This article appeared in a journal published by Elsevier. The attached

This article appeared in a journal published by Elsevier. The attached

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and educat...

512KB Sizes 0 Downloads 5 Views

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright

Author's personal copy Landscape and Urban Planning 108 (2012) 49–56

Contents lists available at SciVerse ScienceDirect

Landscape and Urban Planning journal homepage: www.elsevier.com/locate/landurbplan

Research paper

Unintentional habitats: Value of a city for the wheatear (Oenanthe oenanthe) Peter J. Meffert a,∗ , John M. Marzluff b , Frank Dziock c a

Institute for Community Medicine, University Medicine Greifswald, Walther-Rathenau-Straße 48, D-17475 Greifswald, Germany College of Forest Resources, University of Washington, Box 352100, Seattle, WA 98195, USA c Chair of Animal Ecology, Dresden University of Applied Sciences (HTW Dresden), Faculty of Agriculture/Landscape Management, Pillnitzer Platz 2, D-01326 Dresden, Germany b

h i g h l i g h t s    

An endangered open-land bird species occurs and reproduces well on urban wastelands. Its occurrence depends mainly on the size of the habitat and vegetational features. Location within the city and landscape context did not matter. Urban planners should consider wastelands to enhance urban biodiversity.

a r t i c l e

i n f o

Article history: Received 15 September 2011 Received in revised form 23 July 2012 Accepted 30 July 2012 Available online 24 August 2012 Keywords: Wasteland Urban matrix Breeding success Grassland birds Boosted regression trees

a b s t r a c t Humans are rapidly reducing and isolating the habitats of native species such as the wheatear (Oenanthe oenanthe) through urbanisation and agricultural intensification. The wheatear, a small songbird, has declined dramatically throughout Europe. It is known to live in alpine meadows, in tundra, and rural landscapes as well as in urban areas, but it is unknown if these urban populations reproduce sufficiently. This study aimed to investigate reproductive success and habitat requirements of the wheatear in the city of Berlin, Germany. We analysed occurrence and breeding success in relation to vegetation and surface structure of the settled sites, intensity of direct disturbance by humans and dogs, as well as degree of sealing and residential population density in the surrounding urban matrix. Finally, we compared early-settled to late-settled territories to appraise habitat preferences of wheatears. The proportion of successful nests was high (73%) compared to other regions and habitats. Area size of a site greatly affected the probability of wheatear occurrence; it was much higher on sites larger than five hectare. Factors affecting breeding success differed completely from those explaining variation in occurrence, indicating that breeding success seems not to be related to habitat preferences. There was no influence of the urban matrix at a landscape level (200 m and 2000 m zone) on occurrence or breeding success. To maintain and create habitat for endangered open-land species as the wheatear, we recommend minimum area size of five hectare, sparse vegetation, open soils, short grass, and very few trees and shrubs. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Human domination of terrestrial systems is reducing biological diversity (Vitousek, Mooney, Lubchenco, & Melillo, 1997). In particular, urbanisation reduces and isolates native habitats (Benfield, Raimi, & Chen, 1999), a relationship particularly well demonstrated with respect to avian diversity (Marzluff, 2001). At the same time, agricultural intensification causes declines of open-land birds (Donald, Green, & Heath, 2001). Europe’s farmland bird index decreased by 49% from 1980 to 2008 (EBCC, 2010).

∗ Corresponding author. Tel.: +49 15201749393. E-mail addresses: [email protected] (P.J. Meffert), [email protected] (J.M. Marzluff), [email protected] (F. Dziock). 0169-2046/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.landurbplan.2012.07.013

Migratory species suffer additionally from habitat alterations elsewhere. Most strongly affected are migrants wintering in southern Africa (Ockendon et al., 2012) but also species wintering in dry, open habitats in African Sahel show severe declines within the last decades, since intensified land use, widespread desertification and habitat degradation seem to reduce the quality of wintering and stopover habitats (Sanderson, Donald, Pain, Burfield, & van Bommel, 2006). The wheatear (Oenanthe oenanthe), a trans-Sahara migrant wintering in Western, Central and Eastern Africa is also following this trend (BirdLife International, 2004); in Germany it is threatened with extinction (Südbeck, Bauer, Boschert, Boye, & Knief, 2007). While populations in natural habitats of the mountain regions appear stable, those in man-made habitats of the agricultural landscape declined dramatically during the last decades (Glutz von

Author's personal copy 50

P.J. Meffert et al. / Landscape and Urban Planning 108 (2012) 49–56

Blotzheim, 1985). Wheatears live in open habitats with short grassy vegetation. They are known to inhabit alpine meadows, tundra and rural landscapes like vineyards, as well as urban areas in parts of Europe. Urban habitats differ from non-urban habitats in many respects, such as vegetation, predation risk, or food availability (Chace & Walsh, 2006) that have been found to affect reproduction (Chamberlain et al., 2009). It is unknown if urban wheatear populations are able to reproduce sufficiently to offset mortality. Factors related to urbanisation and affecting bird presence are diverse, including land cover change (Jones & Bock, 2002), vegetation volume (Mills, Dunning, & Bates, 1989), human activities (van der Zande & Vos, 1984), risk of collision with man-made objects (Klem, 1990), food supply (Emlen, 1974), diseases (Brittingham & Temple, 1986), predation risk (Jokimäki & Huhta, 2000), and several others (Chace & Walsh, 2006). Work that focuses on the relationship between the degree of urbanisation and breeding success of birds are contradicting (Marzluff, 2001). Several species show a positive response in reproductive parameters to urbanisation, such as blackbirds (Turdus merula; Luniak & Mulsow, 1988), magpies (Pica pica; Baeyens, 1981), or wrens (Troglodytes troglodytes; Wesolowski, 1983). Others show a negative impact of urbanisation on reproduction, such as great tits (Parus major; Schmidt, 1988) or starlings (Sturnus vulgaris; Savard, Clergeau, & Mennechez, 2000). Most of these studies, however, examine sedentary and omnivorous bird species that could profit from additional artificial food supply (Chamberlain et al., 2009; Marzluff, 2001). In contrast, the wheatear is an insectivore, thus it does not exploit artificial food sources. As suggested by Chamberlain et al. (2009), paucity of natural food may lead to lower productivity in urbanised areas, as shown for the house sparrow (Passer domesticus; Peach, Vincent, Fowler, & Grice, 2008). However, the wheatear was detected in eleven of 16 European cities with more than 100,000 inhabitants; out of eight mega cities it was only missing in one (Kelcey & Rheinwald, 2005), indicating a high potential for urban areas to provide habitat for this species. The wheatear can be thought of as a model species for the group of endangered, insectivorous open-land bird species. Typically, urban-adapted birds are not endangered and rather omnivorous or granivorous and sedentary (Chace & Walsh, 2006). To the best of our knowledge this is the first systematic and detailed study on an open-land bird living in a city. Our objective is to shed light on crucial drivers and mechanisms that influence reproduction as one aspect of fitness and population dynamics. We aim to investigate if the city of Berlin provides habitat for a successfully reproducing wheatear population by comparing reproductive success to data in the literature on populations in non-urban areas. Therefore, we analysed factors that could potentially influence (1) occurrence and (2) reproductive success of wheatears, and (3) we compared early-settled territories with late-settled ones to approximate preferences of wheatears in territory choice since preferred territories are occupied earlier than others (Brooke, 1979).

2. Methods 2.1. Study area and site selection The study was conducted in Berlin, Germany. With an annual mean temperature of 9.8 ◦ C and a precipitation of 582 mm (DWD, 2006), the city of Berlin is situated in the transition zone between continental and oceanic temperate climate within the North German Plain. Berlin has 3.4 million inhabitants and an area of 892 km2 . We selected 55 wastelands including all large accessible wastelands (>10 ha) and chose smaller ones in order to cover the main environmental gradients of interest, namely area of the wasteland, human disturbance on the study site and degree of urbanisation of the surrounding urban matrix. We defined a wasteland as any

area without actual use and with no or few buildings, e.g. former railway areas or vacant lots. The sites had a size of 0.1–25 ha and were spread over the city, but mainly in the eastern part and in the city centre due to historical reasons. After World War II, the enclave of the western part of the city was short of building area and only few wastelands have been left. The eastern part spread by building of outer city council estates. After the reunification of the divided parts of the city in 1990, lots of empty spaces arose near the former border and from deconstruction of industrial plants and railway properties. 2.2. Study species The wheatear (also called northern wheatear, O. oenanthe) is a species with an extensive range, but nevertheless globally and locally endangered (Baillie, Hilton-Taylor, & Stuart, 2004; Südbeck et al., 2007). Natural habitats are tundra and alpine meadows above timberline spread over almost the complete Palearctic and north-western Nearctic. Man-made habitats like pastures, vineyards, stone pits or mining areas are also used (Buchmann, 2001; Grimm, 2004; Tye, 1980). The wheatear is an insectivorous longdistance migrant wintering south of the Sahara desert from Senegal to Sudan. Its population has declined dramatically in Germany as well as in other countries in Europe (BirdLife International, 2004) and it is listed on the IUCN Red List of Threatened Species (Baillie et al., 2004). In Germany it is categorised as ‘threatened by extinction’ (Südbeck et al., 2007). 2.3. Bird survey Our fieldwork included two different approaches. For the analysis of occurrence we determined if wheatears were present or absent from 55 wastelands. In 2007 we conducted a breeding bird survey using territory mapping by visiting each of the 55 sites four times from the end of April to the end of July. Additionally, we surveyed breeding territories of wheatears in the whole area of the city to gather data on breeding success. To gather data on breeding success we searched intensively for occupied territories and nests in the whole area of the city by asking local ornithologists, searching for potential suitable habitats such as large, sparsely vegetated areas on aerial photographs and inspecting sites of known occurrences. We found 54 nests in 2007 and 41 nests in 2008. For a complete list of sites and number of breeding pairs see Appendix B. Number of breeding pairs found was higher than in the previous years (BOA, 2005; BOA, 2006; BOA, 2007) since wheatears often use unusual locations for breeding that are not attractive to ornithologists and thus easily overlooked. We assume to have surveyed the city almost completely, with only few pairs uncovered or inaccessible. Territories were inspected at least once a week. We defined presence of wheatears as at least one breeding attempt and assessed nesting attempts as successful if young fledged, that is have been seen outside the nest. We delineated territories by observing foraging bouts by birds and calculated territory sizes using ArcMap 9.2. 2.4. Environmental data For predicting occurrence and breeding success, we mapped vegetation and surface structure with eleven categories of vegetation height and five categories for different surface materials (see Appendix A for a complete list), since vegetation height is crucial for prey availability and attractiveness of territories (Tye, 1980). Growth of herbaceous plants within the same year was estimated visually in four categories (very low, low, medium, and high). For the occurrence analysis we obtained a measure for

Author's personal copy P.J. Meffert et al. / Landscape and Urban Planning 108 (2012) 49–56 Table 1 Intensity of disturbance within the breeding territories by human intrusion. Code

Disturbance intensity

Frequency of human intrusion

0 1 2 3 4

No disturbance Low disturbance Medium disturbance Heavy disturbance Very heavy disturbance

Never or very rarely On or two times per week About every second day Daily Several times a day

vegetation/surface heterogeneity by mapping 30 randomly distributed one-square-metre subsamples per site, and 50 for areas larger than 20 ha, respectively. We estimated proportions of surface category for each subsample and calculated Simpson’s Index for heterogeneity. For breeding success, location of nests was categorised into five categories (gabion, hole near ground, concrete slabs/railway sleeper, heap of stones/boulder, and other). Numbers of humans and dogs were counted as proxies for direct disturbance intensity. Since each site was also inspected in the afternoon, this gives a good approximation for different recreational uses. We calculated a frequency index for humans as the mean number of humans per visit multiplied by the degree of presence, that is, one to four visits where we were able to detect humans. The same procedure was used to calculate disturbance index for unleashed dogs. Intensity of disturbance within territories of wheatears in terms of human intrusion was estimated in five classes (Table 1). Finally, presence of cats was also mapped simultaneously with bird surveys and encoded as presence/absence. The size of adjacent suitable habitat was estimated according to previous described habitat requirements such as short and sparse vegetation and openness. Our proxies for the degree of urbanisation of the surrounding matrix are degree of sealing, that is, impervious surface and residential population density, since these measurements imply many structural features of an urban matrix. Data were taken from Environmental Atlas from the municipality of Berlin (Senate Department for Urban Development, 2004; Senate Department for Urban Development, 2006). Proportion of sealed area and residential population density was calculated for three buffer zones around each study site (50 m, 200 m, and 2 km) using ArcMap 9.2. Size of adjacent suitable habitat around territories was estimated from aerial photographs (Senate Department for Urban Development, 2005) and site inspections.

51

can fit complex nonlinear relationships, handle interaction effects between predictors automatically, and show a high predictive performance (Elith, Leathwick, & Hastie, 2008). In boosted regression trees high numbers of predictor variables and collinearity do not cause numeric aberrations and thus provide a useful basis for interpretation (Friedman & Meulman, 2003). We used the library gbm (version 1.6-3; Ridgeway, 2007), implemented in the software R (R Development Core Team, 2009). Variables which did not contribute to the prediction were removed using a dropping procedure (Elith et al., 2008). This procedure starts with an initial full model including all variables and then removes variables using ten-fold cross-validation. In every step, it removes the lowest contributing variable, computes the change in predictive deviance and compares it to that obtained when using all predictors. It repeats this process and gives a function that is afterwards used to drop as many variables as do not change predictive deviance towards more positive values. Measure of model performance was cross-validated correlation, proportion of explained deviance, and AUC (area under curve of the receiver operation characteristic, ROC). Tree complexity was set to two, allowing interactions between two predictors to be modelled. Learning rate was set to 0.001 to get some thousand trees and bag fraction was varied in order to maximise predictive performance. We calculated the start of breeding as Julian day (January 1st = 1). Where not directly observed, egg-laying date was estimated by assuming incubation took 14 days and nestling period 12 days (Glutz von Blotzheim, 1985). Data for 2008 were corrected (plus five days), because the median day on which incubation began was five days later compared to 2007, probably due to weather conditions. Re-laying attempts were excluded. Second broods were not detected. To assess habitat preferences of wheatears we analysed date of start of breeding. It is expected generally that preferred territories are occupied earlier than others (Fretwell & Lucas, 1969). In the wheatear, number of young fledged is significantly negatively correlated with hatching date (Brooke, 1979). For the analysis of breeding success combined data from 2007 and 2008 were used, whereas territories with spatial overlap between the two years were excluded to avoid pseudo-replication. There was no spatial autocorrelation in breeding success or occurrence of wheatears (Moran’s I: p = 0.91 and p = 0.77, respectively). For a list of all predictor variables of the three models see Appendix A.

2.5. Data analysis 3. Results Statistical analysis predicting (1) presence, (2) breeding success, and (3) start of breeding were accomplished using boosted regression tree analysis (BRT; Ridgeway, 2007). Regression tree methods split the dataset several times into two parts that are most different. The splitting points strongly depend on the distribution of the data and can change considerably with minor changes in the data. This problem is overcome by boosting. Boosting is a technique that combines many weak classifiers, in this case several hundred to thousand truncated trees, to obtain a much more reliable prediction compared to one single highly branched tree. The truncated trees are obtained by selecting randomly only a part of the data whereas the so-called ‘bag fraction’ remains unconsidered. This is repeated some thousand times to get an ensemble of truncated trees. Thereby, the model is ‘boosted’ towards the unexplained deviance optimizing the predictive performance of the final model. This is controlled by the learning rate, also called shrinkage factor. Overfitting is avoided by the use of cross-validation that checks, how well the submodels that were built on one part of the data can predict the left-over part, the bag fraction. Boosted regression trees are known to combine several advantages – they

3.1. Occurrence of wheatears Out of the 55 sites 14 were occupied by wheatears. The final model predicting wheatear occurrence consists of eight predictor variables (Fig. 1). Area size had the strongest influence on occurrence probability. Wheatears only occurred on sites larger than 3.3 ha. Modelled occurrence probability was zero at sites up to three hectares and increased to more than 70% at seven hectares (Fig. 1a). The other variables caused a change in predicted occurrence probability of about nine to two per cent each (Fig. 1b–h). The model explained 85% of the total deviance (Table 2). Sand cover of more than 40% enhanced wheatears occurrence probability by 6% (Fig. 1b). Wheatears were least likely to occur in sites with high trees, moss, shrubs, and grass of 10–20 cm height (Fig. 1c–e and h). Our proxies for urbanisation of the vicinity around each study site, sealing and residential population density, enhanced the model for occurrence, but relative influences were weak (Fig. 1f and g). Degree of sealing and population density explained only 3.1% and 1.0% each, showing reverse effects: Wheatears were least

Author's personal copy

40

60

8

20 10

20

40

60

80

Proportion sealed within 50 m (%)

0

10

20

30

40

Moss cover (%) h

10 0

10 0

0

3.1 %

0

20

Tree cover > 4 m height (%) g 20

20 10 0

20 10 0

4

Shrub cover > 1 m height (%)

10 0

f

3.2 % 0

0

80

4.9 %

20

20

Sand cover (%)

e

0

Occurence probability (%)

20

10 15 20 25

Size of the site (ha)

6.3 %

1.0 %

0

5

10

20 10 0

7.4 %

0

0

73.2 %

d

c

b

0

100

200

Inhabitants/ha within 50 m

0.9 %

0

a

40

80

P.J. Meffert et al. / Landscape and Urban Planning 108 (2012) 49–56

Occurence proba ability (%)

52

0

10

20

Grass cover 10-20 cm height (%)

Fig. 1. Marginal plots of the modelled occurrence probability of wheatears in relation to environmental factors predicting its presence. Percentage inside the boxes is relative contribution of the variable. Ticks on the top indicate deciles of the data distribution. Note that the scale of the y axis of the first plot (a) differs from the others. For model performance see Table 2.

Table 2 Summaries of the three models; CV, cross-validated; se, standard error; AUC, area under curve of the receiver operating characteristic. Model

Dependent variable Distribution Bag fraction Number of trees Mean total deviance Mean residual deviance Explained deviance Estimated CV deviance (se) Training data AUC Training data correlation CV AUC score (se) CV correlation (se)

Occurrence

Breeding success

Start of breeding

Breeding attempt (yes/no) Bernoulli 0.9 3750 1.13 0.18 84.6% 0.68 (0.15) 1.000 0.97 0.938 (0.050) 0.71 (0.10)

Breeding success (yes/no) Bernoulli 0.7 1900 0.89 0.63 29.5% 0.76 (0.07) 0.910 0.58 0.880 (0.052) 0.46 (0.11)

Julian Day of start of breeding Gaussian 0.5 5650 68.4 41.5 39.3% 55.5 (11.6) – 0.64 – 0.51 (0.10)

likely to occur in sites with high densities of residential population within 50 m around the site, but degree of sealing was positively correlated with occurrence probability. Influence of urbanisation was restricted to the 50 m buffer zone, whereas residential human density and degree of sealing within 200 m and 2 km did not explain the occurrence. Variables for disturbance by humans, dogs, or cats had no measurable influence and were thus excluded. Wheatears bred in a wide array of land covers in Berlin, among them wastelands (often former railway properties), active railway areas, construction sites, parks, and storage yards. A large proportion of pairs were found on a former airport site that was 2002 transformed to a country park that includes a nature reserve. 3.2. Breeding success Overall breeding success was 73% for nests found not later than during incubation. If we consider all nest found at the latest at early nestling’s phase, 79% of nests were successful (Table 3). Breeding success was similar between years (chi-square test, p = 0.9). Human activities were the most frequent cause of loss. Out of 15 nest failures, seven happened on construction sites, where heaps of rubble and stones were altered by human activity; six were caused by animals, from which three can be identified: one each by Eurasian jay (Garrulus glandarius) and grass snake (Natrix natrix; direct observation), another by a small mammal (identified by feeding mark),

one nest collapsed and one loss remained ambiguous. Thus, overall predation rate was 6.9% (6 of 87). Magpies and hooded crows (Corvus cornix) have been observed watching the entrances of nests with nestlings, but a successful pursuit was not noticed. The boosted regression tree model predicts higher breeding success with earlier breeding (Fig. 2a), more than 2% cover of high grass (Fig. 2b), and low disturbance intensities (Fig. 2c).

3.3. Preferences in territory choice We modelled the date of first egg deposition as a proxy for habitat quality. Early breeding indicates territories of high quality. Territories with a minimum proportion of 20% sealed within a 50 m buffer zone around the territory were predicted to breed about five days earlier (Fig. 3a). Similarly, territories with earlier breeding attempts had more than 10% very short grass cover (Fig. 3b).

4. Discussion Wasteland and brownfield sites within urban agglomerations are important for wildlife, which has been demonstrated for insects (Strauss & Biedermann, 2006). As our results show, urban wastelands may also be an important habitat for open-land birds such as the wheatear.

Author's personal copy P.J. Meffert et al. / Landscape and Urban Planning 108 (2012) 49–56

53

Table 3 Numbers of nests found and proportions of successful nests. Prop., proportion. 2007

Nests with eggs found Nests with eggs or chicks found Nests with eggs found, re-laying Nests with chicks found, re-laying Nests with young fledged, found during incubation Nests with young fledged, found during nestlings phase

2008

Total number

Prop. success

Total number

Prop. success

Total number

Prop. success

21 42 4 4 25 46

67% 76% 100% 100% 72% 78%

18 39 2 2 20 41

72% 82% 100% 50% 75% 80%

39 81 6 6 45 87

69% 79% 100% 75% 73% 79%

b

39.7 %

37.4 %

25 April 10 May 25 May

Date of first egg depositio deposition n

0

10

20

Grass cover 20-50 20 50 cm height (%)

22.9 %

70

80

80

90

90

c

70

70

80

90

a Breeding B success (%)

Total

0

1

2

3

4

Disturbance intensity inten sity

b 52.6 %

47.4 %

May 10

14

a

6

Prediccted date of first eg gg deposition n

Fig. 2. Marginal plots of the modelled breeding success in relation to predictor variables. Percentage inside the boxes is relative contribution of the variable. Ticks on the top indicate deciles of the data distribution. For disturbance intensities see Table 1.

0

20 40 60 80

Proportion p sealed within 50 m (%)

0

20

40

60

Grass cover ≤ 5 cm height (%)

Fig. 3. Marginal plots of the predicted start of breeding (date of first egg deposition) as proxy for quality of territories (early breeding = high-quality territory) in relation to (a) proportion of sealed area within a 50 m buffer zone around the territory and (b) vegetation cover. Percentage inside the boxes is relative contribution of the variable. Ticks on the top indicate deciles of the data distribution.

4.1. Occurrence of wheatears – size of grassland matters The size of the study site was of high importance to the occurrence of wheatears (Fig. 1) as found for other grassland species (Davis, 2004). Similar to our results, the wheatear was found to be area-sensitive, both in rural (Caplat & Fonderflick, 2009) and urban (Jokimäki, 1999) areas. The strong influence of size that explained almost three-fourth of the variance may also be caused by the sampling of the study sites, since we selected wasteland sites with an open vegetation structure, many of them were potentially suitable for wheatears with respect to habitat features. Consequently, the demand for area turned out to be the most important factor, exceeding the influence of habitat or matrix variables by far. Additionally, wheatears did not inhabit areas smaller than 3.3 ha. While territories of wheatears have on average an area of 1.5 ha, both in urban and non-urban areas (Conder, 1956; own data; Jokimäki, 1999), smallest occupied wastelands were more than twice as large. Herkert (1994) found the same phenomenon for other grassland species. Therefore, not the size of the feeding territory itself but of the whole open patch seems to be crucial for habitat selection of area-sensitive, open-land birds. The avoidance of small patches could be caused by edge effects (Yahner, 1988) such as higher predation pressure (Paton, 1994) or higher disturbance intensity near edges. Isolation effects could be another ultimate reason for wheatears to avoid small habitat patches. Since small patches are

less likely to be found, mating success and colonisation probability during post-breeding dispersal may be reduced. However, in our study there was no effect of either territory size or size of adjacent suitable habitat on breeding success. Interestingly, factors affecting breeding success differed completely from those explaining variation in occurrence. Thus, breeding success seems not to be related to habitat preferences. Possibly, habitat preferences evolved in natural ecosystems and do not coincide well with habitat quality in human-altered environments. As our results show, although the wheatear needs some minimum area, urban wastelands of that size seem to provide proper habitat for this species. Human population density and degree of sealing had a very small but measurable negative effect on the occurrence probability of wheatears within 50 m around the site (Fig. 1g). Beyond that, there was no effect of the urban matrix on occurrence, breeding success, or start of breeding at the landscape scale of 200 m and 2 km. That contrasts to previous findings that habitat specialists, like open-land insectivores, are scarce in highly urbanised areas (Clergeau, Croci, Jokimäki, Kaisanlahti-Jokimäki, & Dinetti, 2006) and that the landscape scale is of high importance for abundances of grassland birds (Renfrew & Ribic, 2008). One reason for this discrepancy might be that most urban centres do not include large open wastelands, like we found in Berlin or they are rarely explored. Furthermore, the landscape scale reflects matrix suitability which is very low for the wheatear that could not be found in regular urban green spaces like parks in Berlin (Otto & Witt, 2002). Caplat and Fonderflick (2009) found a strong negative influence of a woodland matrix on the occurrence of wheatears on grassland patches, although sample size was small. For the urban setting we can state that the matrix has only a very low influence on wheatears. Increased use of sites with sandy ground, sparse vegetation, short grass, and lower incidence on sites with trees and shrubs (Fig. 1) correspond well to known habitat requirements of this species (Glutz von Blotzheim, 1985). Additionally, we found wheatears to be least likely to occur in sites with moss cover. Mosses grew most on pavement and we assume it to indicate older sites that remained without woods due to sealing. Habitat structure seems to be highly important for the selection of territories of many bird species. This has been shown explicitly for the common redstart (Phoenicurus phoenicurus; Martinez, Jenni,

Author's personal copy 54

P.J. Meffert et al. / Landscape and Urban Planning 108 (2012) 49–56

Wyss, & Zbinden, 2010). As the wheatear, the common redstart is a sit-and-wait predator and it also has declined dramatically (BirdLife International, 2004). In experiments, it favoured sparse vegetation due to higher detectability of prey, even if prey density was much higher in dense vegetation. Common redstarts assess habitat quality by using vegetation structure as predictor of food availability, as Tye (1992) as well found for the wheatear. Also many other bird species prefer short and sparse vegetation for foraging (Atkinson, Buckingham, & Morris, 2004). We did not find an influence of direct disturbance by humans or dogs on occurrence. Wastelands occupied by wheatears had an average size of ten hectares, thus density of disturbances was generally low and birds were able to avoid too many encounters with humans or dogs. Field observations showed that proximity of nest locations of a few metres to railroad tracks, walkways or benches did not prevent the birds from nesting and feeding chicks, and there were no abandoned nests or deaths from starvation in response to human intrusion. Even though they keep a certain wariness, individuals breeding in proximity to humans get used to disturbances (Grimm, 2004). Possibly, urban populations adapt their behaviour as known for other species as the blackbird (Luniak & Mulsow, 1988). Comparison of early-settled and late-settled territories showed that appropriate feeding habitats including short grass are important for the choice of territories (Fig. 3b). This is in accordance with findings of Brooke (1979) from the island of Skokholm (UK). Additionally, high degrees of sealing in the surroundings attracted wheatears (Fig. 3a) similar to occurrence (Fig. 1f), possibly because they are mainly free of vegetation, such as parking lots or left over foundations of destructed buildings. This structural poorness seems to attract the birds even though it is not necessarily connected to suitable feeding grounds (Jones, 2001). Even so, suboptimal habitats seem not to act as ecological traps (sensu Schlaepfer, Runge, & Sherman, 2002) since breeding success was high there too. 4.2. High breeding success The breeding success we observed in Berlin (Table 3) was comparatively high. In two German vineyards the proportion of successful breeding attempts was 56% (2 years, N = 695) and 60%, respectively (two years, N = 103; Buchmann, 2001; most nests have been found during incubation), on a Swedish grassland it was 83% (three years, N = 388; Moreno, 1989; most nests have been found before the full clutch had been laid), and on heathlands in England 65% (two years, N = 98; Tye, 1980; no information if nest have been found before prior or during incubation). Although we had a comparatively small sample size, we assume it to be representative since we surveyed the majority of the existing breeding pairs and differences between years were negligible. The most successful wheatears bred earliest (Fig. 2a). This finding is in line with previous findings on wheatears (Tye, 1992). Breeding success decreased with disturbance intensity (Fig. 2c). This effect is likely to be indirect since many nest losses occurred on construction sites that are by their nature regularly visited by humans. Nest failures, however, occurred as heaps of construction materials containing the nests were moved, thus the destruction of the nests only coincidentally occurred together with the presence of humans. The model for breeding success explained only 29.5% of the deviance. This may on the one hand reflect the fact that we did not include an appropriate variable that reflects the nest losses on construction sites or due to predators. On the other hand, losses by predators seemed to appear randomly and thus caused a high stochastic component. Chamberlain et al. (2009) reviewed avian productivity of rural versus urban areas, but found no consistent pattern. Out of nine species with significant differences, six did better in urban areas.

The only strict insectivore investigated, the Spotted Flycatcher (Muscicapa striata), had less nest failures in urban habitats. Our findings add another example of a food specialist that seems to thrive in urban habitats. Several omnivorous species have a lower breeding success within urban centres (Savard et al., 2000; Schmidt, 1988). It is assumed that highly urban-adapted species such as Great Tits feed ‘wrong’ anthropogenic food to their chicks and as a result have less or no offspring (Chamberlain et al., 2009). That could be a consequence of the comparatively short time period during which birds have lived in cities and a resulting lack of adaptation. This would not affect open-land species like the wheatear that do not exploit anthropogenic food sources. A reduced productivity could also be caused by a lack of arthropods. Peach et al. (2008) found a strong connection between abundance of aphids and nestling’s mortality in an urban population of the house sparrow. Shortage of invertebrates could cause lower survival rates and reproductive success also for insectivores. However, wastelands are often unmanaged and thus not treated with pesticides and provide a wide range of invertebrates (Strauss & Biedermann, 2006). Our results show that wheatears were able to find enough prey to raise their offspring. That emphasises the important role of the habitat features of open wastelands within urbanised areas. Another important factor that influences breeding success is predation. Predation pressure is assumed to be higher in urban areas compared to natural areas (e.g. Jokimäki & Huhta, 2000), but findings are inconsistent (Marzluff, 2001) and there are methodological problems regarding the detection of identity and abundance of predators (Zanette, 2002). Marzluff, Withey, and Whittaker (2007) found that predation had no strong influence on breeding success in 15 forest songbird species in an urbanised landscape. It seems that predation pressure is highly variable and depends mainly on predator species. In Berlin, small mammals and corvids seem to play a role in predation. In one of the investigated areas with a high density of wheatears, magpies have been observed waiting for the chicks to leave the nesting holes, however chicks leaving the nest were soon after able to fly and thus to flee. Mustelidae such as the stoat (Mustela erminea), that are frequent nest predators in semi-natural areas (Buchmann, 2001; Moreno, 1989; Tye, 1980) and able to enter the nesting holes are absent in urban areas of Berlin (Rainer Altenkamp, pers. comm.). That might be the most important reason for the low observed predation rate. Overall fitness of individuals includes not only breeding success but also survival, though; nevertheless reproduction is one of the most important life-history traits. From the facts that (1) predation seems to be lower in comparison to other habitats, (2) survival can be assumed to be the same since birds of different breeding habitats share conditions on migration and during winter, and (3) breeding success in Berlin was high, we conclude that the city of Berlin currently appears to provide habitat for a sustainable wheatear population. More data on survival and migration are needed to be able to assess if the population’s growth rate is equal or greater than one. 4.3. Implications for urban planning None of the wastelands surveyed were managed to be a habitat for the wheatear. However, structural features that allow species as the wheatear to thrive could be adopted by landscape architects. In our particular case, a minimum size of five to seven hectares is needed since occurrence probability was much higher on sites larger than five hectares. Beyond that, open soils, sparse vegetation, short grass, and if at all, then very few trees and shrubs favour the wheatear. Since vegetational succession will proceed, habitats have to be maintained or produced consistently. This might happen anyway since cities are dynamic; buildings will be demolished, mass

Author's personal copy P.J. Meffert et al. / Landscape and Urban Planning 108 (2012) 49–56

meetings or temporary events like circuses, motocross, or Christmas tree selling remove vegetation cover of wastelands. Thus, a dynamic cycle of spatiotemporal shifts between disturbance in terms of removal of vegetation and secondary succession could maintain open spaces and thereby increase biodiversity. These mechanisms have been shown to enhance biodiversity of plants and insects on urban wastelands (Kattwinkel, Strauss, Biedermann, & Kleyer, 2009), and of grassland birds whose habitat was managed by prescribed fire (Madden, Hansen, & Murphy, 1999). Some of the wheatear’s territories were located in parks and show that recreational usage and habitat requirements of this species can be balanced. Diverse mowing regimes create various niches and thus enable different species to persist. Another important aspect for wheatears is to provide nesting structures. In the mentioned parks, gabions with sufficiently sized gaps are a landscaping element that meets the requirements of the wheatear. 5. Conclusions Our findings indicate that the wheatear is able to cope with or even profit from urban wasteland habitats. The wheatear, settling and successfully breeding in open spaces within the city, is another example of an early successional species that, with little special effort from people, can thrive in urban settings as described by Marzluff (2005). This might hold true also for other endangered open-land bird species such as tawny pipit (Anthus campestris), linnet (Carduelis cannabina), tree sparrow (Passer montanus), and crested lark (Galerida cristata) that were also found on the study sites but have not yet been studied in detail. We assume our findings can be transferrable to other cities and urban areas. The sandy soils of Berlin may cause a comparatively slow vegetational succession; therefore areas remain open longer compared to other locations with more productive soils. According to Kelcey and Rheinwald (2005), wheatears breed in most European cities. Together with our findings this shows the potential to provide habitat for the protection of this endangered species. The worldwide phenomenon of shrinking cities that leads to an increase of unused land within urban areas (Fritsche et al., 2007) might be used as an opportunity to enhance biodiversity. Appropriate management of wasteland sites and parks could help to stop the decrease of the wheatear’s population in Europe beyond the necessary protection of natural and agricultural habitats. Acknowledgements We thank Barbara Clucas, Galina Churkina, Leonie Fischer and three anonymous reviewers for improving early drafts of the manuscript and the German Research Foundation (DFG) for funding (Graduate Research Programme on Urban Ecology 780/3). Appendix A. Description of predictor variables used (x = variable used). Dependent variables for the model occurrence are presence/absence and refer to the mapped site, for breeding success and start of breeding the variables refer to single territories. Variable description

Occurrence

Size of the site Size of territory Size of adjacent suitable habitat Intensity of disturbance (see Table 1) Index of presence and consistency of dogs Index of presence and consistency of humans Presence/absence of cats

x

x x x x

Breeding success

Start of breeding

x x x

x x x

55

x

Plant growth within a year (very low, low, medium, high) Inhabitants/ha within a 50 m buffer zone around the site Inhabitants/ha within a 200 m buffer zone around the site Inhabitants/ha within a 2000 m buffer zone around the site Proportion sealed within a 50 m buffer zone around the site (%) Proportion sealed within a 200 m buffer zone around the site (%) Proportion sealed within a 2000 m buffer zone around the site (%) Presence/absence of trees around the site Grass cover up to 5 cm height (%) Grass cover between 5 and 10 cm height (%) Grass cover between 10 and 20 cm height (%) Grass cover between 20 and 50 cm height (%) Grass cover between 50 and 100 cm height (%) Grass cover > 100 cm height (%) Tree cover < 4 m height (%) Tree cover > 4 m height (%) Prop. covered by buildings (%) Moss cover (%) Sand cover (%) Prop. covered by stones, tarmac, concrete or ballast (%) Shrub cover < 1 m height (%) Shrub cover > 1 m height (%) Proportion covered by water (%) Proportion covered by other materials (%) Vegetation/surface heterogeneity (Simpson’s index) Day of first egg deposition (Julian Day) Year (20007 or 2008) Location of the nest in five classes (gabion, hole near ground, concrete slabs/railway sleeper, heap of stones/boulder, other)

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x x

x x

x x

x

x

x

x

x

x

x

x

x

x x x x x x x

x x x x x x x

x x x x x x x

x x x x

x x x x

x x x x

x

x

x

x

x x x

x x

Appendix B. Number of breeding pairs per site. Name of site

2007

2008

Biesenhorster Sand Böhlener Straße Bornholmer Straße Boxberger Straße Cottbusser Platz Elsterwerdaer Platz Flugplatz Gatow Gaußstraße Hamburg-Lehrter Güterbahnhof Hauptbahnhof Hornoer Ring Flugfeld Johannisthal Kohlweißlingstraße Bahnhof Lichtenberg Ostbahnhof Pankow-Heinersdorf Quitzowstraße Rhinstraße Siemensstraße Staaken

4 1 1 1 1 1 1 1 3 1 Not surveyed 20 2 Not surveyed 4 2 1 1 1 1

Storkower Straße Trappenfelder Pfad North-Willys-Straße Wolfener Straße

2 2 1 1

2 1 0 1 0 Site developed Not surveyed Site developed 3 Site developed 2 18 Site developed 1 Site developed 4 Site developed Site developed Not surveyed Nesting structure removed 2 4 2 0

Author's personal copy 56

P.J. Meffert et al. / Landscape and Urban Planning 108 (2012) 49–56

References Atkinson, P., Buckingham, D., & Morris, A. (2004). What factors determine where invertebrate-feeding birds forage in dry agricultural grasslands? Ibis, 146(s2), 99–107. Baeyens, G. (1981). Magpie breeding success and carrion crow interference. Ardea, 69, 125–139. Baillie, J. E. M., Hilton-Taylor, C., & Stuart, S. N. (Eds.). (2004). 2004 IUCN Red List of threatened species. A global species assessment. Switzerland and Cambridge, UK: IUCN, Gland, xxiv+191 pp. Benfield, F. K., Raimi, M., & Chen, D. D. T. (1999). Once there were greenfields: How urban sprawl is undermining America’s environment, economy, and social fabric. New York: Natural Resources Defense Council. BirdLife International. (2004). Birds in the European Union: A status assessment. Wageningen, The Netherlands: BirdLife International. BOA. (2005). Berliner Brutvogelbericht 2004. Berliner ornithologischer Bericht [Breeding bird report of Berlin], 15(2), 246–268 (in German). BOA. (2006). Berliner Brutvogelbericht 2005. Berliner ornithologischer Bericht [Breeding bird report of Berlin], 16, 132–145 (in German). BOA. (2007). Berliner Brutvogelbericht 2006. Berliner ornithologischer Bericht [Breeding bird report of Berlin], 17, 119–132 (in German). Brittingham, M., & Temple, S. (1986). A survey of avian mortality at winter feeders. Wildlife Society Bulletin, 14, 445–450. Brooke, M. D. L. (1979). Differences in the quality of territories held by Wheatears (Oenanthe oenanthe). Journal of Animal Ecology, 48(1), 21–32. Buchmann, M. (2001). Die Brutbiologie des Steinschmätzers (Oenanthe oenanthe) auf intensiv genutzten Flächen in Rheinland-Pfalz (The breeding biology of the wheatear (Oenanthe oenanthe) on intensively used areas in RhinelandPalatinate). Die Vogelwarte, 41, 1–17 (in German). Caplat, P., & Fonderflick, J. (2009). Area mediated shifts in bird community composition: A study on a fragmented Mediterranean grassland. Biodiversity and Conservation, 18(11), 2979–2995. Chace, J., & Walsh, J. (2006). Urban effects on native avifauna: A review. Landscape and Urban Planning, 74(1), 46–69. Chamberlain, D., Cannon, A., Toms, M., Leech, D., Hatchwell, B., & Gaston, K. (2009). Avian productivity in urban landscapes: A review and meta-analysis. Ibis, 151(1), 1–18. Clergeau, P., Croci, S., Jokimäki, J., Kaisanlahti-Jokimäki, M.-L., & Dinetti, M. (2006). Avifauna homogenisation by urbanisation: Analysis at different European latitudes. Biological Conservation, 127(3), 336–344. Conder, P. (1956). The territory of the Wheatear Oenanthe oenanthe. Ibis, 98(3), 453–459. Davis, S. K. (2004). Area sensitivity in grassland passerines: Effects of patch size, patch shape, and vegetation structure on bird abundance and occurrence in southern Saskatchewan. The Auk, 121(4), 1130–1145. Donald, P. F. E., Green, R. E., & Heath, M. F. (2001). Agricultural intensification and the collapse of Europe’s farmland bird populations. Proceedings of the Royal Society B: Biological Sciences, 268, 25–29. DWD. Accessed 22.04.06. EBCC. Accessed September 2010. Elith, J., Leathwick, J. R., & Hastie, T. (2008). A working guide to boosted regression trees. Journal of Animal Ecology, 77(4), 802–813. Emlen, J. (1974). An urban bird community in Tucson, Arizona: Derivation, structure, regulation. The Condor, 76(2), 184–197. Fretwell, S., & Lucas, H. (1969). On territorial behavior and other factors influencing habitat distribution in birds. Acta Biotheoretica, 19(1), 16–36. Friedman, J., & Meulman, J. (2003). Multiple additive regression trees with application in epidemiology. Statistics in Medicine, 22(9), 1365–1381. Fritsche, M., Langner, M., Köhler, H., Ruckes, A., Schüler, D., Zakirova, B., et al. (2007). Shrinking cities—A new challenge for research in urban ecology. In M. Langner, & W. Endlicher (Eds.), Shrinking cities: Effects on urban ecology and challenges for urban development (pp. 17–33). Frankfurt: Peter Lang. Glutz von Blotzheim, U. N. (1985). Handbuch der Vögel Mitteleuropas (Handbook of the birds of Central Europe). Wiesbaden: AULA-Verlag. (in German) Grimm, H. (2004). Der Brutbestand des Steinschmätzers Oenanthe oenanthe im Jahre 2002 in Thüringen mit Anmerkungen zur historischen Entwicklung der thüringischen Kulturlandschaft und ihrer Eignung als Lebensraum für im Offenland brütende Vogelarten (The breeding population of the wheatear Oenanthe oenanthe in 2002 in Thuringia with comments on the historic development of the Thuringian cultural landscape and its suitability as habitat for bird species breeding in open landscapes). Anzeiger des Vereins Thüringer Ornithologen, 5, 85–104 (in German). Herkert, J. (1994). The effects of habitat fragmentation on midwestern grassland bird communities. Ecological Applications, 4(3), 461–471. Jokimäki, J. (1999). Occurrence of breeding bird species in urban parks: Effects of park structure and broad-scale variables. Urban Ecosystems, 3(1), 21–34. Jokimäki, J., & Huhta, E. (2000). Artificial nest predation and abundance of birds along an urban gradient. The Condor, 102(4), 838–847. Jones, J. (2001). Habitat selection studies in avian ecology: A critical review. The Auk, 118(2), 557–562. Jones, Z., & Bock, C. (2002). Conservation of grassland birds in an urbanizing landscape: A historical perspective. The Condor, 104(3), 643–651. Kattwinkel, M., Strauss, B., Biedermann, R., & Kleyer, M. (2009). Modelling multi-species response to landscape dynamics: Mosaic cycles support urban biodiversity. Landscape Ecology, 24(7), 929–941.

Kelcey, J. G., & Rheinwald, G. (2005). Birds in European cities. St. Katharinen: GinsterVerlag. Klem, D. (1990). Collisions between birds and windows: Mortality and prevention. Journal of Field Ornithology, 61(1), 120–128. Luniak, M., & Mulsow, R. (1988). Ecological parameters in urbanization of the European blackbird. In L. Tomialojc, & F. R. Gehlbach (Eds.), Avian population responses to man-made environments. Acta XIX congressus internationalis ornithologica, Ottawa, Canada (pp. 1787–1793). Madden, E. M., Hansen, A. J., & Murphy, R. K. (1999). Influence of prescribed fire history on habitat and abundance of passerine birds in Northern mixed-grass prarie. The Canadian Field-Naturalist, 113, 627–640. Martinez, N., Jenni, L., Wyss, E., & Zbinden, N. (2010). Habitat structure versus food abundance: The importance of sparse vegetation for the common redstart Phoenicurus phoenicurus. Journal of Ornithology, 151(2), 297–307. Marzluff, J. (2005). Island biogeography for an urbanizing world: How extinction and colonization may determine biological diversity in human-dominated landscapes. Urban Ecosystems, 8, 157–177. Marzluff, J. M. (2001). Worldwide urbanization and its effects on birds. In J. M. Marzluff, R. Bowman, & R. Donnelly (Eds.), Avian ecology and conservation in an urbanizing world (pp. 19–47). Norwell, MA: Kluwer Academic Publishers. Marzluff, J., Withey, J., & Whittaker, K. (2007). Consequences of habitat utilization by nest predators and breeding songbirds across multiple scales in an urbanizing landscape. The Condor, 109(3), 516–534. Mills, G., Dunning, J., Jr., & Bates, J. (1989). Effects of urbanization on breeding bird community structure in southwestern desert habitats. The Condor, 91(2), 416–428. Moreno, J. (1989). The breeding biology of the Wheatear Oenanthe oenanthe in South Sweden during three contrasting years. Journal of Ornithology, 130(3), 321–334. Ockendon, N., Hewson, C. M., Johnston, A., & Atkinson, P. W. (2012). Declines in British-breeding populations of Afro-Palaearctic migrant birds are linked to bioclimatic wintering zone in Africa, possibly via constraints on arrival time advancement. Bird Study, 59(2), 111–125. Otto, W., & Witt, K. (2002). Verbreitung und Bestand Berliner Brutvögel (Distribution and populations of the breeding birds of Berlin). Berliner Ornithologischer Bericht, 12 (special edition), 1–256 (in German). Paton, P. W. C. (1994). The effect of edge on avian nest success: How strong is the evidence? Conservation Biology, 8(1), 17–26. Peach, W., Vincent, K., Fowler, J., & Grice, P. (2008). Reproductive success of house sparrows along an urban gradient. Animal Conservation, 11(6), 493–503. R Development Core Team. (2009). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Renfrew, R., & Ribic, C. (2008). Multi-scale models of grassland passerine abundance in a fragmented system in Wisconsin. Landscape Ecology, 23(2), 181–193. Ridgeway, G. (2007). Generalized boosted regression models. Documentation on the R Package ‘gbm’, version 1. 6-3. Sanderson, F. J., Donald, P. F., Pain, D. J., Burfield, I. J., & van Bommel, F. P. J. (2006). Long-term population declines in Afro-Palearctic migrant birds. Biological Conservation, 131(1), 93–105. Savard, J.-P. L., Clergeau, P., & Mennechez, G. (2000). Biodiversity concepts and urban ecosystems. Landscape and Urban Planning, 48(3–4), 131–142. Schlaepfer, M., Runge, M., & Sherman, P. (2002). Ecological and evolutionary traps. Trends in Ecology and Evolution, 17(10), 474–480. Schmidt, K.-H. (1988). Site fidelity and isolation of Great Tits (Parus major) in urban habitats. In XIX Congressus Internationalis Ornithologica Ottawa, (pp. 1794–1801). Senate Department for Urban Development. (2004). Berlin digital environmental atlas, 01.02 impervious soil coverage (sealing of soil surface) (2007 ed.). Senate Department for Urban Development. (2005). DOP025-C/DVD 04.3, Orthophotomosaike Berlin – Bildflug August 2004, Berlin. Senate Department for Urban Development. (2006). Berlin digital environmental atlas, 06.06 population density (2006 ed.). Strauss, B., & Biedermann, R. (2006). Urban brownfields as temporary habitats: Driving forces for the diversity of phytophagous insects. Ecography, 29(6), 928–940. Südbeck, P., Bauer, H. G., Boschert, M., Boye, P., & Knief, W. (2007). Rote Liste der Brutvögel Deutschlands (Red List of the breeding birds of Germany). Berichte zum Vogelschutz, 44, 23–81 (in German). Tye, A. (1980). The breeding biology and population size of the wheatear (Oenanthe oenanthe) on the Breckland of East Anglia, with implications for its conservation. Bulletin d’écologie, 11, 559–569. Tye, A. (1992). Assessment of territory quality and its effects on breeding success in a migrant passerine, the Wheatear Oenanthe oenanthe. Ibis, 134(3), 273–285. van der Zande, A. N., & Vos, P. (1984). Impact of a semi-experimental increase in recreation intensity on the densities of birds in groves and hedges on a lake shore in The Netherlands. Biological Conservation, 30(3), 237–259. Vitousek, P. M., Mooney, H. A., Lubchenco, J., & Melillo, J. M. (1997). Human domination of Earth’s ecosystems. Science, 277(5325), 494. Wesolowski, T. (1983). The breeding ecology and behaviour of Wrens Troglodytes troglodytes under primaeval and secondary conditions. Ibis, 125(4), 499–515. Yahner, R. H. (1988). Changes in wildlife communities near edges. Conservation Biology, 2(4), 333–339. Zanette, L. (2002). What do artificial nests tells us about nest predation? Biological Conservation, 103(3), 323–329.