Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

The traits that predict the magnitude and spatial scale of forest bird responses to urbanization intensity

  • Grant D. Paton ,

    Roles Conceptualization, Formal analysis, Methodology, Visualization, Writing – original draft

    gpaton@uncc.edu

    Affiliation Department of Geography and Earth Sciences, The University of North Carolina at Charlotte, Charlotte, North Carolina, United States of America

  • Alexandra V. Shoffner,

    Roles Data curation, Methodology

    Affiliation Department of Fisheries and Wildlife, Michigan State University, East Lansing, Michigan, United States of America

  • Andrew M. Wilson,

    Roles Data curation, Investigation, Writing – review & editing

    Affiliation Environmental Studies Department, Gettysburg College, Gettysburg, Pennsylvania, United States of America

  • Sara A. Gagné

    Roles Conceptualization, Methodology, Project administration, Supervision, Writing – review & editing

    Affiliation Department of Geography and Earth Sciences, The University of North Carolina at Charlotte, Charlotte, North Carolina, United States of America

The traits that predict the magnitude and spatial scale of forest bird responses to urbanization intensity

  • Grant D. Paton, 
  • Alexandra V. Shoffner, 
  • Andrew M. Wilson, 
  • Sara A. Gagné
PLOS
x

Abstract

As humans continue moving to urban areas, there is a growing need to understand the effects of urban intensification on native wildlife populations. Forest species in remnant habitat are particularly vulnerable to urban intensification, but the mechanisms behind these effects are poorly understood. An understanding of how species traits, as proxies for mechanisms, mediate the effects of urban intensification on forest species can help fill this knowledge gap. Using a large point count dataset from the Second Pennsylvania Breeding Bird Atlas, we tested for the effects of species traits on the magnitude and spatial scale of the responses of 58 forest bird species to urbanization intensity in landscapes surrounding count locations. Average urbanization intensity effect size across species was -0.36 ± 0.49 (SE) and average scale of effect of urbanization intensity was 4.87 ± 5.95 km. Resident forest bird species that are granivorous or frugivorous, cavity-nesting, and have larger clutch sizes and more fledglings per clutch had more positive associations with increasing urbanization intensity in landscapes. In addition, the effect of urbanization intensity on forest birds manifested most strongly at larger spatial scales for granivorous, frugivorous, or omnivorous species that are cavity-nesting, have larger clutch sizes and longer wingspans, and flock in larger numbers. To our knowledge, the present study represents the first direct tests of the effects of species traits on both the magnitude and spatial scale of the effect of urbanization on forest birds, as well as the first evidence that migratory status, clutch size, wingspan, and fledglings per clutch are important determinants of the responses of forest birds to urbanization. We discuss the possible mechanisms underlying our results and their implications for forest bird conservation in urbanizing landscapes.

Introduction

The 21st century will be defined by the expansion of urban development. Fifty-four percent of the world’s population lived in urban areas in 2014, with 66% expected to be doing so by 2050 [1]. As a result, total urban area is projected to grow by at least 430,000 km2 from 2000 to 2030, an increase of 139% [2]. Expanding urban centers have significant impacts on species in remnant habitat, such as forest breeding birds. The abundance of generalist species in forest patches is affected by proximity to and intensity of surrounding urban development [3, 4, 5]. For example, blue jay (Cyanocitta cristata) abundance is over three times greater in forests surrounded by medium-intensity housing than in those surrounded by low-intensity housing [6]. Urbanization also has particularly strong negative effects on the abundances of specialist and forest interior species [7, 8, 9, 10]. For example, based on data from the French Breeding Bird Survey, populations of specialist species become more unstable and prone to collapse as urbanization increases [11].

The mechanisms by which urbanization affects forest bird species include reduced movement success, altered resource availability, and increased disturbance [12]. More intense urbanization increases the effective isolation of forest patches [13, 14] and likely increases dispersal mortality by means of collisions with buildings and vehicles, and predation by domestic cats [15, 16]. Conversely, urban areas can be a source of additional resources to birds in forest patches, favoring species that can take advantage of these opportunities [17, 18, 19]. However, the replacement of native habitat with developed land may make native food sources rarer, imperiling species that have specific dietary or nesting needs [20]. Finally, urban development is associated with increases in anthropogenic noise, air temperature, pollution, parasitism, and predation, all of which significantly impact habitat quality for birds [21, 22, 23, 24, 25, 26, 27, 28].

Species traits are useful in elucidating the importance of these mechanisms in structuring forest bird communities. Forest bird traits found to be associated with high urban abundances highlight the altered resource availability and/or increased disturbance of urban settings: an omnivorous diet, a medium or high nesting height, sociality, biparental involvement in nest construction, long duration in nest after hatching, and cavity nesting [8, 20, 29]. Omnivory may indicate a more generalized feeding behavior, potentially enabling a forest species to utilize anthropogenic food sources, such as trash, non-native vegetation, and bird feeders and/or feeding tables [30]. Medium to high nesting heights may ensure that forest bird nests are less exposed to disturbances from human activity [31, 32] and nest predation [33, 34]. Cavity nesting may also similarly reduce the exposure of nests to human disturbance and predation [33]. Species that live in medium- or large-sized groups may benefit from urban survival strategies shared among group members [35] and biparental involvement in nest construction may result in more stable nests that can better withstand urban disturbances [36]. Species that spend longer periods of time in the nest before fledging require lower daily resource needs than faster-growing species, making them more suited to a potentially resource-depleted urban area [20, 37].

Despite these insights, there is much to gain in our understanding of forest bird trait variation associated with urbanization. First, we are not sure whether traits associated with urbanization success in birds in general, such as being a resident rather than a migratory species, large clutch size, multiple clutches per year, and large body size, wingspan, and brain size [20, 38, 39, 40, 41, 42, 43], also characterize abundant forest bird species in urban landscapes. Particularly lacking is evidence of associations between forest bird success in urban landscapes and traits that infer differential movement success, such as wingspan. While prior research on birds has included some forest species in analyses, this is the first study to our knowledge to focus specifically on the assemblage. Second, we have no direct evidence that traits characterizing abundant forest bird species in urban landscapes are in fact associated with positive responses to urbanization. In other words, no study to date has statistically tested the effects of species traits on the responses of forest birds to urbanization. Third, we do not know how species traits influence the spatial scale at which urbanization affects forest birds despite the likelihood that mechanisms underlying the effects of urbanization operate at a variety of spatial scales [44].

The objective of the present research was to begin to address these three gaps in our knowledge of forest bird trait variation in urban landscapes. Using a large point count dataset from the Second Pennsylvania Breeding Bird Atlas, we first estimated the effect of urbanization intensity in landscapes of multiple scales on the individual occurrences of 58 forest bird species, controlling for other potentially important explanatory variables such as forest amount and configuration, local habitat quality, and species detectability. For each species, we selected the coefficient for urbanization intensity with the largest absolute magnitude, and the spatial scale at which this occurred. These data were then used as the response variables in a second set of analyses that used species traits as predictor variables. Traits represented species characteristics linked to the mechanisms by which urbanization affects forest birds, such as: a greater capacity for movement leading to high movement mortality and reduced movement success (wingspan, migratory status, and territory size); differential resource affinity or need for resources for growth and reproduction (body size, dietary habit, foraging height, clutch size, number of fledglings per clutch, number of clutches per year, and duration in the nest); and differential vulnerability to disturbance agents such as domestic cats and anthropogenic noise (nesting height, parental involvement in nest construction, cavity nesting, flock size, lifespan, and song attributes). To our knowledge, this is the first study to use this approach.

Materials and methods

Study area

Pennsylvania is approximately 120,000 km2 in area between 39.0° and 42.5°N and 74.0° and 81°W in the northeastern United States (Fig 1). Elevations are typically 300 to 400 m, except where the state is bisected by the Appalachian and Allegheny mountain ranges (maximum elevation = 980 m), and the coastal southeastern corner of the state [45]. The climate is Humid Continental in the northern half of the state and along the Appalachian Mountains and Humid Subtropical elsewhere [46]. Pennsylvania is predominantly covered by forest– 58% of total land area [47]–dominated by red (Quercus rubra) and white oak (Quercus alba), with agricultural activity occurring in the lower elevations [48]. Average population density is 100 people/km2, with roughly 65% of the population residing in the Philadelphia or Pittsburgh metropolitan areas [49]. By 2040, Pennsylvania’s population is expected to grow by 11% to approximately 14.1 million people, with most of the growth expected in the Philadelphia metropolitan area and the center of the state [50].

thumbnail
Fig 1. Second Pennsylvania Breeding Bird Atlas point count locations within the estimated range, delineated by dashed lines, of an example species, the Kentucky Warbler (Geothlypis formosa).

The inset depicts landscapes of multiple spatial scales centered on a count location. Land cover is from the 2006 National Land Cover Database [51].

https://doi.org/10.1371/journal.pone.0220120.g001

Second Pennsylvania Breeding Bird Atlas

The Second Pennsylvania Breeding Bird Atlas (PBBA) was a formal breeding bird survey conducted statewide between 2004 and 2009 in a grid of 4,937 blocks, each approximately 25 km2 in area [52]. Within each survey block, trained staff conducted up to eight point counts, depending on block size and road accessibility for a total of 33,763 counts in the state. Some completed surveys were omitted from the final analysis due to quality control measures. Count locations were selected by random placement of a point within a block, with adjustments made to locate points on the nearest non-highway road. Point locations across the state were at least 400 m apart. Each point was surveyed once within the 5-year PBBA period in the spring or summer between May 25 and July 4. Twenty-two paid observers received extensive training before each field season. Observers recorded the species of each bird heard or seen within an unlimited distance from the point for 6 minutes and 15 seconds on weekday or weekend mornings, between 5 and 10 am, provided there were no adverse weather conditions.

Forest bird species

From the 176 species detected during PBBA point counts, we excluded species that had the following characteristics: 1) hybrid species; 2) irregular breeder in the state; 3) raptor; 4) waterfowl; 5) nocturnal species and 6) occurrence at fewer than 30 count locations. The forest dependency of each of the remaining 101 species was assessed by comparing the cumulative distribution of counts of the species, across all 33,763 count locations, to the cumulative distribution of forest amount in landscapes at each scale using Kolmogorov-Smirnov tests of no difference (S1 Fig). If a large proportion of the individuals of a species occurred in the landscapes with the most forest or if individuals of a species accumulated in landscapes at the same rate that forest cover increased in landscapes, then the species was classified as "forest-dependent" (S1 Table). Otherwise, the species was classified as “forest-independent”, i.e., if a large proportion of the individuals of a species occurred in the landscapes with the least forest (S1 Table).Two species, the Black-billed Cuckoo (Coccyzus erythrophthalmus) and the Yellow-rumped Warbler (Setophaga coronata), were too rare to permit estimation of the effect of urbanization intensity on their occurrences, resulting in a final count of 58 forest-dependent species.

Landscape definition and selection

We defined landscapes as circular areas centered on point count locations that were surrounded by >50% forest cover–composed of the Deciduous, Mixed, and Evergreen Forest classes of the 2005 National Land Cover Database (NLCD) [51]–within a 0.2 km radius (Fig 1). Deciduous forest was defined as forest cover in which deciduous species accounted for 75% or more of the cover. Evergreen forest was defined as forest cover in which evergreen species accounted for 75% or more of the cover. The Mixed Forest classification included forest cover in which neither deciduous nor evergreen species accounted for more than 75% of the cover. We restricted our selection of count locations in this way to increase the odds that species classified as forest-dependent were in fact breeding in forest. We analyzed landscapes of ten different radii surrounding point count locations: 0.2, 0.5, 1, 2, 4, 6, 8, 10, 12, and 16 km. Landscapes that overlapped with the Pennsylvania state border were omitted from analysis. Landscape sizes ranged over one order of magnitude at a high density in order to increase the likelihood of accurate identification of each species’ urbanization intensity scale of effect [53].

For each forest bird species, we selected the landscapes within the species’ range (Fig 1). To determine species ranges, we used PBBA block occurrence probabilities calculated using simple conditional autoregressive models accounting for the survey hours completed by each observer and land cover class proportions [54]. Block occurrence probabilities were assigned to the center of each block and then spatially interpolated for each species using kriging in ArcGIS to account for variation within survey blocks [55]. We classified land area with an interpolated occurrence probability ≥ 0.2 as lying within the species’ range. An occurrence probability of 0.2 can be interpreted as seeing the species in a given location once every five years on average–the duration of the PBBA.

Urbanization intensity and other explanatory variables

We used the urbanization intensity variables and measures of landscape heterogeneity, habitat quality, and species detectability measured for a companion study [56]. For each landscape scale, urbanization intensity was quantified as the first and/or second components with eigenvalues > 1 [57] from a principal component analysis of six variables: the proportional area of landscapes in each of the four Developed classes of the NLCD, area-weighted average population density in landscapes, and area-weighted average housing density in landscapes. The first and second components, accounting for 62–76% of total variance, were selected at the 0.2 and 0.5 km scales while only the first component, accounting for 67–90% of total variance, was selected at larger scales. The first component was highly correlated with population and housing density (r > 0.90) and the areas of Developed land covers (r ≥ 0.70). The second component was highly correlated with High Intensity Developed cover (r ≥ 0.75).

We included six measures of landscape heterogeneity (forest amount, forest configuration, agriculture amount, mean elevation, elevation range, and land cover diversity), two measures of local habitat quality, and four measures of species detectability in the urbanization intensity analysis. Forest amount was the proportional area of landscapes in the NLCD Forest cover class, Deciduous, Mixed, or Evergreen, that best matched each forest species’ habitat affinity [52, 58]. For generalist forest species or those for which habitat affinity was unknown, the total of all Forest cover classes was used to estimate forest amount. Forest configuration was quantified for each species using the Forest class used to calculate forest amount for the species and the patch density and clumpiness index metrics [59]. Agriculture amount was the proportional area of landscapes in the NLCD Cultivated Crops and Pasture/Hay classes. The mean elevation and elevation range of landscapes were calculated using a 3.2-ft digital elevation model from the PAMAP Program [60]. We measured land cover diversity as Shannon’s diversity index of all NLCD land cover classes in landscapes. During point counts, PBBA staff recorded two indicators of local habitat quality: evidence of recent or active local land use change (presence or absence of change, such as logging or construction) and the identity of the dominant habitat type, e.g., coniferous forest, deciduous forest, or mixed forest, within 75m of the count location. Staff also recorded four metrics that could affect bird detectability: observer identity, survey start time, Julian date, and year. These variables represented variation in observer efficacy in counting birds, daily variation in the occurrence and vocalizations of birds due in part to weather conditions, and yearly variation in climatic conditions and species population sizes.

Species traits

We collected data on the wingspan, body mass, dietary habits (frugivory, granivory, omnivory), foraging height, migratory status, nesting height, clutch size, number of fledglings per clutch, number of clutches per year, duration in the nest, cavity nesting, parental involvement in nest building, flock size, lifespan, song attributes, and territory size of each forest bird species (S2 File). We selected traits that were most commonly included, and found to be important, in previous studies of birds in urban environments. As such, our list does not include all previously studied traits such as habitat association. With the exception of song attributes, we gathered species trait data using the information in respected field guides or other sources (e.g. [58, 61]). We averaged values between sexes and across sources for continuous traits. If species could be classified into multiple categories, such as species with granivorous and omnivorous dietary habits, we assigned them to all applicable categories. Instead of classifying nesting species as ground, shrub, or tree nesters, we collected data on average nesting height in meters. We classified species as migratory or resident, with partially migratory species (species for which some but not all individuals migrate) falling into the latter category. We characterized the song attributes of each species by calculating the mean and range of song frequencies and song length from recordings for eastern and central North America [62] using the Praat software [63].

Analyses

We tested for the influence of species traits on the magnitude and scale of effect of forest bird responses to urbanization intensity in two steps. First, we used logistic regressions and multi-model inference to estimate the association of urbanization intensity with the occurrence of each forest bird species, carrying out a separate analysis for each landscape scale. There was a degree of landscape overlap at all scales except the 200m scale. However, there is evidence that this overlap does not significantly impact results [64]. Although species occurrences do not imply population-level responses to urbanization intensity, we chose to use these data rather than abundances as our response variables because some species were too rare for meaningful analysis of their abundance data. By using occurrences, we maximized the number of species for which we could obtain urbanization intensity effect sizes and scales of effect and the sample size for the second step in our analyses. Models also included additional measures of landscape heterogeneity, local habitat quality, and species detectability. The absolute values of correlations between urbanization intensity and other explanatory variables were below the threshold of 0.70; above this value, collinearity may impair model fit [65]. We divided explanatory variables by their partial standard deviations–standard deviations divided by variable inflation factors, sample size, and number of predictor variables in models–so urbanization intensity effect sizes could be sensibly averaged across models [66]. We also standardized explanatory variables.

For each species and each landscape scale, we used Akaike's Information Criterion (AIC) to evaluate models that included every possible combination of explanatory variables. We calculated the full model-averaged effect size of the first component of urbanization intensity and its unconditional variance using models with ΔAIC ≤ 2 [67] for species and scales with unimodal effect sizes (97% of cases; [68]). The scale of effect of urbanization intensity for each species was the scale at which urbanization intensity had the largest absolute effect on species occurrence.

Second, we determined the effects of species traits on species’ urbanization intensity effect sizes and scales of effect using generalized linear models and multi-model inference. Species effect sizes, i.e., the model-averaged β coefficients for urbanization intensity at its scale of effect, were weighted by their inverse variance to account for the varying precision of estimates [68]. In order to determine whether species relatedness would violate model assumptions, we first performed univariate phylogenetic regressions of urbanization intensity effect sizes and scales of effect on each trait using randomly selected Ericson backbone trees from the online service BirdTree [69, 70]. No traits exhibited a phylogenetic signal (Pagel’s Lambda (λ)) significantly different from 0, indicating no influence of species relatedness on trait effects (S2 and S3 Tables).

Some species and traits were missing many values so we modeled the effects of traits in two groups: 1) traits with positive adjusted R2 values in univariate regressions (urbanization intensity effect size: clutch size, fledglings per clutch, frugivory, granivory, omnivory, cavity nesting, and migratory status (N = 30 species); urbanization intensity scale of effect: clutch size, duration in the nest, frugivory, granivory, omnivory, cavity nesting, migratory status, flock size, foraging height, lifespan, and wingspan (N = 26 species)), and 2) traits that had values for all species (body mass, clutch size, frugivory, granivory, omnivory, cavity nesting, and migratory status (N = 58 species)). With two exceptions (clutch size x fledglings per nest, and frugivory x omnivory in the urbanization intensity response analysis with the best-performing traits (|r| = 0.72 for both pairings)), the absolute values of correlations between pairs of traits in each analysis were low. Traits were transformed in the same manner as explanatory variables in logistic regression models by dividing trait values by their partial standard deviations and standardizing the resulting values.

We used AIC to evaluate models that included all possible combinations of traits for each group and response variable (urbanization intensity effect size and scale of effect) and calculated the full model-averaged effect size and its unconditional variance for each trait in best models. All averaged effect sizes were unimodal. We performed analyses in R, version 3.4.3 [71] using the ape [72], caper [73], fmsb [74], and MuMIn [75] packages.

Results

Forest bird responses to urbanization intensity

The average urbanization intensity effect size across species was -0.36 ± 0.49 (SE), with effect sizes for individual species ranging between -1.49 ± 0.50 and 0.74 ± 0.06 (Table 1). Of the 58 species analyzed, only 14 exhibited positive associations with urbanization intensity, whereas 44 species had negative associations. The average scale of effect of urbanization intensity across species was 4.87 ± 5.95 km (Table 1). Species exhibited scales of effect at each of the ten scales we examined, although half were most sensitive to urbanization intensity at scales ≤ 1 km.

thumbnail
Table 1. The magnitude and spatial scale of the effect of urbanization intensity on the occurrences of forest bird species in Pennsylvania (USA).

Models of occurrence also accounted for other aspects of landscape heterogeneity, local habitat quality, and species detectability. Effect size was calculated as the model-averaged β coefficient of urbanization intensity at its scale of effect using models with ΔAIC ≤ 2. The sample size (N) for each species is the number of point locations surrounded by landscapes with radii equal to the scale of effect. Landscapes that intersected the state border were excluded from analyses, resulting in different numbers of landscapes at each scale.

https://doi.org/10.1371/journal.pone.0220120.t001

Effects of species traits on urbanization intensity responses

Clutch size, fledglings per clutch, frugivory, granivory, cavity nesting, and migratory status were included in the best models of the effects of species traits on the magnitude of species responses to urbanization intensity (Tables 2 and 3). All traits had a positive effect, with that of granivory being the largest, followed by fledglings per clutch, being a resident rather than a migratory species, clutch size, cavity nesting, and frugivory (Fig 2).

thumbnail
Fig 2. The effects of species traits on the magnitude of forest bird responses to urbanization intensity in Pennsylvania, USA.

Traits are those in the best models among models that included only traits with positive adjusted R2 values in univariate regressions or models that included only traits with values for all species. The larger effect is shown for traits included in both sets of best models.

https://doi.org/10.1371/journal.pone.0220120.g002

thumbnail
Table 2. The best models of the effects of species traits on the magnitude of forest bird responses to urbanization intensity in Pennsylvania, USA.

Evaluated models included traits with positive adjusted R2 values in univariate regressions.

https://doi.org/10.1371/journal.pone.0220120.t002

thumbnail
Table 3. The best models of the effects of species traits on the magnitude of forest bird responses to urbanization intensity in Pennsylvania, USA.

Evaluated models included traits with values for all species.

https://doi.org/10.1371/journal.pone.0220120.t003

Clutch size, frugivory, granivory, omnivory, cavity nesting, wingspan, and flock size were included in the best models of the effects of species traits on the spatial scale of species responses to urbanization intensity (Tables 4 and 5). All traits had a positive effect on the spatial scale at which species occurrences were most affected by urbanization intensity, with that of granivory being the largest, followed by frugivory, omnivory, cavity nesting, clutch size, flock size, and wingspan (Fig 3). There was no change to measured effect sizes if an apparent outlier (the Cedar Waxwing (Bombycilla cedrorum)) was excluded from the analysis.

thumbnail
Fig 3. The effects of species traits on the spatial scale of forest bird responses to urbanization intensity in Pennsylvania, USA.

Traits are those in the best models among models that included only traits with positive adjusted R2 values in univariate regressions or models that included only traits with values for all species. The larger effect is shown for traits included in both sets of best models.

https://doi.org/10.1371/journal.pone.0220120.g003

thumbnail
Table 4. The best models of the effects of species traits on the scale of effect of forest bird responses to urbanization intensity in Pennsylvania, USA.

Evaluated models included traits with positive adjusted R2 values in univariate regressions.

https://doi.org/10.1371/journal.pone.0220120.t004

thumbnail
Table 5. The best models of the effects of species traits on the scale of effect of forest bird responses to urbanization intensity in Pennsylvania, USA.

Evaluated models included traits with values for all species.

https://doi.org/10.1371/journal.pone.0220120.t005

Discussion

We show that resident forest bird species that are granivorous or frugivorous, cavity-nesting, and have larger clutch sizes and more fledglings per clutch are more likely to occur with increasing urbanization intensity in landscapes. In addition, the effect of urbanization intensity on forest birds manifested most strongly at larger spatial scales for granivorous, frugivorous, or omnivorous species that are cavity-nesting, have larger clutch sizes and longer wingspans, and flock in larger numbers. In the following, we discuss the possible mechanisms underlying these results and their implications for forest bird conservation in urbanizing landscapes.

In general, dietary traits had strong associations with both urbanization intensity effect size and scale of effect. Granivory had the largest positive effect on both the magnitude and spatial scale of forest bird responses to urbanization intensity. Urbanized landscapes–in our study, those with higher population and housing densities and covers of developed open space and low-, medium-, and high-intensity development–are typified by an abundance of seed-producing ornamental plants [76] as well as feeding tables and bird feeders [77]. Our results indicate that seed-eating forest bird species, such as the Black-capped Chickadee (Poecile atricappilus), are able to capitalize on these supplemental resources. In particular, several of the species that exhibited positive urbanization intensity effect sizes are known to make use of bird feeders and/or feeding tables, such as the White-breasted Nuthatch (Sitta carolinensis), at least intermittently. The frugivorous species we examined, such as the Purple Finch (Haemorhous purpureus) also appear to benefit from the proliferation of fruit-bearing shrubs and trees in urbanizing landscapes [78, 79]. The positive effect of granivory on the magnitude of forest bird responses to urbanization intensity has a corollary in studies of birds in general, in that particularly successful urban exploiters such as the Rock Pigeon (Columba livia) and the House Sparrow (Passer domesticus) are typically granivorous species [80]. The fact that granivorous forest species, as well as those with other dietary habitats, were most affected by urbanization intensity at larger spatial scales indicates that supplementary food resources in urban environments are not evenly distributed [80], resulting in species having to travel farther to obtain them. In addition, the positive association between omnivory and scale of effect indicates that the clumped distribution of resources in urban environments may not be restricted to supplementary foods but may be a general phenomenon characteristic of all food types, including prey.

We did not test for associations of insectivory with the magnitude and spatial scale of species responses to urbanization intensity because the species counted in the PBBA were in their nesting season, and would thus all be engaging in some degree of insectivory to feed hatchlings. Even so, species that are insectivorous in the non-breeding season may be characterized by traits in the breeding season that distinguish them from species with other dietary habits, such as nest shape and vulnerability to air pollution. The effects of these additional characteristics on urbanization intensity effect size and scale of effect remain to be investigated.

The reproductive traits of clutch size and fledglings per clutch were also relatively important determinants of the magnitude and spatial scale of forest bird responses to urbanization intensity in landscapes. Forest bird species that have higher reproductive output, such as the Northern Flicker (Colaptes auratus), occurred more frequently with increasing urbanization intensity. The additional eggs and fledglings of such species may buffer the effects of losses to predation and parasitism–the latter by the Brown-headed Cowbird (Molothrus ater), which is positively associated with suburbanization [81, 82] and the sole avian nest parasite in our study area–suffered by less prolific species. Species with higher reproductive output have also been found to be less vulnerable to the negative effects of roads [83]. In addition, forest bird species with larger clutch sizes were most strongly affected by urbanization intensity at larger spatial scales. We speculate that species that produce larger clutches must move over greater distances in urbanizing landscapes to obtain sufficient resources for their young. This hypothesis remains to be tested.

The positive effect of being a resident rather than a migratory species on the magnitude of forest bird responses to urbanization intensity was similar in size to that of fledglings per clutch. Being a resident species is also known to be associated with urbanization success in birds in general [20, 30, 42], and is represented in our sample by species such as the American Crow (Corvus brachyrhynchos) and the White-breasted Nuthatch (S. carolinensis). Possible explanations for the positive effect of being a resident species include resident species occupying high-quality nesting habitat in urbanizing landscapes before migratory species arrive [20, 38]; resident species having a more comprehensive knowledge of the distribution of urban resources, and thereby garnering a larger share of them, than migratory species [20]; resident species preferentially benefiting from the increasing occurrence of winter bird-feeding in urbanizing landscapes [84]; and, the lack of migratory mortality among resident species [41], although the latter may be offset by higher winter mortality in our study area.

Forest bird species in Pennsylvania that nest in cavities, such as the Tufted Titmouse (Baeolophus bicolor), were more positively associated with increasing urbanization intensity and exhibited a larger scale of effect of urbanization intensity than non-cavity-nesters. Recent research in Europe has also found that cavity-nesters, including threatened species, do relatively well in urban settings [85]. The positive effect of cavity-nesting on urbanization intensity effect size may not be due to an increase in the occurrence of natural cavities in trees, which typically declines in urbanizing landscapes [86]. Instead, cavity-nesters may be less vulnerable to urban-associated disturbances such as increased predation, although support for this hypothesis is equivocal [38, 85, 87, 88]. Alternatively, the positive effect of cavity-nesting on urbanization intensity effect size may be explained by the fairly large proportion of cavity-nesting species in our analysis that nest in nest boxes and other human-made structures, such as the Black-Capped Chickadee (P. atricapillus). While we did not assess whether there was a difference in urbanization intensity effect size between primary and secondary cavity nesters, it is possible that this distinction may be impactful in that the occurrence of secondary nesters would be more positively affected by urbanization intensity than that of primary nesters, although further research is needed [89]. As with the distribution of supplementary food resources, the availability of cavity-bearing trees in urbanizing landscapes appears to be unevenly distributed, resulting in cavity-nesting forest bird species being most strongly affected by urbanization intensity at larger spatial scales. The uneven distribution of cavity-bearing trees may be particularly pronounced in urbanizing landscapes if such trees are restricted to spatially disjunct nature reserves [90].

Finally, flock size and wingspan had relatively small positive effects on the spatial scale at which urbanization intensity most strongly affected forest bird species. Species with larger flock sizes, such as the Cedar Waxwing (B. cedrorum) need larger landscapes to adequately feed all members of the flock [91] and thus would exhibit larger scales of effect of urbanization intensity. Similarly, bird species with longer wingspans, such as the Common Raven (Corvus corax) disperse over longer distances [92] over which they would be most exposed to the effects of urbanization intensity. Alternatively, the positive association between wingspan and urbanization intensity scale of effect may be driven by the spatial clustering of food resources in urban environments, which longer-winged species would be better able to take advantage of.

Our results point to several management strategies that may improve the conservation of forest bird species in urbanizing landscapes should the underlying mechanisms that we propose be confirmed by future research. The positive effects of dietary habit, particularly granivory, being a resident species, and cavity nesting on the magnitude of forest bird responses to urbanization intensity imply that public and private land managers in urbanizing landscapes maintain or increase the year-round availability of food and nesting resources for forest birds. Municipal urban forest management plans that emphasize the safe preservation and creation of cavity-bearing trees and public outreach and educational efforts focused on winter bird-feeding, nest box availability, and backyard wildlife habitat certification are ways in which this goal may be accomplished. If feasible, these initiatives should be implemented in a manner that increases the spatial dispersion of resources across landscapes, thereby shortening the distances over which granivorous, frugivorous, omnivorous, and cavity-nesting forest birds must move and consequently reducing their probability of movement mortality.

Although important to maintain or increase the populations of some forest bird species, increasing the availability and accessibility of food and nesting resources only targets species that are already doing relatively well in urbanizing landscapes and in our study, this was a minority: only 14 of 44 species, or 32%, had positive urbanization intensity effect sizes. Urban land managers should consider prioritizing strategies that benefit forest bird species that are more negatively affected by urbanization intensity, i.e., migratory species with small clutches and few fledglings per clutch. Specifically, we suggest that public land managers and private landowners limit the disproportionally negative impact of nest predation on forest bird species with low reproductive productivity by restricting the activity of domestic cats [93] and corvids [94] in urbanizing landscapes and employ strategies, such as reduced night lighting, that minimize bird collisions with buildings, which may disproportionally affect migratory species [95]. The cultivation of mature vegetation composed of native tree and shrub species may also favor migratory species if used as stopover habitat. Finally, we note that none of the management strategies that we propose here have been empirically tested. This remains an important frontier in urban forest bird species conservation.

In conclusion, our results indicate that the resident forest bird species that are most positively associated with urbanization intensity are those that are granivorous or frugivorous, have larger clutches and more fledglings per clutch, and are cavity-nesting. In addition, urbanization intensity appears to affect granivorous, frugivorous, or omnivorous species, cavity-nesters, species with larger clutch sizes or longer wings, and species that flock in greater numbers most strongly at larger spatial scales. Based on their trait characteristics, the forest bird species in our study expected to have the strongest positive association with urbanization intensity is the Northern Flicker (C. auratus), while the species expected to have the most negative association is the Ruby-throated Hummingbird (A. colubris). And, based on their trait characteristics, the forest bird species in our study expected to be influenced by urbanization intensity at the largest scale is the Common Raven (C. corax), while the species expected to be influenced by urbanization at the smallest scale is the Golden-winged Warbler (Vermivora chrysoptera). To our knowledge, our study represents the first direct tests of the effects of species traits on the magnitude and spatial scale of the effect of urbanization on forest birds, as well as the first evidence that migratory status, clutch size, wingspan, and fledglings per clutch are important determinants of the responses of forest birds to urbanization. Future research should focus on testing the mechanisms we propose to explain the effects of species traits and the effectiveness of the management strategies they imply.

Supporting information

S1 File. Methods used to classify forest-dependent bird species.

https://doi.org/10.1371/journal.pone.0220120.s002

(DOCX)

S2 File. Forest bird species trait values, codes, and sources.

https://doi.org/10.1371/journal.pone.0220120.s003

(XLSX)

S3 File. Results of univariate phylogenetic regressions of urbanization intensity effect sizes and scales of effect on forest bird species traits.

https://doi.org/10.1371/journal.pone.0220120.s004

(DOCX)

S2 Table. Results from general linear models of the effects of traits on forest bird species’ responses to urbanization intensity in Pennsylvania, USA.

https://doi.org/10.1371/journal.pone.0220120.s006

(DOCX)

S3 Table. Results from general linear models of the effects of traits on scales of effect of forest bird species’ responses to urbanization intensity in Pennsylvania, USA.

https://doi.org/10.1371/journal.pone.0220120.s007

(DOCX)

Acknowledgments

We thank the Second Pennsylvania Breeding Bird Atlas staff: Robert Mulvihill, Daniel Brauning, Michael Lanzone, and Tricia Miller, the 22 field technicians, and the 2,000 atlas volunteers, whose work made this research possible.

References

  1. 1. United Nations. World Urbanization Prospects: The 2014 Revision-Highlights. United Nations; 2014.
  2. 2. Seto KC, Fragkias M, Güneralp B, Reilly MK. A meta-analysis of global urban land expansion. PloS One. 2011 Aug 18;6(8): e23777. pmid:21876770
  3. 3. Klingbeil BT, Willig MR. Matrix composition and landscape heterogeneity structure multiple dimensions of biodiversity in temperate forest birds. Biodiversity and Conservation. 2016 Dec 1;25(13): 2687–708.
  4. 4. Posa MR, Sodhi NS. Effects of anthropogenic land use on forest birds and butterflies in Subic Bay, Philippines. Biological Conservation. 2006 Apr 1;129(2): 256–70.
  5. 5. Suarez-Rubio M, Leimgruber P, Renner SC. Influence of exurban development on bird species richness and diversity. Journal of Ornithology. 2011 Apr 1;152(2): 461–71.
  6. 6. Kluza DA, Griffin CR, DeGraaf RM. Housing developments in rural New England: effects on forest birds. Animal Conservation Forum. 2000 Feb;3(1): 15–26. Cambridge University Press.
  7. 7. Dunford W, Freemark K. Matrix matters: effects of surrounding land uses on forest birds near Ottawa, Canada. Landscape Ecology. 2005 Jul 1;20(5): 497–511.
  8. 8. Pidgeon AM, Radeloff VC, Flather CH, Lepczyk CA, Clayton MK, Hawbaker TJ, Hammer RB. Associations of forest bird species richness with housing and landscape patterns across the USA. Ecological Applications. 2007 Oct;17(7): 1989–2010. pmid:17974337
  9. 9. Vignoli L, Scirè S, Bologna MA. Rural–urban gradient and land use in a millenary metropolis: how urbanization affects avian functional groups and the role of old villas in bird assemblage patterning. Web Ecology. 2013 Jul 8;13(1): 49–67.
  10. 10. Minor E, Urban D. Forest bird communities across a gradient of urban development. Urban Ecosystems. 2010 Mar 1;13(1): 51–71.
  11. 11. Devictor V, Julliard R, Couvet D, Lee A, Jiguet F. Functional homogenization effect of urbanization on bird communities. Conservation Biology. 2007 Jun;21(3): 741–51. pmid:17531052
  12. 12. Kennedy CM, Marra PP, Fagan WF, Neel MC. Landscape matrix and species traits mediate responses of Neotropical resident birds to forest fragmentation in Jamaica. Ecological Monographs. 2010 Nov;80(4): 651–69.
  13. 13. Chace JF, Walsh JJ. Urban effects on native avifauna: a review. Landscape and Urban Planning. 2006 Jan 1;74(1): 46–69.
  14. 14. Meffert PJ, Dziock F. The influence of urbanisation on diversity and trait composition of birds. Landscape Ecology. 2013 May 1;28(5): 943–57.
  15. 15. Lepczyk CA, Mertig AG, Liu J. Landowners and cat predation across rural-to-urban landscapes. Biological Conservation. 2004 Feb 1;115(2): 191–201.
  16. 16. Erickson WP, Johnson GD, David P Jr. A summary and comparison of bird mortality from anthropogenic causes with an emphasis on collisions. US Dept. of Agriculture, Forest Service, Pacific Southwest Research Station. 2005;191: 1029–1042.
  17. 17. Fuller RA, Warren PH, Armsworth PR, Barbosa O, Gaston KJ. Garden bird feeding predicts the structure of urban avian assemblages. Diversity and Distributions. 2008 Jan;14(1): 131–7.
  18. 18. Planillo A, Kramer-Schadt S, Malo JE. Transport infrastructure shapes foraging habitat in a raptor community. PLoS One. 2015 Mar 18;10(3): e0118604. pmid:25786218
  19. 19. Rodewald AD, Kearns LJ, Shustack DP. Anthropogenic resource subsidies decouple predator–prey relationships. Ecological Applications. 2011 Apr;21(3): 936–43. pmid:21639056
  20. 20. Croci S, Butet A, Clergeau P. Does urbanization filter birds on the basis of their biological traits? The Condor. 2008 May;110(2): 223–40.
  21. 21. Aronson MF, La Sorte FA, Nilon CH, Katti M, Goddard MA, Lepczyk CA, Warren PS, Williams NS, Cilliers S, Clarkson B, Dobbs C. A global analysis of the impacts of urbanization on bird and plant diversity reveals key anthropogenic drivers. Proc. R. Soc. B. 2014 Apr 7;281(1780): 20133330. pmid:24523278
  22. 22. Bayne EM, Habib L, Boutin S. Impacts of chronic anthropogenic noise from energy‐sector activity on abundance of songbirds in the boreal forest. Conservation Biology. 2008 Oct;22(5): 1186–93. pmid:18616740
  23. 23. Ditchkoff SS, Saalfeld ST, Gibson CJ. Animal behavior in urban ecosystems: modifications due to human-induced stress. Urban Ecosystems. 2006 Jan 1;9(1): 5–12.
  24. 24. Goodwin SE, Shriver WG. Effects of traffic noise on occupancy patterns of forest birds. Conservation Biology. 2011 Apr;25(2): 406–11. pmid:21410531
  25. 25. Jackson AK, Evers DC, Etterson MA, Condon AM, Folsom SB, Detweiler J, Schmerfeld J, Cristol DA. Mercury exposure affects the reproductive success of a free-living terrestrial songbird, the Carolina Wren (Thryothorus ludovicianus). The Auk. 2011 Oct;128(4): 759–69.
  26. 26. Lewis CA, Cristol DA, Swaddle JP, Varian-Ramos CW, Zwollo P. Decreased immune response in zebra finches exposed to sublethal doses of mercury. Archives of Environmental Contamination and Toxicology. 2013 Feb 1;64(2): 327–36. pmid:23229191
  27. 27. Loss SR, Ruiz MO, Brawn JD. Relationships between avian diversity, neighborhood age, income, and environmental characteristics of an urban landscape. Biological Conservation. 2009 Nov 1;142(11): 2578–85.
  28. 28. Mason J, Moorman C, Hess G, Sinclair K. Designing suburban greenways to provide habitat for forest-breeding birds. Landscape and Urban Planning. 2007 Mar 28;80(1–2): 153–64.
  29. 29. Møller AP. Behavioural and ecological predictors of urbanization. Avian Urban Ecology. 2014: 54–68.
  30. 30. Jokimäki J, Suhonen J. Distribution and habitat selection of wintering birds in urban environments. Landscape and Urban Planning. 1998 Jan 1;39(4): 253–63.
  31. 31. Kosinski Z. Effects of urbanization on nest site selection and nesting success of the Greenfinch Carduelis chloris in Krotoszyn, Poland. Ornis Fennica. 2001 Jan 1;78(4): 175–83.
  32. 32. Wang Y, Chen S, Jiang P, Ding P. Black-billed Magpies (Pica pica) adjust nest characteristics to adapt to urbanization in Hangzhou, China. Canadian Journal of Zoology. 2008 Jun 13;86(7): 676–84.
  33. 33. Jokimäki J, Huhta E. Artificial nest predation and abundance of birds along an urban gradient. The Condor. 2000 Nov;102(4): 838–847.
  34. 34. Wilcove DS. Nest predation in forest tracts and the decline of migratory songbirds. Ecology. 1985 Aug;66(4): 1211–4.
  35. 35. Liker A, Bókony V. Larger groups are more successful in innovative problem solving in house sparrows. Proceedings of the National Academy of Sciences. 2009 May 12;106(19): 7893–8.
  36. 36. Jose J, Møller AP, Soler M. Nest building, sexual selection and parental investment. Evolutionary Ecology. 1998 May 1;12(4): 427–41.
  37. 37. Hemborg C, Sanz J, Lundberg A. Effects of latitude on the trade-off between reproduction and moult: a long-term study with pied flycatcher. Oecologia. 2001 Oct 1;129(2): 206–12. pmid:28547598
  38. 38. Kark S, Iwaniuk A, Schalimtzek A, Banker E. Living in the city: can anyone become an ‘urban exploiter'?. Journal of Biogeography. 2007 Apr;34(4): 638–51.
  39. 39. Brown LM, Graham CH. Demography, traits and vulnerability to urbanization: can we make generalizations?. Journal of Applied Ecology. 2015 Dec 1;52(6): 1455–64.
  40. 40. Conole LE, Kirkpatrick JB. Functional and spatial differentiation of urban bird assemblages at the landscape scale. Landscape and Urban Planning. 2011 Mar 30;100(1–2): 11–23.
  41. 41. Evans KL, Chamberlain DE, Hatchwell BJ, Gregory RD, Gaston KJ. What makes an urban bird?. Global Change Biology. 2011 Jan 1;17(1): 32–44.
  42. 42. Jokimäki J, Suhonen J, Jokimäki-Kaisanlahti ML, Carbó-Ramírez P. Effects of urbanization on breeding birds in European towns: Impacts of species traits. Urban Ecosystems. 2016 Dec 1;19(4): 1565–77.
  43. 43. Møller AP. Successful city dwellers: a comparative study of the ecological characteristics of urban birds in the Western Palearctic. Oecologia. 2009;159(4): 849–858. pmid:19139922
  44. 44. Wheatley M, Johnson C. Factors limiting our understanding of ecological scale. Ecological Complexity. 2009 Jun 1;6(2):150–9.
  45. 45. Carpenter A, Provorse C. The World Almanac of the USA. World Almanac Books; 1996 Apr 1.
  46. 46. Kottek M, Grieser J, Beck C, Rudolf B, Rubel F. World map of the Köppen-Geiger climate classification updated. Meteorologische Zeitschrift. 2006 Jun 1;15(3): 259–63.
  47. 47. Albright TA. Forests of Pennsylvania, 2016. Resource Update FS-132. Newtown Square, PA: US Department of Agriculture, Forest Service, Northern Research Station. 4 p. 2017;132:1–4.
  48. 48. Rhoads AF, Block TA. Trees of Pennsylvania: a complete reference guide. University of Pennsylvania Press; 2005.
  49. 49. United States Census Bureau. Annual Estimates of the Resident Population for the United States, Regions, States, and Puerto Rico: April 1, 2010 to July 1, 2016; 2016. Database. Available from https://www.census.gov/data/tables/2016/demo/popest/nation-total.html.
  50. 50. Behney M, Copella S, Shultz J, Bowalick D, Koontz A, Meyers L, Kotovsky M. Pennsylvania Population Projections 2010–2040. The Center for Rural Pennsylvania. The Institute of State and Regional Affairs, Penn State Harrisburg; 2014.
  51. 51. Fry J, Xian GZ, Jin S, Dewitz J, Homer CG, Yang L, Barnes CA, Herold ND, Wickham JD. Completion of the 2006 national land cover database for the conterminous United States. Photogrammetric engineering and remote sensing. 2011;77(9): 858–64.
  52. 52. Wilson AM, Brauning DW, Mulvihill RS. Second atlas of breeding birds in Pennsylvania. The Pennsylvania State University Press; 2012.
  53. 53. Jackson HB, Fahrig L. Are ecologists conducting research at the optimal scale?. Global Ecology and Biogeography. 2015 Jan;24(1): 52–63.
  54. 54. Wilson AM, Brauning DW, Carey C, Mulvihill RS. Spatial models to account for variation in observer effort in bird atlases. Ecology and Evolution. 2017 Aug;7(16): 6582–94. pmid:28861259
  55. 55. ESRI. ArcGIS Desktop: Release 10.4. Redlands, CA: Environmental Systems Research Institute; 2016.
  56. 56. Shoffner A, Wilson AM, Tang W, Gagné SA. The relative effects of forest amount, forest configuration, and urban matrix quality on forest breeding birds. Scientific Reports. 2018 Nov 20;8(1): 17140. pmid:30459396
  57. 57. Legendre P, Legendre L. Numerical Ecology. (24): Developments in Environmental Modeling. 1998.
  58. 58. Rodewald P. The Birds of North America: https://birdsna.org. Cornell Laboratory of Ornithology, Ithaca. New York. 2015.
  59. 59. McGarigal K, Cushman SA, Ene E. FRAGSTATS v4: spatial pattern analysis program for categorical and continuous maps. Computer software program produced by the authors at the University of Massachusetts, Amherst. http://www.umass.edu/landeco/research/fragstats/fragstats.html. 2012 Jul.
  60. 60. Pennsylvania Department of Conservation and Natural Resources. PAMAP program 3.2-ft digital elevation model of Pennsylvania. Dataset at http://www.pasda.psu.edu/uci/DataSummary.aspx?dataset=1247. 2006.
  61. 61. Kaufman K. Lives of North American Birds. Houghton Mifflin Harcourt; 2001.
  62. 62. Peterson RT, Peterson VM, editors. Peterson Field Guide to the Birds of Eastern and Central North America. Houghton Mifflin Harcourt; 2002 Apr 4.
  63. 63. Boersma P, Weenink, D. Praat: doing phonetics by computer [Computer program]. Version 6.0.37, retrieved 3 February 2018 from http://www.praat.org/. 2018.
  64. 64. Zuckerberg B, Desrochers A, Hochachka WM, Fink D, Koenig WD, Dickinson JL. Overlapping landscapes: a persistent, but misdirected concern when collecting and analyzing ecological data. The Journal of Wildlife Management. 2012 Jul;76(5): 1072–80.
  65. 65. Dormann CF, Elith J, Bacher S, Buchmann C, Carl G, Carré G, Marquéz JR, Gruber B, Lafourcade B, Leitão PJ, Münkemüller T. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography. 2013 Jan;36(1): 27–46.
  66. 66. Cade BS. Model averaging and muddled multimodel inferences. Ecology. 2015 Sep;96(9): 2370–82. pmid:26594695
  67. 67. Anderson DR, Burnham KP. Avoiding pitfalls when using information-theoretic methods. The Journal of Wildlife Management. 2002 Jul 1: 912–8.
  68. 68. Cooper H, Hedges LV, Valentine JC, editors. The handbook of research synthesis and meta-analysis. Russell Sage Foundation; 2009 Feb 5.
  69. 69. A global phylogeny of birds. In: birdtree.org [Internet]. Available at http://birdtree.org. 2018.
  70. 70. Jetz W, Thomas GH, Joy JB, Hartmann K, Mooers AO. The global diversity of birds in space and time. Nature. 2012 Nov;491(7424): 444. pmid:23123857
  71. 71. Team RC. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2016.
  72. 72. Paradis E, Claude J, Strimmer K. APE: analyses of phylogenetics and evolution in R language. Bioinformatics. 2004 Jan 22;20(2): 289–90. pmid:14734327
  73. 73. Orme D, Freckleton R, Thomas G, Petzoldt T, Fritz S, Isaac N, Pearse W. caper: Comparative Analyses of Phylogenetics and Evolution in R, version 0.5.2.: 2013.
  74. 74. Nakazawa M. Package ‘Fmsb’. Functions for Medical Statistics Book with Some Demographic Data. 2017.
  75. 75. Barton K. MuMIn: Multi-Model Inference. R package version 1.40.0.: 2017.
  76. 76. McKinney ML. Urbanization, biodiversity, and conservation. BioScience 2002; 52: 883–890.
  77. 77. Robb GN, McDonald RA, Chamberlain DE, Bearhop S. Food for thought: supplementary feeding as a driver of ecological change in avian populations. Frontiers in Ecology and the Environment. 2008 Nov;6(9):476–84.
  78. 78. Belaire JA, Whelan CJ, Minor ES. Having our yards and sharing them too: the collective effects of yards on native bird species in an urban landscape. Ecological Applications. 2014 Dec;24(8): 2132–43. pmid:29188686
  79. 79. Suhonen J, Jokimäki J. Fruit removal from rowanberry (Sorbus aucuparia) trees at urban and rural areas in Finland: a multi-scale study. Landscape and Urban Planning. 2015 May 1;137: 13–9.
  80. 80. Hunter MCR, Brown DG. Spatial contagion: gardening along the street in residential neighborhoods. Landscape Urban Plan. 2012; 105: 407–416.
  81. 81. Tewksbury JJ, Hejl SJ, Martin TE. Breeding productivity does not decline with increasing fragmentation in a western landscape. Ecology. 1998 Dec;79(8): 2890–903.
  82. 82. Thorington KK, Bowman R. Predation rate on artificial nests increases with human housing density in suburban habitats. Ecography 2003; 26: 188–196.
  83. 83. Rytwinski T, Fahrig L. Reproductive rate and body size predict road impacts on mammal abundance. Ecol Appl. 2011; 21: 589–600. pmid:21563588
  84. 84. Wilcoxen TE, Horn DJ, Hogan BM, Hubble CN, Huber SJ, Flamm J, et al. Effects of bird-feeding activities on the health of wild birds. Conserv Physiol. 2015; 3: cov058. pmid:27293740
  85. 85. Jokimäki J, Jukka S, Marja-Liisa KJ. Urban core areas are important for species conservation: A European-level analysis of breeding bird species. Landscape and urban planning. 2018 Oct 31;178: 73–81.
  86. 86. Davis A, Major RE, Taylor CE. Distribution of tree-hollows and hollow preferences by parrots in an urban landscape. Emu 2014; 114: 295–303.
  87. 87. Blewett CM, Marzluff JM. Effects of urban sprawl on snags and the abundance and productivity of cavity-nesting birds. The Condor. 2005 Aug;107(3): 678–93.
  88. 88. Reale JA, Blair RB. Nesting success and life-history attributes of bird communities along an urbanization gradient. Urban Habitats. 2005;3(1): 1–24.
  89. 89. Tomasevic JA, Marzluff JM. Cavity nesting birds along an urban-wildland gradient: is human facilitation structuring the bird community?. Urban Ecosystems. 2017 Apr 1;20(2):435–48.
  90. 90. Le Roux DS, Ikin K, Lindenmayer DB, Manning AD, Gibbons P. The future of large old trees in urban landscapes. PLoS ONE 2014; 9(6): e99403. pmid:24941258
  91. 91. Maldonado-Coelho M, Marini MA. Mixed-species bird flocks from Brazilian Atlantic forest: the effects of forest fragmentation and seasonality on their size, richness and stability. Biological Conservation. 2004 Mar 1;116(1):19–26.
  92. 92. Tittler R, Villard M-A, Fahrig L. How far do songbirds disperse? Ecography 2009; 32: 1051–1061.
  93. 93. Marra PP, Santella C. Cat Wars: The Devastating Consequences of a Cuddly Killer. Princeton: Princeton University Press; 2016.
  94. 94. Luginbuhl JM, Marzluff JM, Bradley JE, Raphael MG, Varland DE. Corvid survey techniques and the relationship between corvid relative abundance and nest predation. Journal of field Ornithology. 2001 Oct;72(4): 556–73.
  95. 95. Loss SR, Will T, Loss SS, Marra PP Bird-building collisions in the United States: estimates of annual mortality and species vulnerability. Condor 2014; 116: 8–23.