Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Biological Conservation 142 (2009) 2578–2585 Contents lists available at ScienceDirect Biological Conservation journal homepage: www.elsevier.com/locate/biocon Relationships between avian diversity, neighborhood age, income, and environmental characteristics of an urban landscape Scott R. Loss a,*, Marilyn O. Ruiz b, Jeffrey D. Brawn a a b University of Illinois, Department of Natural Resources and Environmental Sciences, Shelford Vivarium, 606 E. Healey St., Champaign, IL 61820, USA University of Illinois, Department of Pathobiology, 2001 South Lincoln Avenue, Urbana, IL 61802, USA a r t i c l e i n f o Article history: Received 17 December 2008 Received in revised form 8 May 2009 Accepted 4 June 2009 Available online 5 July 2009 Keywords: Birds Biodiversity conservation Chicago Housing age Landscape ecology Urbanization a b s t r a c t Maintaining biodiversity in urbanizing landscapes has become a top conservation priority. We examined variation in bird communities across a diverse array of urban and suburban neighborhoods in the Chicago, Illinois, metropolitan region. Rather than taking the usual approach of focusing solely on natural features of the urban landscape, we investigated how urban bird communities were related to neighborhood age and income, as well as environmental characteristics. We found that median housing age was strongly related to avian species richness, with newer neighborhoods supporting more species. Housing age was an important correlate of abundance for several species as well as abundance of exotic, migratory, and non-migratory species groups. Per capita income was inversely related to richness of native bird species and positively related to exotic richness. Total richness was higher in urban sites with undeveloped patches and heterogeneous land cover types; moreover, richness decreased with increasing distance from natural areas greater than 1 km2. Our findings suggest that bird richness is enhanced both by small patches of natural land within the urban matrix and by close proximity to large natural preserves. Furthermore, these results suggest that investigating a combination of abiotic and environmental features of the built landscape, rather than focusing solely on environmental features, may provide a more complete understanding of the factors influencing avian diversity in human-dominated landscapes. Ó 2009 Elsevier Ltd. All rights reserved. 1. Introduction Conservation of biodiversity in urban landscapes has emerged as a top priority (Miller and Hobbs, 2002). Seventy percent of the world’s population is expected to live in urban areas by the year 2050 (United Nations, 2007), and 80% of the United States population now lives in urban and suburban areas (U.S. Census Bureau, 2000; Globalis, 2009). The rate of land conversion to urban uses is occurring faster than human population growth. The population of Chicago, Illinois, for example, increased by 38% from 1950 to 1990, while its area of land cover increased by 124% (O’Meara, 1999). These trends are contributing to biotic homogenization, the increased dominance of a few human-adapted species at the expense of locally unique biota (McKinney and Lockwood, 1999; McKinney, 2006). Compared to adjacent natural areas, bird communities in urban settings are generally characterized by low species richness and high total density or biomass (Emlen, 1974; Beissinger and * Corresponding author. Present address: University of Minnesota, Conservation Biology Graduate Program, 1980 Folwell Avenue, St. Paul, MN 55108, USA. Tel.: +1 011 612 624 4796; fax: +1 011 625 5299. E-mail addresses: [email protected] (S.R. Loss), [email protected] (M.O. Ruiz), [email protected] (J.D. Brawn). 0006-3207/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocon.2009.06.004 Osborne, 1982; Crooks et al., 2004; Sandstrom et al., 2006; Chapman and Reich, 2007; Caula et al., 2008). Intermediate levels of suburban development are often associated with a peak in bird richness, due to the presence of both urban-avoiding native species and urban-adapted exotics (Blair, 1996; Blair and Johnson, 2008). Species common to urban settings, such as exotics, are thought to have plentiful food and experience low predation pressure, thus leading to high reproductive success and higher density (Marzluff, 2001; Shochat et al., 2004). The habitat loss and increased human density associated with urbanization leads to homogenization of avifaunas among urban areas in regions that are otherwise ecologically distinct (Blair, 2001; Clergeau et al., 2006; Blair and Johnson, 2008; Sorace and Gustin, 2008). Studies of urban bird communities typically examine the influence of natural environmental factors, such as small-scale vegetation features (Sewell and Catterall, 1998; Daniels and Kirkpatrick, 2006a,b) and the size and spatial arrangement of natural habitat patches (Donnelly and Marzluff, 1998; Mortberg, 2001; Melles et al., 2003; Campbell, 2009; Evans et al., 2009). These types of studies have contributed greatly to our understanding of avian habitat relationships in urban areas. However, non-environmental factors associated with urban areas, such as landscape planning, structural and architectural features of neighborhoods, and socioeconomic status of urban residents, also have an important influ- S.R. Loss et al. / Biological Conservation 142 (2009) 2578–2585 ence on urban habitat. For example, both housing age and socioeconomic status are related to the amount and structure of native and ornamental vegetation (Hope et al., 2003; Martin et al., 2004; Smith et al., 2005). Despite these observations, studies of the relationship between bird communities and non-environmental factors are rare (but see Kinzig et al., 2005; Melles, 2005). Furthermore, no such studies have been conducted in the Midwestern United States. Clarification of such relationships will be useful to conservation planners, urban planners, and land managers, since databases of non-environmental and socio-economic data are often accessible via the internet. In addition, these data present an opportunity to extend understanding of the urban characteristics (both environmental and abiotic) that influence urban bird diversity. We investigated the possible relationships among environmental landscape characteristics (e.g., land cover heterogeneity), housing age and income, and variation in avian richness, by studying a set of neighborhoods ranging from high density urban to outer suburban in the Chicago, Illinois, metropolitan region. This region includes more than 90,000 hectares of preserved natural land (Chicago Wilderness, 2009) and is ideal for an analysis of urban and suburban bird communities in relation to a human-dominated landscape that contains an extensive network of preserves. We sought to: 1) determine whether urban bird diversity is related to neighborhood age and income in addition to environmental characteristics of the landscape, 2) assess whether neighborhood age and income are useful predictors for other aspects of avian community structure (i.e., relative density of individual species and richness and density of species groups), and 3) investigate whether the amount and proximity of surrounding natural habitat influences the richness of bird communities within the urban/suburban matrix. The impetus for these objectives was to inform avian conservation efforts in human-dominated landscapes while more broadly contributing to an understanding of the relationship between human-built environments and biological systems. 2. Materials and methods 2.1. Study area We selected 42 total study sites in the Chicago metropolitan area to represent a diverse set of urban neighborhoods with varying proximity to undeveloped lands (Fig. 1). All sites were originally developed for a study of broad and fine-scale urban and ecological factors influencing West Nile virus transmission (Loss et al., 2009). The wide array of urban typologies sampled also lent itself to an analysis of the abiotic and environmental landscape determinants of avian community structure. Below we describe how broad and fine-scale sites were selected. Sites were different between broad and fine-scale categories with regard to size and the process of site selection. However, all other methods, including bird surveys and landscape analysis, were equivalent across all sites. Sixteen of the sites (hereafter, ‘‘metro” sites) represent the twocounty Chicago metropolitan area (Cook and DuPage Counties) and were selected using an urban classification technique from a previous broad-scale study of West Nile virus (WNV; Ruiz et al., 2007). Random selection of census tracts, stratified by general urban classifications (outer suburban, inner suburban, high density urban upper income, high density urban lower income, and urban open space) resulted in a diverse representation of urban neighborhoods across the metropolitan area. This suite of sites was supplemented by two more sites in the WNV high risk areas of 2002 (Ruiz et al., 2004). These sixteen sites (mean area = 4406 ha) comprised from one to three census tracts. 2579 We also selected 26 smaller sites (hereafter, ‘‘south-side” sites) to represent Chicago’s south-side urban/suburban interface in an intensive fine-scale study of WNV ecology (Loss et al., 2009). Twenty-one of these sites were classified as residential and were made up of 20–30 contiguous and relatively homogeneous 150 m hexagons (mean area of sites = 585 ha). The hexagons were classified as residential if >35% of land use was ‘‘low/medium density urban” using a published digital land cover map of 30 m resolution from the years 1999 to 2000 (Illinois Dept. of Agriculture, 2009). The residential hexagons were then classified as low, medium or high housing unit density (U.S. Census Bureau, 2000) and as <0.5 km or >0.5 km from an undeveloped area (i.e., >40% in a non-urban land use class). The final set of south-side sites represented combinations of the three urban types (Ruiz et al., 2007) present in this area, varieties of housing unit density, and distances to undeveloped areas. In addition, we selected five predominantly undeveloped sites (e.g., cemeteries) that were surrounded by an otherwise urban matrix. The goal of this entire process was to select a broad range of sites representing an array of human population densities, neighborhood ages, income levels, and landscape contexts. 2.2. Estimation of bird richness and abundance We conducted point surveys (Bibby et al., 2000) along routes consisting of 5–10 survey points, depending on size of the site, following previously published survey methods (Loss et al., 2009). We used street-level maps prior to site visits to evenly distribute survey points across sites. Survey points were established at least 0.5 km apart to prevent double counting of birds, in accordance with Breeding Bird Survey protocols (U.S. Geological Survey, 1998). We then verified point locations and relocated points that were inaccessible or noisy, owing to traffic. We conducted 5-min unlimited radius point counts (Reynolds et al., 1980) and estimated distance to each bird. Counts of longer duration were not used in order to allow for sampling of all survey points (185 total points in 2005; 145 points in 2006) twice during each breeding season. All censuses were conducted between 0.5 h before sunrise and 4.0 h after sunrise (05:30–10:00) on relatively calm days with no precipitation. We surveyed each study site once in June and once in July to coincide with the peak avian breeding season in the Chicago area. South-side sites were surveyed in 2005 and 2006 while metro sites were surveyed only in 2005, owing to logistic constraints. 2.3. Measurement of variables for landscape analysis We estimated median housing age, per capita income, and environmental variables using ArcMap 9.0 geographical information system (Environmental Systems Research Institute, Redlands, California). Landscape measurements encompassed the study site scale and did not include buffer analysis of surrounding land. Buffer measurements would not have been independent, due to the close proximity of some sites to each other (Hostetler and Knowles-Yanez, 2003). We selected candidate landscape variables (Table 1) based on results of previous studies that reported important associations with bird species richness, local densities, or both (Mortberg, 2001; Melles et al., 2003; Martin et al., 2004; Palomino and Carrascal, 2005). Environmental variables measured for each site included proportion of undeveloped land within the site, land cover heterogeneity, and average distance of sites to natural habitat preserves with area greater than 1 km2. We also measured percent imperviousness (i.e., amount of impermeable surface, such as pavement or buildings), which is inversely proportional to vegetative cover (U.S. Geological Survey, 2009). We used Illinois land cover data (Illinois Dept. of Agriculture, 2009) to classify cover types within each site. 2580 S.R. Loss et al. / Biological Conservation 142 (2009) 2578–2585 Fig. 1. Map of study area, including locations of ‘‘metro” (gray polygons) and ‘‘south-side” (black polygons) study sites, used for bird community landscape analysis, Chicago metropolitan area, Cook and DuPage Counties, Illinois, 2005–2006. Table 1 Mean, range, standard deviation, and source of independent landscape variables measured for bird community landscape analysis, Chicago, Illinois, 2005–2006. Variable Mean Range St. Dev. Source Land cover heterogeneity Imperviousness (%) Median housing age (years) Per capita income ($) Proportion undeveloped Distance to 1 km2 natural preserve (m) 1.01 45.36 45.64 26912.99 0.20 2014.43 0.16–7.69 0.45–86.41 20.00–67.00 10081.00–85196.00 0.00–0.87 15.61–12150.20 1.20 17.24 13.33 12495.85 0.25 2589.53 Illinois Dept. of Agriculture U.S. Geological Survey (EROS) U.S. Census Bureau U.S. Census Bureau Illinois Dept. of Agriculture Illinois Dept. of Agriculture Classifications included agriculture, woodland, wetland, savanna, and low, medium, and high density urban. Using these classifications, we defined undeveloped land within sites to include agriculture, woodland, wetland, and savanna cover types. We estimated land cover heterogeneity within each site using all seven cover types, and by calculating Shannon’s diversity index. This metric simultaneously considers the number of land cover types and the evenness of distribution of cover types within a study site (Magurran, 2004). Since large sites are likely to include more land cover types, we standardized heterogeneity values by dividing by site area. Finally, we calculated average distance of study sites to the nearest natural area (forest, savanna, or grassland cover type) with area greater than 1 km2 using ArcMap’s Spatial Analyst extension. Whereas, study sites contained relatively small patches (typically <2 ha) of undeveloped land, such as municipal parks or railroad rights-of-way, the natural preserves were characterized by unbroken tracts of protected forest and parkland that provide significant habitat for breeding birds. S.R. Loss et al. / Biological Conservation 142 (2009) 2578–2585 To understand how urban morphology, landscape planning, and socio-economic status influence urban bird communities, we also measured median housing age and per capita income for each site (U.S. Census Bureau, 2000). Since sites comprised of greater than one census tract, we averaged census data across tracts for each site. Thus, the housing age and per capita income values used for the remainder of the analysis represent the study site scale. These measurements have been used in previous studies of urban vegetation and bird communities and have been suggested to vary predictably with avian diversity (Kinzig et al., 2005; Melles, 2005). Such metrics are readily available via the internet (U.S. Census Bureau, 2000) and may be indirectly related to environmental landscape measures, such as amount of natural area. 2.4. Data analyses We used ‘‘site” as the unit of replication rather than individual survey points, and we restricted analyses to bird species that were likely breeding in the region (i.e., migrants or rare species were not considered). Nor did we consider waterfowl, gulls, herons, raptors, and shorebirds, owing to their wide ranging habits and tendency for extensive daily movements in the region (See Appendix A for a full list of species included in the study). We derived density estimates based on detection functions using the program Distance 5.0 (Thomas et al., 2005). Estimated densities were natural log-transformed to meet distributional assumptions. Annual species richness was estimated by combining all bird species observed in the June and July surveys. We standardized species richness by number of survey points, since larger study sites contained more survey points and were thus likely to produce higher species richness estimates. Since there were no significant differences between 2005 and 2006 for any of the richness and density estimates of southside sites, we pooled data across years for these sites. We used multiple regression models to analyze factors affecting total species richness and total density. Multiple regression was also used to identify factors affecting densities of six common species that were present at all sites. To determine responses of different species groups to urbanization, we analyzed richness and density of native, exotic, migratory, and non-migratory species groups. Chicago’s avifauna includes several non-native bird species. Those that met the above criteria for inclusion in the study included rock pigeons (Columba livia), Eurasian collared-doves (Streptopelia decaocto), monk parakeets (Myiopsitta monachus), European starlings (Sturnus vulgarus), house sparrows (Passer domesticus), and house finches (Carpodacus mexicanus). We also used logistic regression models to examine factors contributing to the presence of 12 species that were frequently absent from sites. We used an information-theoretic approach to determine the best model(s) for species richness, density, and presence. Akaike’s Information Criteria (AIC) is based on the principle of parsimony, whereby the best model or model set is selected from a set of candidate models based on how well it explains the relationship with the observed data, while using relatively few parameters (Burnham and Anderson, 2002). To minimize multi-collinearity, we identified correlated independent variables using Pearson’s correlations. Each candidate model was therefore comprised of variables that were not strongly correlated (r < 0.5). Nine candidate models were developed for each dependent variable (total richness, total density, species group density and richness, and density or presence of individual species) based on previous analyses of urban bird density and richness (Mortberg, 2001; Melles et al., 2003; Kinzig et al., 2005; Palomino and Carrascal, 2005). We conducted preliminary data analysis to arrive at a set of likely candidate models for each dependent variable. A global model comprising all variables was also used for each analysis. 2581 We used mean square errors from multiple regression models and log-likelihood from logistic regression models to calculate Akaike’s Information Criterion, corrected for small sample sizes (AICc). We report DAIC values (the difference in AIC value between the best model and the model of interest) and AIC weights (the relative support for a model). If more than one model was strongly supported (i.e., DAIC = 0–2), we considered variables from all strong models, and averaged regression coefficients for independent variables that were represented in more than one strong model (Burnham and Anderson, 2002). We also compared variables by calculating variable importance weights (sum of AIC weights for all models containing the variable of interest; Burnham and Anderson, 2002). 3. Results We observed 76 bird species during the 2 years of the study. Fifty-four of these were probable breeders in the region. On average, 17 species (Range = 8–32, SE = 1.01) and 67 individuals per hectare (Range = 29.80–152.24, SE = 4.51) were observed at each site. The most common and widespread species were the house sparrow, American robin (Turdus migratorius), European starling, mourning dove (Zenaida macroura), and common grackle (Quiscalus quiscula). 3.1. Associations between species richness and housing age, income, and environmental measurements Spatial variation in total species richness was related to both housing age and to environmental measurements of the landscape, but not to per capita income (Table 2; also see Appendix A for full model selection results). Notably, richness was more strongly related to median housing age than to any of the environmental variables (Table 2). Newer neighborhoods were found to support greater richness than old developments (Fig. 2). The AIC importance weight for housing age (0.89; maximum possible weight = 1) further identified an important relationship between housing age and richness. Species richness was also greater in sites containing more undeveloped land and greater land cover heterogeneity, and richness decreased with increasing distance from the natural preserves with area greater than 1 km2 (Table 2, Fig. 2). AIC importance weights for land cover heterogeneity (0.34), undeveloped land (0.13), and distance to natural preserves (0.12), suggest that these variables were less important than housing age for explaining variation in total richness. Correlation results indicate that median housing age summarized a combination of environmental features of the urban landscape. This result may provide an explanation for the relationship between avian richness and housing age. For example, older neighborhoods had more impervious surface (r = 0.61), less undeveloped area (r = 0.43) and were further from the large natural preserves (r = 0.39). Housing age is thus positively related to intensity of urban development and isolation from natural habitats. Per capita income was not strongly related to any of the environmental variables we investigated, but declined slightly with increasing imperviousness (r = 0.22). 3.2. Associations between species groups and individual species and housing age, income, and environmental measures Richness and density of avian species groups (i.e., exotic vs. native and migratory vs. non-migratory) were related to housing age and income as well as to environmental measurements (Table 2). Densities of exotic and non-migratory groups were higher in areas 2582 S.R. Loss et al. / Biological Conservation 142 (2009) 2578–2585 Table 2 Estimated regression coefficients for landscape variables and model R2 values for best multiple regression models selected using AICc. Coefficients and R2 values for variables that appeared in more than one strongly supported model (DAIC 6 2) for a single dependent variable were averaged across models. Full AIC results are presented in Appendix A. Regression coefficients Variable Intercept Richness Total Native Exotic Migratory Nonmigratory 2.25 1.71 0.30 1.21 0.70 Density Total Native Exotic Migratory Nonmigratory 3.31 2.41 1.99 2.05 3.05 Land cover heterogeneity 0.10 Imperviousness – – 2.16 0.02 0.01 0.06 0.02 – 0.08 0.06 – – 0.12 – – – – Median housing age Per capita income – – – 0.01 1.03 0.30 – – – 0.04 2.12 – 0.89 Proportion undeveloped Distance to natural 0.29 0.50 0.50 1.40 – 0.02 – – 0.16 0.43 0.17 – – – – – 0.45 0.84 2.56 0.92 1.46 0.01 – – – – – 0.12 – Model R2 0.36 0.32 0.18 0.35 0.06 0.07 0.39 0.44 0.46 0.50 Fig. 2. Relationship between avian species richness and land cover heterogeneity, median housing age, proportion of undeveloped land, and distance to natural preserves >1 km2, across 42 sites in the Chicago metropolitan area, 2005–2006. Richness indicates average number of species recorded per survey point. with older housing. Richness of migratory species was lower in older developments. Neighborhoods with high per capita income had more exotic species (imp. wt. = 0.19) while lower income sites were characterized by more native (imp. wt. = 0.27) and migratory species (imp. wt. = 0.35; Table 2). Variable importance weights from best models for species groups were lower than for the total richness model, suggesting less support for these relationships. The best models for richness of non-migratory species were uninformative (R2 = 0.06). Housing age and income also explained variation in density and presence of individual bird species. Model R2 values were greater than 0.30 for most species (Table 3). Per capita income was represented in the best models for 10 species and median housing age was represented in the best models for 9 species, when considering 2583 S.R. Loss et al. / Biological Conservation 142 (2009) 2578–2585 Table 3 Estimated regression coefficients for landscape variables and model R2 values for best individual species models selected using AICc. Multiple regression was used for density models of common species and logistic regression was used for presence/absence models of uncommon species. Coefficients and R2 values for variables that appeared in more than one strongly supported model (DAIC 6 2) for a single dependent variable were averaged across models. Regression coefficients Variable Intercept Density American robin House sparrow Northern cardinal European starling Common grackle Mourning dove 1.72 2.86 0.91 1.16 0.09 0.30 Presence/absence Blue jay Black-capped chickadee Gray catbird Baltimore oriole Red-eyed vireo Indigo bunting Red-winged blackbird Rock pigeon Chipping sparrow Downy woodpecker House wren Brown-headed cowbird Song sparrow 0.46 1.83 4.16 0.41 6.61 7.62 3.38 7.43 7.99 4.97 1.89 1.42 7.16 Land cover heterogeneity 0.04 – – 0.76 0.01 – Imperviousness Median housing age Per capita income – – – – – – – – – – 0.01 0.01 0.30 0.28 – – 1.44 3.37 – – – 0.15 0.77 0.07 – 0.09 0.12 0.05 – 0.79 – – 1.13 0.28 0.10 – – – species models with R2 P 0.20. Densities of native species, including black-capped chickadees (Poecile atricapillus), Baltimore orioles (Icterus galbula), red-winged blackbirds (Agelaius phoeneceus), song sparrows (Melospiza melodia) and chipping sparrows (Spizella passerina) tended to be greater in new developments. Density of the exotic house sparrow, the most abundant species recorded (comprising 49% of total bird density), was highest in older neighborhoods. 4. Discussion We found that total species richness and the abundances of several common species were strongly related to median housing age. Moreover, housing age was related to other aspects of avian community structure, including richness of the migratory species group and density of exotic and non-migratory groups. In general, as neighborhoods aged, species richness eroded and bird communities shifted from being primarily composed of native and migratory species to increased dominance by exotic and non-migratory species. We also observed that proximity to large natural preserves enhances avian richness. For example, distributions of some common species, including mourning doves and common grackles were independent of distance, while most neotropical migrants, including scarlet tanagers (Piranga olivacea), eastern wood-pewees (Contupus virens), and great crested flycatchers (Myiarchus crinitus) occurred exclusively at sites within 0.5 km of large natural preserves. 4.1. Value of housing age and income metrics for urban bird studies Our results suggest the potential utility of housing age for describing and predicting avian diversity in urban areas. In contrast to our findings, previous studies have reported that older neighborhoods contain more bird species, owing to more diverse and complex vegetation cover compared to new developments (Vale and Vale, 1976; Munyenyembe et al., 1989; Palomino and Carrascal, 2005; but see Hope et al., 2003; Martin et al., 2004). The nature of the association between avian richness and housing age may depend on landscape and ecoregional contexts and the ex- – – – 0.11 Distance to natural 0.27 2.92 2.00 1.08 – 0.07 0.22 0.24 – 0.14 0.12 0.05 – – – – 0.51 0.41 0.09 – – 1.13 Proportion undeveloped 0.31 4.21 3.61 – – – 4.14 8.63 0.34 0.34 0.41 0.15 0.18 0.41 0.09 0.28 – 0.14 0.11 0.11 0.18 0.30 0.66 0.09 0.26 0.01 0.06 0.01 0.25 0.29 0.23 0.42 0.44 0.39 0.30 0.24 0.54 0.27 0.31 0.30 0.45 – – – – – – – 6.05 0.06 – 5.41 – 0.01 – 2.21 3.35 2.33 Model R2 0.20 0.11 0.21 tent to which development alters natural elements of avian habitats. In the Chicago area, newer developments (20–30 years old) were established in previously forested or mixed woodland-agricultural settings. Newer suburban developments that retained components of this natural habitat appear to support relatively great avian diversity. Subsequent urbanization, however, leads to increased habitat loss and erosion of habitat diversity (structurally and floristically, see also Walcott, 1974). We observed this phenomenon in Chicago, as the oldest neighborhoods are generally typified by high human density and few remaining patches of undeveloped land. In contrast, developments in formerly forested landscapes that do not retain natural habitat are characterized by low initial structural complexity and gradual maturation of forest structure over time. This pattern was noted in Baltimore, Maryland (Grove et al., 2006) and in isolated Chicago neighborhoods greater than 50 years old that now contain rich vegetative communities with complex structure. The implication of ours and previous findings is that planners and habitat managers that seek to retain significant natural components of vegetative habitat during all stages and ages of urban development may be able to minimize the decay of species richness and diversity that may otherwise occur. As the Chicago metropolitan area continues to expand into surrounding agricultural land, the newest housing developments (<20 years old) are established in landscapes dominated by rowcrop agriculture. Since our study did not include these exurban areas, the relationship between housing age, vegetation, and avian richness remains unclear for them. However, we would expect to observe increased avian richness over time, as the initially sparsely vegetated neighborhoods would likely develop greater vegetation cover and structure due to planting of ornamental shrubs and trees (Hope et al., 2003). Further research is needed to determine how the relationship between housing age and avian richness varies with regional and landscape context. Though per capita income was unrelated to total avian richness, the result was likely due to offsetting relationships of native and exotic species groups. We found that high-income areas contain more exotic bird species, while low-income neighborhoods support more native and migratory species. Income is likely related 2584 S.R. Loss et al. / Biological Conservation 142 (2009) 2578–2585 to environmental characteristics that were not measured in our study. For example, fine-scale vegetation characteristics, such as proportion of native and exotic plant species, density and structure of the understory, and plant diversity, have been linked to neighborhood socio-economic status (Hope et al., 2003; Martin et al., 2004). Importantly, our results differ from previous studies that detected a positive relationship between socio-economic status and avian richness (Kinzig et al., 2005; Melles, 2005). Mechanisms responsible for our conflicting findings are unclear. Future research should continue to identify and measure intermediate variables to clarify the mechanistic link between income and avian richness. 4.2. Importance of small undeveloped patches and large habitat preserves in an urban landscape Even though most study sites were distant from large nature preserves (greater than 2 km, on average) and contained only small undeveloped patches (less than 20% of total land cover, on average), such as municipal parks or railroad rights-of-way, it was clear that these patches contributed to enriching bird richness. In addition to a positive relationship with total richness, and relationships with several species groups, the proportion of undeveloped land was positively related to density of twelve individual species (when considering species models with R2 P 0.20), more than any other variable. Species group responses to this measure varied, with density of native and migratory species increasing and density of exotic and non-migratory species decreasing with increased undeveloped area. The inverse relationship between undeveloped area and density of house sparrows, the most abundant species, was especially strong (Table 3). Undeveloped patches often contributed to increased avian richness, supporting species not typically associated with predominantly urban settings, such as indigo buntings (Passerina cyanea) and Baltimore orioles. Higher proportions of undeveloped land (40% or greater) tended to occur in newer neighborhoods. Notably, urban sites with the highest richness values were comparable to nature reserves in rural parts of north-central and northwestern Illinois (Robinson et al., 1997), when analyzing the same orders of birds considered here (i.e., Columbiformes, Coraciformes, Piciformes, Passeriformes). Though some species differed between these two areas, with more exotic species occurring in urban sites, native species richness of urban sites still approached that of natural preserves. Preservation efforts are usually focused on relatively pristine natural areas; however, these results suggest the potential value of small patches of appropriate habitat in urbanized landscapes for avian conservation. Though small habitat patches can function as ecological sinks, resulting in negative local scale population growth rates within a patch (Pulliam, 1988), some species associated with highly disturbed habitats have been found to breed successfully in small fragments of natural habitat (Brawn et al., 2002; Leston and Rodewald, 2006). Thus, further research should address nest success and productivity in urban habitat patches to determine the importance of urban sites for conservation of viable bird populations. Bird species have been suggested to ‘‘spillover” from small patches of natural habitat (15–136 ha) into directly adjacent urban neighborhoods in Chicago (Lussenhop, 1977). Our result that richness increased with increasing proximity to large patches of natural habitat suggests the possibility of a larger scale spillover effect from natural preserves into surrounding urban sites greater than one-half kilometer away. For example, densities of blue jays (Cyanocitta cristata), red-winged blackbirds, chipping sparrows, house wrens (Troglodytes aedon), and song sparrows, were highest within 0.5 km of large natural areas. Densities of these species decreased with increasing isolation from the reserves. Thus, large blocks of natural area may influence bird communities of surrounding urban areas. These findings agree with previous work suggesting that birds in urban settings respond to both local and landscape resource cues (Hostetler, 1999; Bakermans and Rodewald, 2006). 5. Conservation implications Vast databases of social and economic metrics are easily accessible online through the U.S. Census Bureau (http://www.census.gov), and they can be merged with spatial data using geographic information systems software. Used in combination with vegetation and habitat data, as well as existing avian point count data, these variables may facilitate identification of urban biodiversity hotspots and inform management that seeks to enhance or eradicate certain species groups (e.g., to increase diversity of migratory species or reduce density of exotics). Though our findings are likely applicable to cities in the Midwestern U.S., our results contrast previous similar work in other regions (Munyenyembe et al., 1989; Kinzig et al., 2005; Melles, 2005). Thus, further investigation is warranted to clarify the relationship between housing age, income, and avian community structure in other biomes and regions in the U.S. and globally. While large nature preserves may support large numbers of species, budgetary and spatial constraints often prevent conservation at such large scales. Small preserves, however, can also be useful for conserving naturally patchy habitats and species with isolated distributions (Schwartz, 1999). Moreover, a series of small preserves that captures a variety of land cover types can be especially favorable for conservation of many species (Schwartz and Van Mantgem, 1997). Our results further suggest that bird richness in urban areas is enriched by small habitat patches as small as 1–10 hectares. In addition to providing habitat for breeding birds, urban patches of natural and undeveloped habitat are important as stop-over sites for long-distance migrant birds (Brawn and Stotz, 2001; Pennington et al., 2008). This study suggests that assessment of a combination of nonenvironmental (e.g., housing age and income) and environmental measurements, rather than focusing solely on environmental factors, may provide a more complete understanding of the factors influencing avian diversity in an urban setting. Investigating relationships between these types of variables and biological patterns, including community structure of birds, as well as plants and other animal taxa, will determine whether such non-environmental variables are generally useful for understanding and predicting patterns of biological diversity in human-dominated landscapes. Acknowledgements We thank the Village of Oak Lawn for generously providing field lab space during the study, as well as the municipalities who cooperated with us during field research (Alsip, Blue Island, Burbank, City of Chicago, Dolton, Evanston, Evergreen Park, Harvey, Indian Head Park, Orland Park, Palos Hills, Western Springs). We also thank Judy Pollock of Audubon Chicago Region, who helped coordinate citywide counts, and Eric Secker, who assisted with avian surveys. We thank Robert Blair, Tony Goldberg, and Uriel Kitron, who provided valuable comments on earlier versions of the manuscript. This work was supported by the NSF/NIH program in the Ecology of Infectious Diseases (Grant 04-29124) and by the Jonathan Baldwin Turner Graduate Fellowship from the University of Illinois – College of Agriculture, Consumer, and Environmental Sciences. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.biocon.2009.06.004. S.R. Loss et al. / Biological Conservation 142 (2009) 2578–2585 References Bakermans, M.H., Rodewald, A.D., 2006. Scale-dependent habitat use of Acadian Flycatcher (Empidonax virens) in central Ohio. Auk 123, 368–382. Beissinger, S.R., Osborne, D.R., 1982. Effects of urbanization on avian community organization. Condor 84, 75–83. Bibby, C., Burgess, N.D., Hill, D.A., Mustoe, S.H., 2000. Bird census techniques, second ed. Academic Press, London. Blair, R.B., 1996. Land use and avian species diversity along an urban gradient. Ecol. Appl. 6, 506–519. Blair, R.B., 2001. Creating a homogeneous avifauna. In: Marzluff, J.M., Bowman, R., Donnelly, R.E. (Eds.), Avian Ecology and Conservation in an Urbanizing World. Kluwer Academic, Norwell, Massachusetts, pp. 459–486. Blair, R.B., Johnson, E.M., 2008. Suburban habitats and their role for birds in the urban-rural habitat network: points of local invasion and extinction? Landscape Ecol 23, 1157–1169. Brawn, J.D., Stotz, D.F., 2001. The importance of the Chicago region and the ‘‘Chicago Wilderness” initiative for avian conservation. In: Marzluff, J.M., Bowman, R., Donnelly, R.E. (Eds.), Avian Ecology and Conservation in an Urbanizing World. Kluwer Academic, Norwell, Massachusetts, pp. 509–522. Brawn, J.D., Heske, E.J., Herkert, J., Robinson, S.K., 2002. Restoration of small reserves may be valuable for certain types of birds. Ecol. Restor. 20, 63–64. Burnham, K.P., Anderson, D.R., 2002. Model Selection and Multimodel Inference: A Practical Information-theoretic Approach, second ed. Springer-Verlag, New York. Campbell, M.O., 2009. The impact of habitat characteristics on bird presence and the implications for wildlife management in the environs of Ottawa, Canada. Urban Forestry & Urban Greening 8, 87–95. Caula, S., Marty, P., Martin, J.L., 2008. Seasonal variation in species composition of an urban bird community in Mediterranean France. Landsc. Urb. Plan. 87, 1–9. Chapman, K.A., Reich, P.B., 2007. Land use and habitat gradients determine bird community diversity and abundance in suburban, rural and reserve landscapes of Minnesota, USA. Biol. Conserv. 135, 527–541. Chicago Wilderness. Chicago wilderness: A regional nature reserve. [online] URL: http://www.chicagowilderness.org/index.cfm. Accessed 4-30-2009. Clergeau, P., Croci, S., Jokimaki, J., Kaisanlahti-Jokimaki, M-L., Dinetti, M., 2006. Avifauna homogenisation by urbanisation: Analysis at different European latitudes. Biol. Conserv. 127, 336–344. Crooks, K.R., Suarez, A.V., Bolger, D.T., 2004. Avian assemblages along a gradient of urbanization in a highly fragmented landscape. Biol. Conserv. 115, 451–462. Daniels, G.D., Kirkpatrick, J.B., 2006a. Does variation in garden characteristics influence the conservation of birds in suburbia? Biol. Conserv. 133, 326–335. Daniels, G.D., Kirkpatrick, J.B., 2006b. Comparing the characteristics of front and back domestic gardens in Hobart, Tasmania, Australia. Landsc. Urb. Plan. 78, 344–352. Donnelly, R., Marzluff, J.M., 1998. Importance of reserve size and landscape context to urban bird conservation. Conserv. Biol. 18, 733–745. Emlen, J.T., 1974. An urban bird community in Tucson, Arizona: Derivation, structure, regulation. Condor 76, 184–197. Evans, K.L., Newson, S.E., Gaston, K.J., 2009. Habitat influences on urban avian assemblages. Ibis 151, 19–39. Globalis. An interactive world map, United States. [online] URL: http:// globalis.gvu.unu.edu/indicator_detail.cfm?IndicatorID=30&Country=US. Accessed 4-30-2009. Grove, J.M., Troy, A.R., O’Neille-Dunne, J.P.M., Burch Jr., W.R., Cadenasso, M.L., Pickett, S.T.A., 2006. Characterization of households and its implications for the vegetation of urban ecosystems. Ecosystems 9, 578–597. Hope, D., Gries, C., Zhu, W., Fagan, W.F., Redman, C.L., Grim, N.B., Nelson, A.L., Martin, C., Kinzig, A., 2003. Socioeconomics drive urban plant diversity. Proc. Natl. Acad. Sci. U.S.A. 100, 8788–8792. Hostetler, M., 1999. Scale, birds, and human decisions: a potential for integrative research in urban ecosystems. Landsc. Urb. Plan. 45, 15–19. Hostetler, M., Knowles-Yanez, K., 2003. Land use, scale, and bird distributions in the Phoenix metropolitan area. Landsc. Urb. Plan. 62, 55–68. Illinois Department of Agriculture. Illinois Gap Analysis Program, Land Cover Classification. [online] URL: http://www.agr.state.il.us/gis/pass/gapdata. Accessed 4-30-2009. Kinzig, A.P., Warren, P.S. Martin, C., Hope, D., Katti, M., 2005. The effects of human socioeconomic status and cultural characteristics on urban patterns of biodiversity. Ecol. And Soc. 10, 23. [online] URL: http:// www.ecologyandsociety.org/vol10/iss1/art23/. Leston, L.F.V., Rodewald, A.D., 2006. Are urban forests ecological traps for understory birds? An examination using Northern Cardinals. Biol. Conserv. 131, 566–574. Loss, S.R., Hamer, G.L., Walker, E.D., Ruiz, M.O., Goldberg, T.L., Kitron, U.D., Brawn, J.D., 2009. Avian host community structure and prevalence of West Nile virus in Chicago, Illinois. Oecologia 159, 415–424. Lussenhop, J., 1977. Urban cemeteries as bird refuges. Condor 79, 456–461. Magurran, A.E., 2004. Measuring Biological Diversity. Blackwell Publishing, Oxford. Martin, C.A., Warren, P.S., Kinzig, A.P., 2004. Neighborhood socioeconomic status is a useful predictor of perennial landscape vegetation in residential 2585 neighborhoods and embedded small parks of Phoenix, AZ. Landsc.Urb. Plan. 69, 355–368. Marzluff, J.M., 2001. Worldwide urbanization and its effects on birds. In: Marzluff, J.M., Bowman, R., Donnelly, R.E. (Eds.), Avian Ecology and Conservation in an Urbanizing World. Kluwer Academic, Norwell, Massachusetts, pp. 19–47. McKinney, M., 2006. Urbanization as a major cause of biotic homogenization. Biol. Conserv. 127, 247–260. McKinney, M.L., Lockwood, J.L., 1999. Biotic homogenization: a few winners replacing many losers in the next mass extinction. Trends in Ecology and Evolution 14, 450–453. Melles, S., Glenn, S., Martin, K., 2003. Urban bird diversity and landscape complexity: species-environment associations along a multi-scale habitat gradient. Conservat. Ecol. 7, 5. [online]. URL: http://www.consecol.org/vol7/ iss1/art5. Melles, S.J., 2005. Urban bird diversity as an indicator of human social diversity and economic inequality in Vancouver, British Columbia. Urb. Hab. 3, 25–48. Miller, J.R., Hobbs, R.J., 2002. Conservation where people live and work. Conserv. Biol. 16, 330–337. Mortberg, U.M., 2001. Resident bird species in urban forest remnants; landscape and habitat perspectives. Landsc. Ecol. 16, 193–203. Munyenyembe, F., Harris, J., Hone, J., Nix, H., 1989. Determinants of bird populations in an urban area. Austral Ecol. 14, 549–557. O’Meara, M., 1999. Reinventing cities for people and the planet. Worldwatch Paper 147. Worldwatch Institute, Washington, D.C. Palomino, D., Carrascal, L.M., 2005. Urban influence on birds at a regional scale: A case study with the avifauna of northern Madrid province. Landsc. Urb. Plan. 77, 276–290. Pennington, D.N., Hansel, J., Blair, R.B., 2008. The conservation value of urban riparian areas for landbirds during spring migration: Land cover, scale, and vegetation effects. Biological Conservation 141, 1235–1248. Pulliam, H.R., 1988. Sources, sinks, and population regulation. Am. Nat. 132, 652– 661. Reynolds, R.T., Scott, J.M., Nussbaum, R.A., 1980. A variable circular-plot method for estimating bird numbers. Condor 82, 309–313. Robinson, S.K., Brawn, J.D., Hoover, J.P., 1997. Effectiveness of small nature preserves for breeding birds. In: Schwartz (Ed.), Conservation in Highly Fragmented Landscapes. Chapman and Hall, New York, pp. 154–188. Ruiz, M.O., Tedesco, C., McTighe, T.J., Austin, C., Kitron, U., 2004. Environmental and social determinants of human risk during a West Nile virus outbreak in the greater Chicago area, 2002. Int. J. Health Geogr. 3, 8. Ruiz, M.O., Walker, E.D., Foster, E.S., Haramis, L.D., Kitron, U.D., 2007. Association of West Nile illness and urban landscapes in Chicago and Detroit. Int. J. Health Geogr. 6, 10. Sandstrom, U.G., Angelstam, P., Mikusinski, G., 2006. Ecological diversity of birds in relation to the structure of urban green space. Landsc. Urb. Plan. 77, 39–53. Schwartz, M.W., Van Mantgem, P.J., 1997. The value of small preserves in chronically fragmented landscapes. In: Schwartz (Ed.), Conservation in Highly Fragmented Landscapes. Chapman and Hall, New York, pp. 379–394. Schwartz, M.W., 1999. Choosing the appropriate scale of reserves for conservation. Annu. Rev. Ecol. Syst. 30, 83–108. Sewell, S.R., Catterall, C.P., 1998. Bushland modification and styles of development: their effects on birds in southeast Queensland. Wildlife Research 25, 41–63. Shochat, E., Lerman, S.B., Katti, M., Lewis, D.B., 2004. Linking optimal foraging behavior to bird community structure in an urban-desert landscape: Field experiments with artificial food patches. Am. Nat. 164, 232–243. Smith, R.M., Gaston, K.J., Warren, P.H., Thompson, K., 2005. Urban domestic gardens (V): relationships between landcover composition, housing and landscape. Landscape Ecol. 20, 235–253. Sorace, A., Gustin, M., 2008. Homogenisation processes and local effects on avifaunal composition in Italian towns. Acta Oecologica 33, 15–26. Thomas, L., Laake, J.L., Strindberg, S., Marques, F.F.C., Buckland, S.T., Borchers, D.L., Anderson, D.R., Burnham, K.P., Hedley, S.L., Pollard, J.H., Bishop, J.R.B., Marques, T.A., 2005. Distance 5.0. Release 5. Research Unit for Wildlife Population Assessment, University of St. Andrews, U.K. URL: http:// www.ruwpa.stand.ac.uk/distance/. United Nations. 2007. World Urbanization Prospects: The 2007 Revision Population Database. [online] URL: http://esa.un.org/unup/. Accessed 4-30-2009. U.S. Census Bureau. American Factfinder: 2000 Census of Population and Housing, Summary File 3. [online] URL: http://factfinder.census.gov. Accessed 4-30-2009. U.S. Geological Survey: North American Breeding Bird Survey. 1998 BBS Instructions. [online] URL: http://www.pwrc.usgs.gov/bbs/participate/ instructions.html. Accessed 4-30-2009. U.S. Geological Survey. Seamless Data Distribution System, Earth Resources Observation and Science. [online] URL: http://eros.usgs.gov/products/ landcover.html. Accessed 4-30-2009. Vale, T.R., Vale, G.R., 1976. Suburban bird populations in west-central California. J. Biogeogr. 3, 157–165. Walcott, C.F., 1974. Changes in bird life in Cambridge, Massachusetts from 1860 to 1964. Auk 91, 151–160.