Content uploaded by Gary W. Luck
Author content
All content in this area was uploaded by Gary W. Luck on Nov 29, 2017
Content may be subject to copyright.
Environmental and socio-economic factors related to
urban bird communitiesaec_2383 111..120
GARY W. LUCK,1* LISA T. SMALLBONE1AND KATHRYN J. SHEFFIELD2
1Institute for Land,Water and Society, Charles Sturt University, PO Box 789, Albury, NSW 2640,
Australia (Email: galuck@csu.edu.au), and 2Future Farming Systems Research Division, Department
of Primary Industries, Carlton,Victoria, Australia
Abstract Urban fauna communities may be strongly influenced by environmental and socio-economic factors,
but the relative importance of these factors is poorly known. Most research on urban fauna has been conducted in
large cities and it is unclear if the patterns found in these locations coincide with those from smaller human
settlements. We examined the relative importance of environmental and socio-economic factors in explaining
variation in urban bird communities across 72 neighbourhoods in 18 regional towns in south-eastern Australia.
Native bird species richness varied from 6 to 32 across neighbourhoods and was higher in neighbourhoods with
more nectar-rich plants. Variation in bird species diversity across neighbourhoods was also strongly positively
related to the density of nectar-rich plants, but was higher also in neighbourhoods with higher socio-economic
status (reflecting higher levels of disposal income, education and home ownership). The density of native birds
across neighbourhoods per season varied from 1 to 15 birds per hectare and was lower in neighbourhoods with a
greater cover of impervious surfaces. The density of exotic birds (introduced to Australia) per season also varied
across neighbourhoods (0–13 birds per hectare) and was lower in neighbourhoods with more nectar-rich plants and
higher in neighbourhoods with greater impervious surface cover. Our results demonstrated that the vegetation
characteristics of household gardens, along streetscapes and in urban parklands had a strong influence on the
richness and diversity of urban bird communities.The density of native and exotic birds varied primarily in response
to changes in the built environment (measured through impervious surface cover). Socio-economic factors had
relatively little direct influence on urban birds, but neighbourhood socio-economics may influence bird commu-
nities indirectly through the positive relationship between socio-economic status and vegetation cover recorded in
our study area.
Key words: avian ecology, urban ecology, urban ecosystem, urbanization.
INTRODUCTION
The world is becoming increasingly urbanized. The
rate of growth in the number of households and urban-
ized land is substantially outpacing population growth
in many countries (Liu et al. 2003). By 2050, approxi-
mately 70% of the global population will live in urban
areas (United Nations 2010). Researchers are only just
beginning to understand the impact of this rapid
urbanization on ecological communities (Grimm et al.
2008; Goddard et al. 2009; Gaston 2010). While many
studies find that highly urbanized locations support
few fauna species (McKinney 2008), the structure
of fauna communities varies spatially and temporally
within and across urban areas.
Variation in urban fauna communities has been
explored largely as a factor of the natural features of
urban landscapes such as parkland and streetscape
vegetation, waterways and remnant vegetation patches
(e.g. Fernández-Juricic 2000; Gehrt & Chelsvig 2003;
Miller et al. 2003; White et al. 2005; Daniels & Kirk-
patrick 2006; Garden et al. 2006; Gaublomme et al.
2008). However, there is increasing recognition of the
role of human factors in dictating patterns and pro-
cesses in urban ecosystems. In particular, neighbour-
hood socio-economic characteristics such as income,
home ownership and education levels may have strong
relationships with features of the urban environment,
especially vegetation cover and diversity (e.g. Hope
et al. 2003; Martin et al. 2004; Luck et al. 2009).
Understanding the relative importance of natural,
built and socio-economic factors in influencing urban
fauna communities is crucial to promoting species
persistence within urban settlements (Kinzig et al.
2005; Loss et al. 2009). Such information may help
guide urban planners regarding the social, economic or
landscape architectural policies that could be imple-
mented to improve conservation outcomes in urban
areas. Given the substantial similarity in urban
*Corresponding author.
Accepted for publication March 2012.
Austral Ecology (2013) 38, 111–120
bs_bs_banner
© 2012 The Authors doi:10.1111/j.1442-9993.2012.02383.x
Austral Ecology © 2012 Ecological Society of Australia
settlement design, particularly within countries, and
the predominance of certain types of species in urban
areas (Chace & Walsh 2006), it is possible that the
relationships between settlement characteristics and
fauna communities are quite predictable from one loca-
tion to another. If so, this suggests that implementing
the same management strategies across urban areas will
yield similar biological outcomes.
However, most research on urban fauna, particularly
birds, has been conducted in large cities (e.g. Parsons
et al. 2006; Young et al. 2007; Palmer et al. 2008; Loss
et al. 2009). Fewer studies have focused on smaller
human settlements (e.g. Jokimäki & Kaisanlahti-
Jokimäki 2003; Kath et al. 2009). Therefore, knowl-
edge is lacking regarding species responses to the full
spectrum of urbanization levels and whether relation-
ships recorded in major metropolitan centres are con-
sistent with those found in smaller human settlements.
Comparisons across settlements of differing size
should yield a broader understanding of the impacts of
varying levels of urbanization on native ecosystems.
Here, we assess spatial variation in bird communities
across residential neighbourhoods located in regional
towns and cities in south-eastern Australia. We
examine relationships between birds and the environ-
mental and socio-economic characteristics of neigh-
bourhoods, and compare our results to studies
conducted in major cities.
METHODS
Study area
Our study was conducted in 18 regional towns and cities
across the states of Victoria and New South Wales (NSW) in
south-eastern Australia (Fig. 1). We surveyed four neigh-
bourhoods in each town (total 72 neighbourhoods) with
neighbourhood boundaries defined by census collection dis-
tricts (approx. 200 houses; see Luck et al. 2009). Neighbour-
hoods were selected for surveying using stratified random
sampling to capture the full range of variation in neighbour-
hood types, which ranged from relatively sparsely populated
peri-urban neighbourhoods at the fringes of towns to more
densely populated suburban neighbourhoods near town
centres. Each neighbourhood was within 10 km of the town
centre, predominant land use was residential and housing
density ranged from 0.5 to 11 houses per hectare. We
Fig. 1. The location of the 18 survey towns in south-eastern Australia.The population size for each town is from the Australian
Bureau of Statistics 2006 Census (http://www.abs.gov.au/websitedbs/censushome.nsf/home/Data).
112 G. W. LUCK ET AL.
© 2012 The Authorsdoi:10.1111/j.1442-9993.2012.02383.x
Austral Ecology © 2012 Ecological Society of Australia
controlled for differences in neighbourhood age, which may
affect, for example, vegetation cover (Hope et al. 2003;
Martin et al. 2004), by selecting neighbourhoods that were at
least 20 years old (as at 2007). Neighbourhoods in the same
town were separated by at least 1 km.
Bird surveys
Species richness and diversity
In Victoria (and the border town of Albury, which is in
NSW), all neighbourhoods were surveyed for birds over two
consecutive days in each of four seasons (summer, autumn,
winter and spring) between December 2007 and October
2008. In NSW, we surveyed in each neighbourhood over two
consecutive days in both spring and summer in 2009. All
surveys were conducted by LTS during suitable weather
conditions (e.g. avoiding strong wind and heavy rain). As we
are examining relationships across all towns in this analysis,
we combined the species data for Victoria and NSW based
only on the spring and summer surveys.
To record species richness, we used an active search
method and employed a results-based stopping rule that was
defined by a pilot study (Miller & Cale 2000; Watson 2003).
All neighbourhoods were actively searched for birds between
dawn and 10.00 hours by visiting all of the microhabitats in
the survey area including private gardens (by looking into the
garden from the footpath), streetscapes, urban parks and
remnant vegetation. All bird species seen and heard were
recorded unless they were flying >10 m overhead or outside
of the neighbourhood boundary. Our stopping rule was
based on a survey duration–result outcome whereby we sur-
veyed for a minimum of 20 min and a maximum of 40 min in
each neighbourhood, but halted the survey prior to 40 min if
no new species was recorded in a continuous 5-min period
after surveying for 20 min. We calculated bird species diver-
sity using the Shannon–Wiener diversity index (Krebs 1999).
Bird density
To measure bird abundance, we used the point transect
method and a distance sampling procedure to account for
differences in detectability among species and sites when
converting abundance measures to densities (Buckland et al.
2001). Four sampling points were randomly located in each
neighbourhood and surveyed for 3 min each. Surveys began at
dawn and were completed by 10.00 hours. A short sur vey
duration was utilized to avoid double-counting the same
individuals.The radial distance of each bird from the observer
was recorded up to a distance of 50 m, although the maximum
distance was truncated during data analysis (Appendix S1).
Birds heard, but not seen, that were clearly <50 m from the
observer were included in abundance estimates as a single
individual. Any birds that were flushed when the observer
moved into the sample point at the beginning of the survey
were recorded at the original distance before flushing. Bird
abundance in each neighbourhood was converted to density
after calculating a detection probability using Distance
sampling software (version 5.0) that accounts for variation in
the detectability of species (Thomas et al. 2006, 2010; see
Appendix S1).
Neighbourhood environment and
socio-economic profile
We measured natural, built and socio-economic characteris-
tics of each neighbourhood that were hypothesized to have an
influence on urban bird communities (Table 1).
Vegetation cover
The proportional cover of woody and non-woody vegetation
was measured in each neighbourhood and within a 1-km
radius of the neighbourhood boundary to examine the poten-
tial for broader landscape effects.We developed a land cover
classification for each town using ALOS satellite imagery at a
10-m pixel resolution (Appendix S1). The following land
cover classes were defined in each case: water; bare soils or
surfaces (including cleared land and bare paddocks); imper-
vious surfaces (e.g. roads); and green vegetation cover incor-
porating woody and non-woody vegetation (e.g. annual
vegetation, including lawns, that was green at the time of the
satellite survey). The latter was used to represent ‘vegetation
cover’ in our study, while the proportional cover of impervious
surfaces was included in our measure of the built environ-
ment. Water availability was not considered, as most neigh-
bourhoods (60 of 72) had <1% of their land area under water.
Vegetation density
Vegetation characteristics were measured in each neighbour-
hood at four randomly located quadrats (20 m ¥100 m,
and nested quadrats of 20 m ¥50 m) in parks or built-up
areas. In the built-up areas, quadrats were aligned along the
road edge and projected 20 m into house blocks. In the
20m¥100 m quadrats, we measured the total number of
trees >20 cm in diameter at breast height and the number of
native trees (e.g. eucalypts). In the nested 20 m ¥50 m
quadrats, we measured the number of nectar-rich plants
from the families Myrtaceae and Proteaceae (an important
food source for native birds). This gave an indication of the
‘nativeness’ of gardens, as species in these families comprised
>90% of all native species recorded.
Built environment
To represent the built environment, we measured housing
density and impervious surface cover (including paved areas
and the roofs of buildings). Housing density values were
obtained from the Australian Bureau of Statistics (ABS)
2006 Census.The proportional cover of impervious surfaces
was calculated from ALOS satellite imagery. Housing density
per hectare (square-root transformed) and impervious
surface cover (arcsine(square-root) transformed) were posi-
tively correlated (r=0.60) and we created the composite
variable ‘urban intensity’ from these measures using principal
components analysis (Appendix S1).
URBAN BIRD COMMUNITIES 113
© 2012 The Authors doi:10.1111/j.1442-9993.2012.02383.x
Austral Ecology © 2012 Ecological Society of Australia
Socio-economic profile
The socio-economic profile of each neighbourhood was rep-
resented by the following variables: disposable income; edu-
cation level; home ownership; and a composite variable
‘socio-economic status’. Data for all variables were obtained
from the ABS 2006 Census.To calculate disposable income
per household, we subtracted weekly housing loan repay-
ments and rent from weekly household income. This value
was divided by an equivalization factor that adjusts income
relative to the number of residents per household and
their age (see Luck et al. 2009). The proportion of all resi-
dents in a neighbourhood with a tertiary degree (i.e. at least
a bachelor degree) was used to represent education level,
while home ownership was measured as the proportion of
residents that owned outright or had a mortgage on their
home.
Disposable income (log10 transformed), education level
and home ownership (both arcsine(square-root) trans-
formed) were positively intercorrelated (mean r=0.62) and
the composite variable socio-economic status was derived
by combining these measures using principal components
analysis (Appendix S1). Higher socio-economic status
reflected neighbourhoods with more disposable income and
higher levels of tertiary education and home ownership.
The socio-economic profile of neighbourhoods may influ-
ence bird communities through relations with vegetation
cover (e.g. in our study area, neighbourhoods with higher
socio-economic status have greater vegetation cover and
more native vegetation; Luck et al. 2009) or through a link
with householder behaviour. For example, Fuller et al. (2008;
see also Parsons et al. 2006) demonstrated that the provision
of supplementary food by householders can influence the
structure of urban bird communities and that residents of a
higher socio-economic status were more likely to feed birds
than residents of a lower status. However, this relationship
may be context-dependent (see Discussion).
Hypotheses, explanatory and response variables
We established four a priori hypotheses regarding the rela-
tionships between the neighbourhood environment and
socio-economic profile and urban bird communities. These
hypotheses were based on evidence from the literature and
are labelled throughout as ‘vegetation cover’, ‘vegetation
density’, ‘built environment’ and ‘socio-economic profile’.We
hypothesized that bird species richness and diversity, and the
density of native birds would increase with increasing vegeta-
tion cover, vegetation density and socio-economic profile of
Ta b l e 1 . A description of the explanatory variables measured in each neighbourhood
Variable (transformation) Description Source
Vegetation cover
Neighbourhood vegetation
cover (arcsine-square-root)
The proportion of the neighbourhood covered in woody and
non-woody vegetation. Per cent cover ranged from 1% to 78%.
Satellite imagery
1-km vegetation cover
(arcsine-square-root)
The proportional cover of vegetation within a 1-km radius of the
neighbourhood boundary. Per cent cover ranged from 9% to
60%.
Satellite imagery
Vegetation density
Density of trees (log10) The density of all trees per 2000 m2in each neighbourhood,
ranging from 1 to 11.
Field survey
Density of native trees
(square-root)
The density of native trees per 2000 m2, ranging from 0 to 11. Field survey
Density of nectar plants
(square-root)
The density of nectar-rich plants per 1000 m2, ranging from 0 to
15.
Field survey
Built environment
Housing density
(square-root)
Number of houses per hectare in each neighbourhood, ranging
from 0.5 to 11.
Australian Bureau of
Statistics
Impervious surface cover
(arcsine-square-root)
The proportion of the neighbourhood covered in impervious
surfaces. Per cent cover ranged from 1% to 88%.
Satellite imagery
Urban intensity A composite variable derived through principal components
analysis combining housing density and impervious surface
cover (see Methods).
Socio-economic profile
Home ownership
(arcsine-square-root)
The proportion of total houses in the neighbourhood that were
owned or where residents had a mortgage. Per cent values
ranged from 20% to 100%.
Australian Bureau of
Statistics
Education level
(arcsine-square-root)
The proportion of the total neighbourhood population with a
tertiary degree. Per cent values ranged from 1% to 33%.
Australian Bureau of
Statistics
Income (log10) Average disposable income across all residents in the
neighbourhood, ranging from $322 to $831 per week.
Australian Bureau of
Statistics
Socio-economic status A composite variable derived through principal components
analysis combining the positively correlated measures of
income, home ownership and education level (see Methods).
114 G. W. LUCK ET AL.
© 2012 The Authorsdoi:10.1111/j.1442-9993.2012.02383.x
Austral Ecology © 2012 Ecological Society of Australia
residents, and decrease with increasing cover of impervious
surfaces.We expected the exact opposite relationships for the
density of exotic birds (introduced to Australia).
We tested for multi-collinearity among the explanatory
variables (Table 1) using the approach based on variance
inflation factors (VIF) described by Zuur et al. (2010). Here,
one explanatory variable is modelled against all other
explanatory variables using linear regression. A high r2value
and high VIF values suggest collinearity among the explana-
tory variables. To reduce collinearity, we sequentially
removed each explanatory variable with the highestVIF score
and recalculated the r2and VIF values for the remaining
variables until all VIF values were <3 (see Zuur et al. 2010).
We modified this approach slightly by ensuring also that any
pairwise correlation among the remaining variables was <0.6
and that we retained at least one variable to represent each of
our four hypotheses. Each hypothesis was represented by the
following variables: vegetation cover – the proportional cover
of vegetation within a 1-km radius of the neighbourhood
boundary (1-km vegetation); vegetation density – the density
of nectar-rich plants per 1000 m2in each neighbourhood
(nectar density); built environment – the proportion of the
neighbourhood covered in impervious surfaces (impervious
surfaces); and socio-economic profile – the socio-economic
status of neighbourhood residents (socio-economic status).
We considered the following response variables in our
analysis: native species richness; species diversity; native bird
density; and exotic bird density. Total species richness and
native species richness were highly correlated across neigh-
bourhoods (r=0.99) and we focused on the latter to better
represent the conservation implications of our results. Exotic
species richness and total bird density (native and exotic bird
density combined) varied little across neighbourhoods and
were not correlated with any of our explanatory variables;
therefore, we did not model these variables separately.
Data analysis
We compared alternative explanatory models derived from
the four hypotheses using a generalized linear mixed model.
As data were collected in different years (which corre-
sponded to surveys in either Victoria (including Albury) or
NSW; see above) and neighbourhoods confined to particular
towns, we included ‘year’ and ‘town’ as random effects in all
models. We included also the variable ‘neighbourhood area’
in models of species richness and diversity as larger neigh-
bourhoods may encompass more microhabitats and hence
contain more bird species (influencing measures of
diversity). Relationships with native species richness were
modelled with a Poisson distribution (a negative binomial
distribution did not yield a better fit to the data) and all other
relationships were modelled with a Gaussian distribution.
Overall model fit was assessed using the likelihood ratio
chi-square, which represents the change in deviance between
the fitted model and the constant-only model. All models
included here significantly improved model fit (P<0.01 in all
cases). The relative fit of competing explanatory models was
compared using Akaike’s information criterion corrected for
small sample sizes (AICc) based on an information theoretic
approach (Burnham & Anderson 2002). Smaller values of
AICcindicate a better fit. The difference in AICcvalues was
compared between the best ranked model and model i(Di).
Models where Diis <2 are usually considered to have substan-
tial empirical support, values between 2 and 4 suggest some
support, while values >10 indicate little support in the suite of
models being considered (Burnham & Anderson 2002).
We also calculated AICcweights (wi) for each model and
these represent the relative likelihood of the model and can
be interpreted as the probability that any given model is the
best model based on the data at hand. Summed AICcweights
for each explanatory variable (i.e. summing wiacross the
models that the explanatory variable occurred in) were cal-
culated as a measure of the relative importance of the
variable. All statistical analyses were completed using S-plus
8.2 and spss 17.0.
RESULTS
Species richness and diversity
We recorded a cumulative total of 96 bird species
across the 18 towns: 86 native species and 10 exotic
species. Species richness ranged from 6 to 32 across
neighbourhoods ( x15.6, SD ⫾6.0). The most wide-
spread species (occurring in >80% of neighbourhoods)
were the introduced common starling (Sturnus vul-
garis) and house sparrow (Passer domesticus), and the
native Australian magpie (Cracticus tibicen) and red
wattlebird (Anthochaera carunculata). Mean native
species richness differed between years, being higher
in 2009 (x17.7 ⫾4.7) compared to 2007/2008
(x13.5 ⫾6.5; Z=3.83, P<0.001, Mann–Whitney).
Conversely, species diversity was higher in 2007/2008
(x2.2 ⫾0.3) compared to 2009 (x1.9 ⫾0.3;
t=3.94, P<0.001).
Mean species richness per town (averaged across the
four neighbourhoods in each town) varied significantly
(
χ
17
236 94=.,P=0.003; Kruskal–Wallis), being high-
est in Nowra (23.5) and lowest in Warrnambool (9;
Fig. 2). Similarly, species diversity varied at the town
level (F17 =2.51, P=0.005).
The highest ranked model explaining variation
in native species richness across neighbourhoods
included only the density of nectar-rich plants (with
‘year’, ‘town’ and ‘area’ included in the model), and
this was well supported as the best model among
those considered (wi=0.51). The model including
the density of nectar-rich plants and socio-economic
status was also highly ranked, although the summed
Akaike weights indicated nectar density was clearly the
most influential variable (Table 2). Native bird species
richness was higher in neighbourhoods with more
nectar-rich plants. Parameter estimates were positive
for vegetation cover and socio-economic status, and
negative for impervious surfaces, but confidence inter-
vals often encompassed zero (Appendix S2).
The density of nectar-rich plants was also strongly
related to species diversity, although the models including
URBAN BIRD COMMUNITIES 115
© 2012 The Authors doi:10.1111/j.1442-9993.2012.02383.x
Austral Ecology © 2012 Ecological Society of Australia
only nectar density and nectar density and socio-economic
status had similar support. Summed Akaike weights also
suggested a stronger influence of socio-economic status on
species diversity than species richness, although nectar
density clearly had the highest weight (Table 2). Bird
species diversity increased in neighbourhoods with
more nectar-rich plants and higher socio-economic
status (Appendix S2).
Bird density
The density of native birds across neighbourhoods
per season varied from 1 to 15 birds per hectare ( x
4.9 ⫾2.9) and the density of exotic birds varied from
0 to 13 birds per hectare (x4.4 ⫾2.8). Mean density
of native birds per hectare was higher in 2007/2008
(x5.7 ⫾3.1) than 2009 (x4.1 ⫾2.6; t=2.67,
P=0.009), as was the mean density of exotic birds (x
5.8 ⫾3.2 vs. x3.1 ⫾1.7; t=4.48, P<0.001). The
mean density of native birds per hectare varied across
towns (F17 =2.09, P=0.02), ranging from 1.4 (⫾0.6)
in Shellharbour to 8.3 (⫾4.8) in Albury. Similarly,
there were town-level differences in the mean density
of exotic birds per hectare (F17 =3.30, P<0.001),
which was highest in Traralgon (9.5 ⫾3.5) and lowest
in Queanbeyan (1.7 ⫾1.4).
For the density of native birds, only one model
had substantial support (Di<2; wi=0.47) and this
included only impervious surfaces (Table 2). Two
models with moderate support (Di<2) included also
Fig. 2. Variation in native bird species richness across towns. Error bars are 95% confidence intervals.
Ta b l e 2 . The highest ranked models (Diⱕ2) explaining variation in urban bird communities across south-eastern Australia
(see Appendix S2 for all models)
Response Model†AICcDiAIC wic2
Species
richness
Year +To w n +Area‡(0.38, 0.15 to 0.61) +Nectar§(0.24, 0.13 to 0.35) 415.9 0 0.51 126.4
Year +To w n +Area (0.36, 0.13 to 0.59) +Nectar (0.21, 0.10 to
0.33) +SE¶(0.07, -0.21 to 0.15)
417.9 2.0 0.19 128.7
Summed wi: Nectar (0.99); SE (0.26); 1 km Veg†† (0.16); IS‡‡ (0.15)
Species
diversity
Year +To w n +Area (0.05, -0.15 to 0.25) +Nectar (0.22, 0.12 to 0.31) 39.8 0 0.48 67.0
Year +To w n +Area (0.03, -0.16 to 0.23) +Nectar (0.19, 0.10 to
0.29) +SE (0.07, -0.01 to 0.14)
40.7 0.9 0.31 70.2
Summed wi: Nectar (1.00); SE (0.39); IS (0.11);1 km Veg (0.11)
Native bird
density
Year +To w n +IS (-3.83, -5.98 to -1.67) 326.8 0 0.47 47.6
Summed wi: IS (0.77); Nectar (0.33); 1 km Veg (0.16); SE (0.13)
Exotic bird
density
Year +To w n +Nectar (-0.98, -1.67 to -0.30) +IS (2.85, 0.44 to 5.27) 313.1 0 0.39 79.4
Year +To w n +Nectar (-1.48, -2.04 to -0.92) 314.3 1.2 0.21 74.2
Summed wi: Nectar (0.90); IS (0.63); SE (0.23); 1 km Veg (0.13)
Numbers in brackets are the parameter estimates and their 95% confidence intervals. †AICc=Akaike’s information criterion
corrected for small sample sizes, Di=difference in AICcvalues, wi=AICcweights, and c2=the likelihood ratio chi-square, which
represents the change in deviance between the fitted model and the constant-only model (all values are significant; P<0.01).
‡Neighbourhood area. §The density of nectar-rich plants. ¶Neighbourhood socio-economic status. ††1-km vegetation cover.
‡‡Impervious surface cover.
116 G. W. LUCK ET AL.
© 2012 The Authorsdoi:10.1111/j.1442-9993.2012.02383.x
Austral Ecology © 2012 Ecological Society of Australia
the density of nectar-rich plants (Appendix S2), but
impervious surface cover appeared to be the most
influential variable (Swi=0.77). Native bird density
declined with increasing cover of impervious surfaces.
Two models had similar support in explaining varia-
tion in the density of exotic birds and these included
the density of nectar-rich plants and impervious sur-
faces, which were the two most influential variables
(Swi=0.90 and 0.63, respectively). Exotic bird density
was lower in neighbourhoods with greater cover of
nectar-rich plants and higher in neighbourhoods with
greater impervious surface cover.
DISCUSSION
Support for hypotheses
There was varied support for our hypotheses about the
relationships between urban bird communities and
the natural, built and socio-economic characteristics of
neighbourhoods. Native species richness increased in
neighbourhoods with a greater density of nectar-rich
plants, which reflected the presence of native vegetation
in gardens, streetscapes and urban parklands.However,
there was greater uncertainty about the influence of
overall vegetation cover, urbanization level and socio-
economic status on species richness. Similarly, species
diversity was most strongly positively related to the
density of nectar-rich plants. However, neighbourhood
socio-economic status was also positively related to this
bird measure likely reflecting the association between
vegetation type and cover and socio-economic status
that exists in our study area (see below).
The impact of urbanization level per se on bird commu-
nities was most pronounced when examining variation in
bird density. The cover of impervious surfaces had a sub-
stantial negative impact on the density of native birds and
this was the only variable strongly related to this bird
measure. Conversely, the density of exotic birds increased
with increasing impervious surface cover (suggesting a
preference for more urbanized areas), although nectar-
plant density contributed to limiting the number of exotic
birds.This likely reflects the fact that nectar-plant density
was higher in more sparsely populated neighbourhoods
and was also positively related to the level of‘nativeness’ of
residential gardens and streetscapes. Such locations appear
more suitable for native bird species. In sum, only nectar-
plant density and impervious surface cover had strong
relationships with bird communities in our study area.
The importance of local-scale and
broad-scale factors
There is growing recognition that urban bird commu-
nities are influenced more by local factors measured at
small scales than broader-scale regional factors (Cat-
terall et al. 1989; Clergeau et al. 2001; Daniels & Kirk-
patrick 2006; Evans et al. 2009; Kath et al. 2009). The
results of our study support this finding, showing that
bird species richness and diversity, and bird density
(particularly exotics) were more strongly influenced
by vegetation characteristics within neighbourhoods
(i.e. the density of nectar-rich plants), measured at the
scale of square metres, than total vegetation cover
measured more broadly (i.e. within 1 km of the neigh-
bourhood boundary).
The characteristics of household gardens and
streetscape vegetation appear to be important deter-
minants of urban bird communities (Chamberlain
et al. 2004; Daniels & Kirkpatrick 2006; Young et al.
2007; Burghardt et al. 2009; Goddard et al. 2009).
In Australia, native birds tend to favour native plant
species (Green 1984; White et al. 2005), and our
results support the conclusions of past studies. For
example, in Sydney, French et al. (2005) found that
native nectarivores preferentially fed on native Banksia
and Grevillea species rather than introduced plants.
In the streets of Adelaide, Young et al. (2007) found
that native nectarivores preferred to use the native
tree Eucalyptus camaldulensis compared to other tree
species. The importance of native plants in increasing
bird species richness and abundance in urban environ-
ments has also been recorded in the northern hemi-
sphere (e.g. Burghardt et al. 2009).
While local-scale factors are important, landscape
characteristics at larger scales may also influence the
structure and dynamics of fauna communities, par-
ticularly for highly mobile species (Chamberlain et al.
2004; Warren et al. 2008; Oneal & Rotenberry 2009).
A single household garden is unlikely to provide all the
resources required for most vertebrates and the influ-
ence of landscape characteristics such as remnant
native vegetation cover and riparian areas should be
considered (Hennings & Edge 2003; Hodgson et al.
2006; Pennington et al. 2008). Indeed, studies of
urban bird communities in Australia that confine their
surveys to remnant vegetation patches within urban
areas generally find that larger remnants support
more species (e.g. Palmer et al. 2008; Fitzsimons et al.
2011). Therefore, the presence and amount of
remnant native vegetation, coupled with the vegetation
characteristics of gardens and streetscapes, likely inter-
acts to influence the richness and composition of
urban bird communities.
Regional-level factors will also affect the composi-
tion of bird communities in any particular urban
location. This is underlined by the fact we recorded
significant differences in bird richness, diversity and
density across the 18 survey towns. Influential factors
may include, for example, regional variation in native
vegetation cover, geographical location (e.g. coastal vs.
inland), productivity (e.g. temperate vs. semi-arid),
URBAN BIRD COMMUNITIES 117
© 2012 The Authors doi:10.1111/j.1442-9993.2012.02383.x
Austral Ecology © 2012 Ecological Society of Australia
topographical variation or land-use history. Yet, the
strength of region-level influences is dampened by the
fact that urbanization tends to homogenize ecological
communities (i.e. urban areas may support similar
suites of species; McKinney 2006). This occurs in
our study area, where bird communities across the 18
survey towns are more similar than would be expected
from the variation in regional species pools (Luck &
Smallbone 2011). Studies confined to a single urban
area are unable to address the complex interactive
effects of local and regional-level factors on urban
ecosystems and more research is required that exam-
ines variation across widely separated urban localities.
Relevance of the built environment and
neighbourhood socio-economic profile
The built environment of urban landscapes may
impact fauna through, for example, changes in land
cover affecting biophysical processes (e.g. impervious
surface cover altering water infiltration and run-off)
or the reduction in vegetation cover with increasing
urban density (White et al. 2005). In our study,
increasing urban development as measured through a
greater cover of impervious surfaces was most strongly
related to bird density, negatively impacting on native
birds and favouring exotic species.This suggests that a
small suite of introduced species appear to be highly
adapted to urban environments (see below).
We did not record a strong relationship between the
socio-economic characteristics of neighbourhoods and
urban bird communities. Socio-economic factors are
unlikely to influence birds directly (although see Fuller
et al. 2008); however, they may influence urban fauna
indirectly through the link between socio-economic
profile and vegetation cover (Hope et al. 2003; Martin
et al. 2004). In our study area, neighbourhoods with
high socio-economic status were generally character-
ized by greater vegetation cover and a high proportion
of native plant species (Luck et al. 2009).These neigh-
bourhoods were mostly on the fringes of towns in more
sparsely populated ‘peri-urban’ locations.
We hypothesize that the degree of association
between neighbourhood socio-economic status and
vegetation cover and type is likely dependent on the
overall size of urban areas and site-specific context. As
urban areas grow, it can become less attractive to live
on the fringes of settlements distant from the central
business and entertainment district. In these circum-
stances, inner city suburbs become more attractive to
residents of higher socio-economic status (who have
greater financial capacity to choose where they live)
and these suburbs may be more densely populated
and unlikely to support substantial cover of native
vegetation. The interrelationships among the socio-
economic, built and natural features of neighbour-
hoods are also likely to be context-specific, com-
plicating attempts at generalization. For example, Loss
et al. (2009), in their study of urban birds in the
Chicago metropolitan region, reported a negative rela-
tionship between neighbourhood income level and
native bird species richness, whereas other studies have
reported the opposite (e.g. Kinzig et al. 2005; Melles
2005).
Comparisons with studies in large cities
The results of our study generally concur with previ-
ous research on urban bird communities in major
cities in Australia. For example, large and aggressive
birds such as the red wattlebird and Australian magpie
are widespread and common in urban areas across
Australia in both native vegetation remnants and resi-
dential areas (our study; Lenz 1990; White et al. 2005;
Garden et al. 2006; Fitzsimons et al. 2011). Similarly,
the common starling and house sparrow were wide-
spread in our residential neighbourhoods and have
also been commonly recorded in residential areas in
Canberra (Lenz 1990), while the starling was also
common in remnant vegetation in Melbourne (Antos
et al. 2006) and the sparrow the most abundant
species recorded in Hobart suburbs (Daniels & Kirk-
patrick 2006).
Despite some similarities, there were also important
differences between our results and those from large
cities. For example, in their Sydney study, Parsons
et al. (2006) recorded the widespread occurrence of
the pied currawong (Strepera graculina) and noisy
miner (Manorina melanocephala) – species that are
likely to suppress the populations of smaller birds
through predation and competition. Yet, we rarely
recorded these species in our neighbourhoods. While
some towns (e.g. Mildura) are outside the distribution
of the pied currawong, its infrequent occurrence in
other locations is surprising, although may reflect a
lack of tree cover in more urbanized neighbourhoods.
Similarly, noisy miners appeared to be confined to the
few neighbourhoods in our study that contained a
vegetation structure similar to open woodland. The
result for noisy miners is significant because this
species is common in other major urban centres
throughout eastern Australia (e.g. Brisbane – Sewell &
Catterall 1998; Hobart – Daniels & Kirkpatrick 2006;
Melbourne – Fitzsimons et al. 2011) and in urbanizing
regions (e.g. south-east Queensland – Kath et al.
2009) and is considered to exclude smaller birds
from these locations through aggressive defence of
territories. The presence of shrub vegetation may limit
the abundance of noisy miners as they prefer to forage
on open ground, and also mitigate their impact on
smaller birds by providing these birds with cover (Kath
et al. 2009). Therefore, the management of native
118 G. W. LUCK ET AL.
© 2012 The Authorsdoi:10.1111/j.1442-9993.2012.02383.x
Austral Ecology © 2012 Ecological Society of Australia
vegetation in urban areas should focus on providing
adequate tree and shrub cover to promote a diverse
bird community.
Managing urban environments for native fauna
The importance of small-scale vegetation characteris-
tics in influencing bird communities, as found in our
study and elsewhere, highlights the important role of
garden and streetscape management in conserving
urban birds. This suggests that both local government
and individual householders can contribute to manag-
ing urban environments to benefit native animals.
Householder behaviour has a strong influence on
garden vegetation characteristics and is increasingly
recognized as an important influential factor for neigh-
bourhood biodiversity (Daniels & Kirkpatrick 2006;
Goddard et al. 2009). However, residential lots are
mostly small and a single property generally cannot
support a diverse and viable fauna community.
Goddard et al. (2009) argued for the need to synchro-
nously manage a collection of neighbouring proper-
ties, while also considering the place of household
gardens in the broader landscape (e.g. as stepping
stones between conservation reserves). Promoting
desirable conservation outcomes through landholder
behaviour may be facilitated through incentive pro-
grammes promoted by local governments and non-
governmental organizations or by taking advantage of
social trends and the desire of individuals to conform
to social norms. For example, Nassauer et al. (2009)
demonstrated that in ex-urban landscapes in the USA,
an individual’s preference regarding their household
surrounds was influenced by neighbourhood cultural
norms and neighbourhood appearance. Warren et al.
(2008) suggested that the land management actions
of an individual householder may be influenced by
neighbourhood norms and a desire to ‘keep up with
the Joneses’ (also see Grove et al. 2006) resulting in
spatial autocorrelation of gardening behaviour across
neighbourhoods.
While these results suggest management agencies
may be able to target particular landholders willing to
adopt desirable garden management approaches, who
in turn may influence the behaviour of their neigh-
bours, this phenomenon requires much more research.
Understanding the complex interactions among the
natural, built and socio-economic characteristics of
urban neighbourhoods, and householder behaviour, is
important to achieving successful conservation out-
comes in urban areas and improving neighbourhoods
for both human and non-human residents.
ACKNOWLEDGEMENTS
This project was funded by an Australian Research
Council Discovery Grant (DP0770261) to GWL.
Thanks to Simon McDonald from the Charles Sturt
University Spatial Data Analysis Network for assistance
with the geographic information systems analyses,Tara
Martin and two referees for constructive comments on
the manuscript, and the local councils and residents of
Albury, Ballarat, Bathurst, Bendigo, Dubbo,Goulbur n,
Griffith, Mildura, Nowra, Orange, Queanbeyan, Shell-
harbour, Shepparton,Traralgon,WaggaWagga,Wanga-
ratta,Warrnambool and Wodonga for their support.
REFERENCES
Antos M. J., Fitzsimons J. A., Palmer G. C. & White J. G. (2006)
Introduced birds in urban remnant vegetation: does size
really matter? Austral Ecol. 31, 254–61.
Buckland S. T., Anderson D. R., Burnham K. P., Laake J. L.,
Borchers D. L. & Thomas L. (2001) Introduction to Distance
Sampling: Estimating Abundance of Biological Populations.
Oxford University Press, Oxford.
Burghardt K.T., Tallamy D.W. & Shriver W. G. (2009) Impact of
native plants on bird and butterfly biodiversity in suburban
landscapes. Conserv. Biol. 23, 219–24.
Burnham K. P. & Anderson D. R. (2002) Model Selection
and Multimodel Inference: A Practical Information-Theoretic
Approach, 2nd edn. Springer-Verlag, New York.
Catterall C. P., Green R. J. & Jones D. N. (1989) Occurrence of
birds in relation to plants in a subtropical city. Aust.Wildl.
Res. 16, 289–305.
Chace J. F. & Walsh J. J. (2006) Urban effects on native avifauna:
a review. Landsc. Urban Plan. 74, 46–69.
Chamberlain D. E., Cannon A. R. & Toms M. P. (2004) Asso-
ciations of garden birds with gradients in garden habitat and
local habitat. Ecography 27, 589–600.
Clergeau P., Jokimäki J. & Savard J.-P. L. (2001) Are urban bird
communities influenced by the bird diversity of adjacent
landscapes? J. Appl. Ecol. 38, 1122–34.
Daniels G. D. & Kirkpatrick J. B. (2006) Does variation in
garden characteristics influence the conservation of birds in
suburbia? Biol. Conserv. 133, 326–35.
Evans K. L., Newson S. E. & Gaston K. J. (2009) Habitat
influences on urban avian assemblages. Ibis 151, 19–39.
Fernández-Juricic J. (2000) Avifaunal use of wooded streets in an
urban landscape. Conserv. Biol. 14, 513–21.
Fitzsimons J. A., Antos M. J. & Palmer G. C. (2011) When more
is less: urban remnants support high bird abundance but
diversity varies. Pac. Conserv. Biol. 17, 97–109.
French K., Major R. & Hely K. (2005) Use of native and exotic
garden plants by suburban nectarivorous birds. Biol.
Conserv. 121, 545–59.
Fuller R. A., Warren P. H., Armsworth P. R., Barbosa O. &
Gaston K. J. (2008) Garden bird feeding predicts the struc-
ture of urban avian assemblages. Divers. Distrib. 14, 131–7.
Garden J., McAlpine C., Peterson A., Jones D. & Possingham H.
(2006) Review of the ecology of Australian urban fauna:
a focus on spatially explicit processes. Austral Ecol. 31, 126–
48.
Gaston K. J., ed. (2010) Urban Ecology. Cambridge University
Press, Cambridge.
Gaublomme E., Hendrickx F., Dhuyvetter H. & Desender K.
(2008) The effects of forest patch size and matrix type on
changes in carabid beetle assemblages in an urbanized
landscape. Biol. Conserv. 141, 2585–96.
URBAN BIRD COMMUNITIES 119
© 2012 The Authors doi:10.1111/j.1442-9993.2012.02383.x
Austral Ecology © 2012 Ecological Society of Australia
Gehrt S. D. & Chelsvig J. E. (2003) Bat activity in an urban
landscape: patterns at the landscape and microhabitat scale.
Ecol. Appl. 13, 939–50.
Goddard M. A., Dougill A. J. & Benton T. G. (2009) Scaling
up from gardens: biodiversity conservation in urban
environments. Trends Ecol. Evol. 25, 90–8.
Green R. J. (1984) Native and exotic birds in a suburban habitat.
Aust.Wildl. Res. 11, 181–90.
Grimm N. B., Faeth S. H., Golubiewski N. E. et al. (2008) Global
change and the ecology of cities. Science 319, 756–60.
Grove J. M., Troy A. R., O’Neil-Dunne J. P. M., Burch W. R.,
Cadenasso M. L. & Pickett S.T. A. (2006) Characterization
of households and its implications for the vegetation of
urban ecosystems. Ecosystems 9, 578–97.
Hennings L. A. & Edge W. D. (2003) Riparian bird community
structure in Portland, Oregon: habitat, urbanization, and
spatial scale patterns. Condor 105, 288–302.
Hodgson P., French K. & Major R. E. (2006) Comparison of
foraging behaviour of small, urban-sensitive insectivores in
continuous woodland and woodland remnants in a subur-
ban landscape. Wildl. Res. 33, 591–603.
Hope D., Gries C., Zhu W. et al. (2003) Socioeconomics drives
urban plant diversity. Proc. Natl Acad. Sci. U.S.A. 100, 8788–
92.
Jokimäki J. & Kaisanlahti-Jokimäki M.-L. (2003) Spatial similar-
ity of urban bird communities: a multiscale approach.
J. Biogeogr. 30, 1183–93.
Kath J., Maron M. & Dunn P. K. (2009) Interspecific competi-
tion and small bird diversity in an urbanizing landscape.
Landsc. Urban Plan. 92, 72–9.
Kinzig A. P., Warren P., Martin C., Hope D. & Katti M. (2005)
The effects of human socioeconomic status and cultural
characteristics on urban patterns of biodiversity. Ecol. Soc.
10, 23. [online] [Cited 31 March 2012.] Available from
URL: http://www.ecologyandsociety.org/vol10/iss1/art23/
Krebs C. J. (1999) Ecological Methodology, 2nd edn. Addison
Wesley Longman, Menlo Park, CA.
Lenz M. (1990) The breeding bird communities of three Can-
berra suburbs. Emu 90, 145–53.
Liu J., Daily G. C., Ehrlich P. R. & Luck G. W. (2003) The effect
of household dynamics on resource consumption and
biodiversity. Nature 421, 530–3.
Loss S. R., Ruiz M. O. & Brawn J. D. (2009) Relationships
between avian diversity, neighbourhood age, income, and
environmental characteristics of an urban landscape. Biol.
Conserv. 142, 2578–85.
Luck G. W. & Smallbone L. T. (2011) The impact of urbaniza-
tion on taxonomic and functional similarity among bird
communities. J. Biogeogr. 38, 894–906.
Luck G. W., Smallbone L. T. & O’Brien R. (2009) Socio-
economics and vegetation change in urban ecosystems: pat-
terns in space and time. Ecosystems 12, 604–20.
McKinney M. L. (2006) Urbanization as a major cause of biotic
homogenization. Biol. Conserv. 127, 247–60.
McKinney M. L. (2008) Effects of urbanization on species rich-
ness: a review of plants and animals. Urban Ecosyst. 11,
161–76.
Martin C. A., Warren P. S. & Kinzig A. P. (2004) Neighborhood
socio-economic status is a useful predictor of perennial
landscape vegetation in residential neighborhoods and
embedded small parks of Phoenix, AZ. Landsc. Urban Plan.
69, 355–68.
Melles S. (2005) Urban bird diversity as an indicator of social
diversity and economic inequality in Vancouver, British
Columbia. Urban Habitats 1, 25–48.
Miller J. R. & Cale P. (2000) Behavioral mechanisms and habitat
use by birds in a fragmented agricultural landscape. Ecol.
Appl. 10, 1732–48.
Miller J. R., Wiens J. A., Hobbs N. T. & Theobald D. M. (2003)
Effects of human settlement on bird communities in
lowland riparian areas of Colorado (USA). Ecol. Appl. 13,
1041–59.
Nassauer J. I., Wang Z. & Dayrell E. (2009) What will the
neighbors think? Cultural norms and ecological design.
Landsc. Urban Plan. 92, 282–92.
Oneal A. S. & Rotenberry J. T. (2009) Scale-dependent habitat
relationships of birds in riparian corridors in an urbanizing
landscape. Landsc. Urban Plan. 92, 264–75.
Palmer G. C., Fitzsimons J. A., Antos M. J. & White J. G. (2008)
Determinants of native avian richness in suburban remnant
vegetation: implications for conservation planning. Biol.
Conserv. 141, 2329–41.
Parsons H., Major R. E. & French K. (2006) Species interactions
and habitat associations of birds inhabiting urban areas of
Sydney, Australia. Austral Ecol. 31, 217–27.
Pennington D. N., Hansel J. & Blair R. B. (2008) The conser-
vation value of urban riparian areas for landbirds during
spring migration: land cover, scale, and vegetation effects.
Biol. Conserv. 141, 1235–48.
Sewell S. R. & Catterall C. P. (1998) Bushland modification and
styles of urban development: their effects on birds in south-
east Queensland. Wildl. Res. 25, 41–63.
Thomas L., Buckland S.T., Rexstad E.A. et al. (2010) Distance
software: design and analysis of distance sampling surveys
for estimating population size. J. Appl. Ecol. 47, 5–14.
Thomas L., Laake J. L., Strindberg S. et al. (2006) Distance 5.0.
Release ‘2’1. Research Unit for Wildlife Population Assess-
ment, University of St. Andrews, Fife. [Cited 31 March
2012.] Available from URL: http://www.ruwpa.st-and.ac.uk/
distance/
United Nations (2010) World Urbanization Prospects: The 2009
Revision. Department of Economic and Social Affairs,
United Nations, New York.
Warren P. S., Lerman S. B. & Charney N. D. (2008) Plants of a
feather: spatial autocorrelation of gardening practices in
suburban neighbourhoods. Biol. Conserv. 141, 3–4.
Watson D. M. (2003) The ‘standardized search’: an improved
way to conduct bird surveys. Austral Ecol. 28, 515–25.
White J. G., Antos M. J., Fitzsimons J. A. & Palmer G. C. (2005)
Non-uniform bird assemblages in urban environments: the
influence of streetscape vegetation. Landsc. Urban Plan. 71,
123–35.
Young K. M., Daniels C. B. & Johnston G. (2007) Species of
street tree is important for southern hemisphere bird trophic
guilds. Austral Ecol. 32, 541–50.
Zuur A. F., Leno E. N. & Elphick C. S. (2010) A protocol for
data exploration to avoid common statistical problems.
Methods Ecol. Evol. 1, 3–14.
SUPPORTING INFORMATION
Additional Supporting Information may be found in
the online version of this article:
Appendix S1. Supporting methods and results.
Appendix S2. All models explaining variation in
urban bird communities across south-eastern
Australia.
120 G. W. LUCK ET AL.
© 2012 The Authorsdoi:10.1111/j.1442-9993.2012.02383.x
Austral Ecology © 2012 Ecological Society of Australia
