger areas of hardened surfaces were outlined. Road lanes divided by a green verge were also defined. Every visible railway track was marked on the RAIL layer, except for tracks embedded in a hard surface (e.g. tramlines in a roadway), which were treated as roads. The areas covered by low (LOWBUILD) and tall buildings (TALLBUILD) were measured using polygonal vector layers covering the roof areas of both types of buildings in the buffer zones. Buildings whose walls were visible on the orthophotograph, as well as those throwing a shadow longer than 4 m, were classified as tall (TALLBUILD), all other buildings were categorized as low (LOWBUILD). The number of trees within the buffer zones (TREES) was counted using a point vector layer, in which a dot marked every tree within a buffer zone that had a crown contrasting with its surroundings that was broader than 4 m or that had a visible shadow. Clusters of trees were divided based on the texture and colour of their crowns. All of the variables were mapped on the basis of the orthophotograph taken in 2009 and accessed via the WMS (Web Map Service) server in Geoportal resources (GUGiK 2009).

The areas of garden vegetation (GARDENVEG) and woody vegetation (WOODVEG) were calculated on the basis of polygonal vector layers in the vegetation atlas of Kraków (UMK 2012), a floristic inventory carried out in 2006 (Bartoszek et al. 2008). Woody vegetation (WOODVEG) was taken to include areas described as deciduous/mixed woodland, coniferous forest, natural scrub, other tree stands, various stages of forest succession on abandoned land, parks, greenery, squares, public green areas, playgrounds, sports grounds and greenery in cemeteries. All these include areas dominated with woodland-like vegetation. Garden vegetation (GARDENVEG) was assumed to include areas described as orchards, gardens and allotments, which are dominated with fruit-bearing trees and shrubs.

The distance to the nearest watercourse or water body (WATER) was the shortest distance between the buffer zone boundary and the boundary of the nearest body of surface water as defined in the Topographic Database (WODGiK 2009). It was measured with an accuracy of 2.5 m (MIIP 2012). Surface waters include all watercourses and reservoirs of the local watershed. The map does not, however, show all bodies that could be of importance to birds, such as drinking troughs and fountains.

Parameter NOISE was the maximum value in the buffer zone taken from the road noise map (Mrugała 2009). This map, produced by the Provincial Environmental Protection Inspectorate, Kraków, in cooperation with the Kraków City Hall, shows mean daily noise levels in 2007. It is accurate to 5 dB and has a resolution of 18 m.

Parameter POLLUTANT was the dominant surface value in the buffer zone read off the layer of pollutant emission with the resolution of 500 m (MIIP 2012). In the data source, pollutant emissions were divided into five classes (1: < 100 kg/year; 2: 100–500 kg/year; 3: 500–1000 kg/year; 4: 1000–5000 kg/year and 5: > 5000 kg/year). An inventory of ground emission sources was carried out in 2010 and covered residential buildings, service buildings, public facilities, natural sources, agriculture and fugitive (unintended) emissions (Chyra et al. 2010). The dominant sources of pollution in Kraków are domestic heating systems and road traffic (WIOS´ 2012).

Statistical analyses

Mean values and standard deviations of each environmental variable were calculated for Syrian Woodpecker breeding territories and the random points. The differences between the variables were analysed using Student's t-test. Variables that did not have a normal distribution were expressed in logarithmic form (LOWBUILD, TALLBUILD, WOODVEG, GARDENVEG and WATER).

To eliminate multicollinearity, a Pearson correlation matrix of all variables was prepared. Strongly correlating pairs of variables (r > 0.4) of lesser biological significance were excluded from further analysis. From the initial set of variables, ROAD + RAIL, TREES, LOWBUILD + TALLBUILD, WOODVEG + GARDENVEG and POLLUTANT were used for constructing the model. Owing to the dichotomous nature of the dependent variable (territory occupied/unoccupied), logistic regression was used to determine variables predicting the probability of Syrian Woodpecker occurrence. Akaike's information criterion (AIC) was used for model selection (Burnham & Anderson 2002).

Because of the relatively small sample size (n/K ratio < 40), a small-sample version of AIC with bias adjustment (AICc) was applied in the modelling. The resulting models were subsequently ranked in order of increasing AICc. The differences between the models with the lowest AICc were calculated (ΔAICc) for each of the resulting models. Model likelihoods were normalized according to Akaike weights (w) to illustrate the weight of evidence of each model. The set of models with ΔAICc < 15 is presented.

Multimodel inference for all the candidate models was applied to evaluate the importance of each model predictor. AICc weights were summed for models containing a given variable. The predictor with the largest weight was considered the most important (Burnham & Anderson 2002). Logistic regression for the two most important variables was run to detect threshold values determining the presence of the species. Statistical procedures were performed using statistica 8.0 software (StatSoft 2008).

RESULTS

Significant differences were found between Syrian Woodpecker breeding territories and random points with respect to the number of trees (TREES), woody vegetation cover (WOODVEG), total vegetation cover (WOODVEG + GARDENVEG) and pollutant emissions (POLLUTANT) (Table 2). All these parameters had higher values in Syrian Woodpecker breeding territories. In contrast, there was a trend for the total lengths of roads and railway tracks (ROAD + RAIL) to be greater in the random points.

Table 2. Descriptive statistics and Student's t-test results for all variables analysed in the Syrian Woodpecker breeding territories and random points in an urbanized environment (Kraków, S Poland; for parameters, see Table 1).

Parameter	Occupied		Random		t	P
	Mean	± sd	Mean	± sd
ROAD	606.8	371.4	724.0	532.3	−1.3	0.20
RAIL	57.5	141.7	113.8	278.2	−1.3	0.21
ROAD + RAIL	664.3	359.9	837.9	606.2	−1.7	0.08
LOWBUILD	3917.5	3331.6	3900.4	4351.8	−0.3	0.74
TALLBUILD	2135.6	3928.6	3108.4	4820.9	−0.7	0.47
LOWBUILD + TALLBUILD	6053.1	4978.5	7008.9	6397.0	−1.2	0.22
TREES	109.1	43.0	69.7	36.4	4.9	<0.001
WOODVEG	27 547.9	24 462.1	18 412.9	21 002.6	2.5	0.013
GARDENVEG	19 511.6	25 104.5	11 400.0	15 791.9	1.0	0.33
WOODVEG + GARDENVEG	47 059.6	24 991.2	29 812.9	21 573.3	3.4	0.001
WATER	301.0	371.1	257.8	310.1	−0.8	0.41
POLLUTANT	3.5	1.1	2.7	1.1	3.7	<0.001
NOISE	7.7	2.8	7.8	2.8	−0.2	0.83

The model best describing the probability of Syrian Woodpecker occurrence (Table 3) consisted of four environmental parameters: the number of trees (TREES), the coverage of total built-up area (LOWBUILD + TALLBUILD), total vegetation cover (WOODVEG + GARDENVEG) and pollutant emissions (POLLUTANT). Two other useful models (ΔAICc = 4.2–4.3) consisted of combinations of three of these parameters. The parameters best explaining the probability of Syrian Woodpecker occurrence were the number of trees and pollutant emissions (Table 4): the higher the values of these two parameters, the more likely the occurrence of the species (Fig. 2).

Figure 2. Logistic regression models of the probability of occurrence of the Syrian Woodpecker and the number of trees and pollutant emission (classes – see Methods) in an urban environment (Kraków, S Poland).

Table 3. Sets of candidate models explaining the occurrence of Syrian Woodpeckers in an urbanized environment (Kraków, S Poland; for parameters, see Table 1).

No.	Model	AICc	ΔAICc	w
1	TREES + (LOWBUILD + TALLBUILD) + (WOODVEG + GARDENVEG) + POLLUTANT	100.2	0.0	0.745
2	TREES + (LOWBUILD + TALLBUILD) + POLLUTANT	104.4	4.2	0.092
3	TREES + (WOODVEG + GARDENVEG) + POLLUTANT	104.5	4.3	0.086
4	TREES + POLLUTANT	106.3	6.1	0.035
5	TREES + (ROAD + RAIL) + (LOWBUILD + TALLBUILD) + POLLUTANT	106.5	6.3	0.031
6	TREES + (LOWBUILD + TALLBUILD) + (WOODVEG + GARDENVEG)	110.3	10.1	0.005
7	TREES + (WOODVEG + GARDENVEG)	111.6	11.4	0.002
8	TREES + (ROAD + RAIL) + (LOWBUILD + TALLBUILD) + (WOODVEG + GARDENVEG)	111.9	11.8	0.002
9	(LOWBUILD + TALLBUILD) + (WOODVEG + GARDENVEG) + POLLUTANT	114.2	14.1	0.001

Akaike's information criterion for small samples (AICc), difference between the given model and the most parsimonious model (Δ) and Akaike weight (w) are reported for each model.

In urban environments, the Syrian Woodpecker requires a lower number of trees in areas of higher pollution. According to the logistic regression, a high occurrence probability (0.8) occurred in both low polluted (emissions < 100 kg/year) areas with ca. 200 trees within the core area of the territory and highly polluted areas (emissions > 5000 kg/year) with ca. 100 trees (Fig. 2).

DISCUSSION

In the temperate latitudes, the Syrian Woodpecker mainly inhabits a farmed landscape, that is, orchards, clusters of trees among fields, roadside tree lines, parks, cemeteries, gardens and, to a lesser extent, natural habitats like damp woodlands in river valleys or woodland margins (Cramp 1985, Buczek 2004, 2007, Michalczuk & Michalczuk 2011). Even when it colonizes new areas, the Syrian Woodpecker avoids closed canopy woodlands, a feature that distinguishes it from other Dendrocopos species in Europe (Cramp 1985).

In towns and cities, the occurrence of birds is governed by the type and age of built-up areas, which support a particular type of vegetation (Melles et al. 2003, Sandström et al. 2006, Stagoll et al. 2012). In Sweden, the number of woodpeckers in the municipality of Örebro was found to depend on the availability of dying trees, the number of which increased with distance from the city centre (Sandström et al. 2006). In an urban environment, there is little dead wood, because this is quickly removed by the inhabitants or the city services. Moreover, the relatively few trees in such an environment tend to be scattered, and the open-canopied tree cover of this kind possibly resembles a forest steppe or the Mediterranean-type vegetation typical of the Middle East, which is, of course, from where the Syrian Woodpecker began its expansion.

The influence of the number of trees and of pollutant emission on the probability of Syrian Woodpecker occurrence points to the key role played by food resources in the choice of breeding territory. Air pollution directly and indirectly (through soil) (Sharma et al. 2012, WIOS´ 2012) affects the health of trees (Bach & Pawłowska 2007), weakening them and rendering them more vulnerable to pathogens. The changes brought about in an environment by air pollution may alter the species composition and increase the size of its insect populations (Hain 1987, Koricheva & Haukioja 1992, Koricheva & Haukioja 1995, Connor et al. 2002, Yun et al. 2002, Gao et al. 2008). Air pollution may act in favour of species inhabiting the cambium of trees (Grodzki et al. 2004), leading to an increase in body size of phytophages such as the caterpillars of certain moths (Koricheva & Haukioja 1992, Koricheva & Haukioja 1995, Yun et al. 2002). And such insects living beneath the tree bark and foraging on leaves are the principal constituents of Syrian Woodpecker diet (Cramp 1985).

The Syrian Woodpecker's preference for areas with a greater degree of pollution may turn out to be an ecological trap, however. The pollutants could enter and accumulate in woodpeckers' bodies due to direct exposure and by the consumption of harmful elements or compounds in food (Eeva et al. 2009, 2012, Berglund et al. 2011). Birds inhabiting urban environments may be sink populations, which only persist because of immigration of birds from non-urban environments (Pulliam 1988, Pulliam & Danielson 1991). But the degree of pollution, breeding success or survival rate of urban populations of Syrian Woodpeckers is not known, and so the sink nature of their urban populations is uncertain.

The influence of the availability of tree-covered areas on the probability of Syrian Woodpecker occurrence points to the importance of the spatial distribution of trees. In an urban environment, trees are present in almost every type of built-up area, but for Syrian Woodpeckers it is the trees growing in clusters (parks, gardens and cemeteries) that are probably important. A low or even moderate building coverage does not appear to have a significant effect on the quality of Syrian Woodpecker habitat, in contrast to other bird species (Palomino & Carrascal 2006). But the presence of Syrian Woodpeckers in densely built-up areas is not very likely and the marginal significance of total transport network within a territory indicates that this species tends to avoid locations with extensively developed city infrastructure.

The lack of any preference on the part of the Syrian Woodpecker regarding distance from water may be due to the species' adaptation to arid environments. The urban ecosystem is relatively dry and, in this respect, resembles the ecosystems from which this species began its expansion. In arid areas, Syrian Woodpecker has been observed to perforate irrigation pipes (Moran 1977, Moran 1981, Barnea & Yom-Tov 1984), which shows that it can make use of anthropogenic sources of water. An additional water source may be fruit, an item that is more frequent in the diet of this species than in other woodpecker species (Cramp 1985, Michalczuk & Michalczuk 2005).

Table 4. The AICc weights for each variable used in the model selection procedure (see Table 3).

Parameter	P
TREES	0.999
POLLUTANT	0.990
LOWBUILD + TALLBUILD	0.875
WOODVEG + GARDENVEG	0.841
ROAD + RAIL	0.034

The values are the probabilities of a given predictor being in the best approximating model.

A high level of noise may be a factor determining how birds communicate in urban environments: they may produce sounds of higher frequency or sing for longer periods of time (Nemeth et al. 2013, Proppe et al. 2013, Ríos-Chelén et al. 2013). But the effect of noise has been studied mainly in songbirds. In the case of woodpeckers, which communicate by drumming, noise may not have any significant effect, or at least may not govern the choice of territory. The lack of any preference in the Syrian Woodpecker regarding acoustic conditions or the presence of roads or railway lines in their territories may be a sign of its tolerance to strongly urbanized and noisy environments.

The Syrian Woodpecker finds suitable living conditions in highly urbanized environments and its urban populations may make up a sizeable proportion of regional or national populations (Fröhlich & Ciach 2013). In Poland, this species inhabits mainly river valleys and farmland, and its population is estimated at 1000–2000 breeding pairs (Buczek 2004, Michalczuk & Michalczuk 2011). The largest urban populations in Poland consist of 10–40 pairs (Luniak et al. 2001, Tomiałojc´ & Stawarczyk 2003, Biaduń & Stachyra 2005), but these numbers are probably substantially underestimated (Fröhlich & Ciach 2013). The high density in towns may be due to the numbers of parks and allotments with relatively large numbers of trees exposed to air pollution. As urban populations have never yet been monitored, their dynamics remains an unknown quantity. However, the vocal activity of urban populations of Syrian Woodpecker may differ from that in the mosaic of rural and farming areas (Michalczuk & Michalczuk 2006). Syrian Woodpeckers are relatively shy and silent in urban areas and low vocal activity may contribute to monitoring difficulties. Moreover, the Syrian Woodpecker may form a mixed pairs with the Great-spotted Woodpecker Dendrocopos major (Dudzik & Polakowski 2011), and the lack of diagnostic features in the hybrid offspring makes assessing the state of the population problematic (Randler 2004).

The Syrian Woodpecker has become another bird species breeding in the towns and cities of central and south-eastern Europe as a result of its spectacular expansion in the past 100 years. Its densities in various kinds of urban habitat may be relatively high, and urban populations can make up a significant proportion of regional populations. In towns and cities, the Syrian Woodpecker may exploit tree-covered areas that are seriously affected by air pollution. One reason for the ongoing expansion of this species may be its colonization of a niche hitherto unoccupied by other woodpecker species. On the other hand, one should not rule out the possibility that its urban populations are merely a sink for the fast-growing populations in agricultural areas. The habitat preferences of Syrian Woodpeckers suggest that the most likely scenario in the coming years will be its continuing expansion and colonization of urban areas in Western Europe. The quality of the environment and its contamination may be playing a significant albeit equivocal part in this phenomenon.

ACKNOWLEDGEMENTS

We wish to express our gratitude to M. Albrycht, B. Binkiewicz, B. Brożek, M. Bujoczek, A. Chwierut, K. Czajowski, B. Czerwiński, S. Czyżowicz, M. Ejsmond, F. Erhart, M. Faber, A. Felger, P. Filimowski, T. Folta, P. Guzik, Ł. Kajtoch, M. Kata, M. Keppert, G. Kopka, W. Król, K. Kucharska, D. Kurlej, K. Kus, J. Lachman, P. Malczyk, J. Marciniak, R. Martyka, S. Mazgaj, W. Mrowiec, K. Paciora, W. Solarz, K. Stępniewski, T. Szewc, E. Szumakowicz, D. Świta, S. Tworek, K. Walasz, D. Wiehle, T. Wilk, J. Wójcik and S. Wójcik, who provided records of the Syrian Woodpecker from Kraków. We are grateful to the anonymous referees for their constructive comments on the manuscript.