Content uploaded by Cagan H Sekercioglu
Author content
All content in this area was uploaded by Cagan H Sekercioglu on Dec 05, 2022
Content may be subject to copyright.
Community science data suggest the most common
raptors (Accipitridae) in urban centres are smaller,
habitat-generalist species
DANIEL S. COOPER,*
1,2
ALLISON J. SHULTZ,
2,3
C
ßA
GAN H. S
ßEKERCIO
GLU,
4,5
FIONA M. OSBORN
1
&
DANIEL T. BLUMSTEIN
1
1
Department of Ecology and Evolutionary Biology, University of California, Los Angeles, CA, USA
2
Ornithology Department, Natural History Museum of Los Angeles County, Los Angeles, CA, USA
3
Urban Nature Research Center, Natural History Museum of Los Angeles County, Los Angeles, CA, USA
4
School of Biological Sciences, University of Utah, Salt Lake City, UT, USA
5
Department of Molecular Biology and Genetics, Koc
ßUniversity, Istanbul, Turkey
As the world becomes more urbanized, identifying traits that allow some species to
thrive in cities will be key to predicting which species will probably remain common and
which may require conservation attention. Large, diverse, widely distributed and readily
documented raptors represent an ideal taxonomic group to understand how species per-
sist and thrive in urban areas. Global community science datasets can reveal patterns that
might be obscured in studies limited to a small number of locations, those relying on
presence/absence data or those conducted by a small number of observers. We analysed
127 species of raptors (hawks and related species; Family: Accipitridae) using recent
community-science (eBird) records from 59 cities on five continents, modelling two
indices of occurrence with five ecological and life history traits, and incorporating phylo-
genetic relatedness. Based on previous studies of avian traits in urban vs. rural popula-
tions, and well as our casual observations of birds in cities across the USA and around
the world, we hypothesized that urban raptor communities would be dominated by
smaller, ecological-generalist species regardless of the regional species pool. We defined
urban occurrence in two ways: urban abundance (the frequency of breeding season
reports within 10 km of a city centre) and species proportion (the relative abundance of
each species in the local raptor community). We did not detect a strong phylogenetic
signal for either urban occurrence index, suggesting that various unrelated raptor species
may become common in cities of the world. In the best-performing models, both urban
indices were significantly negatively associated with body mass, and significantly posi-
tively associated with habitat breadth; species proportion was also significantly associated
with nest substrate breadth. Our analysis suggests that there may be an ‘archetypal urban
raptor’and that species lacking these traits (e.g. large, specialist taxa) may be at greater
conservation risk as global urbanization increases.
Keywords: avian ecology, cities, eBird, generalist, global, hawks, ornithology.
Diurnal raptors (Family Accipitridae, including
eagles, hawks, kites and related species) exhibit a
variety of sizes and morphological traits, contain
species ranging from diminutive sparrowhawks
(Accipiter spp.) to massive Old World vultures
(Gyps spp.) and occupy a broad range of ecologi-
cal niches on every continent except Antarctica.
Many raptor species are clearly thriving in urban
landscapes, nesting in built structures and planted
introduced trees, feeding on human-subsidized
urban prey and providing predation ecosystem ser-
vices (S
ßekercio
glu 2006, McCabe et al. 2018,
Rosenfield et al. 2018, Mak et al. 2021). Other
*Corresponding author.
Email: dan@cooperecological.com
© 2022 British Ornithologists' Union
Ibis (2022) doi: 10.1111/ibi.13047
species are restricted to wildland habitats, and for
many tropical species and single-island endemics,
their biology is comparatively poorly known
(McClure et al. 2018, Buechley et al. 2019). While
studies of wildland raptor communities (e.g. Marti
et al. 1993) have long outnumbered those investi-
gating urban ones, as the footprint of global urban-
ization expands (Seto et al. 2011), many raptors
will need to adapt to some level of human distur-
bance to survive. Those species that are less toler-
ant of disturbance might be assigned a higher
priority for conservation, as their populations will
receive increasing pressure from the effects of
expanding urbanization.
Many studies have investigated ecological traits
associated with urban life in birds (see reviews by
Chace & Walsh 2006, Marzluff 2016), but the
few that have examined raptors explicitly have
focused on single species or single cities (e.g. Cade
et al. 1996; Kopij 2018; White et al. 2018, but see
Kettel et al. 2018). Boal (2018) analysed eight
traits associated with raptor occurrence in 14 US
state capitals (all with human popula-
tions >100 000), finding that both diet breadth
and preferred ‘normal’(non-urban) habitat type
were strong predictors of presence in urban areas
during both winter and summer. Yet this study
was limited to the USA, and did not take species’
abundance into account.
Prey type and availability are understood to be
key to the development and maintenance of raptor
communities, including those along an urban gra-
dient (e.g. Rullman & Marzluff 2014). Estes and
Mannan (2003) found stark differences in prey
type in a study of urban vs. rural-nesting Cooper’s
Hawks Accipiter cooperii in Tucson, AZ, USA, with
urban birds more restricted in their feeding
(mainly doves, vs. a wide range of prey types). Yet
these patterns may not be universal, as Suri et al.
(2017) found no change in diet breadth or prey
composition with increasing urban cover in a study
of Black Sparrowhawks Accipiter melanoleucus
around Cape Town, South Africa. Still, having a
broad diet may enable bird species to thrive in
urban areas, as suggested in a recent review of sev-
eral hundred taxa (Palacio, 2020).
Prey type may favour certain traits in raptors in
urban areas, including migratory status and body
mass, but the direction may vary based on the
type of urbanized habitat, the regional (raptor)
species pool and the particular prey base. Powers
(1996) suggested that sedentariness in an urban
Sharp-shinned Hawk Accipiter striatus was related
to the year-round availability of this food source in
Idaho, USA. Rullman and Marzluff (2014) linked
raptor abundance in urbanized habitats to higher
densities of urban prey –particularly rodents and
birds associated with humans –than those found
in wildland areas. Thus, in cases where the most
common urban prey items tend to be small (com-
pared with the available range of prey items con-
sumed by raptors outside urban areas), raptor
species associated with those cities might also be
small. These patterns may vary based on local
practices, such as the provisioning of ‘predictable
anthropogenic food subsidies’(Shochat 2004, Oro
et al. 2013), which may in turn affect predator
body size, but in a direction depending on the par-
ticular food resource. For example, larger sizes of
urban carnivorous mammals have been docu-
mented in parts of Israel that receive a ‘garbage
subsidy’unavailable to non-urban populations
(Yom-Tov 2003), and large garbage dumps within
urban areas may have enabled massive Old World
Vultures to persist in cities of Africa and India.
Yet this may come with a serious cost associated
with garbage, namely poisoning (see Cuthbert
et al. 2011), which may then drive down the size
of urban raptors in those cities. Rodenticide use
may also influence raptor body size, in that species
that primarily consume rodents may decline due
to poisoning (Nakayama et al. 2019), whereas
those that can shift to a ‘safer’food item (such as
birds) might be smaller raptor species buffered
from its effects. Other costs associated with urban
life could more directly reduce body size, such as
increased parasite load in urban-dwelling raptors
affecting nestling growth (Boal et al. 1998).
Morphological traits may also play a role in
allowing birds to thrive in urban areas. Change in
body size has been found to be associated with
urban occurrence in mammals (e.g. Santini et al.
2019) but, for birds, this link is less clear, as is its
ecological and evolutionary advantage (see Croci
et al. 2008, Sol et al. 2014). Evans et al. (2009)
found no significant size difference between urban
and rural European Blackbirds Turdus merula, yet
Meill
ere et al. (2015) reported smaller body size
and reduced juvenile fat scores in urban-dwelling
vs. rural House Sparrows Passer domesticus, and
Caizergues et al. (2018) reported that urban Great
Tits Parus major had shorter tarsus, lower body
mass, and smaller wing and tail lengths relative to
body mass. Comparing multispecies assemblages of
© 2022 British Ornithologists' Union
2D. S. Cooper et al.
urban-associated birds with those in a larger ‘re-
gional species pool’, Hensley et al. (2019) found
no association with body mass in the urban spe-
cies. Yet for raptors, White et al. (2018) found the
largest species in Reno, Nevada (USA), Golden
Eagle Aquila chrysaetos, to be the least tolerant of
urban land use there, and found the two smallest
hawks (both Accipiters) among the most urban-
tolerant at various landscape scales of eight taxa
examined. In a recent analysis of a raptor commu-
nity around Los Angeles, California (USA),
Cooper et al. (2020a) showed that as the region
urbanized over five decades, nests of one smaller
raptor, the Cooper’s Hawk, had greatly increased
in number, whereas the largest diurnal raptor
(Golden Eagle) had vanished. However, small size
does not guarantee urban adaptation; Cooper et al.
(2020a) also found that the smallest raptors in
their study area (White-tailed Kite Elanus leucurus
and American Kestrel Falco sparverius) had also
become extirpated, or nearly so.
Based on these previous studies, as well as our
own field observations, we speculated that the
most common nesting raptors in cities around the
world may have a particular array of shared traits,
including small size, broad diet (e.g. birds and
mammals, including non-native, urban-associated
prey items), sedentary (vs. migratory) populations,
and a tendency to utilize a variety of habitats.
Using sightings from birders and casual observers
around the world entered into the community-
science platform eBird (www.ebird.org), we iden-
tify breeding-season records of raptors in and
around cities of various sizes and settings. We cal-
culated urban occurrence in two ways for each
focal species, and modelled these values using five
morphological and ecological traits, along with
phylogenetic data, to explore the characteristics of
an ‘archetypal urban raptor’–one that benefits
from a predictable set of life history characteristics
to thrive in urban environments. Not only would
this identify a suite of traits that might confer suc-
cess with future urbanization, it could, conversely,
highlight species that may be at conservation risk
(i.e. those that lack these urban-associated traits).
METHODS
Data preparation
We used GBIF (GBIF 2020a) to download sight-
ings of hawks (Family Accipitridae, here referred
to as ‘raptors’) from the eBird Observational Data-
set (an edited, simplified version of the complete
eBird database; see Auer et al. 2020) hosted by
GBIF from nine countries encompassing a range of
global biomes (USA, Mexico, Colombia, Brazil,
Spain, UK, South Africa, India and Australia)
within a recent 5-year time period (2014–2018;
GBIF 2020b, 2020c, 2020d, 2020e, 2020f, 2020g,
2020h, 2020i, 2020j, 2020k, 2020l; see Support-
ing Information Table S1 for the list of cities and
Table S2 for the raptor species evaluated). We
selected countries with a high number of sightings
during the study period (2014–2018) submitted to
eBird, and attempted to include a broad range of
geography and ecological variation, while recogniz-
ing that certain biomes would necessarily be
excluded due to lower rates of participation in
eBird (e.g. equatorial Africa, southeast Asia). To
maximize the diversity of species analysed, we did
not select adjacent countries where the overlap in
breeding species was likely to be very high (e.g.
Canada and the USA). We selected a subset of
records to include only the months when our tar-
get species would probably be breeding, which
varied by country. We then selected the largest
urban areas within each country based on the total
population of the largest cities (generally popula-
tion >1 million). This was done to maximize the
number of raptor records in a framework where
we had no control over effort (mean =6045 rap-
tor records/city, range 28–38 881; mean =13.4
species/city, range 3–24). We assumed that these
cities, while not representative of every global
biome, adequately represent a broad subset of
ecological and biogeographical attributes found
globally.
For each city, we visually estimated the ‘urban
centre’(i.e. the centre of the urban extent of each
city) using the most recent aerial imagery on the
Apple Maps application on the iPhone (ver. 14.2).
We then used the Geosphere package (Hijmans
et al. 2015) in R (R Core Team 2020; version
4.0.0) to find records at two radial distances: a 10-
km radial distance, which we considered the ‘ur-
ban core’, and a 10- to 50-km radial distance,
which we considered the ‘peri-urban band’. See
Supporting Information Figure S1 for aerial ima-
gery of example cities, and Table S1 for a list of
the coordinates used.
For landcover data, we used satellite imagery
data (300 m resolution) from the Copernicus Cli-
mate Change Service Climate Data Store (2021)
© 2022 British Ornithologists' Union
Global urban raptors 3
to examine land cover types surrounding our study
areas. Using ESRI ArcMap 10.8 equipped with a
spatial analyst licence, we converted the Coperni-
cus NetCDF file into a raster format. From there,
we created 10- and 50-km buffers around our
coordinate points and used the tabulate area tool
to calculate the total area of each land cover type
within the specified buffer zone. We simplified the
classifications into five major categories –cropland,
tree cover, shrub/grasslands, urban areas and water
features –and converted the values into percent-
ages, which we used to calculate amounts of each
land cover category within the 10-km urban core,
and the 10- to 50-km peri-urban band. For each
species found within 50 km of a particular city,
we calculated a rate of detection in both the urban
core and the peri-urban band by dividing the num-
ber of records by the amount of terrestrial habitat
within each (i.e. the number of records divided by
the non-water area in each).
To avoid overlapping data, we dropped overlap-
ping large cities, opting to retain whichever was
the larger one (e.g. San Jose, California, was
retained over San Francisco, California). We recog-
nize that the urban core may also include a mix of
agricultural lands and fragments of scrub and for-
est, depending on the city, but confirmed that the
urban core consistently had a higher amount of
urban cover than the peri-urban band (71% urban
cover; range 20–100% in urban core, vs. 12%
urban cover; range 0–48% in peri-urban band;
n=59). Our final dataset of 127 focal species
found within 50 km of the urban centre of at least
one focal city represents about 45% of the world’s
285 widely recognized raptor (Accipitridae) spe-
cies and 67% of Accipitridae genera (45 of 67; Del
Hoyo et al. 2013, BirdLife International 2019).
This represented the ‘species pool’, from which
we derived urban abundance and species propor-
tion values (measured within the 10-km radial
band) used in the analysis (several of which had
values at or near zero; see Table S2 for a complete
list of species and values).
Measuring urban occurrence
We used two indices of occurrence to assess the
presence of raptors in urban areas: urban abun-
dance (number of records within 10 km of the
urban centre) and species proportion (percentage
of records of each species within 10 km of the
urban centre; Table 1). Although several species
consistently rank highly in each metric (e.g. the
Shikra Accipiter badius), we believe that both met-
rics are useful to express a species’urban associa-
tion, as the number of records varies widely by
species and by city for a given species. Because we
treated georeferenced records of individual birds,
rather than nests, we did not attempt to estimate
land cover surrounding the locations, with the
assumption that many of the raptors used would
have been observed in flight, and not necessarily
associated with the habitat where the observer was
standing (locations of nests would have been
preferable, but a comparable global dataset of
nests does not exist).
Traits
We evaluated traits most likely to influence urban
occurrence based on previous research on urban
birds, and urban raptors in particular (e.g. Samia
et al. 2015, Boal 2018, Cooper et al. 2020b,
Table 2). Trait values were taken from ‘BirdBase’,
a dataset maintained and continuously updated by
S
ßekercio
glu (S
ßekercio
glu et al. 2004, 2019), which
we cross-referenced using additional sources (in-
cluding Ferguson-Lees & Christie 2001 and Glob-
alraptors.org 2020) to insert estimated values
where needed due to missing data (see Supporting
Information Table S3). Due to high collinearity
between artificial nest substrate and nest substrate
Table 1. Urban index variables used to calculate raptor occurrence in urban areas.
Index Measures Calculation
Urban abundance Numerical abundance of
each species in urban core.
Density (n/terrestrial land area) of eBird
reports within 10 km of urban centre.
Species proportion Relative abundance of
each species in urban core.
Percentage of eBird reports of given
species relative to the number of reports
of other raptor species, within 10 km of urban centre.
© 2022 British Ornithologists' Union
4D. S. Cooper et al.
breadth (r>0.6) using Spearman’s rank correla-
tion, we eliminated the former, both because its
correlation with urban occurrence is already well
established (e.g. Cooper et al. 2020b) and because
substrate breadth seemed more informative for a
wider range of species (few species we analysed
are known to nest on artificial structures). Ulti-
mately, we selected five trait variables for the
models: mass, diet breadth, habitat breadth, migra-
tory status and nest substrate breadth (see Sup-
porting Information Figure S2 for a correlation
matrix of urban indices and traits).
Statistical analysis
To account for phylogenetic relatedness among
species in our analyses, we used the latest phy-
logeny of Accipitridae from the Open Tree of Life
(2019, ver. 3.1), which represents a synthetic tree
derived from multiple sources of phylogenetic
information. We first tested for phylogenetic signal
in each urban index individually by fitting a series
of generalized least squares (GLS) models (with-
out trait variables) that employed three different
modes of evolution: Brownian motion (BM),
Pagel’s lambda, Ornstein–Uhlenbeck (OU) and a
non-phylogenetic model. This phylogenetically
informed GLS (PGLS) framework is useful for
data where the dependent variable lacks a normal
distribution (M€
unkem€
uller et al. 2012).
For the Brownian motion, or random-walk,
model we used a Blomberg’sKtest (Blomberg
et al. 2003), which compares the variance of phy-
logenetically independent contrasts with what we
would expect under a BM model. Here, K=1
means that relatives resemble one another as much
as we should expect under BM (i.e. non-
relatedness), K<1 means that there is less phylo-
genetic signal than expected under BM, and K>1
means that there is more. For Pagel’s lambda
(Pagel 1999), if our estimated lambda =0, then
the traits would be inferred to have no phyloge-
netic signal. Lambda =1 corresponds to a BM
model, and 0 <lambda <1 is intermediate. The
OU mode of evolution incorporates ‘stabilizing
selection’wherein the trait is drawn toward a fit-
ness optimum, or long-term mean, rather than
being completely random and directionless (Mar-
tins 1994). This model has two terms: alpha,
which represents the strength of the pull toward
the fitness optimum (where alpha =0 indicates no
pull, as in a BM model, the larger the alpha value,
the stronger the pull), and sigma
2
, which is the
dispersion of the data (Martins 1994). Finally, to
test for no phylogenetic signal, we used a ‘no-
signal’GLS model where lambda was set to 0. We
used the phytools package in R (ver. 0.7-70; Revell
2012) for the BM, Pagel’s lambda and OU models,
and the nlme package in R (ver. 3.1-147; Pinheiro
et al. 2019) for one non-phylogenetic general lin-
ear model, and compared adjusted Akaike infor-
mation criterion (AICc) values of each to select
the model that best explained variation in the
data.
We repeated this process to test associations
separately between each urban index and the five
traits, using the same three phylogenetic models
and one non-phylogenetic model described above.
Table 2. The traits used in our analyses. Body mass, diet breadth, habitat breadth and nest substrate values were taken from C
ß.H.
S
ßekercio
glu (unpubl. data). Migratory status values were inferred from descriptions in Ferguson-Lees and Christie (2001). Various
sources were used to fill in missing values (e.g. Globalraptors.org 2020).
Variable Type Description
Body mass Numeric; grams Ln transformed; up to four reported values (various sources) were averaged.
Diet breadth Numeric; 1–6 Calculated from nine major food categories: invertebrate, fruit, nectar, seeds, land
vertebrates, fish, carcasses/garbage, vegetation and miscellaneous items.
Habitat breadth Numeric; 1–10 Calculated from 15 major habitat types: forest, bamboo, dry forest/woodland,
shrubland, savannah, grassland, dry/open, rocky areas, desert/dunes, agricultural/
artificial, sea coast, riparian, wetland, pelagic and ‘other’.
Nest substrate breadth Numeric; 1–6 Calculated from 12 categories: bamboo, building, stump, ground, cactus,
invertebrate nest, pole, rock, shrub, tree, water and grass.
Migratory status Factor; 1–3Defined as fully migratory: (1) vacating most of breeding range during the non-
breeding season; (2) partially migratory: engaging in short-distance movements
during non-breeding season, facultatively migratory and/or nomadic; (3) largely
sedentary/non-migratory throughout the year.
© 2022 British Ornithologists' Union
Global urban raptors 5
For each analysis, best-fit parameters of the phylo-
genetic model were estimated with maximum like-
lihood. Lastly, we selected the analysis with the
lowest AIC values as the best model for each
urban index tested, and compared correlations
using that model.
We checked residuals from the full models
using QQ tests, finding a somewhat skewed pat-
tern for both urban occurrence indices (Supporting
Information Figures S3 and S4). However, this
pattern was not changed by log-transforming the
dependent variables, and probably reflects ‘reality’,
in that many raptor species are simply rare in
urban areas. We intentionally assembled a large
sample size of cities and species to improve model
performance (see discussion in Mundry 2014).
RESULTS
Modelling the urban occurrence indices alone (ur-
ban abundance and species proportion) without
the life history or environmental variables, we
found little evidence of phylogenetic signal for
either (Table 3). While this suggests that phy-
logeny alone is unrelated to urban occurrence, we
still incorporated phylogenetic relatedness into
modelling the indices using the five trait variables.
In doing so, we found the OU, Pagel’s lambda and
non-phylogenetic models returned similarly low
AIC values for each index, again suggesting little
phylogenetic signal in the data for models incorpo-
rating phylogeny (Table 4). We found two vari-
ables significantly associated with both urban
abundance and species proportion: body mass
(negative) and habitat breadth (positive; Table 5).
Additionally, we found that nest substrate breadth
was significantly positively associated with species
proportion. This suggests that both smaller raptor
species and habitat generalist species were more
abundant in the urban core of the cities examined,
and were more relatively dominant in their local
raptor communities. Neither urban index was
found to be significantly correlated with diet
breadth or migratory status.
The most common urban species (Fig. 1a) and
dominant species (Fig. 1b) appear to fall within a
fairly narrow window of body mass (150–1000 g),
with those at the extreme ends of body mass hav-
ing lower values for the two urban occurrence
indices. Considering that most of the 127 raptor
species are rather rare in the urban core of cities
(Table S2), the most abundant/dominant urban
species globally represent a handful of smaller (but
not the smallest) species that have managed to
achieve abundance in many cities, such as Red-
tailed Hawk Buteo jamaicensis, Black Kite Milvus
migrans and Roadside Hawk Rupornis magnirostris.
Several of these same species were among those
with the highest habitat breadth values, including
Red-tailed Hawk, Black Kite and Brown Goshawk
Accipiter fasciatus (Fig. 2).
Table 3. Comparison of the phylogenetical signal of each of the two urban indices alone, using three modes of evolution (BM, OU,
Pagel’s lambda) and one non-phylogenetically informed model. Models with the lowest AICc values for each index are in bold.
KSigma squared Alpha Lambda AICc
Index: urban abundance
BM 0.0357, P=0.190 290.121
OU 9.589 2.718 259.835
Pagel’s lambda 0.0110; P=0.721 187.849
Non-phylogenetic 0.190 184.770
Index: species proportion
BM 0.0233, P=0.057 –161.251
OU 0.274 2.718 –191.749
Pagel’s lambda 0.730; P=0.008 –250.802
Non-phylogenetic 0.00587 –257.418
Table 4. Comparison of the AIC scores of four models used
to test two urban indices against five traits (see Table 2). The
lowest scores for each variable are indicated in bold text. Note
that the scores of OU, Pagel’s lambda and the non-
phylogenetic model are all very close for each index, suggest-
ing minimal influence of phylogeny in explaining variation.
Model Urban abundance Species proportion
BM 348.295 68.744
OU 216.170 207.407
Pagel’s lambda 213.115 208.281
Non-phylogenetic 214.188 209.407
© 2022 British Ornithologists' Union
6D. S. Cooper et al.
Yet, counter-examples abound, including sev-
eral small raptors found to have little to no repre-
sentation in the urban core (see Table S2; e.g.
Rufous-thighed Kite Harpagus diodon, Gabar
Goshawk Micronisus gabar, Bicoloured Hawk
Accipiter bicolor and Little Sparrowhawk Accipiter
minullus), as well as raptors with high habitat
breadth values that were rare in the urban areas
studied (e.g. Bonelli’s Eagle Aquila fasciata and
Buteogallus spp.). Several common urban raptors
were found to have fairly low habitat breadth,
including Cooper’s Hawk, Red-shouldered Hawk
Buteo lineatus and Eurasian Sparrowhawk Accipiter
nisus. While a few common urban species were
somewhat large-bodied (e.g. Buteo spp.), none was
larger than Red-tailed Hawk, a mid-sized raptor
(Fig. 1).
DISCUSSION
Raptors are well established in urban areas
throughout the world, and the literature on their
ability to adapt to our human-centred environment
continues to expand (e.g. Mak et al. 2021). The
lack of phylogenetic signal in our urban occurrence
indices suggests that a variety of unrelated raptor
taxa are able to thrive in cities, a result evident in
the broad range of hawk genera found most com-
monly in the world’s urban areas. So, while several
are in the genus Accipiter, there appears to be no
single taxonomic group of raptors found most
commonly in urban areas globally. In many cities
of India and Australia, for example, this role
appears to be filled by members of the genus
Accipiter (Shikra and Brown Goshawk, respec-
tively), whereas in many Latin American cities,
two unrelated, non-Accipiter species, Grey Hawk
Buteo plagiatus and Roadside Hawk, are the most
common and/or dominant raptor. Perhaps not
coincidentally, these latter two species happen to
resemble most species of Accipiter, being smallish
species with greyish and brownish plumage, a
banded tail and rapid wingbeats –characters, we
suggest, of an ‘archetypal urban raptor’.
The significant positive associations we found
with habitat breadth and nest substrate breadth
suggest that the most common raptors in urban
areas, both in absolute numbers (urban abun-
dance) and in relative abundance (species propor-
tion), are generalists –utilizing a variety of
vegetation and terrain types for both foraging
(habitat breadth) and breeding (nest substrate
breadth). These associations reflect previous find-
ings that generalists thrive in cities, whereas urban-
avoiders show a narrower habitat tolerance (Croci
et al. 2008, Sol et al. 2014). Multiple studies over
decades have documented the utilization of habi-
tats in and around urban areas by the same urban
raptors we found to be most common where
patches of woodland and other habitat elements
persist in urban areas (e.g. Stout et al. 2006 for
Red-tailed Hawks, Kumar et al. 2014 for Black
Kites, and Rosenfield et al. 2018 for Cooper’s
Hawks). Urban habitat use of tropical raptors has
Table 5. Results from the best model for each urban index using generalized least squares tests, fitted to explain variation based on
the five traits analysed. *Values significant at P<0.05.
Variable Value 95% CI P
Urban abundance (Pagel’s lambda)
(Intercept) 0.390 [–0.153 to 0.933] 0.158
log(Mass) –0.105 [–0.185 to –0.025] 0.010*
Diet breadth –0.022 [–0.102 to 0.057] 0.578
Habitat breadth 0.146 [0.086–0.205] 0.000*
Nest substrate breadth 0.110 [0.004–0.146] 0.043*
Migratory status (Partial/Sedentary) –0.125 [–0.479 to 0.228] 0.483
Migratory status (Fully/Sedentary) –0.132 [–0.455 to 0.192] 0.422
Species proportion (Non-phylogenetic)
(Intercept) 0.139 [0.030–0.139] 0.013
log(Mass) –0.026 [–0.041 to 0.011] 0.001*
Diet breadth 0.001 [–0.013 to 0.015] 0.891
Habitat breadth 0.018 [0.007–0.028] 0.001*
Nest substrate breadth 0.011 [–0.007 to 0.030] 0.229
Migratory status (Partial/Sedentary) 0.009 [–0.053 to 0.071] 0.783
Migratory status (Fully/Sedentary) 0.001 [–0.057 to 0.059] 0.973
© 2022 British Ornithologists' Union
Global urban raptors 7
received far less attention, but we note a survey of
Brown Goshawk in Darwin, Australia (Riddell
2015), and mentions of urban occurrence of
Shikra in Singapore (Ward 1968) and Brahminy
Kite Haliastur indus in Java (Van Balen et al.
1993). A few urban-associated, habitat-generalist
species that were not among the most abundant
species in our dataset include several that are
restricted to tropical and subtropical areas, where
eBird use is lower. Their use of urban habitats has
received scant research attention, but we note pre-
vious studies of urban occurrence of Roadside
Hawk (Dos Santos & Rosado 2009) and Ovambo
Sparrowhawk Accipiter ovampensis (McPherson
et al. 2021).
Raptor species scoring highly in urban occur-
rence with low habitat breadth values, such as Eur-
asian Sparrowhawk and Red-shouldered Hawk,
may be utilizing some particular habitat present in
urban areas, such as a localized or super-abundant
food source (see Bell et al. 2010) or the presence
of an appealing microhabitat such as riparian
woodland (see Preston et al. 1989). While body
mass is fairly straightforward to measure, a poten-
tial difficulty in interpreting trait breadth as a vari-
able is that it does not distinguish between a
Figure 1. Small to mid-sized species are among the most abundant (a) and dominant (b) of 127 focal raptor species across 59
cities. Mean body mass is plotted on a natural log scale, and a value of 7 roughly corresponds to 1 kg. The most abundant and dom-
inant species are labelled, and the grey shading indicates a 95% CI.
© 2022 British Ornithologists' Union
8D. S. Cooper et al.
species that is ubiquitous and flexible, and one
that requires a diversity of a particular resource,
such as a mosaic of multiple habitat types (or prey
types). More granular investigation into actual
habitat usage by urban raptors within urban areas
(particularly in the tropics) would elucidate some
of these patterns, given how variable habitats can
be from city to city (see Dykstra 2018 for discus-
sion). Understanding what allows for the persis-
tence of these species, and large-yet-urban-tolerant
raptors such as Red-tailed Hawk, in urban areas
could guide conservation decisions in cities work-
ing to promote a diversity of raptors.
Further research into the rarest smaller and
mid-sized raptors in the study may yield important
information about why certain (smaller) taxa are
threatened by urbanization, as our results suggest
these would be more common (see Poos & Jackson
2012). It could also elucidate why the largest rap-
tors appear be absent or very rare in urban areas
(Fig. 1). Is this simply because more small raptors
are abundant (everywhere) than large ones, or
might there a mechanistic explanation for having a
small body size in an urban area? Several authors
have identified a decline in body mass over time
linked to climatic warming (e.g. Lurgi et al. 2012,
Figure 2. Both species abundance (a) and species dominance (b) in urban areas are positively associated with habitat breadth,
based on 127 focal raptor species across 59 cities. The most abundant and dominant species are labelled, and the grey shading
indicates a 95% CI.
© 2022 British Ornithologists' Union
Global urban raptors 9
Weeks et al. 2019, but see Salewski et al. 2014),
and Merckx et al. (2018) suggested that because
urban areas are influenced by the heat island
effect, smaller body size in animals favours species
with increased dispersal capability and reduced
metabolic needs. While most raptors would be
adept at dispersal, it seems logical that the high
mobility of smaller species such as Accipiters and
similar genera makes them ‘pre-adapted’to urban
life (see Johnson & Munshi-South 2017). Future
work could investigate the relationship between
urban occurrence and other aspects of body size
such as wing loading, and the role of types of flight
or foraging methods conducive to life in urban
areas.
Further work could also investigate inter- and
intra-species effects and competition, as the pres-
ence of particular species in a given raptor com-
munity could affect which other species are
excluded or included via competition. Perhaps the
sheer abundance of the most common species in a
given region might enable them to find mates
readily and thus occupy a broader geographical
area (including urban areas) than a scarce species
would, regardless of their preferred habitat type or
body size. This spillover effect into the city from
peripheral areas could be examined by comparing
species’abundance in the peri-urban band around
cities with that within the urban core.
Community science datasets can reveal patterns
that might be obscured by studies limited to a
small number of locations, or those using a simple
binary classification of occurrence such as range
maps or presence/absence (Adler et al. 2020).
Nevertheless, as a source of data, we recognize
that eBird reports have potential limitations that
could not be totally controlled in our study,
including observer bias (over-reporting the same
individual, under-reporting a familiar species due
to its abundance, or overlooking a shy or incon-
spicuous species). We deliberately selected coun-
tries with high levels of eBird participation, and
analysed multiple cities from each country to max-
imize occurrences of each species in an effort to
reduce this bias. Future investigations could always
use more cities and more species, as the popularity
of online community-science platforms such as
eBird and iNaturalist (www.inaturalist.org) grows.
Still, although these platforms are excellent for
determining seasonal status and distribution (and
are comparable to existing standardized survey
methods; see Horns et al. 2018 and Neate-Clegg
et al. 2020), they cannot be used for assessing
demographics or nesting success. Increased partici-
pation in eBird across all countries outside the
USA and Canada, particularly in urban areas,
would refine future analyses, as would a compar-
ison with patterns found in winter raptor commu-
nities in cities, though this may be unlikely to
change overall results, as most raptors are non-
migratory (Horns & S
ßekercio
glu 2018; C
ß.H.S
ßek-
ercio
glu unpubl. data).
We caution against equating urban occurrence
of any wildlife species (including urban raptors;
see Dwyer et al. 2008) with conservation success.
Sol et al. (2020) and Bregman et al. (2016) discuss
the loss of functional diversity in urban species
assemblages, which in the long term may lead to
the loss of global biodiversity as particularly spe-
cialist species fail to adapt faster than their habitats
are urbanized (S
ßekercio
glu 2011). Consequently,
the selection pressure toward generalist bird spe-
cies in urban areas means that most of the threat-
ened and near-threatened raptor species may not
survive in these human-dominated landscapes.
Furthermore, our study did not compare the traits
of individuals within the same species (where, for
example, smaller individuals of the same species
might have reduced fitness; see Liker et al. 2008).
Thus, we cannot draw any conclusions about the
long-term outlook for the population health or
productivity of urban raptors through this analysis.
Repeating the study for other taxonomic groups
(including non-avian taxa) would be worthwhile
to test whether the patterns observed for raptors
are universal. Finally, more research into the
mechanisms affecting raptor occurrence in urban
areas, incorporating diet studies, nest-searching
and monitoring, and demographic research (such
as nesting success) would help fill the gaps in our
knowledge of urban wildlife and allow better plan-
ning for future ecological changes.
We thank Dan Chamberlain, Richard Fuller, David
Douglas and three anonymous reviewers for insightful
edits that greatly improved the manuscript. The Blum-
stein lab (in particular, Rachel Blakey and Watcharapong
‘Win’Hongjamrasslip) and Ryan Harrigan at UCLA
assisted Cooper in analytical methods, and members of
the Urban Nature Research Center at the Natural His-
tory Museum of Los Angeles County (including Kayce
Bell, Adam Clause, Greg Pauly and Jann Vendetti) pro-
vided helpful editorial comments on earlier drafts. We
would like to thank the Cornell Lab of Ornithology and
the thousands of eBird participants around the world
© 2022 British Ornithologists' Union
10 D. S. Cooper et al.
who continue to improve the database through their
sightings and volunteer review process. This study was
funded in part by the UCLA Grand Challenge as part of
Daniel S. Cooper’s doctorate programme.
FUNDING
None.
ETHICAL NOTE
None.
AUTHOR CONTRIBUTIONS
Daniel S. Cooper: Conceptualization (lead); For-
mal analysis (lead); Investigation (lead); Methodol-
ogy (supporting); Supervision (lead); Writing –
original draft (lead); Writing –review & editing
(lead). Allison J. Shultz: Conceptualization (sup-
porting); Formal analysis (supporting); Methodol-
ogy (supporting); Writing –review & editing
(supporting). Cagan H. Sekercioglu: Data curation
(equal); Writing –review & editing (supporting).
Fiona M. Osborn: Formal analysis (supporting);
Methodology (supporting); Software (supporting).
Daniel T. Blumstein: Supervision (lead); Writing –
review & editing (supporting).
Data availability statement
The authors confirm that the data supporting the
findings of this study are available within the
Appendix S1 associated with this paper, and via
download from Auer et al. (2020).
REFERENCES
Adler, F.R., Green, A.M. & S
ßekercio
glu, C
ß.H. 2020. Citizen
science in ecology: a place for humans in nature. Ann. N. Y.
Acad. Sci. 1469:52–64.
Auer, T., Barker, S., Borgmann, K., Charnoky, M., Childs,
D. & Curtis, J., Curtis, J., Davies, I., Downie, I., Fink, D.,
Fredericks, T., Ganger, J., Gerbracht, J., Hanks, C.,
Hochachka, W., Iliff, M., Imani, J., Johnston, A., Lenz, T.,
Levatich, T., Ligocki, S., Long, M.T., Morris, W., Morrow,
S., Oldham, L., Padilla Obregon, F., Robinson, O.,
Rodewald, A., Ruiz-Gutierrez, V., Strimas-Mackey, M.,
Wolf, H. & Wood, C. 2020. EOD –eBird Observation
Dataset. Cornell Lab of Ornithology. Occurrence dataset
accessed via GBIF.org in May 2020 (see GBIF citations
below).
Bell, C.P., Baker, S.W., Parkes, N.G., de Brook, M.L. &
Chamberlain, D.E. 2010. The role of the Eurasian
Sparrowhawk (Accipiter nisus) in the decline of the House
Sparrow (Passer domesticus) in Britain. Auk 127: 411–420.
Birdlife International. 2019. HBW & BirdLife Taxonomic
Checklist v4. December 2019. Available at: http://datazone.
birdlife.org/species/taxonomy [Accessed 6th May 2020].
Blomberg, S.P., Garland, T. & Ives, A.R. 2003. Testing for
phylogenetic signal in comparative data: behavioral traits are
more labile. Evolution 57: 717–745.
Boal, C.W. 2018. Urban raptor communities: Why some
raptors and not others occupy urban environments. In: Boal,
C.W. & Dykstra, C.R. (eds.) Urban Raptors: Ecology and
Conservation of Birds of Prey in Cities:36–50. Berkeley:
Island Press.
Boal, C.W., Mannan, R.W. & Hudelson, K.S. 1998.
Trichonomiasis in Cooper’s Hawks from Arizona. J. Wildl.
Dis. 34: 590–593.
Bregman, T.P., Lees, A.C., MacGregor, H.E.A., Darski, B.,
de Moura, N.G., Aleixo, A., Barlow, J. & Tobias, J.A.
2016. Using avian functional traits to assess the impact of
land-cover change on ecosystem processes linked to
resilience in tropical forests. Proc. R. Soc. B 283: 20161289.
Buechley, E.R., Santangeli, A., Girardello, M., Neate-Clegg,
M.H.C., Oleyar, D., McClure, C.J.W. & S
ßekercio
glu, C
ß.H.
2019. Global raptor research and conservation priorities:
tropical raptors fall prey to knowledge gaps. Divers. Distrib.
25: 856–869.
Cade, T.J., Martell, M., Redig, P., Septon, G.A. & Tordoff,
H.B. 1996. Peregrine Falcons in Urban North America. In:
Bird, D., Varland, D. & Negro, J. (eds.) Raptors in Human
Landscapes: Adaptations to Built and Cultivated
Environments:3–14. San Diego, CA: Raptor Research
Foundation and Academic Press.
Caizergues, A.E., Gr
egoire, A. & Charmantier, A. 2018.
Urban versus forest ecotypes are not explained by divergent
reproductive selection. Proc. Biol. Sci. 285: 20180261.
Chace, J.F. & Walsh, J.J. 2006. Urban effects on native
avifauna: a review. Landsc. Urban Plan. 74:46–69.
Cooper, D.S., Shultz, A.J. & Blumstein, D.T. 2020b.
Temporally separated data sets reveal similar traits of birds
persisting in a USA Megacity. Front. Ecol. Evol. 8: 251.
Cooper, D.S., Yeh, P.J. & Blumstein, D.T. 2020a. Tolerance
and avoidance of urban cover in a southern California
suburban raptor community over five decades. Urban
Ecosyst. 24: 291–300.
Copernicus Climate Change Service Climate Data Store.
2021. Land cover classification gridded maps from 1992 to
present derived from satellite observations. Available at,
https://cds.climate.copernicus.eu/cdsapp#!/dataset/satellite-
land-cover?tab=overview [Accessed 15th October 2021].
Croci, S., Butet, A. & Clergeau, P. 2008. Does urbanization
filter birds on the basis of their biological traits? Condor 110:
223–240.
Cuthbert, R., Taggart, M.A., Prakash, V., Saini, M., Swarup,
D., Upreti, S., Mateo, R., Chakraborty, S.S., Deori, P. &
Green, R.E. 2011. Effectiveness of action in India to reduce
exposure of Gyps vultures to the toxic veterinary drug
Diclofenac. PLoS One,6, e19069.
Dos Santos, W.M. & Rosado, F. 2009. Preliminary data on
the Roadside Hawk (Rupornis magnirostris, Gmelin, 1788)
biology in the northeast of Paran
a, Brazil. Rev. Agroneg
ocio
Meio Ambiente 2: 421–430.
© 2022 British Ornithologists' Union
Global urban raptors 11
Dwyer, J.F., Hindmarch, S. & Kratz, G.E. 2008. Raptor
mortality in urban landscapes. In: Boal, C.W. & Dykstra,
C.R. (eds.) Urban Raptors: Ecology & Conservation of Birds
of Prey in Cities: 199–213. Berkeley: Island Press.
Dykstra, C.R. 2018. City lifestyles: behavioral ecology of
urban raptors. In: Boal, C.W. & Dykstra, C.R. (eds.) Urban
Raptors: Ecology and Conservation of Birds of Prey in
Cities:18–36. Berkeley: Island Press.
Estes, W.A. & Mannan, R.W. 2003. Feeding behavior of
Cooper’s Hawks at urban and rural nests in southeastern
Arizona. Condor 105: 107–116.
Evans, K.L., Gaston, K.J., Sharp, S.P., McGowan, A. &
Hatchwell, B.J. 2009. The effect of urbanization on avian
morphology and latitudinal gradients in body size. Oikos
118: 251–259.
Ferguson-Lees, J. & Christie, D. 2001. Raptors of the World.
New York: Houghton Mifflin.
GBIF.org 2020a. GBIF Home Page Available from: https://
www.gbif.orgf [Accessed 30th January 2022].
GBIF.org 2020b. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.8snenr [Accessed 17th August
2020] (Spain records).
GBIF.org 2020c. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.retqje [Accessed 6th May 2020]
(U.S. records).
GBIF.org 2020d. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.4kaqm8 [Accessed 9th May 2020]
(Mexico records).
GBIF.org 2020e. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.h57yfd [Accessed 9th May 2020]
(Australia records I).
GBIF.org 2020f. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.4k7s5c [Accessed 9th May 2020]
(India records).
GBIF.org 2020g. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.mcym3m [Accessed 9th May
2020] (Colombia records).
GBIF.org 2020h. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.akpw4r [Accessed 9th October
2020] (United Kingdom, Taiwan & Brazil records).
GBIF.org 2020i. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.4u8xaw [Accessed 9th October
2020] (Mexico records II).
GBIF.org 2020j. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.m4g8kt [Accessed 9th October
2020] (India records II).
GBIF.org 2020k. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.qyaa8s [Accessed 9th October
2020] (Australia records II).
GBIF.org 2020l. GBIF Occurrence Download. Available from:
https://doi.org/10.15468/dl.tb346k [Accessed 9th October
2020] (Spain records II).
Globalraptors.org 2020. The Global Raptor Information
Network Home Page. Available from: https://www.
globalraptors.org [Accessed 30th January 2022].
Hensley, C.B., Trisos, C.H., Warren, P.S., MacFarland, J.,
Blumenshine, S., Reece, J. & Katti, M. 2019. Effects of
Urbanization on Native Bird Species in Three Southwestern
US Cities. Front. Ecol. Evol. 7: 71. https://doi.org/10.3389/
fevo.2019.00071
Hijmans, R.J., Williams, E. & Vennes, C. 2015.
Geosphere: spherical trigonometry. R package version 1.5-
10. Available online. https://CRAN.R-project.org/package=
geosphere [Accessed 6th May 2020].
Horns, J.J., Adler, F.R. & S
ßekercio
glu, C
ß.H. 2018. Using
opportunistic citizen science data to estimate average
population trends. Biol. Conserv. 221: 151–159.
Horns, J.J. & S
ßekercio
glu, C
ß.H. 2018. Conservation of
migratory species. Curr. Biol. 28: R980–R983.
del Hoyo, J., Elliott, A. & Sargatal, J. (eds) 2013. Handbook
of the Birds of the World, Vol 2, Barcelona: Lynx Edicions.
Johnson, M.C. & Munshi-South, J. 2017. Evolution of life in
urban environments. Science 358: eaam8327.
Kettel, E.F., Gentle, L.K., Quinn, J.L. & Yarnell, R.W. 2018.
The breeding performance of raptors in urban landscapes: a
review and meta-analysis. J. Ornithol. 159:1–18.
Kopij, G. 2018. Ecological distribution and population
densities of raptors in the inner and outer zone of a Central
European City. Ukrain. J. Ecol. 8: 772–779.
Kumar, N., Mohan, D., Jhala, Y., Qureshi, Q. & Sergio, F.
2014. Density, laying date, breeding success and diet of
Black Kites Milvus migrans govinda in the city of Delhi
(India). Bird Study 61:1–8.
Liker, A., Papp, Z., B
okony, V. & Lendvai,
A.Z. 2008. Lean
birds in the city: body size and condition of House Sparrows
along the urbanization gradient. J. Anim. Ecol. 77: 789–795.
Lurgi, M., L
opez, B.C. & Montoya, J.M. 2012. Novel
communities from climate change. Philos. Trans. R. Soc. B
367, 2913–2922.
Mak, B., Francis, R.A. & Chadwick, M.A. 2021. Living in the
concrete jungle: a review and socio-ecological perspective
of urban raptor habitat quality in Europe. Urban Ecosyst. 24:
1179–1194.
Marti, C.D., Korpim€
aki, E. & Jaksi
c, F.M. 1993. Trophic
structure of raptor communities: a three-continent
comparison and synthesis. In: Power, D.M. (ed.) Current
Ornithology:47–137 10. Boston: Springer.
Martins, E.P. 1994. Estimating the rate of phenotypic
evolution from comparative data. Am. Nat. 144: 193–209.
Marzluff, J.M. 2016. A decadal review of urban ornithology
and prospectus for the future. Ibis 159:1–13.
Meill
ere, A., Brischoux, F., Parenteau, C. & Angelier, F.
2015. Influence of urbanization on body size, condition, and
physiology in an urban exploiter: a multi-component
approach. PLoS One 10: e0135685. https://doi.org/10.1371/
journal.pone.0135685
McCabe, J.D., Yin, H., Cruz, J., Radeloff, V., Pidgeon, A.,
Bonter, D.N. & Zuckerberg, B. 2018. Prey abundance and
urbanization influence the establishment of avian predators
in a metropolitan landscape. Proc. R. Soc. B 285:
20182120.
McClure, C.J.W., Westrip, J.R.S., Johnson, J.A., Schulwitz,
S.E., Virani, M.Z., Symes, A., Wheatley, H., Thorstrom,
R., Amar, A., Buij, R., Jones, V.R., Williams, N.P.,
Buechley, E.R. & Butchart, S.H.M. 2018. State of the
world’s raptors: distributions, threats and conservation
recommendations. Biol. Conserv. 227: 390–402.
McPherson, S.C., Sumasgutner, P. & Downs, C.T. 2021.
South African raptors in urban landscapes: a review. Ostrich
92:41–57.
Merckx, T., Souffreau, C., Kaiser, A., Baardsen, L.F.,
Backeljau, T., Bonte, D., Brans, K.I., Cours, M., Dahirel,
M., Debortoli, N., De Wolf, K., Engelen, J.M.T., Fontaneto,
D., Gianuca, A.T., Govaert, L., Hendrickx, F., Higuti, J.,
© 2022 British Ornithologists' Union
12 D. S. Cooper et al.
Lens, L., Martens, K., Matheve, H., Matthysen, E., Piano,
E., Sablon, R., Sch€
on, I., Van Doninck, K., De Meester, L.
& Van Dyck, H. 2018. Body-size shifts in aquatic and
terrestrial urban communities. Nature 558: 113–116.
Mundry, R. 2014. Statistical issues and assumptions of
phylogenetic generalized least squares. Chapter 6. In:
Garamszegi, L.Z. (ed.) Modern Phylogenetic Comparative
Methods and Their Application in Evolutionary Biology: 131–
153. Berlin: Springer. https://doi.org/10.1007/978-3-662-
43550-2
M€
unkem€
uller, T., Lavergne, S., Bzeznik, B., Dray, S.,
Jombart, T., Schiffers, K. & Thuiller, W. 2012. How to
measure and test phylogenetic signal. Methods Ecol. Evol.
3: 743–756.
Nakayama, S.M.M., Morita, A., Ikenaka, Y., Mizukawa, H. &
Ishizuka, M. 2019. A review: poisoning by anticoagulant
rodenticides in non-target animals globally. J. Vet. Med. Sci.
81: 298–313.
Neate-Clegg, M.H.C., Horns, J.J., Adler, F.R., Kemahlı
Aytekin, M.C
ß.&S
ßekercio
glu, C
ß.H. 2020. Monitoring the
world’s bird populations with community science data. Biol.
Conserv. 248: 109653.
Open Tree of Life. 2019. Version 3.2, October 30, 2019.
Available at: https://tree.opentreeoflife.org/opentree/argus/
ottol@1036185/Accipitridae [Accessed 6th May 2020].
Oro, D., Genovart, M., Tavechhia, G., Fowler, M.S. &
Mart
ınez-Abra
ın, A. 2013. Ecological and evolutionary
implications of food subsidies from humans. Ecol. Lett. 16:
1501–1514.
Pagel, M.D. 1999. Inferring the historical patterns of biological
evolution. Nature 401: 877–884.
Palacio, F.X. 2020. Urban exploiters have broader dietary
niches than urban avoiders. Ibis 162:42–49.
Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D. & R Core
Team 2019. nlme: Linear and Nonlinear Mixed Effects
Models. R Pack. Vers. 3.1-143. Available online: https://
CRAN.R-project.org/package=nlme [Accessed 5th June
2021].
Poos, M.S. & Jackson, D.A. 2012. Addressing the removal of
rare species in multivariate bioassessments: The impact of
methodological choices. Ecol. Ind. 18:82–90.
Powers, L.R. 1996. Wintering Sharp-shinned Hawks (Accipiter
striatus) in an urban area in southwestern Idaho. Northwest.
Nat. 77,9–13.
Preston, C.R., Harger, C.S. & Harger, H.E. 1989. Habitat use
and nest-site selection by Red-shouldered Hawks in
Arkansas. Southwest. Nat. 34:72–78.
R Core Team. 2020. R: A language and environment for
statistical computing. Vienna: R Foundation for Statistical
Computing Available from: https://www.r-project.org/
Revell, L.J. 2012. Phytools: an R package for phylogenetic
comparative biology (and other things). Methods Ecol. Evol.
3: 217–223.
Riddell, W.E. 2015. Aspects of breeding ecology of the Brown
Goshawk (Accipiter fasciatus) in an urban environment in
northern Australia. Northwest Terr. Nat. 26:32–43.
Rosenfield, R.N., Mannan, R.W. & Millsap, B.A. 2018.
Cooper’s Hawks: the bold backyard hunters. In: Boal, C.W.
& Dykstra, C.R. (eds.) Urban Raptors: Ecology and
Conservation of Birds of Prey in Cities:93–109. Berkeley:
Island Press.
Rullman, S. & Marzluff, J.M. 2014. Raptor presence along an
urban–wildland gradient: influences of prey abundance and
land cover. J. Raptor Res. 48: 257–272.
Salewski, V., Siebenrock, K.-H., Hochachka, W.M., Woog,
F. & Fielder, W. 2014. Morphological change to birds over
120 years is not explained by thermal adaptation to climate
change. PLoS One 9: e101927.
Samia, D.S.M., Nakagawa, S., Nomura, F., Rangel, T.F. &
Blumstein, D.T. 2015. Increased tolerance to humans
among disturbed wildlife. Nat. Commun. 6: 8877.
Santini, L., Gonz
alez-Su
arez, M., Russo, D., Gonzalez-
Voyer, A., Von Hardenberg, A. & Ancillotto, L. 2019. One
strategy does not fit all: determinants of urban adaptation in
mammals. Ecol. Lett. 22: 365–376.
S
ßekercio
glu, C
ß.H. 2006. Ecological significance of bird
populations. In del Hoyo, J., Elliott, A. & Christie, D.A. (eds.)
Handbook of the Birds of the World:15–51, vol 11.
Barcelona: Lynx Press; Cambridge: BirdLife International.
S
ßekercio
glu, C
ß.H. 2011. Functional extinctions of bird
pollinators cause plant declines. Science 331: 1019–1020.
S
ßekercio
glu, C
ß.H., Daily, G.C. & Ehrlich, P.R. 2004.
Ecosystem consequences of bird declines. Proc. Natl. Acad.
Sci. U.S.A. 101: 18042–18047.
S
ßekercio
glu, C
ß.H., Mendenhall, C.D., Oviedo-Brenes, F.,
Horns, J.J., Ehrlich, P.R. & Daily, G.C. 2019. Long-term
declines in bird populations in tropical agricultural
countryside. Proc. Natl. Acad. Sci. U.S.A. 116: 9903–
9912.
Seto, K.C., Fragkias, M., Guneralp, B. & Reilly, M.K. 2011.
A meta-analysis of global urban expansion. PLoS One 6:
e23777.
Shochat, E. 2004. Credit or debit? Resource input changes
population dynamics of city-slicker birds. Oikos 106: 622–
626.
Sol, D., Gonz
alez-Lagos, C., Moreira, D., Maspons, J. &
Lapiedra, O. 2014. Urbanisation tolerance and the loss of
avian diversity. Ecol. Lett. 17: 942–950.
Sol, D., Trisos, C., M
urria, C., Jeliazkov, A., Gonz
alez-
Lagos, C., Pigot, A.L., Ricotta, C., Swan, C.M., Tobias,
J.A. & Pavoine, S. 2020. The worldwide impact of
urbanization on avian functional diversity. Ecol. Lett. 23:
962–972.
Stout, W.E., Temple, S.A. & Papp, J.M. 2006. Landscape
correlates of reproductive success for an urban-suburban
Red-tailed Hawk population. J. Wildl. Manag. 70: 989–997.
Suri, J., Sumasgutner, P., Hellard, E., Koeslag, A. & Amar,
A. 2017. Stability in prey abundance may buffer Black
Sparrowhawks Accipiter melanoleucus from health impacts
of urbanization. Ibis 159:38–54.
Van Balen, B., Sumelo, I.S., Hadi, D.W., Soepomo, D.,
Marlon, R. & Mutiarina 1993. The decline of the Brahminy
Kite Haliastur indus on Java. Forktail 8:83–88. https://
static1.squarespace.com/static/5c1a9e03f407b482a158da87/
t/5c1fa88a40ec9abd16b871d0/1545578635912/VanBalen-
BrahminyKite.pdf
Ward, P. 1968. Origin of the avifauna of urban and suburban
Singapore. Ibis 110: 239–255.
Weeks, B.C., Willard, D.E., Zimova, M., Ellis, A.A.,
Witynski, M.L., Hennen, M. & Winger, B.M. 2019. Shared
morphological consequences of global warming in North
American migratory birds. Ecol. Lett. 17: 942–950.
© 2022 British Ornithologists' Union
Global urban raptors 13
White, J.H., Smith, J.M., Bassett, S.D., Brown, J.L. &
Ormsby, Z.E. 2018. Raptor nesting locations along an
urban density gradient in the Great Basin, USA. Urban
Ecosyst. 21:51–60.
Yom-Tov, Y. 2003. Body sizes of carnivores commensal with
humans have increased over the past 50 years. Funct. Ecol.
17: 323–327.
Received 3 February 2021;
Revision 3 December 2021;
revision accepted 18 January 2022.
Associate Editor: David Douglas.
SUPPORTING INFORMATION
Additional supporting information may be found
online in the Supporting Information section at
the end of the article.
Table S1. Cities used, including coordinates
used for urban centre and months used for breed-
ing season records (2014–2018).
Table S2. Summary of focal species and trait
values, averaged across the number of cities
recorded. DB—Diet Breadth; HB—Habitat
Breadth; Mass—body mass; Mig.—Migratory sta-
tus, SBS—Number of nest substrate categories,
Urb_Abund—Urban Abundance, Urb_Prop—
Urban Proportion. Refer to Table 2 for descrip-
tions of trait variables.
Table S3. Summary of edits to eBird records
and trait database (S
ßekercio
glu et al. 2004, includ-
ing updates). These include values found in exist-
ing sources (e.g. Ferguson-Lees & Christie 2001),
and those estimated from similar species to replace
missing data (e.g. body mass for similarly sized
species). We also report taxa dropped due to
nomenclatural and spelling discrepancies between
the various databases and R packages used to help
guide future investigators. Finally, we list the taxa
omitted from analysis in the particular countries or
cities where they are unlikely to be breeding in
any of our focal cities (in some cases omitting
them from the entire analysis if occurring only as
non-breeding visitors; e.g. Buteo lagopus).
Figure S1. (a) Aerial imagery of 10- and 50-km
radial bands around cities of Spain used in analysis.
(b) Aerial imagery of Barcelona, Spain, showing
amount of urbanized land within 10- and 50-km
radial bands. Note that urban cover in this particu-
lar city is fairly dispersed, such that areas of urban
and wildland cover are located both within the 10-
km band and within the ‘peri-urban’band. (c)
Aerial imagery of Ahmadebad, India, showing solid
urbanization within urban core of city (10-km
radial band) and comparatively little urban cover
in the ‘peri-urban’band. (d) Aerial imagery of Sao
Paolo, Brazil, showing how the urban core extends
well outside the 10-km radial band, but the ‘peri-
urban’band is still far less developed than the
urban core.
Figure S2. Correlation matrix of variables used
in the study using Spearman’s rank correlation.
Figure S3. QQ plot of residuals for the best-
performing model using Urban Abundance as the
dependent variable (Pagel’s lamdba).
Figure S4. QQ plot of residuals for the best-
performing model using Species Proportion as the
dependent variable (Non-phylogenetic).
© 2022 British Ornithologists' Union
14 D. S. Cooper et al.
