Recent Studies Of Avian Species Or Populations Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Habitats undergo continuous and worldwide increasing modifications due to growing urbanization and agriculture intensification Marzluff 2001; Shochat et al. 2006; Donald et al. 2006. These extensive modifications of the landscape and consecutive restructuring of habitats can impinge on ecological processes that influence the population growth potential of organisms, particularly birds (Lawton 1995). For example, a variety of studies in Europe have shown that nest predation is reduced in human-settled environments compared to forest (Snow 1958; Stjernberg 1979; Tomialojc 1980) while others have suggested that changes in farming practices could adversely affect survival (Robinson et al. 2004) .

In such a context of large scale environmental changes, ecology aims at understanding the scaling of the responses of species and communities to temporal and spatial changes in habitat structure and composition (Bossenbroek et al. 2005). In turn, such ecological data can be used by policy markers and environmental managers to set targets, inform the choice of relevant baselines and to assess the level of natural variability (Froyd & Willis 2008). Such management practices may address declining populations facing habitat deterioration as well as invasive species colonizing new habitat niches.

In assessing the importance of any habitat to breeding birds, it is necessary to determine not only the number of birds present but also the contribution it makes to overall recruitment of new birds to the population, since territory density may not reflect habitat quality as defined by productivity (van Horne 1983, Vickery et al. 1992). Indeed, for mobile species such a birds, presence is not enough to indicate sustainability, and apparently viable population of native birds into a given habitat may be in fact functioning as population sinks, maintained through immigration from patches of better quality habitat (Pulliam 1988).

Although annual reproductive success (young produced per female per year) may be a more desirable metric for examining pop dynamics (Thompson et al. 2001), daily nest survival and consecutive nest success (the probability that a nest will fledge at least one young over the nesting period.) are thought to be an important determinant of reproductive rates and, in turn, population dynamics. Accordingly, a number of studies have highlighted the importance of spatial variation in reproductive performance in maintaining local bird populations from both theoretical (Blondel 1985, Blondel et al. 1990, Hatchwell et al. 1996a) and applied perspectives (Galbraith 1988, Green 1988, Green et al. 1997). Nest success may be influenced by a wide array of factors such as food resources (Duguay et al 2000), predator population size (Smith & Ostfeld 2003), conspecific density (Barber et al. 2001), climatic events (Baiser et al. 2008; Lehman et al. 2008) and may therefore be a more sensitive indicator of the effects of habitat differences (Armstrong et al. 2002).

Recent studies of avian species or populations (Paradis et al. 2000, Driscoll et al. 2005, Reidy et al. 2009) support the idea that multiple scales affect nest survival. These multiple scales range from microhabitat (nature of nest support, nest height and concealment; Hatchwell et al. 1996b, Weidinger 2002) to macrohabitat characteristics (relative importance in habitat patch surface (Møller 1988), edge effects (Chamberlain et al. 1995). The two main group of factors potentially affecting nest success, namely resources and predation are also considered using variable spatial scales: predation is in most case considered at local scale since the impact on nest success will depend mostly on local predator densities (Paradis et al. 2000, Stoate & Szczur 2001, White et al. 2008), and nest conspicuousness (Hatchwell et al. 1996b; Weidinger 2002 ). On another hand, resources are considered from local to larger scale, depending upon potential foraging range radius of birds during breeding season (Fluetsch & Sparling 1994; Boulton et al. 2008).

There is therefore a need to evaluate how processes operate at different spatial scale (Caldow & Racey 2000) which is of interest for widespread species and particularly those which have enlarged their breeding range following the colonization of new habitat niches. For these species, the wide range of environmental conditions they withstand is likely to affect more substantially their breeding performance. However, so far, applied studies have often concentrated on rare and/or declining species, taking into account local context and it is unclear whether their finding apply to common and widespread species. Thus, to date, studies comparing nesting success estimates across habitat gradients, landscape features, and large geographic range (Paradis et al. 2000; Siriwardena et al. 2000) remain relatively scarce. An obvious constraint to that type of study is that it requires a huge logistic support which is seldom available when using volunteers.

In this paper, we report a study of temporal and large-scale spatial variation in nesting success for the woodpigeon, Columba palumbus, throughout a 5 years monitoring study, from 2004 to 2008. Our analysis are based on extensive national data collected by professional technicians.

The woodpigeon is a common breeding species in France, which inhabits all lowland habitats (Yeatman & Jarry 1994), with most birds occurring on farmland, woodland or human associated habitats from village to large cities. While its ancestral habitat was woodland (Cramp 1985), it has taken opportunity from agriculture development to colonize farmland (Murton 1965). Then, since the end of the XIX century, this species has been increasingly present in urban habitat (Cramp 1985; Tomialjoc 1976). From a geographical perspective, the woodpigeon was present in all the northern half of the country (Yeatman & Jarry 1994), but since the last decades, the breeding population has quickly extended southward (Yeatman & Jarry 1994; Boutin 2001). Simultaneously, the French population has dramatically increased, and such increase is still going on to date (Birdlife International 2004; Roux et al. 2009).

Relationship between humans and woodpigeons present contrasted aspects: it is considered as a valuable game species by French hunters and is now the first game species in France with regard to annual bag size (Lormée et al. 2000). On another hand, this granivorous species is increasingly considered as a pest by farmers, particularly in open field area (Murton et al. 1974; Inglis et al. 1994). In such a context, management strategies are expected by both farmers and hunters with regards to breeding woodpigeon population, and nest success might be a key component to take into account into such management practices.

We have identified 2 questions that need to be addressed in order to understand how woodpigeon breeding success is shaped; (i) what is its range of variability through time and space, and (ii) how important are a set of environmental variables, including nesting and foraging habitat structure, climatic events, food resources availability, predation pressure and density effect, for such variation in breeding success. We were particularly interested to compare nest success estimates between urban and non urban population, as we expect the recently urbanisation of woodpigeon to have large impact on reproductive rates and further French breeding population dynamic as a whole.

Material & Methods

Study area

This study rely on data collected by professional technicians within a national monitoring programme developped by ONCFS since 2001. This monitoring was progressively extended to all the national territory from 2001 to 2008. At the time we conducted our analyses, the two-third of the country were covered by this monitoring. To avoid geographical bias in our nest survival estimation, we used data collected over the same portion of the national territory throughout the study period (fig.1).

Nesting data

Nest searching period occurred from March to late September, covering the entire breeding season of woodpigeon. Nests were located by walking randomly or using relevant information in the field such as locations of territorial singing birds, flight displays, nest material transport. Observers were not constrained to prospect a fixed area over the study period so the sites where nests were found could vary every year. However experienced observers prospected most of the time the same sites each year. Each monitored nest was checked at intervals ranging from 5 to 9 days; occasionally, due to logistical constraints, intervals could be as long as 10 to 12 days. Only nests visited at least twice were used into the analysis. Previous studies of the effects of nest-visiting on open-nesting species were reviewed by Mayer-Gross et al. (1997), who concluded that any such effects are unlikely to be important. However, because monitoring of nest may increase the probability of predation and nest desertion, observers took precautionary measures to minimize their impact: they avoided systematically harassing adults sitting on nest and, as often as possible, nests were checked at distance using binoculars. Nests that were too high to see into were checked with a mirrored pole. All nests were relocated through detailed field notes.

We used only nests with confirmed activity (eggs or nestlings) in analyses. Woodpigeon incubation period last 17 days (Cramp 1985). Fledging period is variable and may range from 20 to 35 days (Colquhoun 1951). Throughout our monitoring, the youngest age at which chicks were known to have successfully left the nest was 22 days and once the chicks had left the nest, they could not be followed. The results of this work relate therefore to incubation and the first 22 days post-hatch. Chicks were aged using a growth curve we have previously established, relying upon wing length, length from the back of the skull to the tip of the bill, and tarsus length. We were therefore able to back calculate laying and hatching dates, and the ending date of the rearing period when chicks reached 22 days old.

When possible we assigned a cause of failure (i.e., nest predation, adverse weather conditions, abandon…) to failed nests. We attributed nest failure to weather or other natural events when nests were found destroyed and/or nests content were on the ground below a tilted nest after a storm. Nest were considered to have failed for unknown reasons when nest contents was empty with no visual clues to indicate the cause of failure. During incubation, we did not record the nest fate as abandoned until the eggs remained cold and unattended for at least 2 consecutive visits. During the rearing period, a nest was considered successful if it fledged  1 young and we assumed that a nest had failed if young disappeared > 3 days before the predicted fledging date. Nests with uncertain fate were not included in analysis.

We standardized the beginning and the end of the breeding season across the 5 years by using the earliest date a nest was located in any year as the first day of the season (incubation and rearing periods: 6 march) and the latest failure/hatching or fledging date in any year as the last day of the season (incubation 23 October; rearing 14 November). If several breeding attempts were made in the same nest throughout the breeding season, we randomly selected only one per nest in order to obtain statistical independence of data.

Candidate variables for exploring variation in nest survival

We selected a set of variables (listed in Table 1) that we considered to have potential effect on breeding performance in our species. These variables included: (i) nesting habitat; (ii) foraging habitat; (iii) monthly precipitation amount; (iii) mean monthly minimal temperature; (iv) woodpigeon local abundance; (v) combined local abundance of corvids, the avian main nest predators (jay: Garrulus glandarius, magpie Pica pica, and crow Corvus Corone); (vi) nest location with longitude and latitude; (vii) elevation.

Justification and description and of the choice of variables

Nesting habitat - The characteristics of habitat used by birds for breeding may dramatically shape nest success and in turn population dynamics (O'Connor & Fuller 1985; Martin 1992; Martin & Clobert 1996; Chalfoun & Martin 2007). The impact of habitat on nesting success may act at different spatial-scale: nest success is affected primarily by nest predation (Chalfoun & Martin 2007) and in such a context the geo-physical and vegetational characteristics of proximal habitat (referred to as nesting habitat) are of particular interest. Nesting habitat referred here as the homogenous landscape unit within which the bird breeds. Information on nesting habitat was collected by observers in the field at the end of nesting attempt. We used a substantially simplified version of a coding system implemented by BTO for the Nest Record Scheme in United Kingdom (Crick 1992) and designed to fulfil the following requirements: (i) to require no expert botanical knowledge, (ii) to be based on the structural aspects of bird habitats but including simple floristic categories within the broads habitat types. The coding system consists of a simple four-level hierarchy. In the top level are 4 major types of habitat and within each of the major habitats are three further levels that are used to record increasingly detailed information. In this study, we only used the first level of classification; nest habitat was therefore classified as: Woodland (dominated by trees generally greater than 5 m tall); Farmland (defined by all types of fields, cultivated or grazed, etc…); Urban (densely built-up area, town centres, and suburban area); Villages (small urban area with generally less than 3000 inhabitants, as scattered houses or other buildings).

Foraging habitat - Complementarily, food availability may affect the ability of parents to optimize parental care such as parental attendance at nest, feeding frequency of the nestlings and hence their growth speed, and in turn daily nest success. In this case, the relevant spatial scale to take into account in the description of habitat must integrate the potential foraging range of breeding adults. We therefore described potential foraging habitat used by woodpigeons during breeding using digital land-cover data collected in the CORINE Land Cover database (CLC). CLC includes the main habitats for the whole country in contiguous polygons classified into 44 different land-cover categories (Bossard et al. 2000). Among all the landscape variables available, we considered 4 groups of variables that we considered relevant to test predictions on woodpigeon nest success; within agricultural habitats, we identified a group representative of intensive agricultural lands, including all classes defined as "arable lands", and a group representative of mosaic agricultural landscape, including pasture and all habitats defined as "heterogeneous agricultural areas". Considering that both these habitat groups were associated with high food availability through growing and mature cereals crops, we predicted that an increase in the surface of both these habitats into the woodpigeon foraging radius should be associated with a better nesting success. We also identified a "forest" habitat group, including broad-lived, coniferous and mixed forests and scrub and/or herbaceous vegetations associations. We predicted that forested habitats were associated with a high predation pressure and low food availability, and hence a reduced nesting success. Finally we identified an "urban" habitat group, including all artificial surfaces. Urban area do not offer significant food resources to woodpigeon, however such area are considered to be associated with low densities of predators potentially preying upon woodpigeon nests, we predicted therefore an increase in nest success along with increasing urbanized surface within the foraging radius.

Until 2008, observers did not collect geographical coordinates of every nests, we therefore assigned to the nest the coordinates of the commune within which it was found. To investigate the effect of foraging habitat on nest success, we then extracted habitat composition data within a 5 km radius centred around those coordinates. We assumed that a 5 km radius allowed us to correctly describe foraging habitat composition in the vicinity of the nest, and that it was biologically meaningful with regards to scientific data available on woodpigeon foraging range (Haynes et al. 2003; Lormée et al. 2002 and unpublished data from a radio-tracking study). In order to check that foraging habitat was correctly defined following such a design, we investigated specifically in 2008 the correlation between foraging habitat composition within a 5 km radius using either the commune coordinates or the very coordinates of the nest. The relationship was highly significant (N = 1899 nests; Farmland: R²= 0.93, t = 158.52, P <0.0001, slope = 0.934; Forest: R²= 0.86, t = 106.68, P <0.0001, slope = 0.932; Urban: R²= 0.87, t = 112.96, P <0.0001, slope = 0.974) .

Monthly Precipitation & temperature - Woodpigeons use open and flat nests often built quite rudimentarily with twigs. Consequently, nests are particularly vulnerable to meteorological event such as heavy rain, wind and cold temperature. We therefore predicted a negative impact on nest success from increasing amount of precipitation and decreasing minimal temperature.

Precipitation and temperature values were obtained from MétéoFrance climatic database, relying upon a framework of ca 700 meteorological stations covering all the country and collecting data daily. To capture spatial and temporal difference in precipitation amount or minimal temperature, we calculated each nest value by extracting the monthly precipitation and temperature values measured over the active nesting period. For every nest, we assigned meteorological data from the nearest measure station (mean distance, Incubation: 13.6 ± 0.2 km (SE), n = 982; Rearing: 13.8 ± 0.2 km, n = 1024) . When the active nesting period was totally encompassed within the same month, we directly used the monthly amount of precipitation or the mean monthly minimal temperature. If the active nesting period overlapped two consecutive months, we then calculated an averaged value over the two months period.

Abundance of woodpigeon and corvids during breeding season: we predicted (i) a positive relationship between nesting survival rate and woodpigeon abundance since sites showing a better success should be more attractive for breeding birds and (ii) a negative relationship between nesting success and corvids abundance as an increasing abundance in corvids should lead to a increase in predation pressure. Information on woodpigeon and corvids abundance during breeding were taken from a national point count grid (ACT monitoring program; see Boutin et al. 2001, 2003 for full description of the methodology) where territorial singing birds are counted each year through two 10 minutes sessions long (préciser pour corvids que ce sont surtout ceux qui sont vus?), between early April and mid June. Woodpigeons were surveyed since 1996 but corvids only since 2008. Based on these point count data, a "local" woodpigeon abundance was assigned to each nest monitored into our study using the number of birds recorded in the closest point count and during the same year the nest was monitored (mean distance, Incubation: 9.4 ± 0.1 km (SE), n = 982; Rearing: 9.5 ± 0.1 km, n = 1024). For corvids abundance, we assigned to all nesting attempts the abundance estimated in 2008, considering that it was more or less representative of corvids abundance throughout the study period.

Elevation and location of the nest - As we limited the number of environmental covariates used in this analysis, we could not exclude that other covariates might have a significant effect on spatial variation in nest survival. Thus we also included some geographical covariates such as latitude, longitude, and elevation of the commune within which the nest was found.

Following the recommendation of White and Burnham (1999), we centred all individual covariates around the average and standardized them by the standard deviation in the time series covering the study period.

To minimize auto-correlated observational covariates we created a correlation matrix, detecting those highly correlated (r  0.7; REF?) and in response retaining only that with the lowest P values in an univariate model.

Statistical analysis

We used Mayfield estimates of daily survival probability as a measure of nest success, which is the estimated probability of a nest surviving a day within a defined stage of the nest cycle (see Mayfield 1975 for the principles of Mayfield method). Rates of nest success were estimated using the daily-survival estimator (DSR) available in program Mark (White & Burnham 1999). DSR was then used to estimate cumulative probabilities for nest survival. Because daily survival can vary across nesting stages, separate Mayfield estimates were calculated for each stage (incubation, nestling) and multiplied together to derive the overall estimate of nesting success.

Nesting attempt termination (failure or Hatching/fledging of the young) was assumed to occur half way between the penultimate observation of active nest and first observation of failed or Hatched/fledged youngs.

We used a modified form of Akaike's information criterion, AICc (Burnham and Anderson 1998) for model selection. The models with the lowest AICc value was considered to have the best fit; models with AICc values differing by  2.00 units were considered equally supported, in which case the model with the fewest parameters was chosen (Lebreton et al. 1992). We used the symbol "" to refer to additive effects and the symbol "" to refer to interactions. We used "." to designate constant models. Estimates were constrained using a logit link function. We used the statistical package MARK 5.1 (White and Burnham 1999) to obtain maximum likelihood estimates of nest survival and fit statistics, under various models.

The model selection was conducted following a strategy described in Grobois and Tavecchia (2003) and Grosbois and Thompson (2005). Prior to testing the covariates that might underlie any spatial variation in nest survival, we built and compared the reduced models nested in the departure model. For both breeding stage, the departure model was S(Year ï‚´ Month). This step of the analysis permitted the definition of a "reference model" that captured the most important variation in nest survival, without relying on specific assumptions concerning the covariates underlying their temporal variation (hereafter called "Time model). We performed the selection of the reference model independently for incubation and rearing.

Once the reference model had been defined, we tested the covariates potentially underlying time variation (listed in Table 1). For each breeding stage, we performed a step up procedure in which, at each step, we selected among the covariates considered as potentially underlying variation in nest survival the one whose addition to the model led to the best joined AICc and parsimony criterion. The model including all the covariates selected for describing variation in nest survival probabilities will be hereafter referred to as the "final model". We did not include polynomial or interaction terms in these models because this would have given rise to increased problems of model selection and interpretation. Environmental covariates were all standardized [(values - mean)/sd] to allow a direct comparison of their relative importance. Means and estimates are presented  SE.


Over the whole 2004-2008 period, we monitored a total of 5505 nests during incubation and 5412 nests during nestling stage that were observed respectively for 53285 and 81041 exposure days (see Table 2 & 3 for more details).

Annual failure rate averaged 43.6%  1.6 during incubation and 30.8%  1 during nestling stage (Table 3). Predation was always the main cause of breeding failure and accounted for an average of 50%  3.7 of failed nests during incubation and 44.3%  2.7 during nestling stagerearing.

Temporal variation in nest survival

In both breeding stages, time models performed better than constant model. Selection from the departure model suggested an additive effect of year and month on nest survival during incubation, and an effect from the interaction of year and month during the nestling stage (Table 5 & 6). Annual nest survival rate ranged from 34  1% to 46 1% during incubation, and 58  1% to 63  1% during the nestling stage (Fig. 2). Nest survival during incubation was always below 50% and therefore lower than during the nestling stage but also more variable between years. In both breeding stages, nest survival significantly increased throughout the breeding season, being the lowest from march to may and culminating during September-October (Fig. 3?). Evaluation du succès mensuel sans effet habitat à sortir

Environmental covariates underlying variation in nest survival.

The final model describing for variation in nest survival during incubation included effects of proximal habitat, minimal temperature, latitude, forest cover and corvids abundance (Table 5). During the nestling stage, the final model retained effects from proximal habitat, corvids abundance, longitude and latitude, and arable land cover (Table 6). In both breeding stages, the addition of other covariates to these models did not significantly improve the fit of the final model (AICc < 2).

For both breeding stages, the addition of proximal habitat to the time reference model improved the model considerably, with a decrease of respectively 139.6 and 188.3 AICc (Table 5 & 6). Whatever the breeding stage, nest survival was the lowest in farmland and forest habitats, and the highest in urbanized habitats (Fig. 4). Among the latter ones, nest survival was consistently higher in villages compared to towns (Fig. 5a, b). The difference in nesting survival between rural and urban habitats reached at least 14% (Forest/urban) during incubation and 15% (farmland/urban) during the rearing period.

Models including effect from precipitation, woodpigeon abundance, mosaic and urban habitat cover, received no support. Other covariates included in the final model accounted only for a relatively substantial part in the residual variation of nest survival. Table 7 shows the estimates of the slope for the covariates retained in the final models. During incubation, nest survival was positively affected by minimal temperature, and, to a lesser extent, by corvids abundance, but inversely related to the amount of forest land cover within the foraging radius. Among those covariates, the effect of minimal temperature was stronger than the others. During the rearing period, success was again affected positively by corvids abundance and the amount of arable land cover, with no marked difference in the strength of the relationship between these two covariates.

For both breeding stages, final models also included some effects from latitude only (incubation) or both latitude and longitude (rearing), giving no functional explanation by itself but suggesting that a remaining part of variation in nest success was explained by other spatially structured factors that we did not use in this study. During the incubation period, the effect of latitude was as strong as the effects from forest cover or corvids abundance. During the nesting stage, longitude and latitude even ranked in second position after corvids abundance effect.


The knowledge of time and spatial variation in breeding performance along with the identification of environmental factors underlying such variation is critical for making predictions with respect to the dynamics of populations. Among the various parameters underlying breeding performance, nest success has been suggested to be the most important component (Wiklund 1995).

The measure of temporal variation of nest success is then necessary to evaluate the potential consequences of its variability on population trend. Temporal comparison of nest success may be limited when relying on distinct studies which use incomparable estimators. Our study offers the advantage to use the same estimator and field methodology over the five years study period, making easier such comparison. With respects to spatial variation in nest success, the study of species breeding over large area and using a wide range of breeding habitat types is problematic since the results of local-scale studies might be representative only of a specific habitat and will not systematically apply to the whole population. Further, local-scale process might constrained by habitat characteristics at the landscape scale (Peak et al. 2004, Donovan et al. 1997, Hartley and Hunter 1998, Tewksbury et al. 1998). To understand such large scale patterns, it is therefore necessary to evaluate how processes operate at different spatial scales (Caldow and Racey 2000, Ormerod and Watkinson 2000, Chalfoun and Martin 2007) and for all main breeding habitats used by the targeted population. Consequently to logistical costs inherent to such design, few studies investigated large spatial-scale variation in nest success at intraspecific levels (Paradis et al. 2000; Donovan et al. 1997, Beauchamp et al. 1996, Siriwardena et al. 2000). This study on woodpigeon nest success is in keeping with such approach and offers the opportunity to extend it to a non-passerine and tree-nesting species.

Variation of nest survival between breeding stage

During the study period, hatching probability was lower than fledging probability. Such effect was highly consistent since it was observed for every year and all breeding habitat. Interspecific comparisons are uneasy to perform since other studies have been conducted most of the time in a specific area and during a short time period, and address mainly passerines species. Some studies found a higher success during incubation (Holcomb 1972, Siriwardena et al. 2000, Lantz and Conway 2009), other during nestling stage (Young 1963, Robertson 1972, Schaub et al. 1992, Hatchwell et al. 1996, Paradis et al. 2000, Burhans et al. 2002, Schmidt et al. 2008), or no significant difference (Donald et al. 2002). Interestingly, Siriwardena et al. 2000 analysed temporal variation in nest survival of farmland birds from 1962 to late seventies using "year block data" and showed that for some periods, the difference between breeding stages in nest success could reverse within the same species, which pinpoints the need to work over long periods to correctly estimate the range of variation in nest success. More surprisingly, in both Colombides species that were analysed in this study (stock dove Streptopelia decaocto and turtle dove S. turtur), success during incubation was higher that during nestling stage. That could result from a different context in predation pressure since predator populations in United Kingdom were low and recovering during the period of the study (sixties to early nineties; Marchant et al. 1990).

Temporal variation in nest survival

Annual variation in woodpigeon nest survival was much more pronounced during incubation than during the nestling period: the maximal annual difference in hatching probability reached 35% whereas it reached 9% for nestling stage. In 3 years out of 5, nest survival rates ranked similarly in both breeding stages (2004, 2005, 2006); 2004 and 2005 had respectively the highest and lowest values in both stages. If considering constant all other components underlying breeding performance, then most of the annual variation in this parameter might be driven by nest survival rate during incubation. Whatever the breeding stage, the later the nesting attempt was initiated in the breeding season, the higher was nest survival. Such result might seem equivocal since variation in bird breeding success is often shown to decrease in the late part of the breeding season (ref XXX?). However, some other species show an increase in nest survival as the breeding season progress, like the burrowing owl, Athene cunicularia (Lantz & Conway 2009) or the Moutain plover, Charadrius montanus (Dinsmore et al. 2002). In our study, such seasonal increase might result from a change in predation rate (shift in predator community or in availability of alternative preys). Alternatively, food availability peaks in September-October, when harvested cereals field provide large amount of set aside seeds. In the case of the woodpigeon, this higher nest survival late in the season relies on a much lower number of breeding attempt, raising the possibility that this higher nest success is due to more experienced birds with higher parental skill. One can wonder why breeding should stop in September if nest success is so high at that time but it should be reminded that we measured nest survival only during incubation and nestling stage, not taking into account the post-fledging period, when young birds have left the nest but are still fed by adults. Young woodpigeon body size is only two-third of adult body size when they leave the nest (Murton 1965, Robertson 1986) and selection pressures might arise late in autumn and winter, mainly through variation in food availability, to negatively affect survival rates of late born woodpigeons (RefXXX).

Effect of environmental variables on nest survival

In both incubation and nestling stages, final models included a few number of environmental effect, with limited support for our hypotheses concerning foraging habitat effect on nest survival. Contrarily, the environmental variable accounting for most of the variation in nest survival was nesting habitat, suggesting that local scale process such as habitat patch type effect may be predominant over processes acting at landscape levels such as foraging habitat types. The most spectacular result was that nest survival was much higher in urbanised habitats than in forest and farmland. If taking into account nest survival over the whole breeding attempt, by making the product of incubation and nestling survival rates probabilities, then one clutch in rural or forested habitats would have only 10 to 20% of chance to produce at least one offspring while urban would reach 25 to 45% of chance, that is a two-threefold difference between rural and urban habitats.

Greater nest survival in urban than rural habitats have been shown for other species having recently colonized urban habitats on various continents (Kentish et al. 1995, van Heezik et al. 2008). Tomialjoc (1980) had previously shown that woodpigeon productivity in urban could be much higher in some polish cities compared to adjacent rural populations. Food availability within urban area is unlikely to explain the higher nesting success of woodpigeon since most birds feed outside of the towns (Tomialojc 1980). This higher urban nest survival has often been related to a lower predation pressure resulting from lower predator abundance (Snow 1958, Stjernberg 1979, Tomialojc 1980, Tomialojc et al. 2004, Piotrowski and Wesolowski 1989, Newton 1193) and/or change in predator community (refXXX). However these results are equivocal since some studies have found increasing nest predation in more urbanised areas where there are more generalist predators, both avian and mammalian (Jokimäki & Huhta 2000, Thorington and Bowman 2003), whereas others reported decreasing nest predation and declining predator abundance with increasing human housing density (Wilcove 1985, Donovan et al. 1997, Gering and Blair 1999, Haskell et al. 2001). These differences between studies might also stem from the definition of urbanized habitat they use (see Chamberlain et al. 2009) since urban area may greatly vary in habitat composition (ratio between green space, residential housing, industrial/commercial buildings). It should be noted that in our study, nest survival was higher in villages than in larges cities, particularly during the incubation stage. In our habitat coding classification, "village" encompassed scattered houses to small towns of less than 3000 inhabitants. Several hypotheses might explain the difference in nest survival between villages and towns, (i) woodpigeons breeding inside village have less distance to travel to food resources, allowing shorter absences from the nest, (ii) habitat microstructure in village may offer more opportunities to conceal nests (persistence of hedge and thickets in garden), or to settle them in places less accessible to predators such as beams under roof in opened shelter, (iii) large cities welcome more parks more or less forested which give opportunities to predators to settle more durably and in larger numbers, particularly for corvids but also for mammalian predators such as squirrels.

Nests found both in forested and farmland habitats had the lowest survival rates and during the incubation stage, and nests found in forested habitat had a lower survival rate than in farmland. Studies that investigated nest survival rates in forested habitats found low values (Boulton et al. 2008), in some cases similar to survival rates measured in adjacent farmland habitat (Hatchwell et al. 1996). Forest area might be considered as a low biomass territory for foraging woodpigeons which may need to leave the nest unattended more frequently and therefore more vulnerable to predators (Schmidt 1999, Zanette et al. 2000). It has also been shown that within forested habitat, nest survival decreased from core area to the edge, due to an increasing predation pressure on the edge (Benson et al. 2010, Driscoll et al. 2005, Donovan et al. 1997, Hatchwell et al. 1996), although some works found evidence of an "edge effect on ground-nesting guilds but not shrub or canopy-nesting guilds (Lloyd et al. 2005; Chamberlain et al. 1995). In our study, the "Forest habitat" category encompass thickets to large forest patch. Woodpigeons nests monitored in forested habitats were predominantly found in small patch of a few hectares, and most often detected within the edge of the patch. It is therefore reasonable to think that our estimation of nest survival in forested habitat might be a bit underestimated because nest monitoring occurred most often in edge area of forest patches, where nests face a higher predation risk, particularly in fragmented landscapes (Donovan et al. 1997, Andren 1995), where nest predators and particularly members of the Corvidae family increase in abundance, activity and species richness (Chalfoun et al. 2002). Underestimation of nest survival could also result from a bias in nest height distribution within our data set. Observers found mainly nests between 2 to 6 meters high, that is in a height range where nests are exposed both to mammalian and avian predators (Martin 1995). Nests higher in the canopy suffer predation mostly by corvids (Remes 2005) with, possibly, a consecutive lower failure rate (Murton 1958) .

Woodpigeons nesting in farmland habitats use mainly hedgerows and scattered trees. The low nest survival rate we measured may be unexpected since breeding adults are close to food resources and should attend the nest more efficiently. Again, the habitat classification we use hide a large diversity in habitat structure, particularly with respects to the hedgerow density. It is likely that observers were likely to find more easily badly concealed nests in relatively degraded hedgerows complex than in preserved bocage. As nest failure rate increases along with fragmentation intensity of hedgerows (Hinsley and Bellamy 2000), our estimate of nest failure might have been slightly overestimated. However, it should be noted that French agricultural landscape is now predominantly dominated by openfield system (Agreste 2010), with degraded hedgerows systems, so our results are likely to be representative of the whole farmland habitat.

Foraging habitat

Nest survival was negatively related to importance of forest land cover within the foraging radius during the incubation period, and positively linked to the amount of arable land cover during the rearing stage. However, these effects were very weak within the final model. Such effects are consistent with the results we obtained with regards to the proximal habitat. During incubation, the increase of forest area may probably reduce the area where adults can access to food resources and lead birds to forage longer/farther to find food, leaving nest unattended. Why such relationship does not apply for the nestling stage remains unclear. Pigeons and dove eggs are pure white, and therefore unattended nests with eggs may be more conspicuous and detectable by predators than nests with chicks.

The increase of nest survival during the nestling stage along with the amount of arable land cover suggests that higher food availability may favour shorter foraging trips for adults , a higher feeding frequency and consequently improves chick growth rate. Chicks growing faster are able to spend less time at nest and therefore reduce their period of vulnerability to nest predators.


An unexpected result of our study was the positive relationship in both breeding stages between the 2008 corvids abundance, used as a proxy of predation pressure, and nest survival rate. In Europe, corvids are considered to be the main predators of open-nesting birds (Groom 1993, Del Hoyo et al. 2003) and particularly Colombides species (Murton 1965, Murton and Isaacson 1964). Descriptive and experimental studies have shown a negative impact of predator abundance on nest survival (White et al. 2008, Donald et al. 2002, Anthony et al. 1991, Tapper et al. 1996, Paradis et al. 2000). Our result may appear to be in contradiction with the interpretations of our results on nesting and foraging habitat effect on nest survival which often call on predation argument. However, paradoxes are not uncommon in a multivariate framework and are best explained by looking the slope of each variable to assess relative influence on nest survival. The model including the effect of time and nesting habitat was poorly improved by the addition of environmental. In the incubation stage, the slope of the corvids abundance effect (0.044) was the weakest among all variables selected in the final model. Conversely it ranked first in the nestling stage (0.13). This apparent paradox might be better understood if considering the European context with respect to predation pressure: Martin and Clobert (1996) mentioned that nest predation rates are lower in Europe than in North America for open-nesting birds, and that "large-scale modification of the environment may have reduced evolutionary constraints of nest predation on life-history traits in European systems. Consequently, patterns of nest predation and the importance of food versus nest predation to life-history variation may have changed such that food may be a more important influence in European systems". Accordingly, our study suggests that predation pressure might not be a powerful limiting factor on woodpigeon nest survival, which might be more under the influence of habitats structure and consecutive access to food. The positive relationship we obtained might illustrate that habitat in which woodpigeon nest survival is high are probably also those that are favoured by corvids to breed. Indeed, it should be noted here that the effect of predators on nest survival may be equivocal and may not apply similarly to all open nesting bird species. Average song thrush nest survival have not declined in regions of Britain with increasing magpies densities (Gooch et al. 1991). In some experimental studies which showed a reduction of nest failure following the control of predators populations, the authors recognized that such effect might be confounded by concurrent and concomitant introduction of habitat management measures (White et al. 2008, see also Stoate and Szczur 2001). Andren (1995) mentioned that shifts in predator identity between fragmented and unfragmented landscapes may cause nonlinear relationship between predation rates and landscape structure. Additionally, many studies focused on ground nesting bird species (Donald et al. 2002, Anthony et al. 1991, Tapper et al. 1996), and a majority address passerines species, thus the question remains of their application to non passerine species.

We acknowledge that some caution must be taken in the interpretation of the results about predation impact with respects to our own limits in the way we approximated predation pressure. The main limit stem from the fact that our estimation of corvids abundance index relies on one year only (2008), thus the source of variation in predator abundance was only spatial and not temporal. Furthermore, the main distance between woodpigeon nests in our data set and the closest point count providing abundance value is 9 km; such distance might exceed the foraging radius of corvids and therefore our abundance index might be not as representative of the predator abundance surrounding the nests as we would expect. Finally, we have no information about the abundance of mammal nest predators, particularly with regards to squirrels which are known to prey upon eggs left unattended (Murton 1965, Murton and Isaacson 1964).

Meteorological conditions

Contrary to our prediction, monthly precipitation amount was not selected into final models to explain variation in nest survival. Colombides nest are loosely built structure which are known to be vulnerable to adverse weather conditions such as strong winds and heavy precipitations (refXXXX). The monthly amount of precipitation may be not a well adapted measure of such conditions and further studies should rather use weekly at least amount along with wind strength in order to better take into account adverse conditions.

On another hand, mean monthly minimal temperature appeared to significantly impact on nest survival during incubation. This factor ranked first after nesting habitat effect (slope 0.13) among all environmental factors included in the final model. To our knowledge, no large scale studies had already analysed the impact of such factor on nest survival. Low temperature can affect incubation success either directly by chilling unattended eggs and slowing embryo development (Haftorn 1988), or indirectly by increasing adult metabolic activity required to maintain body temperature (Conway and Martin 2000a, b), and therefore altering the bouts length at nest and the frequency of foraging trips (Conway and Martin 2000a, b). Increased metabolic costs may lead adults to leave more often the nest in order to replenish their body stores, making the nest more vulnerable to predators or adverse weather conditions. Temperatures could therefore constitute one of the most significant limit on nest success in the first part of breeding season. Again, nesting habitats might be not all equivalents with respect to such constraint since urban area may generate a warmer microclimate, compared to other rural habitats (Luniak et al 1990, Sukopp 1998). Urban birds would then benefit from faster embryo and chick growth, and thus reduce the exposure period where they are accessible to nest predators. Urban birds can also start breeding earlier, and it has already been noted that woodpigeons start breeding in city centres 8-10 days earlier than in rural area and finish the broods somewhat later (Tomialojc 1980).

Woodpigeon population abundance

Nest survival was not linked with population abundance. Despite the steady increase in woodpigeon abundance, no density dependence effect was detected on nest survival, suggesting that population abundance has not reached yet a saturation threshold. Because we did not included models with interaction between variables, we can not however rule out possible effect of abundance in specific habitat. Such investigation remains to be done, particularly in habitats where large variations in abundance values are fund, like urban and farmland habitats.


Our study put into evidence that proximal habitat is the main factor driving nest survival rate in woodpigeon population, well ahead foraging habitats and other environmental variables. Such effect appears very consistent over time and space, and at large spatial scale. Although it is initially tempting to first link such differences between habitats to difference in predation rate, our results suggest that the impact of predation on nest survival variation is not as powerful as we expected, at least in a spatial perspective. However we acknowledge that this point deserve more analysis since our proxi of predation pressure needs to be improved. Further, our data do not enable us to infer about the consequence of temporal variation in predation pressure on nest survival. Yet, we think that our results are in agreement with comments made by Martin and Clobert (1996) suggesting that life-history traits might be driven mainly by food.

We voluntarily used a reduced number of variables in our analysis, potentially encompassing main potential factors acting on nest survival. However, in both breeding stages, and particularly in the nestling one, nest survival was also related to latitude and longitude, which gives no functional explanation by itself but suggests that a remaining part of variation in nest survival is explained by other spatially structured factors not considered in this study.

Urban woodpigeon populations show much higher nest survival rates than in any other nesting habitat. These higher values may result from lower predation pressure, proximity of food resources since a majority of cities are surrounded by large cultivated areas, and more favourable climatic conditions. In the context of the dramatically increase in populations observed over the lasts decades, this result raises the question of the contribution of the urban population into the whole woodpigeon population. Urban populations can exhibit a demographic surplus, acting as a source of individuals within a regional network of local populations (Dias 1996). Tomialojc (1980) argued that a high emigration of young hatched in urban parks is indispensable, as a consequence of a very high production of young and simultaneous lack of continuous increase in numbers of these populations. Radio-tracking data also show that woodpigeon born in urban places can start to breed in adjacent agricultural areas (unpublished data). Yet, nest survival estimates provide only one piece of the puzzle with regard to understanding the dynamic of urban bird populations. Information on the number of clutches laid each season, juvenile and adult mortality, emigration and immigration rates is necessary before conclusions can be drawn confidently about whether populations are self-sustaining within the urban environment. Similarly, whether lower nest survival in rural habitats is sufficient to create "sink populations" remains to be proved because it depends on the survival of juveniles and adults in the different habitats. If any of the vital rates mentioned above are under density-dependent control, results of increased nest survival and further nest success may be counterbalanced by reduction in another vital rate.

Woodpigeon productivity is potentially subject to much variation since birds are able to rear multiple broods (up to 5) and produce replacement clutches quickly after failure. Conversely, the rather fixed clutch size (2 chicks maximum) limits the range of productivity per breeding attempt. We suggest however that productivity of urban woodpigeon population is likely to be higher than in other habitats at least because of a longer urban breeding season (earlier start and later ending), giving more renesting opportunities. Our estimation of nest survival during the nestling stage does not take into account the period where juvenile woodpigeons are out of the nest but still dependant of parents to be fed. At this time, birds fly very poorly, and stay in the immediate vicinity of the nest, where they are still accessible to nest predators. This period remains very poorly studied largely because of the cryptic coloration and the secretive behaviour of young birds, making difficult to collect repeated observations and to quantify their survival. Consequently, estimates of spatial and temporal variation pattern of reproductive productivity, and the processes that limit avian productivity are often based strictly on the number of young that are successfully fledged from the nest (reviewed in Anders & Marshall 2005). Some studies have shown that the factors influencing the probability of nest success may not directly apply to the post-fledging period (Schmidt et al 2008). For this reason, extrapolation of nesting data for inferring population-level effects or ecological interactions will not be warranted. Accordingly, our considerations about urban woodpigeon population dynamics need thorough studies on post-fledging period, relying either on radio-tracking data and/or capture-recapture data set on ringed birds.

Armstrong, D. P., E. H. Raeburn, R. G. Powlesland, M. Howard, B. Christensen, and J. G. Ewen. 2002. Obtaining meaningful comparisons of nest success: data from New Zealand Robin (Petroica australis) populations. New Zealand Journal of Ecology 26:1-13.

Baiser, B., R.L. Boulton and J.L. Lockwood. The influence of water flows on nest success of the endangered Cape Sable seaside sparrow in the Florida Everglades. Animal Conservation 11: 190-197.

Barber, D. R., T. E. Martin, M. A. Melchiors, R. E. Thill, and T. B. Wigley. 2001. Nesting success of birds in different silvicultural treatments in southeastern U.S. pine forests. Conservation Biology 15:196-207.

Birdlife International, 2004. - Birds in Europe : population estimates, trends and conservation status. Cambridge, UK : BirdLife International. BirdLife Conservation Series N°12. 374 p.

Blondel, J. 1985. Breeding strategies of the Blue Tit and the Coal Tit (Parus) in mainland and island Mediterranean habitats: a comparison. J. Anim. Ecol. 54: 531-556.

Blondel, J., Gosler, A., Lebreton, J.D. and McCleery R. (Eds.). 1990. Population Biology of Passerine Birds. An integrated Approach. NATO ASI Series G, Vol. 24. 513 Pages. Springer-Verlag, Heidelberg.

Bossard, M., Feranec, J. & Otahel, J. 2000. CORINE land cover technical guide - Addendum 2000. Technical Report No. 40. European Environmental Agency. Copenhagen.

105ppBossenbroek, J. M. Wagner, H.H. & Wiens, J.A. 2005. Taxon-dependent scaling: beetles, birds, and vegetation at four North American grassland sites. - Landscape Ecol. 20: 675-688.

Chalfoun, A.D. & Martin, T.E. (2007) Assesments of habitat preferences and quality depend on spatial scale and metrics of fitness. Journal of Applied Ecology 44: 983-992.

Chamberlain, D.E., Hatchwell, B.J. & Perrins, C.M. 1995. Spaced out nests and predators: an experiment to test the effects of habitat structure. J. Avian Biol. 26: 346-349.

Colquhoun, M.K. 1951. The Woodpigeon in Britain. London.

Cramp, S. 1985. The Birds of the Western Palearctic. Vol. IV - Terns to Woodpeckers. Oxford University Press, Oxford.

Boulton, R.L., Richard, Y. & Armastrong, D.P. 2008. Influence of food availability, predator density and forest fragmentation on nest survival of New Zealand robins. Biological Conservation 141: 580-589.

Boutin, J. M. 2001. Les populations de colombidés nicheurs en France. Faune Sauvage, cahier technique 253 : 26-32.

Boutin, J.-M., Roux, D. & Eraud, C. 2003. Breeding bird monitoring in France:

the ACT Survey. Ornis Hungarica, 12-13, 1-2.

Caldow, R.W.G. & Racey, P.A. (2000) Introduction. Large-scale processes in ecology and hydrology. Journal of Applied Ecology, 37 (Suppl. 1): 6-12.

Chamberlain, D.E., Hatchwell, B.J. & Perrins, C.M. 1995. Spaced out nests and predators: an experiment to test the effects of habitat structure. Journal of Avian Biology 26(4): 346 - 348.

Cramp, S. (1985) The Birds of the Western Palearctic. Vol. IV - Terns to Woodpeckers. Oxford University Press, Oxford.

Donald, P.F., Sanderson, F.J., Burfield, I.J. & van Bommel, F.P.J. (2006) Further evidence of continent-wide impacts of agricultural intensification on European farmland birds, 1990-2000. Agriculture, Ecosystems and Environment, 116, 189-196.

Driscoll, M. J., T. Donovan, R. Mickey, A. Howard, and K. K. Fleming. 2005. Determinants of wood thrush nest success: a multi-scale, model selection approach. Journal of Wildlife Management 69:699-709.

Duguay, J. P., P. B. Wood, and G. W. Miller. 2000. Effects of timber harvests on invertebrate biomass and avian nest success. Wildlife Society Bulletin 28:1123-1131.

Fluetsch, K.M. & Sparling, D.W. 1994. Avian nesting success and diversity in conventionally and organically managed apple orchards. Environ. Toxicol. Chem. 13: 1651-1659.

Froyd, C. & Willis, K. (2008) Emerging issues in biodiversity & conservation management: the need for a palaeoecological perspective Quaternary Science Review, 27, 1723-1732

van Horne, B. 1983. Density as a misleading indicator of habitat quality. J. Wildl. Manage. 47: 893-901.

Galbraith, H. 1988. Effects of agriculture on the breeding ecology of lapwings Vanellus vanellus. J Anim Ecol 25:487-504.

Green, R.E. 1988. Effects of environmental factors on the timing and success of breeding of common snipe Gallinago gallinago (Aves: Scolopacidae). Journal of Applied Ecology 25:79-93.

Green, R.E., Tyler, G.A., Stowe, T.J. & Newton, A.V. 1997. A simulation model of the effect of mowing of agricultural grassland on the breeding success of the corncrake (Crex crex) Journal of Zoology, London 243: 81-115.

Inglis, I.R., Isaacson, A.J. & Thearle, R.J. P. (1994) Long term changes in the breeding biology of the wood pigeon Columba palumbus in eastern England. Ecography 17(2): 182-188.

Hatchwell, B.J., Chamberlain, D.E. & Perrins, C.M. 1996(a) The demography of blackbirds Turdus merula in rural habitats: is farmland a sub-optimal habitat?. Journal of Applied Ecology 33: 1114 - 1124.

Hatchwell, B.J., Chamberlain, D.E. & Perrins, C.M. 1996(b) The reproductive success of Blackbirds Turdus merula in relation to habitat structure and choice of nest site. Ibis 138: 256- 262.

Haynes, P. J., Inglis, I. R., Isaacson, T. J. & Fryday, S. L. (2003) Woodpigeon Columba palumbus movements in eastern England. Bird Study 50: 33-38.

van Horne, B. 1983. Density as a misleading indicator of habitat quality. Journal of Wildlife Management 47: 893 - 901.

Lawton, J.H. 1995. Population dynamic principles. In Extinction rates, J.H. Lawton & R. May, Eds., Oxford University Press, New York, New York, USA, pp 147-163.

Lehman, C.P., Rumble, M.A., Flake, L.D. & Thompson D.J. 2008. Merriam's Turkey Nest Survival and Factors Affecting Nest Predation by Mammals. Journal of Wildlife Management 72(8): 1765-1774.

Lormée, H., Boutin, J. M. & Aubineau, J. (2000) Le pigeon ramier. In « Enquête nationale sur les tableaux de chasse à tir - Saison 1998/1999. Faune sauvage - Cahiers techniques 251 : 62-69.

Lormée, H., Aubineau, J., Trimoreau, R. & Boutin, J. M. 2002. Dispersion et survie juveniles chez le pigeon ramier, Columba palumbus. Proceedings of the XXVII colloque français d'ornithologie. 2-29 september, Strasbourg.

Martin, T. E. (1992) Breeding productivity considerations: what are the appropriate habitat features for management? In: Ecology and conservation of neotropical migrants (eds J.M. Hagan & D.W. Johnston), pp 455-473. Smithsonian Institution press, Washington, D.C.

Martin, T. E. & Clobert, J. (1996) Nest predation and avian life-histor evolution in Europe versus North-America: a possible role of humans? American Naturalist 147(6): 1028-1046.

Marzluff, J.M. 2001. Worldwide urbanization and its effects on birds. In Marzluff, J.M., Bowman, R. & Donelly, R. (eds) Avian Ecology and Conservation in an Urbanizing World: 19-48. New York: Kluwer.

Mayfield, H.F. (1975) Suggestions for calculating nest success. Wilson Bulletin 87:456-466.

Møller, A.P. 1988. Nest predation and nest site choice in passerine birds in habitat patches of different size: a study of magpies and blackbirds. Oikos 53: 215 - 221.

Murton, R. K. (1965) The wood pigeon. The new naturalist, London.

Murton, R. K. Westwood, N. J. & Isaacson, A. J. (1974) A study of wood pigeon shooting: exploitation of a natural animal population. Journal of Applied Ecology. 11(1): 61-81.

O'Connor, R. J. & Fuller, R.J. (1986) Bird population responses to habitat. In: Bird Census and Atlas Studies (eds K. Taylor, R.J. Fuller & P.C. Lack), pp 197-211. BTO, Tring (A relire).

Paradis, E., Baillie, S.R., Sutherland, W.J., Dudley, C., Crick, H.Q.P. & Gregory, R.D. 2000. Large-scale spatial variation in the breeding performance of song thrushes Turdus philomelos and blackbirds T. merula in Britain. J. Appl. Ecol. 37 (Suppl. 1):


Pulliam, H. R. 1988. Sources, sinks, and population regulation. American Naturalist 132:653-661.

Reidy, J.L., Thompson, F.R. & Peak, R.G. 2009. Factors Affecting Golden-Cheeked Warbler Nest Survival in Urban and Rural Landscapes. Journal of Wildlife management 73(3): 407-413.

Robinson, R.A,. Green, R.E., Baillie, S.R., Peach, W.J. & Thomson, D.L. 2004. Demographic mechanisms of the population decline of the song thrush Turdus philomelos in Britain J. Anim. Eco. 73: 670-682

Roux, D., Eraud, C., Boutin, J. M., Landry, P., Dej, F. & Lormée, H. 2009. Suivi des populations des oiseaux nicheurs et de passage en France : année 2009. Réseau national d'observation « Oiseaux de passage », rapport interne ONCFS/FNC/FDC, octobre 2009, 60 p.

Shochat, E., Warren, P.S., Faeth, S.H., McIntyre, N.E. & Hope, D. 2006. From patterns to emerging processes in mechanistic urban ecology. Trends Ecol. Evol. 21: 186-191.

Schmidt, K. A. and R. S. Ostfeld. 2003. Songbird populations in fluctuating environments: predator responses to pulsed resources. Ecology 84:406-415.

Siriwardena, G.M., Baillie, S.R., Crick, H.Q.P. & Wilson, J.D. (2000) The importance of variation in the breeding performance of seed-eating birds in determining their population trends on farmland. Journal of Applied Ecology, 37, 128±148.

Snow, D. 1958. A study of blackbirds. British Museum, London.

Stjernberg, T. 1979. Breeding biology and population dynamics of the scarlet rosefinch Carpodacus erythinus. Acta Zoologica Fennica 157: 1-88.

Stoate, C. & Szczur, J. 2001. Could game management have a role in the conservation of farmland passerines? A case study from a Leicestershire farm. Bird Study 48: 279-292.

Thompson, B. C., G. E. Knadle, D. L. Brubaker, and K. S. Brubaker. 2001. Nest success is not an adequate comparative estimate of avian reproduction. Journal of Field Ornithology 72:527-536.

Tomialjoc, L. (1976) The urban population of the wood pigeon Columba palumbus L. 1758, in Europe - its origin, increase and distribution. Acta Zoologica Cracoviensia 21(18): 585-621.

Tomialojc, . 1980. The impact of predation on urban and rural woodpigeon Columba palumbus (L.) populations. Polish Ecological Studies 5: 141 - 220.

Vickery, P.D., Hunter, M.L. & Wells, J.V. 1992. Is density an indicator of breeding success? Auk 109: 706-710.

Weidinger, K. 2002. Interactive effects of concealment, parental behaviour and predators on t