ASSESSING WILDLIFE HABITAT CONTRIBUTIONS OF GREEN ROOFS IN URBAN LANDSCAPES IN MICHIGAN AND ILLINOIS, U.S.A.: MEASURING AVIAN COMMUNITY RESPONSE TO GREEN ROOF FACTORS By Carly J. Eakin A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Fisheries and Wildlife 2012 ABSTRACT ASSESSING WILDLIFE HABITAT CONTRIBUTIONS OF GREEN ROOFS IN URBAN LANDSCAPES IN MICHIGAN AND ILLINOIS, U.S.A.: MEASURING AVIAN COMMUNITY RESPONSE TO GREEN ROOF FACTORS By Carly J. Eakin Wildlife habitat degradation is a leading cause of biodiversity loss, and largely attributed to urbanization. Green roofs (vegetated roofs) have been identified as a technology having potential to provide wildlife habitat in urban areas by creating vegetation types. Vegetation structure and composition and green space cover were quantified for 12 green roofs and their surrounding landscapes in Michigan and Illinois in 2010 and 2011. Most vegetation variables, including vegetation height and herbaceous cover, were significantly different between intensive green roofs (0-15cm planting media depth) and extensive green roofs (>15cm planting media depth). Herbaceous cover was the dominant cover type on all green roofs. Shrub cover was present on extensive and intensive roofs, and tree and turf cover were only present on some intensive roofs. Green space analysis showed future green roof installations could increase green space area >300% in landscapes immediately surrounding study sites. Twenty-five noninvasive, native bird species were detected on green roofs, and the mean estimated species richness for each green roof (within the range of 36-40 species) was greater than in surrounding landscapes. Our results support the idea that green roof vegetation can contribute to wildlife habitat in urban areas and increase space for wildlife conservation. This information should encourage collaboration of green roof designers and natural resource managers to advance green roof installations towards holistic environmental sustainability that includes wildlife conservation. ACKNOWLEDGMENTS This project would not have been possible without the financial assistance provided by the College of Agriculture and Natural Resources Graduate Student Recruitment and Retention Fellowship Funds, the Sustainable Michigan Endowment Program (SMEP), and the George J. Wallace and Martha C. Wallace Endowed Scholarship. I am grateful to many individuals for support and assistance throughout the duration of this project. I would like to thank all green roof owners and personnel for their assistance and cooperation in arranging roof visits and in providing information pertaining to green roof installation and maintenance. Thanks to the students and faculty of the Green Roof Team who have fostered the enthusiasm and laid the ground work so a project like this could happen. I am grateful to be part of a group of people dedicated to improving the suite of environmental and societal benefits that green roofs and other environmentally sustainable technologies can provide. I am grateful to my committee members Dr. Brad Rowe, Dr. Joanne Westphal, and Dr. Gary Roloff for their advice and guidance throughout the duration of this project. I would like to thank Dr. Dan Linden who helped with data analysis for the second chapter and wrote the script for modeling. To Dr. Rique Campa, thank you for your patience, mentorship, all of your wisdom and for helping me become ‘multidisciplinary’. To my family and friends, thank you for supporting my efforts and believing in my dreams. Most of all, Chadrick, thank you for changing your path to walk with me in this adventure. iii TABLE OF CONTENTS LIST OF TABLES ............................................................................................................ vi LIST OF FIGURES .......................................................................................................... ix INTRODUCTION ...............................................................................................................1 OBJECTIVES ......................................................................................................................6 LITERATURE CITED ........................................................................................................8 CHAPTER 1 - VEGETATION CHARACTERISTICS OF GREEN ROOFS FOR WILDLIFE: HABITAT POTENTIAL IN MICHIGAN AND ILLINOIS, U.S.A. ................................13 Objectives ....................................................................................................................16 Methods .......................................................................................................................17 Site descriptions .........................................................................................17 Vegetation sampling ...................................................................................20 Aerial land cover ........................................................................................21 Data analysis ...............................................................................................21 Results .........................................................................................................................23 Green roof comparisons .............................................................................23 Surrounding landscape comparisons ..........................................................24 Discussion ...................................................................................................................26 Vegetation structure and composition ........................................................26 Bird communities .......................................................................................27 Wildlife conservation .................................................................................31 APPENDIX. .......................................................................................................................87 LITERATURE CITED ....................................................................................................119 CHAPTER 2 AVIAN RESPONSE TO GREEN ROOFS IN URBAN LANDSCAPES IN THE MIDWEST UNITED STATES. ..........................................................................................................125 Objectives ..................................................................................................................128 Methods .....................................................................................................................129 Site descriptions ........................................................................................129 Vegetation measurements..........................................................................130 Aerial land cover .......................................................................................132 Bird surveys ...............................................................................................132 Descriptive analysis and statistics .............................................................134 Results .......................................................................................................................137 Observed and estimated bird community structure and composition .......137 iv Discussion .................................................................................................................141 Bird use......................................................................................................141 Comparison of expected and observed bird species ..................................144 Bird community conservation potential ....................................................145 LITERATURE CITED ......................................................................................................186 CONCLUSIONS..............................................................................................................193 v LIST OF TABLES Table 1.1. Location, landscape classification, and year of installation of green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011 ........................................................34 Table 1.2. Monthly mean temperature (degrees Celsius) for green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011. Data is provided for green roofs study sites during the year each site was sampled ...........................................................................................36 Table 1.3. Monthly total precipitation in centimeters for green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011. Data is provided for green roofs study sites during the year each site was sampled ...........................................................................................38 Table 1.4. Characteristics for building structure, planting media, planted vegetation, maintenance, and human use for green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011 ............................................................................................................................40 Table 1.5. Maintenance regime, primary function, and accessibility of green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011 ........................................................42 Table 1.6. Information sources for characteristics of green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011. All study sites were a Downtown Chicago Park (DCP), Ford Truck Plant (FOR), McCormick Parking Structure (MCC), Aquascape Headquarters (AQU), Haworth Headquarters (HAW), Chicago Cultural Center (CCE), Chicago City Hall (CHA), Nature Museum (NAM), Michigan Avenue Structure (MIA), Schwab Rehabilitation Hospital (SCH), Gary Comer Youth Center (GCY), and the Plant and Soil Science Building at Michigan State University (PSS) ...................................................................................................43 Table 1.7. Means and standard errors for variables characterizing vegetation structure of green roofs in Michigan and Illinois, U.S.A., in 2010 and 2011. Green roofs are arranged by year and type. Probability levels reported were calculated with the Kruskal-Wallis one-way analysis of variance .........................................................................................................................................47 Table 1.8. Means and standard errors for early sampling periods, late sampling periods, and the entire sampling season for variables characterizing vegetation structure of each studied green roof, in Michigan and Illinois, U.S.A., in 2010 and 2011. Probability levels for comparisons between spring and summer within the same year were calculated with the Kruskal-Wallis oneway analysis of variance. ...............................................................................................................48 Table 1.9. Means and standard errors for variables characterizing vegetation structure of the landscape surrounding each green roof during early sampling periods, late sampling periods, and the entire sampling season, in Michigan and Illinois, U.S.A., in 2010 and 2011. Probability vi levels for comparisons between spring and summer within the same year were calculated with the Kruskal-Wallis one-way analysis of variance. .........................................................................60 Table 1.10. Percent cover for land cover variables characterizing surrounding landscapes of green roof study sites in Michigan and Illinois, U.S.A., in 2010 and 2011 ...................................66 Table A.1. List of planted species on the Downtown Chicago Park (DCP) green roof. ...............88 Table A.2. List of planted species on the Ford Truck Plant (FOR) green roof .............................97 Table A.3. List of planted species on the McCormick Parking Structure (MCC) green roof. ......98 Table A.4. List of planted species on the Aquascape Headquarters (AQU) green roof ..............100 Table A.5. List of planted species on the Haworth Headquarters (HAW) green roof .................101 Table A.6. List of planted species on the Chicago Cultural Center (CCE) green roof ................102 Table A.7. List of planted species on the Chicago City Hall (CHA) green roof .........................103 Table A.8. List of planted species on the Nature Museum (NAM) extensive green roof ...........110 Table A.9. List of planted species on the Nature Museum (NAM) intensive green roof. ...........112 Table A.10. List of planted species on the Michigan Avenue Structure (MIA) green roof. .......114 Table A.11. List of planted species on the Schwab Rehabilitation Hospital (SCH) green roof. .115 Table A.12. List of planted species on the Gary Comer Youth Center (GCY) green roof..........116 Table A.13. List of planted species on the Plant and Soil Science Building (PSS) green roof at Michigan State University. ..........................................................................................................118 Table 2.1. The location, building structure characteristics, planting media depth, planted vegetation and year of green roof installation of green roof study sites sampled in 2010 and 2011 in Illinois and Michigan, U.S.A. ..................................................................................................151 Table 2.2. The maintenance regime, primary function, and accessibility of green roof study sites sampled in 2010 and 2011 in Illinois and Michigan, U.S.A. .......................................................153 Table 2.3. Relative abundance of bird species observed on green roofs and in their surrounding landscapes in Michigan and Illinois, U.S.A. in 2010 and 2011. All study sites are a Downtown Chicago Park (DCP), Ford Truck Plant (FOR), McCormick Parking Structure (MCC), Aquascape Headquarters (AQU), Haworth Headquarters (HAW), Chicago Cultural Center (CCE), Chicago City Hall (CHA), Nature Museum (NAM), Michigan Avenue Structure (MIA), vii Schwab Rehabilitation Hospital (SCH), Gary Comer Youth Center (GCY), and the Plant and Soil Science Building at Michigan State University (PSS) .........................................................155 viii LIST OF FIGURES Figure 1.1. Range of scales at which birds may respond to green roof vegetation. The presence of green roof vegetation may invoke a response for some bird species selecting habitat at a broad scale (A – home range), for some at a finer level of patch selection (B - feeding area), or at finer level of microhabitat selection (C - one of many feeding sites) ....................................................68 Figure 1.2. Map of the Northeast United States depicting study site locations in 2010 and 2011. Location A sites are in the greater Chicago area: the western location marker is a site in St. Charles, Illinois and the eastern marker represents all sites within Chicago, Illinois. Location B is a site in Holland, Michigan, location C is a site in East Lansing, Michigan, and location D is a site in Dearborn, Michigan. ...........................................................................................................69 Figure 1.3. Map of northeast Illinois and southern Michigan depicting study site locations in 2010 and 2011. Location A sites are in the greater Chicago, Illinois area, location B is in Holland, Michigan, location C is in East Lansing, Michigan, and location D is in Dearborn, Michigan ........................................................................................................................................70 Figure 1.4. Map of study sites within Chicago, Illinois in 2010 and 2011. Location A is a site at a Nature Museum (NAM), location B is a site at a Michigan Avenue Structure (NAM), location C is at Chicago City Hall (CHA), location D is at Chicago Cultural Center (CCE), location E is at a Downtown Chicago Park (DCP), location F is at McCormick Parking Structure (MCC), location G is at Schwab Rehabilitation Hospital (SCH), and location H is at Gary Comer Youth Center (GCY) .................................................................................................................................71 Figure 1.5. Schematic aerial view of a study site includes the studied green roof (A) and all land types cover within a 200m radius of the green roof. Other land cover types include other nonstudied green roofs (B), woody vegetation (C), herbaceous vegetation (D), water (E), non-green roofs (F), and other impervious surfaces (G). ................................................................................72 Figure 1.6. Land cover composition of the Downtown Chicago Park (DCP) study site. ..............73 Figure 1.7. Land cover composition of the Ford Truck Plant (FOR) study site ............................74 Figure 1.8. Land cover composition of the Haworth Headquarters (HAW) study site. ................75 Figure 1.9. Land cover composition of the Aquascape Headquarters (AQU) study site...............76 Figure 1.10. Land cover composition of the Michigan Avenue Structure (MIA) study site .........77 Figure 1.11. Land cover composition of the Chicago City Hall (CHA) study site........................78 ix Figure 1.12. Land cover composition of the Chicago Cultural Center (CCE) study site ..............79 Figure 1.13. Land cover composition of the McCormick Parking Structure (MCC) study site. ...80 Figure 1.14. Land cover composition of the Nature Museum (NAM) study site ..........................81 Figure 1.15. Land cover composition of the Schwab Rehabilitation Hospital (SCH) study site. .82 Figure 1.16. Land cover composition of the Gary Comer Youth Center (GCY) study site ..........83 Figure 1.17. Land cover composition of the Plant and Soil Science Building (PSS) study site at Michigan State University. ............................................................................................................84 Figure 1.18. Examples of summer green roof vegetation on studied green roofs in 2010 and 2011 in Michigan and Illinois, U.S.A., in order of planting media depth from shallow to deep were (1) Ford Truck Plant, (2) Plant and Soil Science Building, (3) extensive roof on a Nature Museum, (4) Chicago Cultural Center, (5) Haworth Headquarters,(6) Aquascape Headquarters, (7) Michigan Avenue Structure, (8) intensive roof on a Nature Museum, (9) Chicago City Hall, (10) Schwab Rehabilitation Hospital, (11) McCormick Parking Structure, (12) Gary Comer Youth Center, and (13) Downtown Chicago Park. As a general trend, extensive roofs (1-7) have less vegetation structure compared to intensive roofs (8-13) .............................................................. 85 Figure 1.19. Examples of green roof vegetation on (1) extensive and (2) intensive green roofs in 2010 and 2011 in Michigan and Illinois, U.S.A. As a general trend, extensive roofs have less vegetation structure compared to intensive roofs ..........................................................................86 Figure 2.1. Estimated species richness of native bird species, excluding waterfowl and urban associated species, at the point-level for green roof sites in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Mean estimated species richness and quartiles shown are estimated using a multi-species hierarchical Bayes multi-scale model. ..................................................... 171 Figure 2.2. Estimated species richness of non-invasive, native bird species, excluding waterfowl and urban associated species, at the site-level for green roof sites in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Mean estimated species richness and quartiles shown are estimated using a multi-species hierarchical Bayes multi-scale model. ............................... 172 Figure 2.3. Estimated species richness of non-invasive, native bird species, excluding waterfowl and urban associated species, for green roofs and surrounding landscapes in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Mean richness and 95% credible intervals shown are estimated using a multi-species hierarchical Bayes multi-scale model. .....................173 Figure 2.4. Estimated and observed (uncorrected for detection probabilities) species richness of non-invasive, native bird species, excluding waterfowl and urban associated species, from the first year of sampling of green roofs and surrounding landscapes in Illinois and Michigan, U.S.A. x in 2010 and 2011. Mean richness and 95% credible intervals shown are estimated using a multispecies hierarchical Bayes multi-scale model..............................................................................174 Figure 2.5. Occurance probabilities of native waterfowl bird species, excluding urban-associated species, observed on three or more green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are Canada goose (Branta canadensis, CAGO) and mallard (Anas platyrhynchos, MALL). Mean occurrence and 95% credible intervals shown are estimated using a single-species hierarchical Bayes multi-scale occupancy model. ...................................175 Figure 2.6. Occurance probabilities of native bird species, excluding urban-associated and waterfowl species, observed on 3 or more green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are American robin (Turdus migratorius, AMRO), American goldfinch (Carduelis tristis, AMGO), chipping sparrow (Spizella passerina, CHSP), common grackle (Quizcalus quizcula, COGR), northern cardinal (Cardinalis cardinalis, NOCA), red-winged blackbird (Agelaius phoeniceus, RWBL), and white-throated sparrow (Zonotrichia albicollis, WTSP). Mean occurrence and 95% credible intervals shown are estimated using a multi-species hierarchical Bayes multi-scale model.. ..................................................................176 Figure 2.7. Occurence probabilities of declining native bird speciesobserved on green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are Bell’s vireo (Vireo bellii, BEVI), eastern kingbird (Tyrannus tyrannus, EAKI), field sparrow (Spizella pusilla, FISP), and northern flicker (Colaptes auratus, NOFL). Mean occurrence and 95% credible intervals shown are estimated using a multi-species hierarchical Bayes multi-scale model............................................................................................................................................177 Figure 2.8. Mean occurrence probabilities on green roofs compared with mean occurrence probabilities in surrounding landscapes for all non-invasive, native bird species, excluding waterfowl and urban associated species, in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Mean occurrence probabilities shown are estimated using a multi-species hierarchical Bayes multi-scale model. .........................................................................................178 Figure 2.9. Use probabilities of native waterfowl bird species, excluding urban-associated species, observed on three or more green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are Canada goose (Branta canadensis, CAGO) and mallard (Anas platyrhynchos, MALL). Mean use and 95% credible intervals shown are estimated using a single-species hierarchical Bayes multi-scale occupancy model.................................................179 Figure 2.10.Use probabilities of native bird species, excluding urban-associated and waterfowl species, observed on 3 or more green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are American robin (Turdus migratorius, AMRO), American goldfinch (Carduelis tristis, AMGO), chipping sparrow (Spizella passerina, CHSP), common grackle (Quizcalus quizcula, COGR), northern cardinal (Cardinalis cardinalis, NOCA), redwinged blackbird (Agelaius phoeniceus, RWBL), and white-throated sparrow (Zonotrichia xi albicollis, WTSP). Mean use and 95% credible intervals shown are estimated using a multispecies hierarchical Bayes multi-scale model..............................................................................180 Figure 2.11. Use probabilities of declining native bird speciesobserved on green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are Bell’s vireo (Vireo bellii, BEVI), eastern kingbird (Tyrannus tyrannus, EAKI), field sparrow (Spizella pusilla, FISP), and northern flicker (Colaptes auratus, NOFL). Mean use and 95% credible intervals shown are estimated using a multi-species hierarchical Bayes multi-scale model. ......181 Figure 2.12. Mean use probabilities on extensive and intensive green roofs for non-invasive, native bird species, excluding waterfowl and urban associated species, that were observed on at least 25% of green roofs during sampling in Illinois and Michigan, U.S.A. in 2010 and 2011. Mean use probabilities shown are estimated using a multi-species hierarchical Bayes multi-scale model............................................................................................................................................182 Figure 2.13. Mean use probabilities on green roofs organized by size for non-invasive, native bird species, excluding waterfowl and urban associated species, that were observed on at least 25% of green roofs during sampling in Illinois and Michigan, U.S.A. in 2010 and 2011. Mean use probabilities shown are estimated using a multi-species hierarchical Bayes multi-scale model............................................................................................................................................183 Figure 2.14. Mean use probabilities on green roofs organized by green space at study sites for non-invasive, native bird species, excluding waterfowl and urban associated species, that were observed on at least 25% of green roofs during sampling in Illinois and Michigan, U.S.A. in 2010 and 2011. Mean use probabilities shown are estimated using a multi-species hierarchical Bayes multi-scale model ..............................................................................................................184 Figure 2.15. Mean use probabilities on green roofs compared with mean use probabilities in surrounding landscapes for all non-invasive, native bird species, excluding waterfowl and urban associated species, in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Mean use probabilities shown are estimated using a multi-species hierarchical Bayes multi-scale model............................................................................................................................................185 xii INTRODUCTION Wildlife managers, land planners, environmental designers, and policy makers increasingly face challenges incorporating wise management of natural resources with the demands of urban areas. Currently over half of the global population lives in urban areas, and it is projected that by 2050 over 6 billion of the world’s 9 billion inhabitants will live in urban environments (Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, 2007). As developed areas replace green space, urban environmental quality generally decreases and can negatively affect human health and native plant and wildlife communities (Frumkin, 2002; Grimm et al., 2008; McKinney, 2008). One potential design strategy to minimize urban associated problems is establishing vegetation on roofs, also known as green roofs (Green Roofs for Healthy Cities, 2009). Wildlife managers, land planners, environmental designers, and policy makers tasked with improving human health and environmental quality in urban areas need to enhance their understanding of how green roofs can contribute to biodiversity and wildlife habitat conservation. The effects of urbanization can impact wildlife species and communities directly and indirectly (Crooks and Soulé, 1999; Bierwagen, 2007; Evans et al., 2009). Development in urban areas can be destructive to natural communities; wildlife habitat is often degraded and fragmented by roads and lawns (Forman, 2000; Keller and Largiader, 2003), groundwater recharge is prevented by impermeable pavement (Rose and Peters, 2001; Walsh et al., 2005; Feminella and Walsh, 2005), and concentrated pollutants are discharged into the air and water (Relyea, 2005). These actions collectively and individually degrade wildlife habitat quality for many native species (Michigan Natural Features Inventory, 2007). Rapid global population 1 growth compounds these issues, causing biodiversity loss, and places environmental degradation at the forefront of conservation and human survival issues around the world (Pimentel et al., 2007; McMichael et al., 2008). Conservation practices in urban and suburban areas promote wildlife by managing environments that allow ecosystem function to prevail. Even though restoration of native vegetation and wildlife may be impossible after long-term, intense land disturbances (Bakker and Berendse, 1999), environmental remediation projects can return vegetation similar to predevelopment conditions. Planted vegetation can increase species richness and abundance of wildlife (Waltz and Covington, 2004). Green corridors through and between urban areas, such as Emscher Valley in Germany, can transform lands in abandoned industrial areas into greenways for birds and insect communities (Seams, 1995; Miyagi, 2005; Hough, 2007). Human-made corridors can benefit wildlife by creating connections between natural areas (Kohut et al., 2009); however, depending on corridor placement and size, vegetation can also disrupt and fragment landscapes for wildlife movements and ecological processes (i.e., reduction in relative abundance for forest-nesting birds near mowed grass corridors that intersect forest vegetation; Rich et al., 1994). Implementation of green roofs has potential to provide many of the same benefits of other restoration techniques without fragmenting existing vegetation. Humans have long benefited from green roofs. Traditional Scandinavian sod roofs regulated extreme seasonal temperatures (Peck et al., 1999; Coffman and Davis, 2005; Getter and Rowe, 2006), and in 1914 a green roof was constructed in Switzerland on a water filtration plant to control water temperature (Brenneisen, 2006). Green roofs also have the ability to lessen the urban heat island that causes serious heat related health issues such as heat stroke and asthma (Frumkin, 2002; Banting et al., 2005; Getter and Rowe, 2006). The soil and vegetation 2 of green roofs insulates and shades buildings, which regulates internal temperatures, reduces energy used for heating and cooling (Getter and Rowe, 2006; Oberndorfer et al., 2007; Getter et al., 2011), and increases longevity of roofing membranes that result in fewer roofing materials in landfills (Rowe, 2011). Green roof vegetation also intercepts and filters air pollution (Currie and Bass, 2008) and counteracts carbon dioxide emissions through carbon sequestration (Getter et al., 2009). Green roof vegetation and substrates absorb and filter water, which reduces urban stormwater run-off and improves water quality (Peck et al., 1999; Getter and Rowe, 2006; Getter et al., 2007). Green roofs also offer mental health benefits such as noise reduction and therapeutic views (Frumkin, 2001; Oberndorfer et al., 2007; Van Renterghem and Botteldooren, 2009). The organization Green Roofs for Healthy Cities acknowledges that even though the tangible benefits of green roofs are not fully valued by the current market, researching these benefits can help advance green roof technologies to the forefront of high performance green building design, implementation, and maintenance (Green Roofs for Healthy Cities, 2011). Comprehensive study of the effect of green roofs on surrounding landscapes and values to biodiversity conservation will allow government and private sector officials, policy makers, green roof designers, and natural resource managers to make informed decisions about how to better implement green roof management strategies and large scale urban planning. Bird and plant communities can be significant components of biodiversity in urban landscapes. Green roofs may provide habitat for birds in urban landscapes because of the additional green space and fewer disturbances on roof surfaces than at ground level. Nesting attempts by ground nesting birds have been observed on green roofs (Baumann, 2006; Brenneisen, 2006), as have communities of rare and endangered insects affected by land use 3 changes (Jones, 2002; Kadas, 2006). European green roofs designed to promote biodiversity have shown increases in beetle colonization rates, demonstrating the potential for conservation success in green roof designs (Brenneisen, 2006). Also, native grasslands, a rare plant community, can be developed on green roofs without heavy demands on building weight restrictions and structure (Oberndorfer et al., 2007). Native grasslands in North America, which provide critical habitat to many grassland bird species, have been reduced by at least 80% due to land use conversion and urbanization and are the most threatened and degraded vegetation type in North America (Samson and Knopf, 1994; Herkert et al., 1996; Jones and Bock, 2002). Grassland birds are sensitive to land use conversion (Winter and Faaborg, 1999; Johnson and Igl, 2001; Jones and Bock, 2002) and have exhibited the most consistent, widespread, and rapid declines of any North American bird group (Herkert et al., 1996). Perturbations on native grasslands and other early successional vegetation types can affect landscape connectivity and disrupt ecological processes, such as dispersal or migration (Weber et al., 1999; McCallum and Dobson, 2002; Bierwagen, 2007). Many grassland bird species that have declined in abundance and distribution because of urbanization (Herkert et al., 1996) have the potential to benefit from green roofs (Brenneisen, 2006). Several factors of green roof design may influence the conservation value of green roofs. Semi-intensive and intensive roofs have deeper substrates than extensive roofs (e.g., generally >15cm, compared to <15; Rowe, 2011) and potentially support a greater range of vegetation conditions that likely contribute more to biodiversity conservation. Placement of green roofs within the landscape matrix may have varying degrees of conservation value dependent on whether placement, size, or quantity of green space patches through urban landscapes is more influential on connectivity of bird populations (Keitt et al., 1997; Donnelly and Marzluff, 2006; 4 Prugh et al., 2008). A series of green spaces could create a greenway for birds, insects, bats, and other wildlife that perceive habitat from the air (Brenneisen, 2006). Most green roofs are elevated above ground level, which could minimize the effects of ground predators on bird communities (Renfrew et al., 2005; Vergara and Hahn, 2009) and create additional nest and foraging sites that would be beneficial to bird conservation. The first chapter of this thesis describes and quantifies conditions present on green roofs in Michigan and Illinois, U.S.A. The vegetation characteristics and composition and building structure characteristics are quantified and examined for how they could provide contribute to wildlife habitat, specifically for bird communities. The second chapter quantifies bird community structure and composition on green roofs and the relationships between bird communities and green roof characteristics. These chapters provide information that can be used to design, manage, and create policy promoting green roofs that will benefit wildlife in urban areas. 5 OBJECTIVES The following are the objectives of this project: 1) Quantify the composition of bird communities on green roofs and in surrounding landscapes. 2) Quantify the vegetation structure and composition of green roofs and surrounding landscapes and their influence on bird abundance and community composition. 3) Characterize the relationships between green roof and landscape structure on the relative abundance and species composition of bird communities associated with green roofs and the surrounding landscape. 4) Make recommendations for green roof design, composition, and management in relation to the existing landscape context to improve ecosystem function and wildlife habitat quality. 6 LITERATURE CITED 7 LITERATURE CITED Bakker, J., Berendse, F., 1999. Constraints in the restoration of ecological diversity in grassland and heathland communities. Tree. 14(2), 63-68. Banting, D., Doshi, H., Li, J., Missios, P., Au, A., Currie, B.A., Verrati, M., 2005. Report on the environmental benefits and costs of green roof technology for the city of Toronto. City of Toronto and Ontario Centres of Excellence–Earth and Environmental Technologies, 88pp. Baumann, N., 2006. Ground-nesting birds on green roofs in Switzerland: preliminary observations. Urban Habitats. 4(1), 37-50. Bierwagen, B., 2007. Connectivity in urbanizing landscapes: The importance of habitat configuration, urban area size, and dispersal. Urban Ecosystems. 10, 29-42. Brenneisen, S., 2006. Space for urban wildlife: Designing green roofs as habitats in Switzerland. Urban Habitats. 4(1), 27-36. Coffman, R.R., Davis, G., 2005. Insect and avian fauna presence on the Ford assembly plant ecoroof. Third Annual Greening Rooftops for Sustainable Communities Conference, Awards and Trade Show. Washington, DC. Coupland, R. T., 1958. The effects of fluctuations in weather upon the grasslands of the Great Plains. Botanical Review. 24 (5): 273-317. Crooks, K.R., Soulé, M.E., 1999. Mesopredator release and avifaunal extinctions in a fragmented system. Nature. 400, 563–566. Currie, B., Bass, B., 2008. Estimates of air pollution mitigation with green plants and green roofs using the UFORE model. Urban Ecosystems. 11(4), 409-422. Donnelly, R., Marzluff, J.M., 2006. Relative importance of habitat quantity, structure, and spatial pattern to birds in urbanizing environments. Urban Ecosystems. 9, 99–117. Evans, K., Newson, S., Gaston, K., 2009. Habitat influences on urban avian assemblages. Ibis. 151, 19-39. Feminella, J., Walsh, C., 2005. Urbanization and stream ecology: An introduction to the series. Journal of the North American Benthological Society. 24(3), 585-587. Forman, R.T.T., 2000. Estimate of the area affected ecologically by the road system in the United States. Conservation Biology. 14 1, 31–35. 8 Frumkin, H., 2001. Beyond toxicity: the greening of environmental health. American Journal of Preventative Medicine. 20, 47-53. Frumkin, H., 2002. Urban sprawl and public health. Public Health Reports. 117, 201-217. Getter, K.L., Rowe, D.B., 2006. The role of green roofs in sustainable development. HortScience. 41(5), 1276–1286. Getter, K.L., Rowe, D.B., Andresen, J.A., Wichman, I.S., 2011. Seasonal heat flux properties of an extensive green roof in a Midwestern U.S. climate. Energy and Buildings. 43, 35483557. Getter, K.L., Rowe, D.B., Robertson, G.P., Cregg, B.M., Andresen, J.A., 2009. Carbon sequestration potential of extensive green roofs. Environmental Science & Technology. 43(19), 7564–7570. Getter, K.L., Rowe, Andresen, J.A., 2007. Quantifying the effect of slope on extensive green roof stormwater retention. Ecological Engineering. 31(4), 225–231. Green Roofs for Healthy Cities, 2009. Green roofs tree of knowledge. Accessed October 18, 2009 at http://greenroofs.org. Green Roofs for Healthy Cities, 2011. About us. Accessed August 11, 2012 at www.greenroofs.org/index.php/aboutus. Grimm, N.B., Faeth, S.H., Golubiewski, N.E., Redman, C.L., Wu, J., Briggs, J.M., 2008. Global change and the ecology of cities. Science. 319, 756-760. Herkert, J., Sampie, D., Warster, R. E., 1996. Management of Midwestern grassland landscapes for the conservation of migratory birds, in: F. R. Thompson, III (ed.). Management of Midwestern landscapes for the conservation of neotropical migratory birds. U.S. Forest Service, General Technical Report NC-187, St. Paul, Minn. p. 89-116. Hough, M., 2007. Nature as infrastructure: Strategies for sustainable regional landscapes. in The Future Metropolitan Landscape: Ecological Structure of City Regions. pp 54-58. Accessed October 19, 2009 at www.designobserver.com/media/pdf/Section_3:__Co_ 462.pdf Johnson, D.H., Igl, L.D., 2001. Area requirements of grassland birds: a regional perspective. The Auk. 118, 24-34. Jones, R.A., 2002. Tecticolous invertebrates: A preliminary investigation of the invertebrate fauna on ecoroofs in urban London. London: English Nature. 34 pp. 9 Jones, Z., Bock, C.E., 2002. Conservation of grassland birds in an urbanizing landscape: A historical perspective. The Condor. 104(3), 643-651. Kadas, G., 2006. Rare invertebrates colonizing green roofs in London. Urban Habitats. 4(1), 6686. Keitt, T.H., Urban, D.L., Milne, B.T., 1997. Detecting critical scales in fragmented landscapes. Conservation Ecology. 1(1), 4. Accessed March 2, 2012 at www.consecol.org/vol1/iss1/ art4. Keller, I., Largiader, C.R., 2003. Recent habitat fragmentation caused by major roads leads to reduction of gene flow and loss of genetic variability in ground beetles. Proceedings of the Royal Society London: Biology. 270, 417–423. Kohut, S.M., Hess, G.R., Moorman, C.E., 2009. Avian use of suburban greenways as stopover habitat. Urban Ecosystems. 12(4), 487-502. McCallum, H., Dobson, A., 2002. Disease, habitat fragmentation and conservation. Proceedings of the Royal Society London: Biology. 269, 2041–2049. McKinney, M., 2008. Effects of urbanization on species richness: A review of plants and animals. Urban Ecosystems. 11, 161-176. McMichael, A. J., Friel, S., Nyong, A., Corvalan, C., 2008. Global environmental change and health: impacts, inequalities, and the health sector. British Medical Journal. 336, 191-194. Michigan Natural Features Inventory, 2007. Rare species explorer. Accessed Oct 18, 2009 at http://web4.msue.msu.edu/mnfi/explorer. Miyagi, S., 2005. From spatial form to system and process through pattern making in the landscape: can landscape design contribute to nature restoration? Landscape Ecological Engineering. 1, 71–76. Oberndorfer, E., Lundholm, J., Bass, B., Coffman, R.R., Doshi, H., Dunnett, N., Gaffin, S., Kohler, M., Liu, K.K.Y., Rowe, D.B., 2007. Green roofs as urban ecosystems: Ecological structures, functions, and services. BioScience. 57(10), 823-833. Peck, S., Callaghan, C., Kuhn, M., Grass, B., 1999. Greenbacks from green roofs: forging a new industry in Canada. Status report prepared for Canada Mortgage and Housing Corporation. Peck and Associates. 78 pp. Pimentel, D., Cooperstein, S., Randell, H., Filiberto, D., Sorrentino, S., Kaye, B., Nicklin, C., Yagi, J., Brian, J., O’Hern, J., Habas, A., Weinstein, C., 2007. Ecology of increasing diseases: population growth and environmental degradation: Environment, development and sustainability. Human Ecology. 35, 653–668. 10 Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, 2007. World population prospects: The 2006 revision and world urbanization prospects: The 2007 revision. Accessed October 05, 2009 at http://esa.un.org/unup. Prugh, L.R., Hodges, K.E., Sinclair, A.R.E., Brashares, J.S., 2008. Effect of habitat area and isolation on fragmented animal populations. Proceedings of the National Academy of Sciences (PNAS). 105(52), 20770–20775. Relyea, R., 2005. The impact of insecticides and herbicides on the biodiversity and productivity of aquatic communities. Ecological Applications. 15(2), 618–627. Renfrew, R., Ribic, C., Nack, J., 2005. Edge avoidance by nesting grassland birds: a futile strategy in a fragmented landscape. The Auk. 122(2), 618-636. Rich, A., Dobkin, D., Niles, L., 1994. Defining forest fragmentation by corridor width: The influence of narrow forest-dividing corridors on forest-nesting birds in southern New Jersey. Conservation Biology. 8(4), 1109-1121. Rose, S., Peters, N. E., 2001. Effects of urbanization on streamflow in the Atlanta area (Georgia, USA): a comparative hydrological approach. Hydrological Processes. 15, 1441–1457. Rowe, D.B., 2011. Green roofs as a means of pollution abatement. Environmental Pollution. 159(8-9), 2100-2110. Samson, F., Knopf, F., 1994. Prairie conservation in North America. BioScience. 44, 418-421. Seams, R.M., 1995. The evolution of greenways as an adaptive urban landscape form. Landscape and Urban Planning. 33, 65-80. Van Renterghem, T., Botteldooren, D., 2009. Reducing the acoustical façade load from road traffic with green roofs. Building and Environment. 44(5), 1081–1087. Vergara, P., Hahn, I., 2009. Linking edge effects and patch size effects: Importance of matrix nest predators. Ecological Modeling. 220(9-10), 1189-1196. Villalobos, F. J., Ritchie, J. T., 1992. The effect of temperature on leaf emergence rates of sunflower genotypes. Field Crops Research. 29 (1), 37-46. Walsh, C., Roy, A., Feminella, J., Cottingham, P., Groffman, P., Morgan, R., 2005. The urban stream syndrome: Current knowledge and the search for a cure. Journal of the North American Benthological Society. 24(3), 706-723. Waltz, A. E. M., Covington, W. W., 2004. Ecological restoration treatments increase butterfly richness and abundance: Mechanisms of response. Restoration Ecology. 12(1), 85–96. 11 Weber, T.P., Houston, A.I., Ens, B.J., 1999. Consequences of habitat loss at migratory stopover sites: a theoretical investigation. Journal of Avian Biology. 30, 416–426. Williams, E. D., 1983. Effects of temperature, light, nitrate and prechilling on seed germination of grassland plants. Annals of Applied Biology. 103 (1), 161–172. Winter, M., Faaborg, J., 1999. Patterns of area sensitivity in grassland-nesting birds. Conservation Biology. 13, 1424-1436. 12 CHAPTER 1 VEGETATION CHARACTERISTICS OF GREEN ROOFS FOR WILDLIFE: HABITAT POTENTIAL IN MICHIGAN AND ILLINOIS, U.S.A. As developed areas replace green space, urban environmental quality decreases and can negatively affect native plant and wildlife communities (McKinney, 2008). One potential urban design strategy to minimize this degradation is establishing vegetation on roofs, also known as green roofs (Coffman and Davis, 2005). Green roofs are becoming more common in North America, and some local governments and federal agencies have incentivized green roof construction to make them more economically attractive to consumers. Policies in favor of green roofs have been developed because of potential environmental and economic benefits that accompany green roof installation (Carter and Fowler, 2008; City of Chicago, 2011). The potential of green roofs to provide economic and environmental benefits has been documented (Oberndorfer et al., 2007; Rowe, 2011). Green roofs help conserve energy (Getter et al., 2011), reduce air pollution (Currie and Bass, 2008), counteract carbon dioxide emissions through carbon sequestration (Getter et al., 2009), reduce the urban heat island (Banting et al., 2005), increase longevity of roofing membranes that result in fewer roofing materials in landfills (Rowe, 2011), improve water quality of storm water runoff (Peck et al., 1999; Getter et al., 2007), and reduce noise pollution (Van Renterghem and Botteldooren, 2009). Some authors have also concluded that green roofs provide wildlife habitat because of the diversity of plants, invertebrates, and birds that have been observed (Brenneisen, 2006; Kadas, 2006; Coffman, 2007). However, the relationships between green roof characteristics (roof structure, vegetation 13 characteristics, surface area, surrounding land use, maintenance, and age) and wildlife communities (abundance, diversity, and richness) have not been quantified. Green roofs offer potential to develop diverse vegetation types, increase green space in urban areas, increase connectivity with other urban green spaces (reduce ecosystem fragmentation), and provide wildlife habitat. In developed areas, these objectives might otherwise be impossible due to lack of available land, economic impracticality, and public perception of how land should be used. A potential increase in wildlife habitat could result in greater abundance and diversity of wildlife species in urban areas, thus contributing to biodiversity conservation. The ability of green roofs to provide urban wildlife habitat has not been extensively studied (Coffman and Davis, 2005; Kadas, 2006; Fernandez-Canero and Gonzalez-Redondo, 2010). Plant, bird, and insect communities, which have been observed on green roofs, can be significant components of biodiversity in urban landscapes (Savard et al., 2000). Additionally, lizards with ground access (e.g., through rock filled gabions spanning roof and ground; Cantor, 2008) and squirrels and rabbits (personal observation) on roofs at ground level are examples of other wildlife that use green roofs. However, many green roofs are primarily accessible to wildlife taxa such as birds and bats because they are elevated above ground level and do not have specialized wildlife access structures. Birds (29 species) have been observed breeding on green roofs in parts of Europe and North America (Baumann, 2006; Fernandez-Canero and GonzalezRedondo, 2010). Past studies suggest that green roofs may provide habitat for birds in urban landscapes at several spatial scales (i.e., microhabitat to home range scales) (Fig. 1.1) because of the additional green space and less disturbances on roof surfaces (e.g., Brenneisen, 2006; Lundholm, 2006; Oberndorfer et al., 2007). Green space on roofs may not appear contiguous 14 with green space in the landscape; however, green roofs may help increase habitat connectivity and suitability for bird species that perceive habitat at macro scales (e.g., landscape) (Kotliar and Wiens, 1990; Morrison et al., 1992). Increased habitat connectivity may enhance wildlife habitat suitability of landscapes and thereby increase species richness (Goddard et al., 2010). Since the creation of green spaces and adequate vegetation structure in urban areas can influence bird community composition and richness (Fontana et al., 2011), green roofs should have the same effect. Vegetation structure and composition have a strong role in determining bird habitat suitability, and hence the vegetation on green roofs will likely contribute habitat for bird communities, such as early successional songbirds (Coffman, 2007). Several studies have described vegetation on green roofs (Coffman, 2007; Wolf and Lundholm, 2008; Rowe et al., 2012), but there is a lack of information about the ecological contributions of vegetation represented on green roofs in North America (Dvorak and Volder, 2010) and their potential to provide green space in developed areas. 15 OBJECTIVES Objectives for this chapter were to: 1) Describe and quantify vegetation structure and composition for intensive (generally >15cm planting media depth) and extensive (generally <15cm planting media depth) green roofs and their surrounding landscapes. 2) Characterize the types of landscapes where green roofs have been constructed. 3) Describe the potential of green roof implementation to conserve components of biodiversity and fulfill vegetation requirements for wildlife species in the Midwest United States. 16 METHODS Site descriptions We selected 12 green roof study sites described as: 1) Downtown Chicago Park (DCP), 2) the Ford Truck Plant (FOR), 3) McCormick Parking Structure (MCC), 4) Aquascape Headquarters (AQU), 5) Haworth Headquarters (HAW), 6) the Chicago Cultural Center (CCE), 7) the Chicago City Hall (CHA), 8) a Nature Museum (NAM), 9) a Michigan Avenue Structure (MIA), 10) Schwab Rehabilitation Hospital (SCH), 11) Gary Comer Youth Center (GCY), and 12) the Plant and Soil Science Building at Michigan State University (PSS) (Table 1.1 ). Study sites were located in Illinois and the Lower Peninsula of Michigan in the United States (Fig. 1.21.4). All study sites were within the Midwest Broadleaf Forest Province ecoregion; characterized by a climate with “warm to hot summers” that often have brief drought in the late summer, vegetation consisting of “cold-deciduous, hardwood-dominated forests”, and “flat to hilly terrain with features associated with former glaciation” (McNab et al., 2007, p. 10). The range of monthly mean temperatures at each study site during the early to peak bird nesting season (April - June) was between 11.3C and 22.2C in 2010 and between 7.9C and 21.5C in 2011 (Table 1.2; National Oceanic and Atmospheric Administration, 2012). The range of monthly mean temperatures at each study site during post-nesting and brood-rearing season (July-September) was between 16.5C and 26.1C in 2010 and between 15.9C and 26.3C in 2011. The range of monthly total precipitation at each study site during the early to peak bird nesting season was between 5.9cm and 20.0cm in 2010 and between 3.5cm and 18.8cm in 2011 (Table 1.3; National Oceanic and Atmospheric Administration, 2012). The range of monthly total 17 precipitation at each study site during post-nesting and brood-rearing season (July-September) was between 1.1cm and 24.4cm in 2010 and between 9.9cm and 18.3cm in 2011. Each study site included the green roof on a building and the landscape area within 200m of a green roof (Fig. 1.5); a minimum of 14ha for our smallest green roof. This size area was based on the size of species’ home ranges that have been observed using green roofs in the past (FernandezCanero and Gonzalez-Redondo, 2010). Each study site was evaluated to address how wildlife species and communities may respond to the environment. In 2010, eight green roof study sites were selected to sample a wide range of roof sizes, building heights, vegetation types (annual, perennial Sedum, perennial non-Sedum, woody), planting media depth (Table 1.4), urban land use (Table 1.1), and accessibility (Table 1.5). Another criterion considered when selecting study sites was that roofs were accessible through building owners. In 2011, four study sites from 2010 were re-sampled along with four new study sites. Green roofs selected for 2011 were elevated above ground level (not on subterranean structures) since the greatest potential for green roof construction in urban areas is on structures above ground level. Also, because patch size is positively related to species-occurrence and density for several grassland bird species (Johnson and Igl, 2001), we selected the largest green roofs possible that met all other selection criteria. To characterize the attributes of the green roofs for their potential as wildlife habitat (Tables 1.1-1.3), we compiled information from green roof owners and managers, on-line green roof websites, green roof designers and engineers, the NOAA database report (National Oceanic and Atmospheric Administration, 2012), and from personal observation and vegetation sampling (Table 1.6). Vegetation growth and plant species on green roofs are limited by water availability, growing media composition, fertilization rate, slope, and substrate depth 18 (Monterusso et al., 2005; Rowe et al., 2006). Seven sites were classified as extensive green roofs (planting media depth of 3-15cm), and six sites were intensive green roofs (planting media depth of 15-120cm). Three of the intensive sites had small areas (e.g., edges of planting media mounds) of planting media <15cm deep, but the majority of these roofs were covered with planting media >15cm deep. These roofs were categorized as intensive for analyses (Table 1.4). Vegetation cover initially planted on the green roofs was based on the intended primary roof function established by owners. Extensive roofs installed primarily for pollution mitigation and energy savings were usually planted with Sedum (Tables 1.2 and 1.3): this was the only vegetation planted on three roofs (FOR, HAW, PSS). Planted vegetation species included S. album, S. kamtschaticum, and S. spurium on five roofs. Green roofs at six study sites supported native plants for their respective regions (Table 1.4). Native perennials such as little bluestem (Schizachyrium scoparium), blazing star (Liatris sp.), coneflower (Echinacea sp.), and aster (Aster sp.) were planted on green roofs at SCH, AQU, and DCP. Woody vegetation was present on intensive and extensive roofs, but woody plant varieties on extensive roofs were low-growing and tolerant of dry conditions (e.g., Juniperus horizontalis at CCE). Green roof maintenance requirements were based on the original planted vegetation and the intended roof functions. Irrigation systems were present on nine roofs, scheduled weeding on 10 roofs, and fertilization on six roofs (Table 1.5). Other maintenance included periodic controlled burns on one roof with the goal of maintaining prairie vegetation, regular mowing on one roof that was readily accessible by the public, and vegetable harvest on one roof that functioned as a youth center garden. One roof had no planned maintenance regime beyond the establishment of the original vegetation. 19 Vegetation sampling Vegetation sampling was conducted during two periods: one during the early to peak bird nesting season April 22 to May 14, 2010 and April 23 to June 24, 2011; and one post-nesting and brood-rearing season June 30 to August 15, 2010 and July 28 to October 1, 2011. By sampling during the spring (April-June) and summer (July-October), vegetation was representative of that available to bird communities during the beginning and peak of the breeding season and postnesting during brood-rearing (Short, 1985; Basore et al., 1986; Best et al., 1997). The sampled portions of landscapes surrounding green roofs were safe, accessible, and included clearly definable vegetation areas (areas with a minimum requirement of exposed soil with potential to support vegetation). Planter boxes attached to buildings, street median vegetation, street trees in grates, and vegetation on other green roofs in the surrounding landscapes were not sampled in the field, however, were represented in the aerial land cover analysis. The line intercept method (Canfield, 1941) was used to quantify vegetation cover of turf grass, herbaceous perennial cover, and shrub and tree canopy on green roofs and in vegetation areas in surrounding landscapes. One-meter belt transects (Clements, 1905) were used on green roofs and surrounding landscapes to determine species presence and stem density of woody plants. Transects ranged from 3.8-200.0m long and were systematically placed perpendicular to the grain of vegetation types on each roof and landscape area. The length of transects corresponded to the size of green roofs. The length of the transect that intersected mowed lawn, perennial, shrub, and tree cover was recorded and used to calculate the percent cover of each vegetation type. The point intercept method (Heady et al., 1959) was used to calculate percent cover of different vegetation types and quantify mean vegetation height. Every 5m, vegetation 20 intersecting transects was identified by type (perennial sedum, non-sedum perennial, woody vegetation) and the height of the vegetation at the intersecting point was measured. Aerial land cover To characterize land cover for each study site (i.e., green roof and the surrounding landscape) we imported Google Earth (Version 6.1; Google Earth, 2011) images into ArcMap (ArcGIS version 9.2; ESRI, 2006), georeferenced the aerial photographs, and digitized and classified land cover as green space and non-green space (Fig. 1.6-1.17). Green space was further classified as studied green roof, other green roof, woody vegetation, or herbaceous vegetation (i.e., turf and perennials). Non-green space was further classified as water or impervious surface, and impervious surface was further classified as non-green roof or other impervious surface (sidewalk, paved roads, paved plazas). At each site, classifications were then used to quantify percent cover of green roof, green space in the surrounding landscape (i.e., herbaceous and woody vegetation), and conventional roofs. We subsequently calculated percent of total green space attributed to the green roof and the potential green space created if all existing conventional roofs were vegetated. Data analysis Data were analyzed using SAS version 9.2 (SAS Institute Inc., 2008). Vegetation characteristics (mean percent cover, mean stem density) associated with spring and summer were compared to determine potential differences between seasons (early to peak bird nesting season and post-nesting/brood-rearing season). Data for each variable were checked for normality (p<0.10) using the Shapiro-Wilk procedure in PROCUNIVARIATE. Since the data sets for most variables were not normally distributed, the non-parametric Kruskal-Wallis one-way analysis of 21 variance test was used for further comparisons. Spring and summer values of each vegetation variable were compared for all roofs. Spring and summer vegetation variable distributions were not different (p<0.10) for most variables, but because some were different, data from the two sampling periods were not pooled. Green space cover before and after green roof installation phases (pre-green roof green space, current green space, potential green space) was also compared. We used the non-parametric Kruskal-Wallis one-way analysis of variance test to determine if the vegetation characteristics associated with roof type (i.e., intensive vs. extensive) differed. If significant differences were identified, we analyzed the data for differences between roof types. Level of significance was set at 0.10 a priori to better identify ecologically significant differences in vegetation on green roofs and in green space cover to reduce the chances of committing a Type I error. 22 RESULTS Green roof comparisons Green roofs (n=12) ranged in size from 9.91ha to 0.03ha with a mean area of 1.83ha; median roof size was 0.19ha. Green roofs were on structures up to 15 stories high, but the median and modal building height was 3 stories from ground level, approximately 10m high (Table 1.4). Planting media depth leads to inherent differences in vegetation on intensive and extensive green roofs (Fig. 1.18 and 1.19). We found that 83% and 33% of the vegetation variables sampled in 2010 and 2011, respectively, differed between the two roof types (Table 1.7). Tree and shrub cover were generally absent on extensive green roofs, except for some shrub cover on one roof in 2011 (Table 1.7 and 1.8). Vegetation height was only measured in 2011, and we found that perennial vegetation was 208% taller on intensive roofs. On all extensive roofs planted entirely in drought-tolerant Sedum we observed ≥99% mean herbaceous cover, while on all extensive roofs planted with a mixture of Sedum and/or non-Sedum perennials, mean herbaceous cover composed 50-86% of the green roof area (Table 1.8). Mean herbaceous cover on extensive roofs was 40% (p=0.013) and 18% (p=0.462) higher than on intensive roofs in 2010 and 2011, respectively. Mean percent herbaceous cover on extensive roofs was 78-100% in 2010 and 50-100% in 2011, compared to 48-75% in 2010 and 40-92% in 2011 on intensive roofs (Table 1.8). Shrub cover occurred on 50% and 75% of intensive green roofs in 2010 and 2011, and tree cover occurred on 100% and 75% of intensive green roofs in 2010 and 2011. Regardless of roof type, turf cover was absent from all green roofs except on one roof in 2010 which had 25% turf cover. Planting media depth corresponded with differences in vegetation 23 type; greater shrub and tree cover and taller vegetation was observed on intensive roofs. Herbaceous cover was the dominant cover type on all green roofs, but sedum covered roofs (commonly on extensive roofs) had the highest percent cover (>99%). Herbaceous vegetation covered the majority of the roofs, with none of the other three vegetation characteristics having a reoccurring order of dominance (Table 1.8). Several roofs planted with a mixture of perennial species, other than Sedum, showed differences in herbaceous cover between seasons (2010: DCP p=0.10, AQU p=0.13, MCC p=0.05; 2011: CHA p=0.01, CCE p=0.09, GCY p=0.08, SCH p=0.02). No significant difference in shrub or tree cover occurred between seasons (p≥0.26, and p≥0.32, respectively) on any roof. Mean percent herbaceous cover on non-Sedum roofs was 33% greater in summer, with an increase of 14% mean percent herbaceous cover between seasons. Surrounding landscape comparisons All vegetation variables, except percent shrub cover for one landscape, were not significantly different between spring and summer sampling periods. Landscape areas had between 22-78% turf cover, 0-53% herbaceous cover, 0-12% shrub cover, 1-72% tree cover, 53566 tree/ha, and 0-913 shrubs/ha (Table 1.9). Shrub and tree vegetation characteristic values (i.e., percent cover and stem density) for seven of eight intensive green roofs were within the ranges of those variables measured in the landscapes; however, mean turf cover was 48% lower and mean herbaceous cover was 42% higher than in the landscape. Vegetation characteristics values on extensive green roof were not within the value ranges of those characteristic values for the landscape. The range of land use intensity within a 200m radius of each green roof ranged from low (lake), to mid (mid-density residential, urban park), to high (railway, highway, industrial 24 complex) (Table 1.1). Our land cover classification indicated that for 67% (8 of 12) of the study sites, non-green roofs and other impervious surfaces were the two main sources of land cover (Table 1.10). Green space area before and after green roof implementation was not significantly different (p=0.421). However, if all existing non-green roofs were converted into green roofs, mean green space would increase 306% (p=0.002). The difference between green space cover before implementation of any green roofs within a study site and the potential green space cover if all roofs were ‘greened’ would more than double green space cover (p=0.001) for the studied landscapes. This increase in green space does not account for the area occupied by rooftop ventilation utilities not suited to be covered with vegetation. If ventilation utilities halved potential green roof area, green space in the landscape would increase >200%, and at study sites like Chicago City Hall with high percentages of roof cover and low percentages of green space cover, green space would increase at least ten-fold. This dramatic change in the availability of green space in urban areas could provide vegetation with the potential to enhance wildlife habitat. 25 DISCUSSION Vegetation structure and composition Wildlife observations on green roofs have led to conclusions that green roofs provide wildlife habitat, and thus have direct wildlife conservation value (Brenneisen, 2006; Kadas, 2006; Coffman, 2007). However, there has been a lack of quantitative vegetation data available to describe the conditions green roofs may provide as suitable wildlife habitat. Quantifying green roofs’ vegetation characteristics and green space contributions in adjacent landscapes is the first step towards assessing wildlife conservation value of green roofs and implementing green roofs with directed wildlife conservation goals. The objectives of our study were to quantify and describe vegetation structure and composition and green space contributions of green roofs and the surrounding landscapes that may contribute to wildlife habitat. This information is vital to assess the potential of green roof construction to increase ecological function and ultimately to help conserve biodiversity in urban areas. Comparisons of vegetation structure and composition observed on green roofs with those required to support wildlife species can be used to assess green roofs’ wildlife habitat potential. Since green roof soil depth is limited by structural support, plants that can withstand shallower growing media (perennials, small shrubs) are likely to comprise the dominant vegetation type on all green roofs. Special planting conditions can be designed to accommodate large shrubs and trees (pockets of extra deep planting media), but providing structural support for the growing media required to support a forest vegetation type on a green roof would normally be cost prohibitive. A difference in vegetation between roof types was expected as extensive green roofs’ shallow growing media creates more stressful growing conditions (high soil 26 temperatures and low soil moisture) than the deeper growing media on intensive green roofs (Oberndorfer et al., 2007). Studied intensive roofs had taller perennial and woody species, whereas extensive green roofs generally had low-growing, drought-tolerant perennial or shrub species. Since a greater variety of plant species can be established on intensive roofs, it is not surprising that vegetation cover and structure and the variety of native species were greater on this roof type (Table 1.7). The increased niche opportunities in vegetation on intensive roofs likely can support a greater diversity of wildlife species; however, wildlife species that require shorter vegetation and less woody cover may be better supported on extensive roofs. These differences in vegetation between roof types may result in greater differences between wildlife communities on intensive and extensive green roofs than on the same roof type. Bird communities Bird communities comprise the majority of urban wildlife with access to all green roofs, whether at ground level or on top of a high-rise building (Fernandez-Canero and GonzalezRedondo, 2010). In terrestrial systems vegetation structure and composition has been used to predict abundance, species richness, and productivity for bird communities (Cody, 1968; Delisle and Savidge, 1997), and the same should hold true for green roofs and their surrounding landscapes. Vegetation characteristics on green roofs that cover a smaller area than the home range of a bird species may contribute to habitat suitability for that species by providing finer scale habitat requirements. As shown in Habitat Suitability Index (HSI) models, each species has a unique set of vegetation characteristics to which it responds at multiple spatial scales (United States Fish and Wildlife Service, 1981). Vegetation characteristic values within specific species-based models can be compared to vegetation conditions at each respective site to evaluate habitat suitability for a particular species. Assuming vegetation characteristic values 27 can similarly be compared to vegetation conditions on green roofs, vegetation on green roofs can satisfy life requisites for specific bird species. According to the life requisite requirements in the red-winged blackbird (Agelaius phoeniceus) HSI model (Short, 1985), the Aquascape Headquarters Green Roof (with 78% herbaceous cover) provided low quality nesting habitat (>1ha, woody vegetation or dense stands of perennials >1m tall covering >10% of the site, no grazing, mowing, burning or tilling) (Table 1.4 and 1.5). The Gary Comer Youth Center Green Roof had vegetation characteristic values within the range of woody vegetation cover and within 6cm of the mean live vegetation height (45cm) reported for Conservation Reserve Program (CRP) study fields throughout the Midwest United States; These values correspond with the highest abundances of American goldfinch (Carduelis tristis), barn swallow (Hirundo rustica), chipping sparrow (Spizella passerine), and song sparrow (Melospiza melodia) (Best et al., 1997). CRP fields with live herbaceous cover, composed of grasses and forbs that contributed 46.8% and 27.1%, respectively, and a mean live vegetation height of 68cm, and 0.4% woody cover corresponded with the highest abundances of common yellowthroat (Geothlypis trichas) and eastern kingbird (Tyrannus tyrannus) (Best et al., 1997). Green roofs that provided herbaceous cover within the range given for highest common yellowthroat and eastern kingbird abundances were SCH and DCP in 2010, and GCY, MIA, and CCE in 2011 (Table 1.8). The green roof that came closest to providing vegetation with an equivalent height was SCH that provided 88% of CRP live vegetation height. No green roof met all three vegetation characteristic values for common yellowthroat and eastern kingbird abundance; however, green roofs’ vegetation characteristics demonstrate potential to provide suitable habitat for some bird species. 28 The Chicago City Hall Green Roof had herbaceous cover (Table 1.8) within 15% and mean vegetation height within 6% of reported vegetation values in Iowa alfalfa fields with highest observed abundance of common yellowthroat (Frawley and Best, 1991). The Chicago Cultural Center Green Roof had vegetation cover (24% shrub cover) that aligned with the percent shrub cover (15-35%) required for optimal habitat suitability for field sparrow (Spizella pusilla), according to the HSI model developed by Sousa (1983). However, herbaceous cover height on this green roof (2cm) was less than half that described as suitable in the HSI model for field sparrow (>5cm), illustrating the importance of comparing a diversity of vegetation characteristics on green roofs with those required by target species. Cover types indicative of bird communities present in the landscape (Anderson and Shugart, 1974) should also hold true on green roofs. Turf grass and low-growing perennials provide foraging opportunities for bird species such as killdeer (Charadrius vociferus) and common grackle (Quizcalus quizcula), taller perennials such as little bluestem and coneflower provide high perches and dense cover for foraging for species such as red-winged blackbird, and shrub and tree vegetation provide habitat for forest edge species such as blue jay (Cyanocitta cristata) and downy woodpecker (Picoides pubescens). The presence of these cover types may have similar outcomes on green roofs. Turf was less common on green roofs than in surrounding landscapes; this was likely because most turf varieties require large water inputs which are impractical on most green roofs, and turf cover did not align with the environmental focus of most roofs (Table 1.5). Even though shrub and tree cover for green roofs fell within the range of those variables in the surrounding landscapes, considering the level of development in the landscapes where the green roofs were located, the surrounding landscapes were not a high standard of ecological function with which to compare the green roofs. 29 The majority of studied green roofs (9 of 13) were planted mainly (>50%) with nonnative species. Even though from a quantified structural standpoint, non-native species can fulfill vegetation characteristic requirements for wildlife suitability, they may provide a different level of habitat quality because of their unique plant characteristics (e.g., type of fruit or seed produced, insect communities supported, color and texture providing camouflage, etc.). Even though vegetation characteristics on green roofs may fit the description of suitable habitat for a species, without quantifying the relationships between habitat attributes and species responses it is difficult to know the effect that other factors (i.e., roof height, human presence, non-native plant species, landscape matrix, and lack of mesopredators) have on how birds use a vegetation type on green roofs. The vegetation conditions in our study support the idea that design intent can influence bird species’ presence on green roofs (Fernandez-Canero and Gonzalez-Redondo, 2010). Green roofs designed to provide wildlife habitat such as NAM-intensive, MCC, CHA, and AQU had above average perennial cover and vegetation height and were composed predominantly of native species. Unsurprisingly, green roofs with aesthetics driving the design ranged from vegetation types that provided little vegetation structure (a homogenous mixture of a few Sedum species) to roofs with a variety of vegetation structure (areas of trees, shrubs, perennials and turf grass). Where pollution mitigation and energy savings drove green roof design, Sedum roofs seemed to prevail. While Sedum does not provide structure for perching or dense cover, these open areas may be suitable for foraging bird species and provide cover for some insect communities that are beneficial for foraging birds. Irrigation on some green roofs provides a water source that birds may utilize for drinking or bathing. Other roof maintenance activities such as annual vegetation removal and pruning may decrease important structural characteristics 30 for wildlife. Green roofs that maintain dead perennial cover may provide opportunities (cover, insects, seeds, etc.) for birds. Human access is another factor that could negatively affect wildlife suitability on green roofs, as human presence can reduce foraging and breeding opportunities for birds (Fernández-Juricic, 2002). Differences in weather between years and locations may have affected observed vegetation conditions on green roofs, and thus altered bird habitat suitability. Temperature and precipitation can influence time and rate of seed germination (Williams, 1983), leaf emergence and cover (Villalobos and Ritchie, 1992), and seed production (Coupland, 1958). Monthly mean temperatures during bird nesting and brood-rearing seasons in 2010 were generally higher than in 2011. Higher temperatures in 2010 may have contributed to earlier leaf emergence and seed production, and may have provided cover and foraging opportunities for seed-eating birds earlier than in 2011. Wildlife conservation Past studies have focused on green roofs’ ecological contributions separate from the rest of the landscape; however, green roofs are part of complex landscapes and interact with ecological components within landscapes (Oberndorfer et al., 2007). Additional green space provided by green roofs may allow migratory species to traverse barriers typically associated with urban landscapes, and therefore restore habitat connectivity for some species (Goddard et al., 2010). Green space cover can also indicate increased ecological function and biodiversity conservation (Corry and Nassauer, 2005). The mean increase in green space if all roofs within our study sites were ‘greened’ would more than triple current green space. This affect, if extrapolated throughout a large area of development, could substantially increase green space and connectivity for bird communities. In an urban area where there is a low percent green space 31 cover, the comparative effect of a green roof on green space, and thereby on associated benefits such as storm water runoff mitigation and increased wildlife habitat, is much greater than in a rural area with an already high percent green space. This difference illustrates the potential influence of landscape context on the potential of a green roof to contribute to wildlife habitat through available green space. Realization of increased wildlife habitat quantity and quality will depend on the management decisions made regarding those green spaces. Implementing green roofs with vegetation appropriate to a bird community targeted for conservation could dramatically enhance wildlife habitat through an urban area. More research is needed to examine how the distribution of green roofs through urban landscapes may affect wildlife habitat connectivity (Donnelly and Marzluff, 2006; Prugh et al., 2008; Evans et al., 2009). Green roof clusters strategically placed as ‘stepping stones’ throughout landscapes may affect connectivity based on cluster size and distance between clusters, whereas connectivity provided by green roofs in a linear ‘roof-top greenway’ may be affected by green way direction and size. Research is also needed to address the possibility that green roofs may function as ecological traps (sinks: attracting wildlife without increasing fitness) or function as ‘safe havens’ that foster increased fitness. Green roofs may also act as sinks due to their potential small size and isolation, thereby maintaining numbers of individuals by recruiting from a nearby source population (Pulliam, 1988). Conversely, wildlife on green roofs may experience greater survival and fitness due to fewer predators (Renfrew et al., 2005; Vergara and Hahn, 2009), additional potential nest sites, abundant food, and elevated position of green roofs, which may reduce negative edge and patch effects (Burke and Nol, 1998; Bollinger and Switzer, 2002). Insight into the effects of green roof design and vegetation type on 32 wildlife species fitness could be used by city planners, resource managers, and policy makers to increase wildlife habitat conservation through green roof development and management. Our study identified differences between vegetation on intensive and extensive green roofs, demonstrated green roof vegetation’s ability to fulfill wildlife habitat requirements and presented potential increases in urban green space through green roof installation in developed areas where green roofs are already an acceptable building strategy in the Midwest United States. These results support the premise that green roof vegetation can contribute to wildlife habitat and increase urban green space important for wildlife conservation. This information should encourage collaboration of green roof designers, city planners, resource managers, and policy makers to advance green roof installations towards environmental sustainability that includes wildlife conservation. 33 Table 1.1. Location, landscape classification, and year of installation of green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011 Study Site City, State Land Use Classification* Year Installed Year Studied Downtown Chicago Park (DCP) Chicago, IL 2004 2010 Ford Truck Plant (FOR) Dearborn, MI Urban park, high-density residential, urban central business district, museum, railway Industrial complex 2003 2010, 2011 McCormick Parking Structure (MCC) Aquascape Headquarters (AQU) Chicago, IL 2003 2010 2005 2010 Haworth Headquarters (HAW) Holland, MI Conference center, urban park, lake, highway Offices and light manufacturing distribution, residential mid-density, airport Industrial complex, commercial complex 2007 2010, 2011 Chicago Cultural Center (CCE) Chicago, IL Urban central business district, urban park, high-density residential, 2006 2011 Chicago City Hall (CHA) Chicago, IL Urban central business district, urban park, high-density residential 2001 2011 Nature Museum (NAM) Chicago, IL Museum, urban park, lake 2002, 2004** 2010, 2011 Michigan Avenue Structure (MIA) Chicago, IL Urban central business district, residential high-density 2008 2011 Schwab Rehabilitation Hospital (SCH) Chicago, IL Health facilities, urban park, residential mid-density 2003 2010, 2011 St. Charles, IL 34 Table 1.1. (cont’d) Study Site City, State Land Use Classification* Year Installed Year Studied Gary Comer Youth Center (GCY) Chicago, IL School, residential mid-density, commercial strip developments, railway 2006 2011 Plant and Soil Science Building (PSS) College campus, urban park, railway 2004 2010 East Lansing, MI *Land Use Classification based on United States Geological Survey (USGS) land use and land cover classification sytems (Anderson et al., 1976). Driveways and surface roads were not included as a land use class because these transportation routes were present at all sites. **One intensive green roof was installed in 2002 and two extensive green roofs were installed in 2004. 35 Table 1.2. Monthly mean temperature (degrees Celcius) for green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011. Data is provided for green roof study sites during the year each site was sampled. Study Site January 2010 2011 February 2010 2011 March 2010 2011 April 2010 2011 May 2010 2011 June 2010 2011 Downtown Chicago Park (DCP) Ford Truck Plant (FOR) -3.7 - -0.9 - 5.5 - 12.4 - 16.2 - 22.2 - -3.5 -4.6 -2.1 -3.4 4.3 ND 11.6 8.4 16.7 15.3 22.0 21.5 McCormick Parking Structure (MCC) -3.7 - -0.9 - 5.5 - 12.4 - 16.2 - 22.2 - Aquascape Headquarters (AQU) -7.3 - -3.9 - 4.3 - ND - 16.3 - 21.1 - Haworth Headquarters (HAW) -3.8 -5.7 -2.6 -2.9 ND 1.1 11.3 7.9 16.7 ND 20.8 18.0 Chicago Cultural Center (CCE) - -3.8 - -0.7 - 4.5 - 9.6 - 13.9 - 20.9 Chicago City Hall (CHA) - -3.8 - -0.7 - 4.5 - 9.6 - 13.9 - 20.9 Nature Museum (NAM) -3.7 -3.8 -0.9 -0.7 5.5 4.5 12.4 9.6 16.2 13.9 22.2 20.9 Michigan Avenue Structure (MIA) - -3.8 - -0.7 - 4.5 - 9.6 - 13.9 - 20.9 Schwab Rehabilitation Hospital (SCH) -3.7 -3.8 -0.9 -0.7 5.5 4.5 12.4 9.6 16.2 13.9 22.2 20.9 Gary Comer Youth Center (GCY) - -3.8 - -0.7 - 4.5 - 9.6 - 13.9 - 20.9 Plant and Soil Science Building (PSS) -5.1 -3.7 - 4.2 - 11.4 - 15.9 - 20.4 - - 36 Table 1.2. (cont’d) Study Site July 2010 2011 August 2010 2011 September 2010 2011 October 2010 2011 November 2010 2011 December 2010 2011 Downtown Chicago Park (DCP) 26.0 - 26.1 - 20.4 - 15.7 - 7.8 - -2.2 - Ford Truck Plant (FOR) 24.8 26.1 24.2 22.9 18.4 17.9 13.1 12.0 5.4 7.7 -3.2 2.0 McCormick Parking Structure (MCC) Aquascape Headquarters (AQU) 26.0 - 26.1 - 20.4 - 15.7 - 7.8 - -2.2 - ND - 23.5 - 17.5 - 12.2 - 4.7 - -6.8 - 23.9 23.1 23.9 21.2 17.8 15.9 12.0 11.2 6.6 6.4 -2.4 2.1 Chicago Cultural Center (CCE) Chicago City Hall (CHA) - 26.3 - 24.9 - 19.0 - 14.5 - 9.1 - 3.8 - 26.3 - 24.9 - 19.0 - 14.5 - 9.1 - 3.8 Nature Museum (NAM) 26.0 26.3 26.1 24.9 20.4 19.0 15.7 14.5 7.8 9.1 -2.2 3.8 Michigan Avenue Structure (MIA) - 26.3 - 24.9 - 19.0 - 14.5 - 9.1 - 3.8 Schwab Rehabilitation Hospital (SCH) 26.0 26.3 26.1 24.9 20.4 19.0 15.7 14.5 7.8 9.1 -2.2 3.8 Gary Comer Youth Center (GCY) - 26.3 - 24.9 - 19.0 - 14.5 - 9.1 - 3.8 Plant and Soil Science Building (PSS) 23.7 - 23.3 - 16.5 - 11.2 - 4.8 - -4.2 - Haworth Headquarters (HAW) 37 Table 1.3. Monthly total precipitation in centimeters for green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011. Data is provided for green roofs study sites during the year each site was sampled. January 2010 2011 February 2010 2011 March 2010 2011 April 2010 2011 May 2010 2011 June 2010 2011 Downtown Chicago Park (DCP) Ford Truck Plant (FOR) 2.7 - 4.0 - 4.4 - 9.8 - 17.3 - 20.0 - 2.2 3.5 3.9 7.8 4.7 ND 5.9 15.0 12.4 18.8 17.3 ND McCormick Parking Structure (MCC) Aquascape Headquarters (AQU) Haworth Headquarters (HAW) Chicago Cultural Center (CCE) 2.7 - 4.0 - 4.4 - 9.8 - 17.3 - 20.0 - 2.7 - 2.6 - 4.6 - 8.2 - 14.3 - 17.2 - ND ND ND ND ND ND 7.3 ND 16.4 ND 19.4 3.5 - 0.7 - 5.0 - 4.5 - ND - 13.5 - 18.8 Chicago City Hall (CHA) - 0.7 - 5.0 - 4.5 - ND - 13.5 - 18.8 Nature Museum (NAM) 2.7 0.7 4.0 5.0 4.4 4.5 9.8 ND 17.3 13.5 20.0 18.8 Michigan Avenue Structure (MIA) - 0.7 - 5.0 - 4.5 - ND - 13.5 - 18.8 Schwab Rehabilitation Hospital (SCH) 2.7 0.7 4.0 5.0 4.4 4.5 9.8 ND 17.3 13.5 20.0 18.8 Gary Comer Youth Center (GCY) - 0.7 - 5.0 - 4.5 - ND - 13.5 - 18.8 2.2 - 3.4 - 1.1 - 6.3 - 10.6 - 11.6 - Study Site Plant and Soil Science Building (PSS) 38 Table 1.3. (cont’d) Study Site July 2010 2011 August 2010 2011 September 2010 2011 October 2010 2011 November 2010 2011 December 2010 2011 Downtown Chicago Park (DCP) 23.4 - 8.8 - 4.2 - 5.6 - 6.3 - 7.1 - Ford Truck Plant (FOR) 12.6 10.0 1.1 10.3 8.0 16.4 3.5 6.6 8.6 15.4 2.4 6.7 McCormick Parking Structure (MCC) 23.4 - 8.8 - 4.2 - 5.6 - 6.3 - 7.1 - Aquascape Headquarters (AQU) 24.0 - 9.6 - 9.4 - 2.6 - 6.0 - 5.4 - Haworth Headquarters (HAW) 24.4 11.3 4.9 18.3 11.5 10.4 5.5 4.0 5.6 8.2 12.2 5.6 Chicago Cultural Center (CCE) - 13.8 - 10.0 - 9.9 - 5.7 - 9.3 - 6.4 Chicago City Hall (CHA) - 13.8 - 10.0 - 9.9 - 5.7 - 9.3 - 6.4 Nature Museum (NAM) 23.4 13.8 8.8 10.0 4.2 9.9 5.6 5.7 6.3 9.3 7.1 6.4 Michigan Avenue Structure (MIA) - 13.8 - 10.0 - 9.9 - 5.7 - 9.3 - 6.4 Schwab Rehabilitation Hospital (SCH) 23.4 13.8 8.8 10.0 4.2 9.9 5.6 5.7 6.3 9.3 7.1 6.4 Gary Comer Youth Center (GCY) - 13.8 - 10.0 - 9.9 - 5.7 - 9.3 - 6.4 Plant and Soil Science Building (PSS) 5.1 - 1.1 - 12.3 - 7.2 - 5.3 - 4.2 - 39 Table 1.4. Characteristics for building structure, planting media, planted vegetation, maintenance, and human use for green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011. Building Structure Characteristics Height (# story) Slope (%) Media Depth (cm) Vegetation Artificial Water Source** Ornamental and native perennials, turf, shrubs, trees Sedum S, H - Study Site Type* Green Roof Size (ha) Downtown Chicago Park (DCP) E and I 9.91 0 1.0 to 5.0 10 to 122 Ford Truck Plant (FOR) E 4.22 3 1.5 2 McCormick Parking Structure (MCC) Aquascape Headquarters (AQU) Haworth Headquarters (HAW) Chicago Cultural Center (CCE) I 2.43 0 NA 45 to 61 E 2.38 2 to 4 8.3 10 Native prairie perennials, trees Native prairie perennials E 0.42 0 to 4 10.0 to 30.0 10 Sedum S E 0.19 8 1.0 9 to 11 SB Chicago City Hall (CHA) E and I 0.19 11 8 to 46 Nature Museum - 2 roofs (NAM) E 0.14 1.5 to 3 Sculpted terrain 7.0 Sedum, ornamental perennials, low evergreen shrubs Native perennials, shrubs, vines, small trees Sedum, native perennials 40 8 S SB D - Table 1.4. (cont’d) Height (# story) Slope (%) Media Depth (cm) Vegetation Artificial Water Source** Study Site Type* Green Roof Size (ha) Nature Museum - 1 roof (NAM) E and I 0.02 1.5 1.5 5 to 25 Native perennials, one tree - Michigan Avenue Structure (MIA) Schwab Rehabilitation Hospital (SCH) E 0.16 15 1.0 10 to 15 S I 0.09 3 20 to 46 Gary Comer Youth Center (GCY) Plant and Soil Science Building (PSS) I 0.08 3 1.5, raised beds, potted trees 1.0 E 0.03 1.5 1.0 3 to 8 Sedum, ornamental perennials Ornamental and native perennials, annuals, shrubs, small trees Perennials, vegetables, fruits, herbs Sedum 61 * Type: I, intensive; E, extensive; E and I, intensive roofs with shallow media depths in some areas. ** Artificial Water Source: S, sprinkler; SB, subsurface; H, hand-watering; D, drip; W, water feature 41 H, D, W S - Table 1.5. Maintenance regime, primary function, and accessibility of green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011. Study Site Maintenance* Primary Function(s) Accessibility** Downtown Chicago Park (DCP) W, R, P, F, M Recreation P Ford Truck Plant (FOR) F Pollution mitigation, energy savings A McCormick Parking Structure (MCC) B Wildlife habitat creation A Aquascape Headquarters (AQU) W A Haworth Headquarters (HAW) W, F Pollution mitigation, energy savings, wildlife habitat creation Pollution mitigation, energy savings Chicago Cultural Center (CCE) W, R Aesthetics, pollution mitigation, energy savings A Chicago City Hall (CHA) W, R, N Wildlife habitat creation, pollution mitigation, energy savings A Nature Museum - 2 extensive roofs (NAM) W Education, wildlife habitat creation, pollution mitigation, energy savings A Nature Museum - 1 extensive roof (NAM) W, R, P Education, wildlife habitat creation, pollution mitigation, energy savings A Michigan Avenue Structure (MIA) W, R, F Aesthetics, pollution mitigation, energy savings A Schwab Rehabilitation Hospital (SCH) W, P, F, A Therapeutic, aesthetics PR Gary Comer Youth Center (GCY) W, F, H Education, gardening PR Plant and Soil Science Building (PSS) None Education, pollution mitigation, energy savings A A * Maintenance: W, weeding; R, removal of dead plant materials; P, pruning; F, fertilizing; M, mowing; B, controlled burning; A, planting annuals; H, harvesting; N, planting new plant species. ** Accessibility: P, public; A, arranged; PR, private. 42 Table 1.6. Information sources for characteristics of green roof study sites sampled in Michigan and Illinois, U.S.A., in 2010 and 2011. All study sites were a Downtown Chicago Park (DCP), Ford Truck Plant (FOR), McCormick Parking Structure (MCC), Aquascape Headquarters (AQU), Haworth Headquarters (HAW), Chicago Cultural Center (CCE), Chicago City Hall (CHA), Nature Museum (NAM), Michigan Avenue Structure (MIA), Schwab Rehabilitation Hospital (SCH), Gary Comer Youth Center (GCY), and the Plant and Soil Science Building at Michigan State University (PSS). Characteristic Roof Source Average Annual Temperature All Average Annual Precipitation All Size DCP, FOR, AQU, HAW, CCE, CHA, NAM, MIA, SCH, GCY, PSS National Oceanic and Atmospheric Administration, National Climatic Data Center, 2012. National Oceanic and Atmospheric Administration, National Climatic Data Center, 2012. Greenroofs.com, 2011. Size MCC Googlemaps.com, 2010. Slope DCP, FOR, AQU, HAW, CCE, CHA, NAM-1 Intensive/Extensive roof, MIA, SCH, GCY, PSS Greenroofs.com, 2011. Slope NAM - 2 Extensive roofs Soil Depth DCP Steven L. Cantor, 2008. Green roofs in sustainable landscape design. Sylvia Schmeichel, 2010. Personal Correspondence. Soil Depth FOR Soil Depth AQU Steven L. Cantor, 2008. Green roofs in sustainable landscape design. Juana Villagrana, 2010. Personal Correspondence. Soil Depth HAW, MIA, GCY Greenroofs.com, 2011. Soil Depth NAM Steven L. Cantor, 2008. Green roofs in sustainable landscape design. 43 Table 1.6. (cont’d) Characteristic Roof Source Soil Depth CCE Anthony Pacente, 2011. Personal Correspondence. Soil Depth CHA American Society of Landscape Architects, 2002. "ASLA Press Release for 2002 Award Winners". http://www.asla.org/meetings/awards/awds02/chicagocityhall.ht ml. Soil Depth SCH Greenroofs.com, 2011. Soil Depth PSS D. Brad Rowe, 2010. Personal Correspondence. Vegetation DCP http://luriegarden.org/plantlife-list, 2010. Vegetation FOR, NAM Vegetation MCC Steven L. Cantor, 2008. Green roofs in sustainable landscape design. Brendan Daley, 2011. Personal Correspondence. Vegetation AQU Juana Villagrana, 2010. Personal Correspondence. Vegetation HAW Liveroof Original Planting List, 2007. Vegetation CCE Anthony Pacente, 2011. Personal Correspondence. Vegetation CHA City of Chicago, 2011. "Documents: Plants A – C, Plants D – O, Plants P – Z, and Trees, Shrubs, and Vines". http://www.cityofchicago.org/content/city/en/depts /doe/supp_info/chicago_city_hallrooftopgardenplantsandmainte nance.html Vegetation MIA Tom Paulsen, 2011. Personal Correspondence. 44 Table 1.6. (cont’d) Characteristic Roof Source Vegetation SCH Laurie Dettmers, 2010. Personal Correspondence. Vegetation GCY Vegetation PSS http://www.hoerrschaudt.com/rooftop-gardens/gary-comeryouth-center.php#, 2011. D. Brad Rowe, 2010. Personal Correspondence. Year Installed DCP, FOR, HAW, AQU, CCE, CHA, NAM, MIA, SCH, GCY, PSS Greenroofs.com, 2011. Year Installed MCC Chicago Park District, 2002. “Nature Areas, McCormick Place Bird Sanctuary”. http://www.chicagoparkdistrict.com/index.cfm/fuseaction/custo m.natureOasis17. Maintenance DCP Sylvia Schmeichel, 2010. Personal Correspondence. Maintenance FOR Mike Longfellow-Jones, 2010. Maintenance MCC Brendan Daley, 2011. Personal Correspondence. Maintenance AQU Juana Villagrana, 2010. Personal Correspondence. Maintenance HAW Chuck Tubergen, 2010. Personal Correspondence. Maintenance MIA Tom Paulsen, 2011. Personal Correspondence. Maintenance GCY Marjorie Hess, 2011. Personal Correspondence. Maintenance NAM - 1 Intensive/Extensive roof Steven L. Cantor, 2008. Green roofs in sustainable landscape design . 45 Table 1.6. (cont’d) Characteristic Roof Source Maintenance NAM Doug Taron, 2010. Personal Correspondence. Maintenance CCE Jeff Brink, 2012. Personal Correspondence. Maintenance CHA Kevin Carroll, 2012. Personal Correspondence. Maintenance SCH Laurie Dettmers, 2010. Personal Correspondence. 46 Table 1.7. Means and standard errors for variables characterizing vegetation structure of green roofs in Michigan and Illinois, U.S.A., in 2010 and 2011. Green roofs are arranged by year and type. Probability levels reported were calculated with the Kruskal-Wallis one-way analysis of variance. 2010 2011 a b Intensive n=4 Mean (±SE) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Percent Cover Perennial cover (%) Woody cover (%) Mean height (cm) Stem density/ha Tree Shrub Extensive n=5 Mean (±SE) Probability Level Intensive n=4 Mean (±SE) 6 (6) 66* (7) 0 (0) 93 (5) 0.264 0.013 0 (0) 67 (12) 0 (0) 79 (9) 1.000 0.462 2* (1) 0 (0) 0.094 4 (2) 5 (5) 0.283 11* (7) 0 (0) 0.007 4* (3) 0 (0) 0.029 - - - 77 (10) 85 (9) 0.268 - - - 4 (2) 1 (1) 0.180 - - - 50* (4) 12 (5) 0.014 83* (56) 0 (0) 0.094 127* (86.97) 0 (0) 0.029 335* (273) 0 (0) 0.094 802 (392) 897 (897) 0.283 * Indicates a significant difference between intensive and extensive green roofs within the same year. a Two of the same intensive green roofs were sampled in 2010 and 2011. b Three of the same extensive green roofs were sampled in 2010 and 2011. 47 Extensive n=5 Mean (±SE) Probability Level Table 1.8. Means and standard errors for early sampling periods, late sampling periods, and the entire sampling season for variables characterizing vegetation structure of each studied green roof, in Michigan and Illinois, U.S.A., in 2010 and 2011. Probability levels for comparisons between spring and summer within the same year were calculated with the Kruskal-Wallis oneway analysis of variance. Downtown Chicago Park (DCP) 2010 Spring n=9 Summer n=9 Seasonal Mean n=2 Probability Mean (±SE) Mean (±SE) Mean (±SE) Level Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub 24 (12) 33 (9) 0 (0) 25 (12) 25 (13) 62 (16) 0 (0) 38 (13) 25 (1) 48 (15) 0 (0) 31 (6) 0.96 267 (221) 0 (0) 202 (109) 0 (0) 234 (32) 0 (0) 0.70 1 48 a 0.10 1.00 0.54 Table 1.8. (cont’d) Mean (±SE) Ford Truck Plant (FOR) 2010 Summer n=3 Seasonal Mean n=2 Mean (±SE) Mean (±SE) Probability Level 0 (0) 100 (0) 0 (0) 0 (0) 0 (0) 100 (0) 0 (0) 0 (0) 0 (0) 100 (0) 0 (0) 0 (0) 1.00 1.00 1.00 1.00 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 Spring n=3 Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Mean (±SE) Ford Truck Plant (FOR) 2011 Summer n=3 Seasonal Mean n=2 Mean (±SE) Mean (±SE) 0 (0) 100 (0) 0 (0) 0 (0) 0 (0) 98 (0) 0 (0) 0 (0) 0 (0) 99 (1) 0 (0) 0 (0) 1.00 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 - n=9 97 (3) 0 (0) 6 (1) - - - 6 (1) - - - 0 (0) - - Spring n=18 Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Percent Cover Perennial cover (%) Woody cover (%) Mean height (cm) Perennial mean height (cm) Woody mean height (cm) 49 Probability Level a 0.00 1.00 1.00 Table 1.8. (cont’d) McCormick Parking Structure (MCC) 2010 Spring n=3 Summer n=3 Seasonal Mean n=2 Probability Mean (±SE) Mean (±SE) Mean (±SE) Level Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub 0 (0) 53 (6) 0 (0) 0 (0) 0 (0) 98 (1) 0 (0) 1 (1) 0 (0) 75 (23) 0 (0) 0 (0) 1.00 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 a 0.05 1.00 0.32 Aquascape Headquarters (AQU) 2010 Spring n=3 Summer n=3 Seasonal Mean n=2 Mean (±SE) Mean (±SE) Mean (±SE) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Probability Level 0 (0) 67 (13) 0 (0) 0 (0) 0 (0) 89 (5) 0 (0) 0 (0) 0 (0) 78 (11) 0 (0) 0 (0) 1.00 0.13 1.00 1.00 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 50 Table 1.8. (cont’d) Haworth Headquarters (HAW) 2010 Spring n=3 Summer n=3 Seasonal Mean n=2 Mean (±SE) Mean (±SE) Mean (±SE) 0 (0) 100 (0) 0 (0) 0 (0) 0 (0) 100 (0) 0 (0) 0 (0) 0 (0) 100 (0) 0 (0) 0 (0) 1.00 1.00 1.00 1.00 0 (0) 0 (0) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Probability Level 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 Spring n=3 Mean (±SE) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Percent Cover Perennial cover (%) Woody cover (%) Mean height (cm) Perennial mean height (cm) Woody mean height (cm) Haworth Headquarters (HAW) 2011 Summer n=6 Seasonal Mean n=2 Mean (±SE) Mean (±SE) Probability Level 0 (0) 100 (0) 0 (0) 0 (0) 0 (0) 100 (0) 0 (0) 0 (0) 0 (0) 100 (0) 0 (0) 0 (0) 1.00 a 0.09 1.00 1.00 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 - n=3 100 (0) 0 (0) 2 (1) - - - 2 (1) - - - NA - - 51 Table 1.8. (cont’d) Chicago Cultural Center (CCE) 2011 Spring n=6 Summer n=6 Seasonal Mean n=2 Mean (±SE) Mean (±SE) Mean (±SE) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Percent Cover Perennial cover (%) Woody cover (%) Mean height (cm) Perennial mean height (cm) Woody mean height (cm) 0 (0) 57 (6) 23 (7) 0 (0) 0 (0) 43 (5) 24 (6) 0 (0) Probability Level 0 (0) 50 (7) 24 (1) 0 (0) 1.00 0 (0) 4487 (529) 1.00 0.75 - n=3 51 (18) 4 (4) 2 (1) - - - 2 (1) - - - 21 (NA) - - 0 (0) 0 (0) 3958 (1190) 5016 (1175) 52 a 0.09 0.63 1.00 Table 1.8. (cont’d) Mean (±SE) Chicago City Hall (CHA) 2011 Summer n=9 Seasonal Mean n=2 Mean (±SE) Mean (±SE) 0 (0) 0 (0) 0 (0) 88 (3) 8 (3) 1 (1) 97 (2) 5 (3) 0 (0) 92 (4) 6 (2) 1 (1) 59 (40) 39 (39) 49 (10) 951 (506) 185 (185) 568 (383) 0.64 a 0.08 - n=6 92 (8) 8 (8) 54 (18) - - - 44 (14) - - - 169 (NA) - - Spring n=12 Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Percent Cover Perennial cover (%) Woody cover (%) Mean height (cm) Perennial mean height (cm) Woody mean height (cm) 53 Probability Level 1.00 a 0.01 0.24 0.39 Table 1.8. (cont’d) Nature Museum (NAM) Extensive Green Roof 2010 2011 Summer Spring Summer Seasonal n=6 n=12 n=12 Mean n=2 Mean Mean Mean Mean Probability (±SE) (±SE) (±SE) (±SE) Level Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Percent Cover Perennial cover (%) Woody cover (%) Mean height (cm) Perennial mean height (cm) Woody mean height (cm) 0 (0) 86 (3) 0 (0) 75 (5) 0 (0) 74 (6) 0 (0) 74 (1) 1.00 0.98 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 - - n=6 91 (4) 0 (0) 20 (1) - - - - 22 (2) - - - - NA - - 54 Table 1.8. (cont’d) Nature Museum (NAM) Intensive Green Roof 2010 2011 Summer Spring Summer Seasonal n=2 n=4 n=4 Mean n=2 Mean Mean Mean Mean Probability (±SE) (±SE) (±SE) (±SE) Level Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Percent Cover Perennial cover (%) Woody cover (%) Mean height (cm) Perennial mean height (cm) Woody mean height (cm) 0 (0) 78 (3) 2 (2) 3 (3) 0 (0) 82 (1) 3 (1) 2 (1) 0 (0) 80 (2) 4 (2) 3 (2) 0 (0) 81 (1) 4 (1) 3 (0) 1.00 0.46 0.77 0.75 99 (99) 100 (58) 1030 (366) 50 (50) 75 (25) 0.32 197 (197) 504 (205) 767 (263) 0.19 - - - - n=2 78 (12) 5 (5) 49 (19) - - 44 (0) - - - - 284 (NA) - - 55 Table 1.8. (cont’d) Michigan Avenue Structure (MIA) 2011 Spring n=6 Summer n=6 Seasonal Mean n=2 Probability Mean (±SE) Mean (±SE) Mean (±SE) Level Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Percent Cover Perennial cover (%) Woody cover (%) Mean height (cm) Perennial mean height (cm) Woody mean height (cm) 0 (0) 63 (10) 0 (0) 0 (0) 0 (0) 82 (4) 0 (0) 0 (0) 0 (0) 73 (10) 0 (0) 0 (0) 1.00 0.17 1.00 1.00 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 - n=3 89 (11) 0 (0) 27 (14) - - - 35 (20) - - - NA - - 56 Table 1.8. (cont’d) 2010 Summer n=4 Mean (±SE) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Percent Cover Perennial cover (%) Woody cover (%) Mean height (cm) Perennial mean height (cm) Woody mean height (cm) Schwab Rehabilitation Hospital (SCH) 2011 Spring Summer Seasonal n=8 n=8 Mean n=2 Probability Level Mean Mean Mean (±SE) (±SE) (±SE) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 64 (8) 30 (5) 50 (7) 40 (10) 0.02 4 (4) 11 (7) 6 (3) 13 (4) 9 (5) 14 (5) 7 (2) 14 (0) 0.78 0.79 500 (410) 519 (368) 248 (146) 383 (136) 0.76 643 (643) 1606 (965) 2137 (1168) 1872 (265) 0.76 - - n=4 47 (8) 10 (6) 60 (35) - - - - - - - - - 36 (7) 358 (262) - 57 a Table 1.8. (cont’d) Gary Comer Youth Center (GCY) 2011 Spring n=6 Summer n=6 Seasonal Mean Probability n=2 Level Mean (±SE) Mean (±SE) Mean (±SE) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Percent Cover Perennial cover (%) Woody cover (%) Mean height (cm) Perennial mean height (cm) Woody mean height (cm) 0 (0) 48 (5) 0 (0) 0 (0) 0 (0) 60 (3) 0 (0) 0 (0) 0 (0) 54 (6) 0 (0) 0 (0) 1.00 0.11 1.00 1.00 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 - n=3 83 (8) 0 (0) 39 (14) - - - 44 (12) - - - NA - - 58 Table 1.8. (cont’d) Plant and Soil Science Building (PSS) 2010 Spring n=3 Summer n=3 Seasonal Mean n=2 Probability Mean (±SE) Mean (±SE) Mean (±SE) Level Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub a 0 (0) 100 (0) 0 (0) 0 (0) 0 (0) 100 (0) 0 (0) 0 (0) 0 (0) 100 (0) 0 (0) 0 (0) 1.00 1.00 1.00 1.00 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.00 1.00 Indicates a significant difference between Spring and Summer within the same year 59 Table 1.9. Means and standard errors for variables characterizing vegetation structure of the landscape surrounding each green roof during early sampling periods, late sampling periods, and the entire sampling season, in Michigan and Illinois, U.S.A., in 2010 and 2011. Probability levels for comparisons between spring and summer within the same year were calculated with the Kruskal-Wallis one-way analysis of variance Mean (±SE) Downtown Chicago Park (DPC) 2010 Summer n=9 Seasonal Mean n=2 Mean (±SE) Mean (±SE) Probability Level 62 (15) 0 (0) 0 (0) 68 (15) 82 (5) 0 (0) 1 (1) 75 (9) 72 (10) 0 (0) 0 (0) 72 (4) 0.44 1.00 0.84 0.64 674 (303) 79 (79) 299 (189) 34 (34) 486 (187) 57 (23) 0.26 0.69 Spring n=9 Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub 60 Table 1.9. (cont’d) Mean (±SE) Ford Truck Plant (FOR) 2010 Summer n=3 Seasonal Mean n=2 Mean (±SE) Mean (±SE) Probability Level 53 (14) 31 (15) 2 (1) 2 (1) 62 (6) 0 (0) 3 (2) 1 (1) 58 (4) 16 (15) 3 (0) 1 (1) 0.86 0.40 0.73 0.25 82 (41) 293 (145) 46 (46) 370 (173) 64 (18) 332 (38) 0.36 0.65 Spring n=3 Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Mean (±SE) Ford Truck Plant (FOR) 2011 Summer n=3 Seasonal Mean n=2 Probability Mean (±SE) Mean (±SE) Level 72 (8) 17 (8) 2 (1) 2 (1) 57 (15) 27 (13) 3 (2) 3 (2) 64 (7) 22 (5) 2 (0) 2 (1) 0.59 0.86 0.62 0.97 111 (43) 379 (164) 74 (52) 213 (122) 92 (19) 296 (83) 0.76 0.89 Spring n=18 Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub 61 Table 1.9. (cont’d) McCormick Parking Structure (MCC) 2010 Spring n=3 Summer n=3 Seasonal Mean n=2 Probability Mean (±SE) Mean (±SE) Mean (±SE) Level Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub 63 (9) 3 (3) 3 (1) 41 (14) 59 (15) 27 (14) 4 (2) 25 (9) 61 (2) 15 (12) 4 (1) 33 (8) 0.86 0.11 0.81 0.66 127 (65) 405 (173) 74 (42) 193 (83) 101 (26) 299 (106) 0.58 0.54 Aquascape Headquarters (AQU) 2010 Spring n=3 Summer n=3 Seasonal Mean n=2 Probability Mean (±SE) Mean (±SE) Mean (±SE) Level Percent Cover Turf cover (%) 86 (7) Herbaceous cover (%) 7 (4) Shrub cover (%) 2 (2) Tree cover (%) 8 (4) Stem Density/ha Tree 119 (119) Shrub 97 (97) a 0 (0) 43 (43) 0.04 86 (8) 0 (0) 1 (1) 47 (39) 1 (1) 5 (4) 0.05 0.32 0.25 137 (137) 137 (137) 128 (9) 49 (20) 0.80 0.32 62 a Table 1.9. (cont’d) Haworth Headquarters (HAW) 2010 Spring n=3 Summer n=3 Seasonal Mean n=2 Mean (±SE) Mean (±SE) Mean (±SE) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub 37 (14) 44 (14) 0 (0) 2 (1) 33 (4) 48 (4) 0 (0) 5 (3) 0.37 0.50 1.00 0.45 78 (40) 0 (0) 44 (35) 0 (0) 61 (17) 0 (0) 0.08 0.20 Mean (±SE) Shrub 29 (14) 52 (14) 0 (0) 8 (4) Spring n=3 Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Probability Level Haworth Headquarters (HAW) 2011 Summer n=6 Seasonal Mean n=2 Mean (±SE) Mean (±SE) Probability Level 59 (15) 33 (16) 0 (0) 6 (3) 49 (9) 43 (10) 0 (0) 7 (3) 54 (5) 38 (5) 0 (0) 7 (0) 0.42 0.24 1.00 0.80 52 (35) 73 (61) 25 (12) 42 (32) 39 (14) 57 (15) 0.59 0.51 63 a Table 1.9. (cont’d) 2010 Summer n=6 Mean (±SE) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Nature Museum (NAM) 2011 Spring Summer Seasonal n=12 n=12 Mean n=2 Mean Mean (±SE) Mean (±SE) (±SE) Probability Level 25 (12) 53 (13) 8 (5) 24 (7) 21 (8) 47 (8) 3 (1) 31 (7) 23 (12) 44 (13) 9 (3) 25 (9) 22 (1) 46 (1) 6 (3) 28 (3) 0.96 0.95 0.07 0.61 88 (83) 504 (265) 286 (103) 597 (253) 138 (64) 1220 (490) 212 (74) 909 (312) 0.28 0.69 2010 Summer n=4 Mean (±SE) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub Schwab Rehabilitation Hospital (SCH) 2011 Spring Summer Seasonal n=8 n=8 Mean n=2 Probability Level Mean Mean (±SE) Mean (±SE) (±SE) 59 (15) 0 (0) 0 (0) 25 (6) 80 (6) 0 (0) 0 (0) 24 (6) 77 (8) 0 (0) 0 (0) 22 (7) 78 (2) 0 (0) 0 (0) 23 (1) 0.45 1.00 0.48 0.78 93 (43) 0 (0) 112 (45) 13 (13) 48 (48) 0 (0) 80 (32) 7 (7) 0.48 0.29 64 Table 1.9. (cont’d) Gary Comer Youth Center (GCY) 2011 Spring n=6 Summer n=6 Seasonal Mean n=2 Probability Mean (±SE) Mean (±SE) Mean (±SE) Level Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub 41 (8) 6 (5) 9 (3) 34 (12) 49 (11) 6 (6) 8 (5) 37 (17) 45 (4) 6 (0) 8 (0) 35 (2) 0.48 0.74 0.94 0.59 480 (206) 1298 (364) 653 (256) 529 (210) 566 (87) 913 (385) 0.21 0.49 Plant and Soil Science Building (PSS) 2010 Summer n=3 Mean (±SE) Percent Cover Turf cover (%) Herbaceous cover (%) Shrub cover (%) Tree cover (%) Stem Density/ha Tree Shrub 55 (10) 4 (2) 12 (4) 33 (9) 137 (98) 339 (207) a Indicates a significant difference between Spring and Summer within the same year 65 Table 1.10. Percent cover for land cover variables characterizing surrounding landscapes of green roof study sites in Michigan and Illinois, U.S.A., in 2010 and 2011 DCP FOR MCC AQU HAW CCE CHA Studied green roof (%) 10 19 9 5 2 1 2 Other green roof (%) 1 0 3 0 0 9 1 Woody vegetation (%) 10 4 17 26 3 1 1 Herbaceous vegetation (%) 5 35 19 49 22 0 0 Water (%) 0 2 29 6 1 0 0 Non-green Roof (%) 33 37 11 14 36 46 55 Other impervious surface (%) 40 3 12 0 35 42 43 Pre-green roof green space (%) 15 40 36 75 26 1 1 Current green space (%) 26 58 48 80 27 12 33 Potential green space (%) 60 98 59 94 64 58 57 Potential green space (% increase) 125 68 22 17 132 399 1990 66 Table 110. (cont’d) NAM MIA SCH GCY PSS Mean (±SE) Studied green roof (%) 1 1 1 0 0 4 (2) Other green roof (%) 0 0 0 0 0 1 (1) Woody vegetation (%) 36 5 19 15 19 13 (3) Herbaceous vegetation (%) 21 0 19 16 19 17 (4) Water (%) 14 0 0 0 0 4 (3) Non-green Roof (%) 4 57 23 20 21 30 (5) Other impervious surface (%) 24 35 38 48 41 30 (5) Pre-green roof green space (%) 57 5 38 31 38 30 (7) Current green space (%) 58 8 39 32 38 36 (7) Potential green space (%) 62 65 61 52 59 Potential green space (% increase) 6 737 59 64 56 66 (4) 306 (165) a,b Means with different letters are significantly different (p≤0.10). 67 a a b Species 3 common yellowthroat Species 2 red-winged blackbird Species 1 American kestrel Figure 1.1. Range of scales at which birds may respond to green roof vegetation. The presence of green roof vegetation may invoke a response for some bird species selecting habitat at a broad scale (A – home range), for some at a finer level of patch selection (B - feeding area), or at finer level of microhabitat selection (C - one of many feeding sites). 68 Figure 1.2*. Map of the Northeast United States depicting study site locations in 2010 and 2011. Location A sites are in the greater Chicago area: the western location marker is a site in St. Charles, Illinois and the eastern marker represents all sites within Chicago, Illinois. Location B is a site in Holland, Michigan, location C is a site in East Lansing, Michigan, and location D is a site in Dearborn, Michigan. 69 Figure 1.3*. Map of northeast Illinois and southern Michigan depicting study site locations in 2010 and 2011. Location A sites are in the greater Chicago, Illinois area, location B is in Holland, Michigan, location C is in East Lansing, Michigan, and location D is in Dearborn, Michigan. 70 Figure 1.4*. Map of study sites within Chicago, Illinois in 2010 and 2011. Location A is a site at a Nature Museum (NAM), location B is a site at a Michigan Avenue Structure (NAM), location C is at Chicago City Hall (CHA), location D is at Chicago Cultural Center (CCE), location E is at a Downtown Chicago Park (DCP), location F is at McCormick Parking Structure (MCC), location G is at Schwab Rehabilitation Hospital (SCH), and location H is at Gary Comer Youth Center (GCY). 71 Figure 1.5. Schematic aerial view of a study site includes the studied green roof (A) and all land types cover within a 200m radius of the green roof. Other land cover types include other non-studied green roofs (B), woody vegetation (C), herbaceous vegetation (D), water (E), non-green roofs (F), and other impervious surfaces (G). 72 Green Roof Non-Studied Green Roof Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N 0 40 80 Meters 160 Figure 1.6*. Land cover composition of the Downtown Chicago Park (DCP) study site. * Some of the figures in the document are presented in color. For interpretation of the references to color in these figures, the reader is referred to the electronic version of this thesis. 73 Green Roof Water Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N 0 Figure 1.7. Land cover composition of the Ford Truck Plant (FOR) study site. 74 35 70 Meters Meters 140 Green Roof Water Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N 0 Figure 1.8. Land cover composition of the Haworth Headquarters (HAW) study site. 75 25 50 Meters 100 Green Roof Water Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N 0 Figure 1.9. Land cover composition of the Aquascape Headquarters (AQU) study site. 76 25 50 Meters Meters 100 Green Roof Non-Studied Green Roof Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N 0 20 40 Figure 1.10. Land cover composition of the Michigan Avenue Structure (MIA) study site. 77 Meters Meters 80 Green Roof Non-Studied Green Roof Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N 0 Figure 1.11. Land cover composition of the Chicago City Hall (CHA) study site. 78 25 50 Meters 100 Green Roof Non-Studied Green Roof Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N 0 Figure 1.12. Land cover composition of the Chicago Cultural Center (CCE) study site. 79 25 50 Meters 100 Green Roof Non-Studied Green Roof Water Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N 0 Figure 1.13. Land cover composition of the McCormick Parking Structure (MCC) study site. 80 30 60 120 Meters Green Roof Water Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N 0 Figure 1.14. Land cover composition of the Nature Museum (NAM) study site. 81 25 50 100 Meters Green Roof Water Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N Figure 1.15. Land cover composition of the Schwab Rehabilitation Hospital (SCH) study site. 82 0 20 40 80 Meters Green Roof Non-Studied Green Roof Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N Meters 0 Figure 1.16. Land cover composition of the Gary Comer Youth Center (GCY) study site. 83 25 50 100 Green Roof Water Impermeable Surface Non-Green Roof Turf and Perennials Woody Vegetation N 0 20 40 80 Figure 1.17. Land cover composition of the Plant and Soil Science Building (PSS) study site at Michigan State University. 84 Meters 1. 2. 5. 8. 11. 4. 3. 7. 6. 9. 10. 12. 13. Figure 1.18. Examples of summer green roof vegetation on studied green roofs in 2010 and 2011 in Michigan and Illinois, U.S.A., in order of planting media depth from shallow to deep were (1) Ford Truck Plant, (2) Plant and Soil Science Building, (3) extensive roof on a Nature Museum, (4) Chicago Cultural Center, (5) Haworth Headquarters,(6) Aquascape Headquarters, (7) Michigan Avenue Structure, (8) intensive roof on a Nature Museum, (9) Chicago City Hall, (10) Schwab Rehabilitation Hospital, (11) McCormick Parking Structure, (12) Gary Comer Youth Center, (13) Downtown Chicago Park. As a general trend, extensive roofs (1-7) have less vegetation structure compared to intensive roofs (8-13). 85 1. 2. Figure 1.19. Examples of green roof vegetation on (1) extensive and (2) intensive green roofs in 2010 and 2011 in Michigan and Illinois, U.S.A. As a general trend, extensive roofs have less vegetation structure compared to intensive roofs. 86 APPENDIX 87 APPENDIX Table A.1. Planted species on the Downtown Chicago Park (DCP) green roof. Scientific Name Common Name Type (W=woody, P=perennial) ‘Violetta’ New England Aster P Abies concolor White Fir W Acer freemanii 'Jeffsred' Autumn Blaze Maple W Agastache ‘Blue Fortune’ Giant Hyssop P Allium ‘Summer Beauty’ Ornamental Onion P Allium aflatunense ‘Purple Sensation’ Ornamental Allium P Allium atropurpureum Ornamental Allium P Allium christophii Star Of Persia P Allium sphaerocephalon Drumstick Allium P Amorpha canescens Leadplant P Amsonia ‘Blue Ice’ Blue Star P Amsonia hubrichtii Arkansas Blue Star P Amsonia tabernaemontana var. salicifolia Willowleaf Blue Star P Anemone blanda ‘Blue Shades’ Windflower P Anemone hupehensis ‘Praecox’ Japanese Anemone P Anemone hupehensis ‘Splendens’ Japanese Anemone P Anemone japonica ‘Honorine Jobert’ Japanese Anemone P Anemone leveillei Windflower P Arborvitae sp. Arborvitae W Aruncus ‘Horatio’ Goatsbeard P Asclepias incarnata Swamp Milkweed P Asclepias tuberosa Butterfly Weed P Aster ‘October Skies’ Aster P Aster divaricatus White Wood Aster P 88 Table A.1. (cont’d) Scientific Name Scientific Name Common Name Common Name Type Type (W=woody, (W=woody, P=perennial) P=perennial) Aster novae Angliae P Aster oblongifolius ‘October Skies’ October Skies Aster P Aster tataricus ‘Jindai’ Tatarian Aster P Astilbe chinensis var. taquetii ‘Purpurlanze’ Purple Lance Astilbe P Astrantia major ‘Claret’ Masterwort P Astrantia major ‘Roma’ Masterwort P Baptisia ‘Purple Smoke’ Hybrid Wild Indigo P Baptisia leucantha Wild White Indigo P Briza media Quaking Grass P Calamagrostis brachytricha Korean Feather Reed Grass P Calamagrostis x acutiflora ‘Karl Foerster’ Feather Reed Grass P Calamintha nepeta subsp.nepeta Calamint P Camassia cusickii Quamash P Camassia leichtlinii ‘Blue Danube’ Quamash P Campanula glomerata ‘Caroline’ Clustered Bellflower P Carex muskingumensis Palm Sedge P Carex pennsylvanica Pennsylvania Sedge P Carpinus betulus ‘Fastigiata’ Hornbeam W Caryopteris x clandonensis ‘Black Knight’ Bluebeard W Cerastostigma plumbaginoides Plumbago P Cercis canadensis Red Bud W Cercis Canadensis Eastern Redbud W Chasmanthium latifolium Northern Sea Oats P Chionodoxa forbesii ‘Blue Giant’ Glory Of The Snow P Chionodoxa forbesii ‘Violet Beauty’ Glory Of The Snow P Chionodoxa sardensis Glory Of The Snow P 89 Table A.1. (cont’d) Type Common (W=woody, Scientific Name Scientific Name Name P=perennial) Coreopsis verticillata ‘Golden Showers’ Common Name Type (W=woody, P=perennial) Thread Leaf Tickseed P Crataegus sp. Hawthorn W Crocus tommasinianus ‘Barrs Purple’ Crocus P Dalea purpurea Purple Prairie Clover P Datisca cannabina Datisca P Deschampsia caespitosa ‘Goldstaub’ Tufted Hair Grass P Digitalis ferruginea Rusty Foxglove P Dodecatheon ‘Aphrodite’ Shooting Star P Echinacea ‘Orange Meadowbrite’ Coneflower P Echinacea ‘Sunset’ Coneflower P Echinacea pallida Pale Coneflower P Echinacea purpurea ‘Green Edge’ Coneflower P Echinacea purpurea ‘Rubinglow’ Coneflower P Echinacea tennesseensis Tennessee Coneflower P Echinops bannaticus ‘Blue Glow’ Globe Thistle P Epimedium grandiflorum ‘Lilafee’ Longspur Barrenwort P Epimedium x versicolor ‘Sulphureum’ Bishop’s Hat P Eragrostis spectabilis Purple Love Grass P Eryngium bourgatii Mediterranean Sea Holly P Eryngium yuccifolium Rattlesnake Master P Euonymus alatus ‘Compacta’ Burning Bush W Eupatorium maculatum ‘Gateway’ Joe Pye Weed P Eupatorium maculatum ‘Purple Bush’ Joe Pye Weed P Eupatorium rugosum ‘Chocolate’ Joe Pye Weed P Fagus sylvatica European Beech W Filipendula rubra ‘Venusta Magnifica’ Queen Of The Prairie P 90 Table A.1. (cont’d) Common Name Type (W=woody, P=perennial) Fritillaria pallidiflora Fritillary P Gentiana andrewsii Gentian P Geranium ‘Brookside’ Cranesbill P Geranium ‘Dilys’ Cranesbill P Geranium ‘Jolly Bee’ Cranesbill P Geranium phaeum f. album Dusky Cranesbill P Geranium sanguineum ‘Max Frei’ Cranesbill P Geranium soboliferum Cranesbill P Geranium x cantabrigiense ‘Karmina’ Cranesbill P Geranium x oxonianum ‘Claridge Druce’ Cranesbill P Geum rivale ‘Flames of Passion’ Avens P Geum triflorum Prairie Smoke P Gillenia trifoliata Bowman’s Root P Hakenochloa macra Hakone Grass P Helenium ‘Rubinzwerg’ Sneezeweed P Helleborus orientalis Lenten Rose P Hemerocallis ‘Chicago Apache’ – Daylily P Hemerocallis ‘Gentle Shepherd’ Daylily P Heuchera ‘Palace Purple’ Coral Flower P Heuchera richardsonii Coral Bells P Heuchera villosa ‘Autumn Bride’ Coral Bell P Hosta ‘Blue Angel’ Hosta P Hosta ‘Halycon’ Hosta P Hosta ‘Royal Standard’ Hosta P Hosta ‘White Triumphator’ Hosta P Common Scientific Name Name Scientific Name Type (W=woody, P=perennial) 91 Table A.1. (cont’d) Common Name Type (W=woody, P=perennial) Inula magnifica ‘Sonnestrahl’ Fleabane P Jeffersonia diphylla Twinleaf P Kalimeris incisa Cast-iron Plant P Knautia macedonica Knautia P Liatris spicata Blazing Star P Liatris spicata ‘Alba’ White Blazing Star P Limonium latifolium Sea Lavender P Lythrum alatum Loosestrife P Malus 'Sutyzam' Mertensia virginica Sugar Thyme Crabapple Virginia Bluebells W Miscanthus sinensis ‘Malepartus’ Common Eulalia Grass P Molinia caerulea ‘Dauerstrahl’ Moor Grass P Molinia caerulea ‘Moorflamme’ Moor Flame Grass P Molinia litoralis ‘Transparent’ Moor Grass P Mondarda didyma ‘Scorpion’ Bee-balm P Muscari aremeniacum ‘Superstar’ Grape Hyacinth P Narccis poeticus Daffodil P Narcissus ‘Jenny’ Daffodil P Narcissus ‘Lemon Drops’ Daffodil P Narcissus ‘Thalia’ Daffodil P Nepeta faassenii ‘Walker’s Low’ Catmint P Nepeta subsessilis ‘Sweet Dreams’ Catmint P Origanum vulgare ‘Herrenhausen’ Oregano P Paeonia lactiflora ‘Jan Van Leeuwen’ Peony P Paeonia suffruticosa ‘Renkaku’ Tree Peony W Panicum virgatum ‘Shenandoah’ Red Switch Grass P Common Scientific Name Name Scientific Name Type (W=woody, P=perennial) 92 P Table A.1. (cont’d) Common Name Type (W=woody, P=perennial) Pennisetum alopecuroides ‘Cassian’ Fountain Grass P Perovskia ‘Little Spire’ Russian Sage W Persicaria amplexicaulis ‘Firedance’ Knotweed P Persicaria polymorpha White Dragon Knotweed P Phlomis tuberosa ‘Amazone’ Phlomis P Phlox maculata ‘Delta’ Wild Sweet William P Polystichum setiferum ‘Herrenhausen’ Soft Shield Fern P Prunus sargentii Sargent Cherry W Prunus subhirtella ‘Autumnalis’ Higan Cherry W Pycnanthemum muticum Mountain Mint P Pyrus calleryana 'Chanticleer' Chanticleer Pear W Quercus macrocarpa x bicolor ‘Schuettii’ Swamp White Oak W Robinia pseudoacacia ‘Chicago Blues’ Black Locust W Rodgersia pinnata ‘Superba’ Featherleaf Rodgersia P Rudbeckia occidentalis ‘Black Beauty’ Coneflower P Ruellia humilis Wild Petunia P Saliva glutinosa Meadow Sage P Salvia azurea Azure Sage P Salvia pratensis ‘Pink Delight’ Meadow Sage P Salvia verticillata ‘Purple Rain’ Meadow Sage P Salvia x sylvestirs ‘Rugen’ Meadow Sage P Salvia x sylvestris ‘Amethyst’ Meadow Sage P Salvia x sylvestris ‘Blue Hill’ Meadow Sage P Salvia x sylvestris ‘Dear Anja’ Meadow Sage P Salvia x sylvestris ‘May Night’ Meadow Sage P Salvia x sylvestris ‘Wesuwe’ Meadow Sage P Common Scientific Name Name Scientific Name Type (W=woody, P=perennial) 93 Table A.1. (cont’d) Common Name Type (W=woody, P=perennial) Sanguisorba canadensis ‘Red Thunder’ Canadian Burnet P Sanguisorba menziesii Burnet P Saponaria x lempergii ‘Max Frei’ Soapwort P Schizachyrium scoparium ‘The Blues’ Little Bluestem P Scilla mischtschenkoana Squill P Scutellaria incana Skullcap P Sedum ‘Red Cauli’ Stonecrop P Sedum x hybrida ‘Bertram Anderson’ Stonecrop P Sesleria autumnalis Autumn Moor Grass P Sesleria nitida Nest Moor Grass P Silphium laciniatum Compass Plant P Smilacina racemosa False Solomon’s-seal P Solidago ‘Fireworks’ Goldenrod P Sorghastrum nutans ‘Sioux Blue’ Indian Grass P Sporobolus heterolepis Prairie Dropseed P Sporobolus heterolepis ‘Tara’ Prairie Dropseed P Stachys officinalis ‘Hummelo’ Betony Or Hedgenettle P Stachys officinalis ‘Rosea’ Betony Or Hedgenettle P Syringa meyeri ‘Palibin’ Lilac W Taxus cuspidate ‘Capitata’ Yew W Taxus cuspidate ‘Dwarf Bright Gold’ Golden Yew W Taxus x media ‘Hicksii’ Yew W Thalictrum delavayi ‘Elin’ Meadow-rue P Thuja occidentalis ‘Brabant’ Arborvitae W Thuja occidentalis ‘Nigra’ Arborvitae W Thuja occidentalis ‘Pyramidalis’ Arborvitae W Common Scientific Name Name Scientific Name Type (W=woody, P=perennial) 94 Table A.1. (cont’d) Common Name Type (W=woody, P=perennial) Thuja occidentalis ‘Wintergreen’ Arborvitae W Thuja standishii x plicata ‘Spring Grove’ Arborvitae W Tradescantia ‘Concord Grape’ Spiderwort P Tricyrtis formosana Toad-lily P Tricyrtis x ‘Tojen’ Toad-lily P Tulipa ‘Ballade’ Tulip P Tulipa ‘Don Quichotte’ Tulip P Tulipa ‘Ivory Floradale’ Tulip P Tulipa ‘Maureen’ Tulip P Tulipa ‘Purissima’ Tulip P Tulipa ‘Queen of Night’ Tulip P Tulipa ‘Spring Green’ Tulip P Tulipa ‘Tres Chic’ Tulip P Tulipa aucheriana Species Tulip P Tulipa bakeri ‘Lilac Wonder’ Species Tulip P Tulipa hageri ‘Splendens’ Species Tulip P Tulipa polychroma Species Tulip P Tulipa turkestanica Species Tulip P Tulipa urumiensis Species Tulip P Tulipa wilsoniana Species Tulip P Ulmus 'Homestead' W Veronica longifolia ‘Eveline’ Homestead Elm Speedwell Veronica longifolia ‘Lila Karina’ Speedwell P Veronica longifolia ‘Pink Damask’ Speedwell P Veronica spicata ‘Giles Van Hees’ Speedwell P Veronicastrum virginicum ‘Diane’ Culver’s Root P Common Scientific Name Name Scientific Name Type (W=woody, P=perennial) 95 P Table A.1. (cont’d) Common Name Type (W=woody, P=perennial) Veronicastrum virginicum ‘Rosea’ Culver’s Root P Veronicastrum virginicum ‘Temptation’ Culver’s Root P Vitex agnus castus Chaste Tree W Zizia aurea Golden Alexander’s P Common Scientific Name Name Scientific Name Type (W=woody, P=perennial) 96 Table A.2. List of planted species on the Ford Truck Plant (FOR) green roof. Scientific Name Sedum ‘Coccineum’ Type (W=woody, P=perennial) P Sedum acre Common Name Coccineum Two-row Stonecrop Gold Moss Stonecrop Sedum album White Stonecrop P Sedum floriferum Sedum kamschaticum P P Sedum kamschaticum ellacombianum Gold Stonecrop Russian Or Orange Stonecrop Stonecrop Sedum kamschaticum kamtschaticum Stonecrop P Sedum middendorfianum diffusum Diffusum Stonecrop P Sedum pulchellum Lime Stonecrop P Sedum reflexum Blue Stonecrop P Sedum spurium ‘Fulda Glow’ Fulda’s Glow Stonecrop P Sedum spurium ‘Superbum’ Superbum Stonecrop P Sedum telephium Purple Stonecrop P 97 P P Table A.3. List of planted species on the McCormick Parking Structure (MCC) green roof. Scientific Name Common Name Type (W=woody, P=perennial) Aesculus sp. Buckeye W Allium cernuum Nodding Wild Onion P Andropogon scoparius Little Bluestem Grass P Asclepias tuberosa Butterfly Milkweed P Aster azureus Sky-blue Aster P Aster ericoides Heath Aster P Bouteloua curtipendula Side-oats Gramma P Carex annectens P Coreopsis palmata Large Yellow Fox Sedge Copper-shouldered Oval Sedge Prairie Coreopsis Coreopsis tripteris Tall Coreopsis P Deschampsia caespitosa Tufted Hair Grass P Echinacea purpurea Broad-leaved Purple Coneflower P Elymus canadensis Wild Canada Rye P Eryngium yuccifolium Rattlesnake Master P Euphorbia corollata Flowering Spurge P Fragaria virginiana Strawberry P Helianthus occidentalis Western Sunflower P Heliopsis helianthoides False Sunflower P Koeleria cristata June Grass P Monarda fistulosa Wild Bergamot P Parthenium integrifolium Wild Quinine P Pedicularis canadensis Wood Betony P Penstemon digitalis Foxglove Beard Tongue P Petalostemum purpureum Purple Prairie Clover P Phlox pilosa Prairie Phlox P Carex bicknellii 98 P P Table A.3. (cont’d) Scientific Name Common Name Type (W=woody, P=perennial) Phystostegia virginiana False Dragonshead P Potentilla arguta Prairie Cinquefoil P Ratibida pinnata Yellow Coneflower P Rudbeckia hirta Black-eyed Susan P Rudbeckia triloba Brown-eyed Susan P Solidago graminifolia Grass-leaved Goldenrod P Vernonia fasciculata Common Ironweed P Zizia aurea Golden Alexander P 99 Table A.4. List of planted species on the Aquascape Headquarters (AQU) green roof. Scientific Name Common Name Allium cernuum Nodding Onion Type (W=woody, P=perennial) P Aster azureus Sky Blue Aster P Aster ericoides Heath Aster P Aster sericeus Silky Aster P Bouteloua curtipendula Sideoats Gramma P Carex gravida Heavy Sedge P Coreopsis palmata Prairie Coreopsis P Echinacea pallida Pale Purple Coneflower P Heuchera richardsonii Richardson's Alumroot P Koeleria cristata Junegrass P Lespedeza capitata Round-headed Bushclover P Liatris cylindracea Dwarf Blazing Star P Lupinus perennis occidentalis Sundial Lupine P Monarda fistulosa Wild Bergamot P Monarda punctata Spotted Beebalm P Rudbeckia hirta Black-eyed Susan P Rudbeckia subtomentosa Sweet Coneflower P Schizachyrium scoparium Little Bluestem P Solidago nemoralis Gray Goldenrod P Tradescantia ohiensis Ohio Spiderwort P 100 Table A.5. List of planted species on the Haworth Headquarters (HAW) green roof. Sedum acre 'Aureum’ Common Name Goldmoss Stonecrop Type (W=woody, P=perennial) P Sedum album 'Coral Carpet' Coral Carpet Stonecrop P Sedum floriferum 'Weihenstephaner Gold' Bailey's Gold Stonecrop P Sedum hybridum 'Immergrunchen' Stonecrop P Sedum reflexum 'Green Spruce' Spruce Stonecrop P Sedum sexangulare Six-sided Stonecrop P Sedum spurium 'Album Superbum' Caucasian Stonecrop P Sedum spurium 'Dragons Blood' Dragon's Blood Stonecrop P Sedum spurium 'Green Mantle Stonecrop P Sedum spurium 'John Creech' Two-row Stonecrop P Scientific Name 101 Table A.6. List of planted species on the Chicago Cultural Center (CCE) green roof. Alium cernuum Common Name Nodding Onion Type (W=woody, P=perennial) P Dianthus gratianopolitanus ‘Firewitch’ Dianthus P Heuchera richardsonii Alumroot P Juniperus horizontalis Creeping Juniper W Mondara didyma Scarlet Beebalm P Sedum acre Gold Moss Sedum P Sedum cauticola ‘Lidokense’ Lidakense Sedum P Sedum floriferum ‘Weihestaphaner Gold’ Weihestaphaner Gold Sedum P Sedum kamschaticum Russian Stonecrop P Sedum reflexum Blue Stonecrop P Sedum sexangulare Sedum Sexangulare P Sedum spectabile ‘Vera Jameson’ Vera Jameson Sedum P Sedum spurium ‘Fulda Glow’ Fulda’s Glow Stonecrop P Sedum spurium ‘John Creech’ John Creech Sedum P Sedum ternatum Woodland Stonecrop P Sedum x ‘Bertam Anderson’ Bertam Anderson Sedum P Thymus serphyllum ‘Coccineus’ Creeping Thyme P Verbena simplex Narrowleaf Vervain P Scientific Name 102 Table A.7. List of planted species on the Chicago City Hall (CHA) green roof. Achilea millefolium 'Paprika' Common Name Yarrow Type (W=woody, P=perennial) P Achillea millefolium 'Heidi' Yarrow P Achillea sp. 'Schwelienburg' Yarrow P Allium canadense Wild Onion P Amorpha canescens Leadplant P Andropogon scoparius Little Bluestem Grass P Anemone canadensis Meadow Anemone P Anemone patens wolfgangiana Pasque Flower P Anemone virginiana Tall Anemone P Aquilegia canadensis American Columbine P Arabis caucasica 'Flore Pleno' Fiore Pleno Arabis P Artemisia schmidtiana 'Silver Mound' Silver Mound P Artemisia sp. 'Powis Castle' Powis Castle Artemisia P Artemisia stelleriana 'Silver Brocade' Silver Brocade Sage P Asclepias tuberosa Butterfly Weed P Asclepias verticillata Whorled Milkweed P Aster azureus sky Blue Aster P Aster ericoides Heath Aster P Aster laevis Smooth Blue Aster P Aster novae-angliae New England Aster P Aster obiongifolius Aromatic Aster P Aster ptarmicoides Upland White Aster P Aster sericeus Silky Aster P Astragalus canadensis Milkvetch P Baptisia leucophaea Cream Wild Indigo P Blephilia ciliata Ohio Horse Mint P Scientific Name 103 Table A.7. (cont’d) Bouteloua curtipendula Common Name Side-oats Gramma Type (W=woody, P=perennial) P Bouteloua gracilis Blue Gramma P Boutelous hirsuta Hairy Gramma P Buchloe dactyloides Buffalo Grass P Callirhoe involucrata Wine Cups P Campanula poscharskyana Serbian Bellflower P Campanula rotundifolia Harebell P Carex bicknellii Bicknell’s Sedge P Carex cephalophora Woodbank Sedge P Carex gravida Sedge P Carex grayi 'Morning Star’ Morning Star Sedge P Carex pennsylvanica Pennsylvania Sedge P Cassia fasciculata Partridge Pea P Ceanothus americanus New Jersey Tea W Celastrus scandens American Bittersweet Vine Ceraatium tomentosum 'Silberteppich' Silver Carpet P Chrysanthemum leucanthemum Ox-eye Daisy P Clematis virginiana Virgins Bower Vine Coreopsis auriculata 'Nana' Coreopsis P Coreopsis lanceolata Sand Coreopsis P Corydalis flexuasa 'Blue Panda' Blue Corydalis P Corydalis lutea Yellow Corydalis P Crataegus crusgalli Cockspur Hawthorn W Danthonia spicata Poverty Oat Grass P Desmanthus illinoensis Illinois Sensitive Plant P Dianthus allwoodii 'Helen' Pink P Scientific Name 104 Table A.7. (cont’d) Dianthus alpinus Common Name Allwood Pinks Type (W=woody, P=perennial) P Dianthus carthusianorum Dianthus P Dianthus deltoides Dianthus P Dianthus gratianopolitanus ‘spotty’ Cheddar Pinks P Dianthus gratianopolitanus 'Tiny Rubies' Dianthus P Dianthus gratianopolitanus Cheddar Pinks P Dianthus plumarius Dianthus P Diervilla lonicera Dwarf Bush Honeysuckle W Dodecatheon meadii Shooting Star P Echinacea purpurea Purple Coneflower P Echinops bannaticus 'Blue Glow' Globe Thistle P Elymus canadensis Canada Wild Rye P Elymus villosus Silky Wild Rye P Eryngium yuccifolium Rattlesnake Master P Euphorbia polychroma Cushion Spurge P Festuca amethystina 'Bronzegianz' Large Blue Fescue P Festuca glauca 'Elijah Blue' Blue Fescue P Gaillardia 'Kobold' Blanket Flower P Geranium sanguineum Cranesbill P Geranium sanguineum 'Max Frei’ Cranesbill P Geranium sanguineum var. striat Cranesbill P Geum coccineum 'Borisii' Geum P Geum triflorum Prairie Smoke P Gypsophila repens Creeping Baby's Breath P Gypsophila repens 'Rosea' Pink Trailing Baby's Breath P Helianthus mollis Downy Sunflower P Helianthus rigidus Showy Sunflower P Scientific Name 105 Table A.7. (cont’d) Helictotrichon sempervirons Common Name Blue Oat Grass Type (W=woody, P=perennial) P Hemerocalis 'Anzac' Daylily P Hemerocallis 'Little Wine Cup' Daylily P Hemerocallis 'Stella De Oro' daylily Stella De Oro P Heuchera brizoides 'Red Spangif’ Coralbells P Heuchera brizoides 'Huntsman' Coralbells P Heuchera richardsonii Prairie Alum Root P Hieracium pilosella hieracium Mouse-ear Hawkweed P Hystrix patula Bottlebrush Grass P Juncus tenuis Path Rush P Juniperus chinensis 'Sea Greer’ Sea Green Juniper W Kerria japonica Japanese Kerria W Knautia macedonia Knautia P Koeleda glauca Large Blue Hair Grass P Koeleria cristata June Grass P Lavandula angustifolia 'Hidecote' Hidecote Lavender P Lavandula angustifolia 'Munstead’ Munstead Lavender P lberis sempervirens Candytuft P Lespedeza capitata Round-headed Bush Clover P Leymus arenarius Blue Lyme Grass P Liatris aspera Rough Blazing Star P Liatris cylindracea Dwarf Blazing Star P Linaria vulgaris Butter-and-eggs P Lobelia inflata Indian Tobacco P Lychnis 'Flottbeck' Campion P Malus ioensis Prairie Crabapple W Microbiota decussata Russian Arborvitae W Scientific Name 106 Table A.7. (cont’d) Miscanthus sinensis var. gracillin Common Name Maiden Grass Type (W=woody, P=perennial) W Monarda 'Cambridge Scarlet' Bergamot P Monarda fistulosa Wild Bergamot P Opuntia humifusa Prickly Pear Cactus P Origanum vulgare Oregano P Panicum leibergii Prairie Panic Grass P Panicum leibergii Prairie Panic Grass P Panicum virgatum Switch Grass P Papaver orientale 'Brilliant' Poppy P Papaver orientate Poppy P Parthenium integrifolium Wild Quinine P Parthenocissus tricuspidata Boston Ivy Vine Penstemon digitalis Foxglove Beard Tongue P Penstemon pailidus Pale Beard Tongue P Penstemon 'Prairie Dusk' Obedient Plant P Penstemon sp. 'Utahensis' Utah Penstemon P Petalostemum candidum White Prairie Clover P Petalostemum purpureum Purple Prairie Clover P Petrorhagia saxigrage Tunic Flower P Phlox bifida Sand Phlox P Phlox sp. 'Emerald Cushion Blue' Creeping Phlox P Phlox sp. 'Emerald Pink' Creeping Phlox P Polemonium reptans Jacob’s Ladder P Potentilia argute Prairie Cinquefoil P Prenanthes alba Lion's Foot P Ranunculus rhomboides Prairie Buttercup P Ratibida columnifera Mexican Hat P Scientific Name 107 Table A.7. (cont’d) Ratibida pinnata Common Name Yellow Coneflower Type (W=woody, P=perennial) P Rhus aromatica 'Gro-low' Gro-low Sumac W Rosa carolina Pasture Rose P Rudbeckia subtomentosa Sweet Coneflower P Ruellia humilis Hairy Ruellia P Salvia nemorosa 'Mainacht' Purple Salvia P Scutelaria parvula Small Skullcap P Sedum acre Wall Pepper P Sedum album White Sedum P Sedum floriferum Sedum P Sedum hybridum Sedum P Sedum kamtschatcum Orange Stonecrop P Sedum 'Mochren' Mochren Sedum P Sedum reflexum Sedum P Sedum sexangulare Sedum P Sedum spectible 'Matrona' Matrona Sedum P Sedum spurium Two-row Stonecrop P Sedum 'Vera Jameson' Vera Jameson Sedum P Sempervivum arachnoideum Hens And Chicks P Sempervivum-Hybriden sepervivum Hens And Chicks P Smilacina racemosa False Solomon’s Seal P Solidago flexicaulis Broad-leaved Goldenrod P Solidago juncea Early Goldenrod P Solidago nemoraiis Old-field Goldenrod P Solidago speciosa Showy Goldenrod P Solidago ulmifolia Elm-leaved Goldenrod P Sporobolus heterolepis Prairie Dropseed P Scientific Name 108 Table A.7. (cont’d) Stachys byzantina 'Helene von Stein' Common Name Lamb’s Ear Type (W=woody, P=perennial) P Symphoricarpus alba Snowberry W Thymus praecox 'Albiflorus' Albifforus Thyme P Thymus praecox 'Coccineus' Coceineus Thyme P Thymus serpylium thymus Thyme P Tradescantia ohiensis Common Spiderwort P Tradescantia ohiensis 'Zwanenbi’ Spiderwort P Trifolium arvense Rabbitfoot Clover P Viola sororia Common Blue Violet P Scientific Name 109 Table A.8. List of planted species on the Nature Museum (NAM) extensive green roof. Achillea ‘Schwellenburg’ Common Name Schwellenburg Yarrow Type (W=woody, P=perennial) P Achillea millefolium ‘Heidi’ Heidi Yarrow P Allium canadense Wild Onion P Allium cernuum Nodding Wild Onion P Amorpha canescens Leadplant P Andropogon scoparius Little Bluestem P Anemone patens wolfgangiana Pasque Flower P Aquilegia Canadensis American Columbine P Asclepias tuberosa Butterfly Weed P Asclepias verticillata Whorled Milkweed P Aster azurenus Sky Blue Aster P Aster laevis Smooth Blue Aster P Aster ptarmicoides Upland White Aster P Aster sericeus Silky Aster P Baptisia leucophaea Cream Wild Indigo P Bouteloua curtipendula Side Oats Gramma P Buchloe dactyloides Buffalo Grass P Campanula rotundifolia Harebell P Carex bicknellii Bicknell’s Sedge P Coreopsis palmata Prairie Coreopsis P Danthonia spicata Poverty Oat Grass P Dianthus gratianopolitanus Spotty Carnation P Dodecatheon meadii Shooting Star P Geum triflorum Prairie Smoke P Helianthus mollis Downy Sunflower P Helianthus occidentalis Western Sunflower P Scientific Name 110 Table A.8. (cont’d) Heuchera richardsonii Common Name Prairie Alum Root Type (W=woody, P=perennial) P Koeleria cristata June Grass P Lavandula angustifolia ‘Hidcote’ Hidcote Lavender P Liatris aspera Rough Blazing Star P Petalostemum candidum White Prairie Clover P Petalostemum purpureum Purple Prairie Clover P Phlox bifida Sand Phlox P Phlox pilosa Downy Phlox P Sedum ‘Mochren’ Mochren Stonecrop P Sedum ‘Vera Jameson’ Vera Jameson Stonecrop P Sedum acre Goldmoss Stonecrop P Sedum album White Stonecrop P Sedum kamtschaticum Orange Stonecrop P Sedum spurium Two-row Stonecrop P Sempervivium arachnoideum Hens And Chicks P Solidago speciosa Showy Goldenrod P Sporbollus heterolepis Prairie Dropseed P Stachys byzantine Large-leafed Helene Von Stein Lamb’s Ear P Thymus serphllus Creeping Thyme P Scientific Name 111 Table A.9. List of planted species on the Nature Museum (NAM) intensive green roof. Acorus calamus Common Name Sweet Flag Type (W=woody, P=perennial) P Alisma subcordatum Common Water Plantain P Asclepias incarnate Swamp Milkweed P Aster sagitarius drumondii Drummond’s Aster P Blephilia ciliate Ohio Horse Mint P Caltha palustris Marsh Marigold P Carex cristatell Crested Oval Sedge P Carex gravida Common Sedge P Carex lacustris Common Lake Sedge P Carex pennsylvanica Pennsylvania Sedge P Celastrus scandens American Bittersweet Vine Clematis virginiana Virgin’s Blower Vine Echinacea purpurea Purple Coneflower P Elymus villosus Silky Wild Rye P Equisetum arvense Horsetail P Eupatorium maculatum Joe Pye Weed P Geranium sanguineum ‘Max Frei’ Max Frei Cranesbill P Geranium sanguineum var. striatum Bloodred Cranesbill P Helenium autumnale Autumn Sneezeweed P Hemerocallis ‘Little Wine Cup’ Little Wine Cup Daylily P Hystrix patula Bottlebrush Grass P Iris virginica shrevei Blue Flag Iris P Juncus dudleyi Dudley’s Rush P Juncus effuses Common Rush P Juncus torreyi Torry’s Rush P Lobelia cardinalis Cardinal Flower P Scientific Name 112 Table A.9. (cont’d) Lobelia siphilitica Common Name Great Blue Lobelia Type (W=woody, P=perennial) P Panicum virgatum Switch Grass P Parthenium intergrifolium Wild Quinine P Penstemon pallidus Pale Bear Tongue P Polemonium reptans Jacob’s Ladder P Pontedaria cordata Pickerel Weed P Quercus imbricaria Shingle Oak W Rhus aromatic ‘Gro-low’ Gro-low Sumac W Sagittaria latifolia Common Arrowhead P Scirpus atroveriens Dark Green Rush P Scutellaria epilobiifolia Marsh Skullcap P Smilacina racemosa False Solomon’s Seal P Solidago flexicaulis Broad-leaved Goldenrod P Solidago riddellii Riddell’s Goldenrod P Sparganium eurycarpum Bur Reed P Spartina petinata Prairie Cordgrass P Spiraea alba Meadowsweet P Tradescantia ohiensis Common Spiderwort P Vemonia fasciculata Ironweed P Verbana hastata Blue Verbena P Veronicastrum virginicum Culver’s Root P Zizia aurea Golden Alexanders P Scientific Name 113 Table A.10. List of planted species on the Michigan Avenue Structure (MIA) green roof. Allium 'Forescate' Common Name Chive Type (W=woody, P=perennial) P Calamagrostis brachytricha Feather Reed Grass P Calamagrostis 'Karl Forester' Feather Reed Grass P Sedum floriferum 'Weihenstephaner Gold' Sedum P Sedum hybridum 'Immergrunchen' Sedum P Sedum 'Mini Me', Stonecrop P Sedum rupestre 'Angelina' Sedum P Sedum spurium 'Green Mantle' Sedum P Scientific Name 114 Table A.11. List of planted species on the Schwab Rehabilitation Hospital (SCH) green roof. Aster sp. Common Name Aster Type (W=woody, P=perennial) P Buddleia sp. Butterfly Bush P Campanula sp. Bellflower P Echinacea sp. Coneflower P Hemerocallis sp. Daylily P Iris sp. Iris P Liatris sp. Blazing Star P Linum sp. Flax P Pennisetum alopecuroides 'Little Bunny' Bunny Fountain P Perovskia atriplicifolia Russian Sage P Schizachyrium scoparium Little Bluestem P Sumac sp. Staghorn Sumac W Scientific Name 115 Table A.12. List of planted species on the Gary Comer Youth Center (GCY) green roof. Abelmoschus esculentus Common Name Okra Type (W=woody, P=perennial) P Allium schoenoprasum Chives P Anethum graveolens Dill P Aster sp. Aster P Brassica oleracea Broccoli P Capitata sp. Cabbage P Capsicum sp. Yellow Bell Pepper P Capsicum spp. Hot Peppers P Cucumis sativus Cucumber P Cucurbita pepo Zucchini P Daucus carota Carrots P Digitalis spp. Foxglove Mixture P Echinacea sp. Coneflower P Helianthus spp. Sunflowers P Ipomoea batatas Sweet Potato P Lactuca sativa Purple Leaf Lettuce P Lactuca sativa L. var. longifolia Romaine Lettuce P Lactuca sativa var. capitata Butterhead Lettuce P Leucanthemum sp. Daisy P Lilium spp. Lilly Mixture P Liriope spicata Creeping Lilyturf P Narcissus spp. Daffodil Bulbs P Ocimum basilicum Basil P Origanum vulgare Oregano P Petroselinum hortense Parsley P Phaseolus vulgaris Beans P Scientific Name 116 Table A.12. (cont’d) Pisum sativum Common Name Peas Type (W=woody, P=perennial) P Rosmarinus officinalis Rosemary P Solanum lycopersicum Tomato P Solanum tuberosum Potato P Tulipa spp. Tulip Bulbs P Scientific Name 117 Table A.13. List of planted species on the Plant and Soil Science Building (PSS) green roof at Michigan State University. Scientific Name Sedum acre Common Name Gold Moss Stonecrop Type (W=woody, P=perennial) P Sedum album White Stonecrop P Sedum hispanicum P Sedum pulchellum Spanish Stonecrop Russian Or Orange Stonecrop Widowscross Sedum reflexum Blue Stonecrop P Sedum spurium 'Tricolor' Tricolor Stonecrop P Sedum spurium Two-row Stonecrop P Sedum kamtschaticum 118 P P LITERATURE CITED 119 LITERATURE CITED Anderson, S.H., Shugart, H. H., 1974. Habitat selection of breeding birds in an eastern Tennessee deciduous forest. Ecology. 55(4), 828-837. Anderson, J. R., Hardy, E. E., Roach, J. T., Witmer, R. E., 1976. A land use and land cover classification system for use with remote sensor data. Geological Survey Professional Paper 964. United States Government Printing Office, Washington. 41pp. Banting, D., Doshi, H., Li, J., Missios, P., Au, A., Currie, B.A., Verrati, M., 2005. Report on the environmental benefits and costs of green roof technology for the city of Toronto. City of Toronto and Ontario Centres of Excellence–Earth and Environmental Technologies, 88 pp. Basore, N.S., Best, L.B., Wooley, J.B., 1986. Bird Nesting in Iowa No-Tillage and Tilled Cropland. The Journal of Wildlife Management. 50(1), 19-28. Baumann, N., 2006.Ground-nesting birds on green roofs in Switzerland: preliminary observations. Urban Habitats. 4(1), 37-50. Best, L.B., Campa III, H., Kemp, K.E., Robel, R.J., Ryan, M.R., Savidge, J.A., Weeks Jr., H.P., Winterstein, S.R., 1997. Bird abundance and nesting in CRP fields and cropland in the Midwest: A regional approach. Wildlife Society Bulletin. 25(4), 864-877. Bollinger, E.K., Switzer, P.V., 2002. Modeling the impact of edge avoidance on avian nest densities in habitat fragments. Ecological Applications. 12(6), 1567–1575. Brenneisen, S., 2006. Space for urban wildlife: Designing green roofs as habitats in Switzerland. Urban Habitats. 4(1), 27-36. Burke, D.M., Nol, E., 1998. Influence of food abundance, nest-site habitat, and forest fragmentation on breeding Ovenbirds. The Auk 115(1), 96–104. Canfield, R.H., 1941. Application of the Line Interception Method in sampling range vegetation. Journal of Forestry. 39(4), 388-394. Cantor, S.L., 2008. Green roofs in sustainable landscape design, first ed. W. W. Norton and Company, Inc., New York. Carter, T., Fowler, L., 2008. Establishing green roof infrastructure through environmental policy instruments. Environmental Management. 42,151–164. 120 City of Chicago, 2011. Green buildings roofs and homes. Accessed March 6, 2012 at www.cityofchicago.org/city/en/depts/doe/provdrs/green.html. Clements, F.E., 1905. Research methods in ecology. University Publishing Company, Lincoln, Nebraska. Cody, M.L., 1968. On the methods of resource division in grassland bird communities. The American Naturalist. 102(924), 107-147. Coffman, R.R., Davis, G., 2005. Insect and avian fauna presence on the Ford assembly plant ecoroof. Third Annual Greening Rooftops for Sustainable Communities Conference, Awards and Trade Show. Washington, DC. Coffman, R., 2007. Vegetated roof systems: Design, productivity, retention, habitat, and sustainability in green roof and ecoroof technology. Doctoral Dissertation from Ohio State University. Corry, R.C., Nassauer, J.I., 2005. Limitations of using landscape pattern indices to evaluate the ecological consequences of alternative plans and designs. Landscape and Urban Planning. 72, 265–280. Currie, B., Bass, B., 2008. Estimates of air pollution mitigation with green plants and green roofs using the UFORE model. Urban Ecosystems. 11(4), 409-422. Delisle, J., Savidge, J., 1997. Avian use and vegetation characteristics of Conservation Reserve Program fields. Journal of Wildlife Management. 61(2), 318-325. Donnelly, R., Marzluff, J.M., 2006. Relative importance of habitat quantity, structure, and spatial pattern to birds in urbanizing environments. Urban Ecosystems. 9, 99–117. Dvorak, B., Volder, A., 2010. Green roof vegetation for North American ecoregions: A literature review. Landscape and Urban Planning. 96(4), 197–213. Evans, K., Newson, S., Gaston, K., 2009. Habitat influences on urban avian assemblages. Ibis. 151, 19-39. Fernandez-Canero, R., Gonzalez-Redondo, P., 2010. Green roofs as a habitat for birds: A review, Journal of Animal and Veterinary Advances. 9(15), 2041-2052. Fernandez-Juricic, E., 2002. Can human disturbance promote nestedness? A case study with breeding birds in urban habitat fragments. Oecologia. 131(2), 269-278. Fontana, S., Sattler, T., Bontadina, F., Moretti, M., 2011. How to manage the urban green to improve bird diversity and community structure. Landscape and Urban Planning. 101(3), 278-285. 121 Frawley, B.J., Best, L.B., 1991. Effects of mowing on breeding bird abundance and species composition in alfalfa fields. Wildlife Society Bulletin. 19(2), 135-142. Getter, K.L., Rowe, D.B., Andresen, J.A., 2007. Quantifying the effect of slope on extensive green roof stormwater retention. Ecological Engineering. 31(4), 225–231. Getter, K.L., Rowe, D.B., Andresen, J.A., Wichman, I.S., 2011. Seasonal heat flux properties of an extensive green roof in a Midwestern U.S. climate. Energy and Buildings. 43, 35483557. Getter, K.L., Rowe, D.B., Robertson, G.P., Cregg, B.M., Andresen, J.A., 2009. Carbon sequestration potential of extensive green roofs. Environmental Science and Technology. 43(19), 7564–7570. Goddard, M.A., Dougill, A.J., Benton, T.G., 2010. Scaling up from gardens: biodiversity conservation in urban environments. Trends in Ecology and Evolution. 25(2), 90–98. Google Earth, 2011. Google Earth Version 6.1. Accessed November 2, 2011 at www.earth.google.com. Heady, H.F., Gibbens, R.P., Powell, R.W., 1959. A comparison of the Charting, Line Intercept, and Line Point Methods of sampling shrub types of vegetation. Journal of Range Management. 12(4), 180-188. Johnson, D.H., Igl, L.D., 2001. Area requirements of grassland birds: a regional perspective. The Auk. 118, 24-34. Kadas, G., 2006. Rare invertebrates colonizing green roofs in London. Urban Habitats. 4(1), 6686. Kotliar, N.B., Wiens, J.A., 1990. Multiple scales of patchiness and patch structure: a hierarchical framework for the study of heterogeneity. Oikos. 59(2), 253–260. Lundholm, J.T., 2006. Green roofs and facades: A habitat template approach. Urban Habitats. 4(1), 87-101. McKinney, M., 2008. Effects of urbanization on species richness: A review of plants and animals. Urban Ecosystems. 11, 161-176. McNab, W.H., Cleland, D.T., Freeouf, J.A., Keys, J.E., Nowacki, G.J., Carpenter, C.A., 2007. Description of ecological subregions: sections of the conterminous United States. General Technical Report WO-76B. U.S. Department of Agriculture, Forest Service, Washington, DC, 80 pp. Monterusso, M.A., Rowe, D.B., Rugh, C.L., 2005. Establishment and persistence of Sedum spp. and native taxa for green roof applications. HortScience. 40(2), 391-396. 122 Morrison, M.L., Marcot, B.G., Mannan, R.W., 1992. Wildlife-habitat relationships: concepts and applications. University of Wisconsin Press, Madison, WI. National Oceanic and Atmospheric Administration (NOAA), 2012. NOAA National Climatic Data Center. Accessed September 16, 2011 at http://www.ncdc.noaa.gov/cdo-web/ Oberndorfer, E., Lundholm, J., Bass, B., Coffman, R.R., Doshi, H., Dunnett, N., Gaffin, S., Kohler, M., Liu, K.K.Y., Rowe, D.B., 2007. Green roofs as urban ecosystems: Ecological structures, functions, and services. BioScience. 57(10), 823-833. Peck, S., Callaghan, C., Kuhn, M., Grass, B., 1999. Greenbacks from green roofs: forging a new industry in Canada. Status Report prepared for Canada Mortgage and Housing Corporation. Peck and Associates. Prugh, L.R., Hodges, K.E., Sinclair, A.R.E., Brashares, J.S., 2008. Effect of habitat area and isolation on fragmented animal populations. Proceedings of the National Academy of Sciences (PNAS). 105(52), 20770–20775. Pulliam, H.R., 1988. Sources, sinks and population regulation. American Naturalist. 132(5), 652661. Renfrew, R., Ribic, C., Nack, J., 2005. Edge avoidance by nesting grassland birds: a futile strategy in a fragmented landscape. The Auk. 122(2), 618-636. Rowe, D.B., 2011. Green roofs as a means of pollution abatement. Environmental Pollution. 159(8-9), 2100-2110. Rowe, D.B., Getter, K.L., Durhman, A.K., 2012. Effect of green roof media depth on Crassulacean plant succession over seven years. Landscape and Urban Planning. 104, 310-319. Rowe, D.B., Monterusso, M.A., Rugh, C.L., 2006. Assessment of heat-expanded slate and fertility requirements in green roof substrates. HortTechnology. 16(3), 471-477. SAS Institute Inc., 2008. SAS version 9.2. Cary, NC. Savard, J., Clergeau, P., Mennechez, G., 2000. Biodiversity concepts and urban ecosystems. Landscape and Urban Planning. 48(3-4), 131-142. Short, H.L., 1985. Habitat suitability index models: Red-winged blackbird. U.S. Fish and Wildlife Service Biological Report 82(10.95). 20 pp. Sousa, P.J., 1983. Habitat suitability index models: Field sparrow. U.S. Fish and Wildlife Service Report FWS/OBS-82(10.62) 14 pp. 123 United States Fish and Wildlife Service, 1981. 103 ESM standards for the development of Habitat Suitability Index models. 170 pp. Van Renterghem, T. Botteldooren, D., 2009. Reducing the acoustical façade load from road traffic with green roofs. Building and Environment. 44(5), 1081–1087. Vergara, P., Hahn, I., 2009. Linking edge effects and patch size effects: Importance of matrix nest predators. Ecological Modeling. 220(9-10), 1189-1196. Wolf, D., Lundholm, J.T., 2008. Water uptake in green roof microcosms: effects of plant species and water availability. Ecological Engineering. 33(2), 179–186. 124 CHAPTER 2 AVIAN RESPONSE TO GREEN ROOFS IN URBAN LANDSCAPES IN THE MIDWEST UNITED STATES Anthropogenic development of land is a threat to native bird species throughout the world. Increasing urbanization is associated with habitat degradation, fragmentation, and loss of landscape connectivity, resulting in shifts in native bird species distribution, richness, and abundance (Blair, 1996; Miller et al., 2003; Lin, 2006). For example, native grasslands in North America, which provide critical habitat to many bird species, have been reduced by up to 99.9% in some areas of North America since European settlement and are the most threatened and degraded habitat type in North America (Samson and Knopf, 1994). Even though agricultural fields can provide habitat for several grassland bird species (e.g., grasshopper sparrow, Ammodramus savannarum; Savannah sparrow, Passerculus sandwichensis; bobolink, Dolichonyx oryzivorus), conversion of agricultural land to urban land uses corresponds with population decreases of 77% in the Midwest United States (Herkert, 1995; Herkert, 1996). Despite the negative effects of development, bird communities can be significant components of biodiversity in urban landscapes (Savard et al., 2000). Increased development corresponds to declines in species richness of native bird communities (Hohtola, 1978); however, non-native species may prevail despite the lack of native plants because adequate structure is provided by available vegetation and other features (Donnelly and Marzluff, 2004; McKinney, 2006). Vegetated roofs, also known as green roofs, have been identified as a technology capable of providing habitat for bird communities in urban areas (Gedge, 2003; Baumann, 2006; Brennisen, 2006). In addition to energy conservation (Getter et al., 2011) and economic benefits 125 (Peck et al., 1999; Banting et al., 2005; Clark et al., 2008) typically associated with green roofs, these structures present an opportunity to develop large areas of vegetation that could provide bird habitat in urban areas (Eakin et al., in review). Vegetation plays a strong role in determining avian habitat suitability (MacArthur and MacArthur, 1961; Wiens, 1974; Christensen, 1997), hence green roofs that provide appropriate vegetation conditions (e.g., grassland or early successional plant communities) may contribute habitat for the corresponding bird community (e.g., grassland birds). Even though most green roof vegetation is composed of low-growing, shallow-rooted, drought tolerant plant species (e.g. Sedum spp.), plants with greater structural diversity such as taller perennials, shrubs, and trees are available to bird communities on some green roofs (Dvorak and Volder, 2010; Eakin et al., in review). Since birds using green roofs may not be subjected to many of the threats present at ground level (e.g. mesopredators, human disturbance), green roofs may provide greater abundance, productivity, and species richness of bird communities compared to ground level areas with similar vegetation characteristics. Green roofs may also provide elements of bird habitat small (i.e., microhabitat) (Brenneisen, 2006; Oberndorfer et al., 2007) and macro scales (i.e., landscape) by increasing connectivity with other urban green spaces to provide bird habitat systems throughout urban areas (Keitt et al., 1997; Lundholm, 2006; Bierwagen, 2007). By establishing habitat connectivity among green roofs and other urban vegetation patches, cities could pose less of a threat for resident and migratory birds passing through or living in developed landscapes. The potential increase in suitable habitat for bird species due to green roof construction could translate to an increase in species richness and abundance of native bird species in urban areas, indicating improved ecological function. Birds have been observed on green roofs in 126 Europe and North America, with 29 nesting species recorded and even IUCN red-listed species observed (Baumann, 2006; Fernandez-Canero and Gonzalez-Redondo, 2010); however, little quantitative research on bird community composition has been conducted. Roofs designed for specific bird species have resulted in successful nesting attempts by the targeted species (Gedge, 2003; Baumann, 2006), but the contributions of various design elements to bird communities has not been quantified. Quantitative studies relating bird community composition with green roof characteristics could provide insight into how to best implement green roofs for sustainability in urban design while meeting wildlife conservation objectives. This information would help inform the integration of wildlife conservation strategies with green roof technology, urban planning and policy. 127 OBJECTIVES Objectives were to: 1) Quantify the composition of bird communities on green roofs and surrounding landscapes. 2) Quantify the influence of vegetation and non-vegetation structure and composition of green roofs and surrounding landscapes on bird community composition and structure. 3) Quantify the relationship between bird communities observed on green roofs and in landscapes. 4) Make recommendations for green roof design, composition, and management in relation to existing landscapes. 128 METHODS Site descriptions Twelve green roof study sites were located in the Midwest United States in northern Illinois and the Lower Peninsula of Michigan, and were described as: Downtown Chicago Park (DCP), the Ford Truck Plant (FOR), McCormick Parking Structure (MCC), Aquascape Headquarters (AQU), Haworth Headquarters (HAW), the Chicago Cultural Center (CCE), the Chicago City Hall (CHA), a Nature Museum (NAM), a Michigan Avenue Structure (MIA), Schwab Rehabilitation Hospital (SCH), Gary Comer Youth Center (GCY), and the Plant and Soil Science Building at Michigan State University (PSS) (Table 2.1). Green roofs represented a range of conditions for roof area, roof type (i.e., extensive roofs, 0-15cm planting media depth; intensive roofs, >15cm planting media depth; Rowe, 2011), building height, vegetation type, and slope. Land use, accessibility and maintenance, which depended on the intended purpose of the roof (Table 2.2), had the potential to influence bird community response. Each study site was composed of a green roof and the surrounding landscape within 200m of the roof. This landscape area (≥14ha for all green roofs) encompasses the home range size of several bird species (e.g., killdeer, Charadrius vociferus; red-eyed vireo, Vireo olivaceus; Cimprich et al., 2000; Jackson and Jackson, 2000) that may utilize green roofs and their surrounding landscape to fulfill their life requisites (Eakin et al., in review). We sampled 8 sites in 2010 and 8 sites in 2011, with 4 sites being sampled both years. All roofs sampled in 2011 were elevated above ground level (not on subterranean structures); the greatest potential for additional green areas on roofs in urban areas is above ground level. Also, because bird communities are affected by patch size and edge effects (Bollinger and Gavin, 129 1992; Delisle and Savidge, 1997; Johnson and Igl, 2001), care was taken to select the largest green roofs possible that represented a range of vegetation types and conditions and were accessible through the building owner or manager. All study sites sampled in 2010 and 2011were within the administrative and biological Mississippi Flyway Migratory Route (United States Fish and Wildlife Service, 2011). However, Chicago, Illinois, lies on a major flyway, and numerous birds pass through this area during migration compared to other green roof areas in our study. The range of monthly mean temperatures at each study site during the early to peak bird nesting season (April - June) was between 11.3C and 22.2C in 2010 and between 7.9C and 21.5C in 2011 (Table 1.2; National Oceanic and Atmospheric Administration, 2012). The range of monthly mean temperatures at each study site during post-nesting and brood-rearing season (July-September) was between 16.5C and 26.1C in 2010 and between 15.9C and 26.3C in 2011. The range of monthly total precipitation at each study site during the early to peak bird nesting season was between 5.9cm and 20.0cm in 2010 and between 3.5cm and 18.8cm in 2011 (Table 1.3; National Oceanic and Atmospheric Administration, 2012). The range of monthly total precipitation at each study site during post-nesting and brood-rearing season (July-September) was between 1.1cm and 24.4cm in 2010 and between 9.9cm and 18.3cm in 2011. Vegetation measurements Vegetation sampling was conducted during two periods in 2010 and 2011: one during the early to peak bird nesting season and one during post-nesting and brood-rearing season. By sampling during the spring (April-June) and summer (July-October), vegetation was representative of that available to bird communities during both sampling periods (Short, 1985; Basore et al., 1986; Best et al., 1997). 130 The sampled portions of landscapes surrounding green roofs were safe, accessible, and included clearly definable vegetation areas (areas with exposed soil with potential to support vegetation as the minimum requirement of). Planter boxes attached to buildings, street median vegetation, street trees in grates, and vegetation on other green roofs in the surrounding landscapes were not sampled in the field, however, were represented in the aerial land cover analysis. The line intercept method (Canfield, 1941) was used to quantify vegetation cover of turf grass, herbaceous perennial cover, and shrub and tree canopy on green roofs and in vegetation areas in surrounding landscapes. One-meter belt transects (Clements, 1905) were used on green roofs and surrounding landscapes to determine species presence and stem densities of woody plants. Transects ranged from 3.8-200.0m long and were systematically placed perpendicular to the gradient of vegetation types on each roof and landscape area. The length of transects corresponded to the size of green roofs. The length of the transect that intersected mowed lawn, perennial, shrub, and tree cover was recorded and used to calculate the percent cover of each vegetation type. The point intercept method (Heady et al., 1959) was used to calculate percent cover of different vegetation types and quantify mean vegetation height. Every 5m, vegetation intersecting transects was identified by type (perennial sedum, non-sedum perennial, woody vegetation) and the height of the vegetation at the intersecting point was measured. Aerial land cover Existing and potential green space for each study site were determined by importing Google Earth (Version 6.1; Google Earth, 2011) images into ArcMap using ArcGIS version 9.2 (ArcGIS version 9.2; ESRI, 2006). Images were then georeferenced, digitized, and classified by land cover type (i.e., studied green roofs, other green roofs, woody vegetation, herbaceous 131 vegetation, water, conventional roofs, impermeable surfaces). Green space in each landscape consisted of green roofs and woody and herbaceous vegetation cover types. Potential green space within a study site was calculated assuming that all existing conventional roofs could be vegetated and was portrayed as total area covered by current green space (on roofs and on the ground) and non-green roofs. Bird surveys Point count surveys (Bibby et al., 1992) were conducted during the nesting and brood rearing seasons (April - July) during both years on green roofs and in their respective landscapes. In 2010, bird surveys were conducted on four to six sampling dates. The accumulation rate of species detected appeared to level-off near the apparent asymptote at each study site and on each studied green roof after three days of surveys, which indicated that few additional species would be detected regardless of additional sampling effort. In 2011, surveys were decreased to four sampling dates because of a lack of additional species detected after the third sampling date of all green roofs sampled the previous year. One extra sampling date (the fourth day) was included as a contingency for variation between years. Surveys began at dawn and were conducted for up to 3 hours (Robbins, 1981). Sampling order of green roofs and landscapes was rotated each survey date, when feasible according to building owners, to equalize differences in bird presence and detection related to time of day. The location and number of survey points on green roofs and their surrounding landscapes were assigned to maximize observed roof area, avoid overlapping observations between survey points, and minimize observer disturbance. One to three sampling points were located on each green roof, depending if structures were present that obscured a view of the entire roof from a single point. Up to four landscape survey points were located in each cardinal direction around each 132 roof within 200m of the roof. The vegetation area closest to the green roof was selected where another building obscured the view between vegetation areas within 200m. Sampling was preceded by a 2-min settling period and followed by three continuous sampling periods of 7-min each (Sauer et al., 1994). The species, number, and behavior (i.e., resting, foraging, nesting, calling/singing, mating, defending/aggressing) of all birds observed were recorded. Sound identification was not used due to varying background noise levels among sites. Incidental observations such as low-flying birds over green roofs and other wildlife on green roofs were noted for anecdotal reference. We focused on non-invasive, native bird communities (Cornell lab of Ornithology and the American Ornithologists Union, 2012) for data analyses given our study was primarily interested in the conservation opportunities provided by green roofs. Native bird species that present possible harm to other native bird species of greater conservation concern (i.e., brownheaded cowbird, Molothrus ater; Walkinshaw, 1991) or that are urban adapted species (i.e., ringbilled gull, Larus delawarensis; Pollet et al., 2012; American crow, Corvus brachyrhyn; Verbeek and Caffrey, 2002) were excluded from model analysis. Waterfowl and songbirds were analyzed separately because green roofs likely cannot provide a source of water generally required by waterfowl, and songbirds generally do not require large amounts of water (Brewer et al., 1991). Because of limited waterfowl habitat potential on green roofs, waterfowl species were analyzed individually in a single-species model. Comparisons at the sampling point level (i.e., green roof or surrounding landscape) were made for waterfowl species (henceforth WS) observed on 25% or more of the studied green roofs. Because of the potential for green roofs to fulfill life requisites and provide conservation value for several native songbird species (Eakin et al., in review), songbirds were collectively analyzed in a multi-species model. The multi-species 133 model incorporated data for all native bird species, excluding waterfowl and urban adapted species, to estimate species richness and occurrence and use probabilities at the sampling-point and site level on green roofs and surrounding landscapes. Of these bird species, native species (henceforth NS) observed on at least 25% of green roofs and those observed on green roofs that are either listed by the U.S. Fish and Wildlife Service (2010) to be rare or declining species in the Midwest United States, or by Sauer et al. (2011) to have experienced declines from 20002010 for the Lower Great Lakes/St. Lawrence Plain, were included in species specific analyses. All field sampling procedures were exempted by the Michigan State University Institutional Animal Care and Use Committee and did not require animal use form approval. Descriptive analysis and statistics Point-count data were used to calculate summary statistics of bird communities observed on green roofs and in surrounding landscapes for each year sampled. Relative abundance, species richness, and feeding guild during the breeding and nesting periods (De Graff et al., 1985) was presented to demonstrate community composition for each study site. Relative abundances were based on the total number of bird observations at each roof or within each landscape. Because these analyses did not account for detection probabilities (the probability that a species present during sampling would be observed), further data analysis that accounted for detection probabilities was conducted. We applied single-species and multi-species frameworks to the multi-scale occupancy model used by Mordecai et al. (2011) to jointly estimate probabilities of bird species occurrence (ψ, the probability that a species is ever present at a survey site), use (θ, the probability that a species was present during sampling at a survey site), and detection (P, the probability that a species present at a survey site was detected during sampling). Occupancy describes the 134 probability that a species is ever present at a site. Use describes the probability of species presence during a given visit and represents how often a species is present at a site. The model only included data for a single year for each roof. Only the first year of data was included for roofs sampled both years. We constructed detection histories from the point count data that indicated with a 0 or 1 whether a species was detected during a 7-min sampling period for each visit to a point. Logit-linear models were used to estimate the effects of site-level covariates (i.e., sampling point location) on each parameter in the multi-scale model. Single- and multi-species frameworks were applied to the following logistic regression equations: logit (ψi) = α0r + α1greenroofi logit (θij) = β0r + β1greenroofi + β2dateij logit (Pijk) = δ0 + δ1greenroofi + δ2Previj where i represents site, j represents visits, and k represents the 7-minute sampling period within a visit; greenroof is a binary indicator for whether site i is located on a green roof or in the surrounding landscape; date is the standardized ordinal date of survey j; Prev is a binary variable for previous detection; α0r and β0r represent random-effect intercepts for occupancy and use, respectively, which account for spatial dependence between sites located on or near the same roof, r; and δ0 represents the fixed intercept for detection. The single-species and multi-species models included the same covariates for estimating differences in the likelihood of use and occurrence on roofs and in corresponding landscapes. In the multi-species framework, species are treated as a random effect which interacts with each model parameter to produce species-specific parameter estimates for the logit-linear models of 135 occupancy, use, and detection (Dorzaio et al., 2006; Royle and Dorazio, 2008). Combining the species detection data into one model results in a more parsimonious approach to parameter estimation, in addition to allowing for the incorporation of rarely detected species with sparse data (Royle and Dorazio, 2008). This random effect allows for occupancy and use probabilities to be estimated for all observed species at all points, including those points where a species may have gone undetected (Zipkin et al., 2010). We estimated the parameters and calculated estimates using WinBUGS (Spiegelhalter et al., 2003) in program R. Each parameter was assigned a set of hyperparameters (i.e., mean and standard deviation) that were given non-informative prior distributions. We standardized all covariates to have a mean of 0 and a unit variance of 1. We examined the single-species model results based on 3 chains of 5,000 iterations after discarding the first 2,500 iterations and thinning by 10; this process resulted in 750 values forming the posterior distribution for each parameter. The multi-species model results were based on 3 chains of 20,000 iterations after discarding the first 5,000 iterations and thinning by 15; this process resulted in 3,000 values forming the posterior distribution for each parameter. These settings provided an acceptable level of Markov chain convergence ( ̂ statistic, or scale reduction factor, <1.1 for all parameters; Gelman et al., 2003). 136 RESULTS Observed and estimated bird community structure and composition A total of 69 bird species were observed in 2010 and 2011 (Table 2.3). Of those species, 29 were observed on green roofs, with three observed only on green roofs and not in surrounding landscapes (American tree sparrow, Spizella arborea; blackburnian warbler, Setophaga fusca; Nashville warbler, Oreothlypis ruficapilla). The four bird species that were observed on >50% of roofs were American robin (Turdus migratorius, 75% of roofs), European starling (Sturnus vulgaris, 67% of roofs), house sparrow (Passer domesticus, 58% of roofs), and American goldfinch (Carduelis tristis, 50% of roofs). Thirty-eight species observed in surrounding landscapes were not observed on green roofs. Eighty-six percent of the bird species (25 species) observed on green roofs were non-invasive, native species, and two were waterfowl species (Canada goose, Branta canadensis; mallard, Anas platyrhynchos). Of the observed native species, 16 and 22 species represented ground-foraging feeding guilds on green roofs and in the adjacent landscapes, respectively. Six and 15 non-invasive, native species on green roofs and in their surrounding landscapes, respectively, represented shrub-foraging, low-canopy foraging, or bark-gleaning feeding guilds. Most birds observed (>50%) on green roofs were resting and/or foraging. Birds were also observed nesting (Canada goose on the Ford Truck Plant and Nature Museum roofs; mallard on the Nature Museum and Aquascape Headquarters roofs; killdeer on the Ford Truck Plant roof; red-winged blackbird, Agelaius phoeniceus, on the Downtown Chicago Park and McCormick Parking Structure roofs), and Canada goose goslings were reared on the Ford Truck Plant roof. Though not counted as bird species observed on green roofs, barn 137 swallows (Hirundo rustica) and chimney swifts (Chaetura pelagica) were observed to dive over green roofs, presumably to eat bugs attracted to the vegetation. Mean detection probabilities generated using the single-species model were 88% and 63% for Canada goose and mallard. Similarly, the range of mean detection probabilities for NS generated using the multi-species model was between 66% and 82%. These detection probabilities indicated that bird species that occupied green roofs were often not detected during point-count surveys. Median species richness was higher on green roofs than in surrounding landscapes when estimated at the sampling-point level and at the site level (Fig. 2.1 and 2.2). At the site level all survey points were grouped (i.e., a roof with three survey points would be grouped together), and thereby increased the uncertainty in estimated species richness for a landscape or roof at a study site. All roofs had similar mean estimated species richness and higher estimated species richness than their respective landscapes (Fig. 2.3), likely because vegetation present on roofs had similar structure, and vegetation present in surrounding landscapes represented a variety of structures and thus increased variability in estimations. Though estimated species richness is similar among roofs, differences in estimated species richness and the number of species observed on roofs indicated differences between bird occupancy and use on green roofs (Fig. 2.4; Table 2.3). Canada goose and mallard were observed on ≥25% of the studied green roofs with mean occurrence probabilities ≥95% and ≥72% at sampling points on green roofs and in surrounding landscapes, respectively; however, occurrence estimates were highly variable, likely suggesting no biological effect (Fig. 2.5). Mean occurrence probabilities for NS were >94% on green roofs and 59-93% in landscapes (Fig. 2.6). Bird species that were observed on green roofs and are listed by the U.S. Fish and Wildlife Service (2010) to be rare or declining species in the Midwest 138 United States, or by Sauer et al. (2011) to have experienced significant negative trends from 2000-2010 for the Lower Great Lakes/St. Lawrence Plain showed similar trends. Mean occurrence probabilities were >91% on green roofs and were 39-74% in landscapes (Fig. 2.7). These trends suggest that if a species is observed on a green roof, that species will occupy green roofs more consistently than they will the surrounding landscape. The mean occurrence probability for each NS on each green roof was >80% for most species and mean occurrence probabilities in landscapes were >20% (Fig. 2.8). In addition, little difference in mean occurrence probability between species was estimated in the model, and no pattern between roof characteristics (i.e., roof type, roof area, percent green space at the study site) and mean occurrence probability on green roofs was observed; however, during surveys differences in how bird species used green roofs were noted. Similar to occurrence probabilities, use probabilities for Canada goose and mallard showed no differentiation between roofs and landscapes (Fig. 2.9). Mean use probabilities for NS and rare and declining species were lower on green roofs than in surrounding landscapes, and mean use probabilities were lower than mean occurrence probabilities for these species (NS, 111% on green roofs and 5-61% in the surrounding landscapes, Fig. 2.10; rare and declining species, 0-1% on green roofs and 4-17% in the surrounding landscapes, Fig. 2.11). The mean use probabilities of NS for intensive green roofs were distributed more evenly between high and low use probabilities than extensive roofs (Fig. 2.12). Roof size or percent green space of a study site did not have a discernible effect on use probability (Fig. 2.13 and 2.14); however, error margins for all estimates had wide overlaps that demonstrated the uncertainty present in estimates. Mean use probabilities on green roofs remained <30% for most bird species until use probability in the landscape reached 80% (Fig. 2.15). 139 Our results indicate the amount of time various bird species use green roofs, as estimated by use probability, is positively related to the amount of time those same species use the landscape directly surrounding green roofs. Green roof type also influences how much time birds use green roofs; intensive green roofs generally experience higher bird use probabilities than extensive roofs. Roof size appeared to be related to the number of species (uncorrected for detection probabilities) observed on green roofs (Fig. 2.4); however, use and occurrence probabilities that did account for detection probabilities did not concur with this trend. Other factors, such as roof size, may not demonstrate a clear effect on bird use in our study, but are biologically important (e.g., home range requirements correspond to a certain size area of appropriate conditions) and likely contribute to patterns of bird use of green roofs. Roofs with conditions present that provide mean use probabilities >80% for at least one NS species include MIA, SCH, and NAM (Fig. 2.12). These roofs represent intensive and extensive roof types, have building height ranging 1.9-15 stories, and have vegetation ranging from only herbaceous on MIA to a diversity of herbaceous, shrubs, and trees on SCH (Table 2.1). 140 DISCUSSION Bird use Green roofs present an opportunity to provide large areas of vegetation that could serve as bird habitat in urban areas (Eakin et al., in review). Although green roofs provide vegetation conditions that fulfill the habitat requirements for several bird species native to the Midwest United States, little is known about which birds use green roofs. During our study, 25 noninvasive, native bird species were observed on green roofs during 2010 and 2011. Using a multiscale occupancy model, we estimated the mean number of species to occur on each roof to be between 36 and 40 (Fig. 2.3); however, the same species likely do not occur on all roofs. The total number of bird species that occur across all green roofs is greater than the average number for any given roof due to differences in vegetation structure and composition on the roofs and landscape attributes surrounding roofs. Our observations support the concept that green roofs can provide habitat for a diversity of birds including those of conservation value. Birds on green roofs were observed feeding, bathing, using a diversity of vegetation for cover, perching, territory defense, nesting, and rearing young. Ground nesting birds (i.e., Canada goose, mallard, killdeer) were observed nesting on green roofs that were above ground level. The elevated position of green roofs may offer protection from predators typically found at ground level and other ground level disturbances (i.e., human disturbance). Although Canada goose had a relatively high mean probability of occurrence (99%) on green roofs (Fig. 2.5), their low mean use probability (Fig. 2.9) indicates variation in the amount of the time spent on different roofs (i.e., Canada goose were likely constantly present on roofs with successful nesting attempts). Canada geese reared goslings on an extensive green roof established in Sedum spp. 141 three stories above the ground. This roof had several areas of flat, conventional roofing membranes that held water after rain and/or irrigation events that could have provided water goslings. However, Canada geese and mallards nested unsuccessfully on two other extensive roofs, one established with a Sedum spp. mixture and herbaceous perennials, and one established with native perennials. Contributing factors to nesting success may be the occasional presence 2 of water from irrigation and the large size of the roof (42,177m ). Bird species that forage on the ground and in low tree and shrub canopies have also been observed on green roofs (e.g., killdeer, common yellowthroat, Geothlypis trichas; song sparrow, Melospiza melodia). Because extensive green roofs typically have shorter vegetation than intensive roofs (Getter and Rowe, 2006; Eakin et al., in review), bird species in ground foraging guilds are likely more prevalent on extensive roofs, while shrub and low canopy foragers and bark gleaners are more likely to occur on intensive roofs. Of the bird species observed on studied green roofs, those in shrub and low canopy foraging and bark gleaning guilds (6 species) were only observed on intensive roofs with shrub and/or tree cover (Table 2.3). However, ground foraging bird species were also observed on all intensive roofs, likely because shrub or tree cover was not the dominant cover type on any intensive roof. The multi-scale model outputs indicated that green roofs provide bird habitat complementary to that in surrounding landscapes. Point- and site-level species richness estimates for green roofs were higher than in surrounding landscapes, indicating that green roofs are providing bird habitats that attract novel species to urban landscapes. Individual bird species included in the multi-species model also had higher mean occurrence probability on green roofs than in surrounding landscapes (Fig. 2.6, 2.7, and 2.8). This result indicates that more bird species were likely to occur at a green roof point than at a landscape point and some species not 142 present in surrounding landscapes were likely present on roofs (Fig. 2.8). Since mean occurrence probability on green roofs for most observed bird species was >80% (Fig. 2.8), even those not observed on green roofs during the study were likely occurring on the roof at some time. However, the low mean use probabilities on green roof points compared to landscape points (Fig. 2.15) indicate that birds were using green roofs only for a short period of time, assuming use probability is indicative of the proportion of time a bird species is present. Only when a species is present in a landscape >95% of the time is that species present on a green roof >50% of the time. Based on the short time most bird species were on green roofs, it does not appear that these birds use resources on green roofs to fulfill the majority of their life requisites. Since studied green roofs had a relatively small size compared to their surrounding landscapes, it was expected that birds would be present in surrounding landscapes for a greater proportion of time than on green roofs. Since the most observations on roofs were of birds foraging and resting, our observations suggest that birds are attracted to green roofs as temporary foraging and resting sites. Hence, occupancy is high on green roofs even though occupancy in the landscape is low (Fig. 2.8) because birds are attracted to the foraging substrate and other structure on green roofs. However, because bird use is short-term, the likelihood of documenting use is relatively low, even when use of the landscape is relatively high (Fig. 2.15). The high species richness and high occupancy probability on green roofs suggest that green roofs may have potential to increase habitat connectivity for bird species during the breeding season. Further research is necessary to investigate the potential for green roofs to function as stepping stones for migratory birds to more effectively traverse urban areas. Telemetry equipment could be used to further understand how birds move through the landscape, when they are on green 143 roofs, and where they are nesting. These observations could help determine which structural and vegetation green roof conditions various species prefer. Our results identified green roof type as a characteristic that may affect habitat availability for native bird communities. Use probabilities between intensive and extensive roofs indicated the ability of intensive roofs to support some bird species for a greater proportion of time than extensive roofs, which may indicate that intensive roofs can provide the majority of resources needed to support these bird species, whereas extensive roofs appear better suited to provide bird habitat complimentary to that in the surrounding landscapes. To better understand the differences in how birds use green roofs compared to the surrounding landscapes (i.e., reduce credible intervals), future studies could examine green roofs with greater variability in roof size and vegetation structural diversity. Conversely, several roofs with similar structure could be studied to hone in on the bird community that uses specific roof types (e.g., intensive, native prairie vegetation with >60% cover, mean height of 1.1m). In addition, percent cover of vegetation on a roof could be studied for possible effect on bird community species richness, as was noted for insect communities on green roofs in a study by Monsma (2011). Comparisons of expected and observed bird species Generally, bird communities known to use various vegetation types on the ground also used green roofs with similar vegetation composition and structure: Species in ground foraging guilds, such as killdeer (Charadrius vociferus) and common grackle (Quizcalus quizcula) (De Graff et al., 1985), were observed feeding on green roofs established with sedum and turf grass; species associated with tall herbaceous vegetation, such as red-winged blackbird (Short, 1985), were observed on roofs established with perennials such as little bluestem (Schizachyrium scoparium) and coneflower (Echinacea sp.); and forest edge associated species, such as downy 144 woodpecker (Picoides pubescens) (Schroeder, 1982), were observed on roofs established with shrubs and trees. Based on foraging guilds (De Graff et al., 1985), species expected on individual green roofs were observed on green roofs (Table 2.3), and other species that were not expected based on their foraging guild (i.e., upper canopy foraging species) were also observed on green roofs. The presence of these “unexpected” species may be a result of fewer disturbances on green roofs compensating for the lack of suitable vegetation charateristics, or/and the presence of features on the roof that complement the bird habitat provided by the surrounding landscape. Without ground predators and human disturbance, vegetation that would otherwise be unsuitable may increase in suitability for a species. For example, the Aquascape Headquarters Green Roof in St. Charles, Illinois established with native prairie plant species was expected to provide suitable vegetation for red-winged blackbirds, but other species whose habitat requirements were not fulfilled by this green roof (i.e., American goldfinch; eastern kingbird, Tyrannus tyrannus; and song sparrow) were also observed. If these species were present despite that their habitat requirements were only partially fulfilled by the green roof, a green roof and its surrounding landscape likely combine to provide some habitat for species. This and similar findings support the concept that green roofs can contribute to bird species habitat, even if the vegetation does not fulfill all habitat requirements. Bird community conservation potential Because species richness on green roofs was greater than in the surrounding landscapes, green roof construction has potential to increase bird species richness in the area immediately surrounding a green roof. The difference between species richness for green roofs and landscapes may vary because of landscape composition. A landscape that does not provide 145 suitable songbird habitat might experience a ‘spill-over’ affect from a green roof that provides high-quality songbird habitat, or conversely a green roof lacking suitable songbird habitat might function as an ecological ‘sink’ (i.e., an area with low reproductive success that is reliant on immigration to sustain a population; Thompson, 2005) in a landscape that provides high quality songbird habitat. These potential scenarios demonstrate the need for further research on the population dynamics of selected species associated with green roofs to accurately describe the affect green roof construction will have on species richness of various sites. In our study sites, typically, observed bird species on the green roofs in suburban and semi-rural landscapes were a subset of those observed in the landscapes. However, in highly urban landscapes, green roofs often had a greater species richness and abundance of migratory songbirds than the surrounding landscapes, and the native species observed on green roofs were not observed in the surrounding landscapes. This observation also highlights the potential for creating suitable bird habitat in urban areas. None of the species observed on green roofs were listed as threatened or endangered; however, Bell's vireo (Vireo bellii), field sparrow (Spizella pusilla), and northern flicker (Colaptes auratus) were observed on studied green roofs and are reported by the U.S. Fish and Wildlife Service (2010) to be rare or declining species in the Midwest United States. In addition, the eastern kingbird was observed on 17% of our green roofs and has experienced significant negative trends from 2000-2010 for the Lower Great Lakes/St. Lawrence Plain (Sauer et al., 2011). Vegetation and other structural features (i.e., log, rock, gravel) could be provided on green roofs to enhance suitable habitat for these species and others in decline. To enhance foraging opportunities for Bell's vireo in the landscape near riparian areas, roofs could be established with shrubs up to 4m tall and other plants that encourage insects (Franzreb, 1989). 146 Field sparrow and northern flicker are ground foraging species and incorporating plant species that attract insects or produce edible seeds could help provide food for these species. However, because northern flicker are cavity nesters (Lawrence, 1967), unless snags or other artificial structures providing cavities are placed on green roofs, northern flicker will not nest on green roofs. Varying media depth and composition across the roof could help establish these plants. For example, habitat for field sparrow is most suitable when vegetation is comprised of herbaceous cover 16-32cm, 50-90% grass cover, 15-35% shrub cover, and 50-75% of shrubs should be <1.5m tall (Sousa, 1983). Rooftops might also be an ideal location for field sparrow nests since predation by ground level predators has been recorded for up to 78% of nests in central Illinois (Best, 1978). Eastern kingbird numbers may be enhanced by establishing herbaceous vegetation that supports insects on which eastern kingbird feed; however, since eastern kingbird nest in trees, they may not nest on green roofs (Murphy, 1983). Other rare or declining species that have been reported by green roof owners/managers of MIA and CHA to use these green roofs are: brown thrasher (Toxostoma rufum), marsh wren (Cistothorus palustris), olive-sided flycatcher (Contopus cooperi), peregrine falcon (Falco peregrinus), prothonotary warbler (Protonotaria citrea), sedge wren (Cistothorus platensis), and wood thrush (Hylocichla mustelina). Brown thrasher, prothonotary warbler, sedge wren, and wood thrush are all species in ground foraging guilds and thus expected to forage on green roofs. The remaining species are marsh gleaning, air sallier (i.e., birds that forage while in the air during short flights from a perch), and air screener (i.e., birds that fly with their bill open to screen prey from the air) species that may momentarily rest on a green roof between feeding bouts or use a green roof as a stop-over site during migration, however are unlikely to feed directly on a green roof. Dense shrubs up to 6m tall can be planted on green roofs to provide 147 suitable nesting sites for some species, such as the brown thrasher (Brewer, 2010). Shrubs established on two studied green roofs could provide suitable nesting sites for brown thrasher. Green roofs also have potential to provide suitable vegetation types for several bird species that have not been observed on green roofs, but have experienced significant negative trends from 2000-2010 for the Lower Great Lakes/St. Lawrence Plain (Sauer et al., 2011). Green roofs planted in native grassland species, such as pale purple coneflower (Echinacea pallida) and Culver’s root (Veronicastrum virginicum), could increase the proportion of grassland vegetation in the landscape and increase habitat suitability for grassland bird species even though these bird species may not occupy green roofs. Some of these native grassland species have been planted on green roofs; hence more extensive plantings could be developed if birds such as Henslow’s sparrow (Ammodramus henslowii) and eastern meadowlark (Sturnella magna) were desired in a landscape with green roofs. Grassland bird species likely will not be greatly affected by green roofs in high density urban areas; however, establishment of grassland plants on rooftops of buildings near grasslands may reduce effective grassland bird habitat loss. Green roofs may also be able to provide stopover areas for some neo-tropical migratory bird species, such as common nighthawk (Chordeiles minor) by providing open gravel for nesting sites amongst vegetation (Gramza, 1967). Green roofs are a technology with potential to reduce bird habitat degradation and fragmentation and to increase landscape connectivity (Lundholm, 2006; Bierwagen, 2007; Eakin et al., in review). The implementation of green roofs, especially intensive roofs, designed to address the habitat requirements of specific bird communities or species could help conserve native bird species and increase species richness and abundance in urban landscapes if this is a desirable objective. Wide-spread implementation of green roofs focused on creating bird habitat 148 throughout urban areas could help reverse population declines of various grassland and neotropical migratory bird species. Though establishment of grassland vegetation on green roofs is unlikely to compensate for the vast areas of grasslands lost to agriculture during the second half of the twentieth century, any additional grassland vegetation may mitigate some of the effects of grassland bird habitat degradation. Also, several other neo-tropical migratory bird species are subject to urban pressures (Blair 1996), and by implementing green roofs designed for birds (i.e., provide food, perches, cover, nesting sites) these species may be less likely to experience declines in the future. Our research can be used to inform green roof design for bird habitat conservation while meeting the objectives for green roofs to conserve energy and storm water runoff. Pre- and postconstruction bird surveys of these green roofs would help explain how to achieve bird conservation goals (i.e., abundance, species richness, diversity). Further research is needed to better understand how green roofs affect species richness and abundance, nesting success, connectivity through urban areas, and source-sink dynamics for individual species. Our study quantified the composition of bird communities on green roofs and their surrounding landscapes, demonstrated the influence of vegetation and non-vegetation structure and composition of the surrounding landscapes and of green roofs on bird community composition and structure. In addition, our study quantified the relationship between bird communities in the landscapes with those on green roofs. Since bird species occupied green roofs and demonstrated trends indicate that they respond to vegetation and non-vegetation characteristics, prior knowledge of bird species’ habitat requirements should be the basis of recommendations for green roof design, composition, and management in relation to existing landscapes. This information should help green roof designers, city planners, natural resource 149 managers, and policy makers enhance wildlife conservation in urban areas through green technology. 150 Table 2.1. Location, building structure characteristics, planting media depth, planted vegetation and year of green roof installation at study sites sampled in 2010 and 2011 in Illinois and Michigan, U.S.A. Study a Site City, State Building Structure Characteristics Green Roof Size Height b 2 Type (m ) (# story) Slope (%) DCP Chicago, IL E and I 99,145 0 1.0 to 5.0 10 to 122 FOR Dearborn, MI E 42,177 3 1.5 MCC Chicago, IL I 24,276 0 AQU St. Charles, IL E 23,782 HAW Holland, MI E CCE Chicago, IL CHA NAM Year Installed Year Studied Ornamental perennials, turf, shrubs, trees 2004 2010 2 Sedum 2003 2010, 2011 16.7 45 to 61 Native prairie perennials, trees 2003 2010 2 to 4 8.3 10 Native prairie perennials 2005 2010 4,181 0 to 4 10.0 to 30.0 10 Sedum 2007 2010, 2011 E 1,892 8 1.0 9 to 11 Sedum, ornamental perennials, low evergreen shrubs 2006 2011 Chicago, IL E and I 1,886 11 Sculpted terrain 8 to 46 Perennials, shrubs, vines, small trees 2001 2011 Chicago, IL E 1,581 1.9 7.0 and 1.5 8 and 5 to 25 Sedum, native perennials, one tree 2002, 2004c 2010, 2011 151 Media Depth (cm) Vegetation Table 2.1. (cont’d) Green Roof Size 2 (m ) Height (# story) Slope (%) Media Depth (cm) Study a Site City, State Type MIA Chicago, IL E 1,567 15 1.0 10 to 15 SCH Chicago, IL I 929 3 GCY Chicago, IL I 758 3 1.5, raised beds, potted trees 1.0 PSS East Lansing, MI E 325 1.5 1.0 b Year Installed Year Studied Sedum, ornamental perennials 2008 2011 20 to 46 Perennials, annuals, shrubs, small trees 2003 2010, 2011 61 Perennials, vegetables, fruits, herbs 2006 2011 3 to 8 Sedum 2004 2010 Vegetation a Study sites: DCP, Downtown Chicago Park; FOR, Ford Truck Plant; MCC, McCormick Parking Structure; AQU, Aquascape Headquarters; HAW, Haworth Headquarters; CCE, Chicago Cultural Center; CHA, Chicago City Hall; NAM, Nature Museum; MIA, Michigan Avenue Structure; SCH, Schwab Rehabilitation Hospital; GCY, Gary Comer Youth Center; PSS, Plant and Soil Science Building at Michigan State University b Type: I, intensive; E, extensive c One intensive green roof was installed in 2002 and two extensive green roofs were installed in 2004. 152 Table 2.2. Maintenance regime, primary function, and accessibility of green roof study sites sampled in 2010 and 2011 in Illinois and Michigan, U.S.A. Study a Site Maintenance DCP b c Artificial Water d Source e Primary Function(s) Accessibility W, R, P, F, M Recreation P S, H Urban park, high-density residential, urban central business district, museum, railway FOR F A S Industrial complex MCC B Pollution mitigation, energy savings Wildlife habitat creation A - Conference center, urban park, lake, highway AQU W Pollution mitigation, energy savings, wildlife habitat creation A SB Offices and light manufacturing distribution, residential mid-density, airport HAW W, F Pollution mitigation, energy savings A S Industrial complex, commercial complex CCE W, R Aesthetics, pollution mitigation, energy savings A SB Urban central business district, urban park, high-density residential CCH W, R, N Wildlife habitat creation, pollution mitigation, energy savings A D Urban central business district, urban park, high-density residential NAM W, R, P Education, wildlife habitat creation, pollution mitigation, energy savings A - Museum, urban park, lake MIA W, R, F Aesthetics, pollution mitigation, energy savings A S Urban central business district, residential high-density 153 Land cover Classification Table 2.2. (cont’d) Study a Site Maintenance SCH b c Artificial Water d Source e Primary Function(s) Accessibility Land cover Classification W, P, F, A Therapeutic, aesthetics PR H, D, W Health facilities, urban park, residential middensity GCY W, F, H Education, gardening PR S School, residential mid-density, commercial strip developments, railway PSS None Education, pollution mitigation, energy savings A - College campus, urban park, railway a Study sites: DCP, Downtown Chicago Park; FOR, Ford Truck Plant; MCC, McCormick Parking Structure; AQU, Aquascape Headquarters; HAW, Haworth Headquarters; CCE, Chicago Cultural Center; CHA, Chicago City Hall; NAM, Nature Museum; MIA, Michigan Avenue Structure; SCH, Schwab Rehabilitation Hospital; GCY, Gary Comer Youth Center; PSS, Plant and Soil Science Building at Michigan State University b Maintenance: W, weeding; R, removal of dead plant materials; P, pruning; F, fertilizing; M, mowing; B, controlled burning; A, planting annuals; H, harvesting; N, planting new plant species c Accessibility: P, public; A, arranged; PR, private d Artificial water source: S, sprinkler; SB, subsurface; H, hand-watering; D, drip; W, water feature e Land Use Classification based on United States Geological Survey (USGS) land use and land cover classification sytems (Anderson et al., 1976). Driveways and surface roads were not included as a land use class because these transportation routes were present at all sites. 154 Table 2.3. Relative abundance of bird species observed on green roofs and in their surrounding landscapes in Michigan and Illinois, U.S.A. in 2010 and 2011. Relative abundances were based on the total number of bird observations at each roof or landscape. All study sites are a Downtown Chicago Park (DCP), Ford Truck Plant (FOR), McCormick Parking Structure (MCC), Aquascape Headquarters (AQU), Haworth Headquarters (HAW), Chicago Cultural Center (CCE), Chicago City Hall (CHA), Nature Museum (NAM), Michigan Avenue Structure (MIA), Schwab Rehabilitation Hospital (SCH), Gary Comer Youth Center (GCY), and the Plant and Soil Science Building at Michigan State University (PSS). DCP Ford Truck Plant MCC c 2011 Four Letter Abbreviation AMCR Foraging d Guild GF 2010 R L 0.005 0.035 R - L - R - L - R - AMGO GF 0.024 0.001 - 0.001 - 0.008 0.025 0.042 AMKE AH - - 0.004 - 0.008 - AMRO GG 0.128 0.054 0.011 0.059 0.017 0.128 0.015 0.032 ATSP GF 0.001 - - - - - - - BAOR UCF - 0.005 - - - - - - Barn swallow BARS ASC - 0.001 - 0.015 - - - 0.030 Bay-breasted warbler BBWA UCG - - - - - - - - BEVI LCG/SG 0.003 0.001 - - - - - 0.002 BEKI WP - - - - - - - Blackburnian warbler Black-capped chickadee BLBW UCF 0.001 0.001 - - - - - - BCCH LCG - - - - - - - - Black-crowned night heron BCNH WA - - - - - - - - Black-headed grosbeak BHGR UCF - 0.001 - - - - - - Blue jay BLJA GF/UCF - 0.001 - - - - - 0.005 Species American crow a American goldfinch American kestrel American robin a a American tree sparrow Baltimore oriole ab a Bell's vireo Belted kingfisher a 2010 e 155 - - 2010 L 0.019 - Table 2.3. (cont’d) DCP Ford Truck Plant MCC c 2011 Foraging d Guild FD 2010 Species Blue-winged teal Four Letter Abbreviation BWTE R - L - R - L - R - L - R - L - Brown-headed cowbird BHCO GF - - - - - - - 0.002 Canada goose Cedar waxwing CAGO WD - 0.003 0.284 0.200 0.633 0.053 - 0.006 CEDW ASA/UCG - - - - - - - 0.001 Chestnut-sided warbler CSWA LCG - - - - - - - 0.001 Chimney swift CHSW ASC - - - - - - - 0.015 CHSP GF - 0.005 - - - - 0.002 - CLSW ASC - - - - - - - 0.010 COGR GF 0.298 0.118 0.019 0.181 0.242 0.089 - 0.023 COTE WP - - - - - - - - COYE LCG 0.005 - - - - - 0.005 0.004 DEJU GF 0.002 - - - - - - 0.002 DOWO BG/LCG - 0.001 - - - - - - EABL GG - 0.001 - - - - - - EAKI ASA - - 0.002 0.009 - - - - EAWP ASA - - - - - - 0.006 EUST GF 0.089 0.113 0.580 0.511 0.025 0.479 0.002 0.372 FISP GF - - - - - - - - GRCA GF/LCF - 0.003 - - - - - 0.006 a Chipping sparrow Cliff swallow Common grackle Common tern a a Common yellowthroat Dark-eyed junco a Downy woodpecker Eastern bluebird a a Eastern kingbird Eastern wood-pewee European starling Field sparrow Gray catbird ab a a e 156 2010 - 2010 Table 2.3. (cont’d) DCP Ford Truck Plant MCC c 2011 Foraging d Guild WA 2010 Species Great blue heron Four Letter Abbreviation GBHE R - L - R - L - R - L - R - L - Hermit thrush HETH WA - 0.003 - - - - - - House finch HOFI GG - 0.001 - - - - - - HOSP GG 0.005 0.134 - 0.004 - 0.008 - - HOWR GG - - - - - - - 0.001 KILL LCG - - 0.082 0.008 0.067 0.028 - - MAWA GG - - - - - - - MALL LCG 0.005 0.005 0.006 - 0.008 - - 0.010 MODO GG - - - 0.015 - 0.006 - 0.002 MUSW GG - - - - - - - - ab NAWA FD - - - - - - - - a NOCA LCG - 0.005 - - - - 0.002 0.007 Northern flicker Northern waterthrush NOFL GF 0.001 0.001 - - - - - 0.001 NOWA GG - - - - - - - - Palm warbler PAWA BG/GF - - - - - - - - Prothonotary warbler PROW GG - 0.001 - - - - - - Purple finch PUFI LCG/BG - - - - - - - - Red-bellied woodpecker RBWO UCG - 0.001 - - - - - - RWBL BG/GF 0.294 0.025 a House sparrow House wren a a Killdeer Magnolia warbler Mallard a Mourning dove Mute swan a Nashville warbler Northern cardinal a Red-winged blackbird a e 157 2010 - 0.002 0.149 2010 0.008 0.093 0.904 0.225 Table 2.3. (cont’d) DCP Ford Truck Plant MCC c 2010 2011 R L 0.054 0.266 R L 0.015 0.005 R - L 0.028 R - L 0.016 GF 0.057 0.126 - 0.019 - 0.043 - - RCKI LCG - - - - - - - - Scarlet tanager SCTA UCG - - - - - - - 0.001 Snow goose SNGO GG - - - - - - - - SOSP LCF/GF 0.008 0.001 - - - - 0.022 0.002 SWTH GF/LCF - - - - - - - TRES ASC - - - - - 0.026 0.002 0.009 WAVI UCG - 0.001 - - - - - - White-crowned sparrow a WCSP GF 0.003 0.062 - - - - - 0.031 a WTSP GF 0.014 0.023 - - - 0.002 0.02 0.111 WODU GG/FSG - - - - - - - - Yellow warbler YWAR LCG - - - - - - - - Yellow-bellied flycatcher YBFL ASA - - - - - - - 0.002 Yellow-rumped warbler YRWA LCG - 0.001 - - - - - 0.001 Species Richness 19 32 9 14 7 14 10 32 No. Survey Dates 7 Four Letter Abbreviation RBGU Foraging d Guild GG 2010 Rock pigeon Ruby-crowned kinglet ROPI Species Ring-billed gull a a a Song sparrow Swainson's thrush a Tree swallow Warbling vireo White-throated sparrow Wood duck e 7 158 4 2010 6 - Table 2.3. (cont’d) AQU Four Letter Abbreviation AMCR Foraging d Guild GF 2010 R - AMGO GF AMKE Haworth Headquarters CCE c L - 2011 R - L - 2011 R L - 0.012 - 0.037 - 0.010 - 0.003 - - - - - - 0.400 0.086 - 0.006 - - - - - - - - - - - 0.019 - - - - - - - - - - - - - - LCG/SG - - - - - - - - BEKI WP - - - - - - - - Blackburnian warbler Black-capped chickadee BLBW UCF - - - - - - - - BCCH LCG - 0.005 - - - - - 0.003 Black-crowned night heron BCNH WA - - - - - - - - Black-headed grosbeak BHGR UCF - - - - - - - - Blue jay BLJA GF/UCF - - - 0.003 - - - - Blue-winged teal BWTE FD - - - - - - - - Brown-headed cowbird BHCO GF - 0.002 - 0.003 - - - - CAGO WD 0.030 0.086 - - - - - - CEDW ASA/UCG - - 0.014 - - - - 2010 R - 0.071 0.016 AH - AMRO GG 0.006 0.079 0.041 0.162 ATSP GF - - - BAOR UCF - 0.005 Barn swallow BARS ASC - Bay-breasted warbler BBWA UCG BEVI Species American crow a American goldfinch American kestrel American robin a a American tree sparrow Baltimore oriole ab a Bell's vireo Belted kingfisher a a Canada goose Cedar waxwing L - 0.012 159 Table 2.3. (cont’d) AQU Species Chestnut-sided warbler Four Letter Abbreviation CSWA Foraging d Guild LCG 2010 R - Chimney swift CHSW ASC CHSP Haworth Headquarters CCE c L - 2011 R - - - - - - - L - 2011 R - L - - - - - 0.028 - - - - - - - - - 0.280 0.038 - 0.028 - - - - - - - - - - - - - 0.009 - - - - - - - - - - - - GG - 0.002 - - - - - - EAKI ASA 0.006 - - - - 0.038 - - EAWP ASA - - - - - - - - EUST GF - 0.121 0.429 0.145 0.120 0.314 0.250 0.021 FISP GF - - - - - - 0.500 - GRCA GF/LCF - - - - - - - - Great blue heron GBHE WA - - - - - - - - Green heron GRHE WA - - - - - - - - Hermit thrush HETH GG - - - - - - - - House finch HOFI GG - - - 0.003 - - - 0.009 Common grackle Common tern Dark-eyed junco a Downy woodpecker Eastern bluebird a a Eastern kingbird Eastern wood-pewee European starling Field sparrow Gray catbird ab a a - ASC - GF 0.006 0.124 0.408 0.210 WP - - - COYE LCG - - DEJU Common yellowthroat GF COGR a - CLSW a - COTE Chipping sparrow Cliff swallow L - 2010 R - GF - DOWO BG/LCG EABL 160 Table 2.3. (cont’d) AQU Four Letter Abbreviation HOSP Foraging 2010 d Guild R GG HOWR LCG KILL Haworth Headquarters CCE c L 0.005 2010 R L 0.020 0.003 2011 R - L 0.019 2011 R L 0.250 0.098 - - - - - - - GG - 0.002 0.102 0.023 0.200 0.029 - - MAWA LCG - - - - - - - - MALL GG 0.024 0.042 - 0.009 - 0.029 - 0.003 MODO GG 0.018 0.051 - 0.023 - 0.067 - - MUSW FD - - - 0.065 - 0.086 - - ab NAWA LCG - - - - - - - - a NOCA GF - - - - - - - - Northern flicker Northern waterthrush NOFL GG - 0.002 - - - - - - NOWA BG/GF - - - - - - - - Palm warbler PAWA GG - - - - - - - - Prothonotary warbler PROW LCG/BG - - - - - - - - Purple finch PUFI UCG - 0.007 - - - - - 0.003 Red-bellied woodpecker RBWO BG/GF - - - - - - - - RWBL GF 0.821 0.404 - 0.233 - 0.286 - 0.009 RBGU GG - - - - - - - 0.009 ROPI GF - - - - - - - 0.715 RCKI LCG - - - - - - - - Species a House sparrow House wren a a Killdeer Magnolia warbler Mallard a Mourning dove Mute swan a Nashville warbler Northern cardinal a Red-winged blackbird Ring-billed gull a a a Rock pigeon Ruby-crowned kinglet 161 - Table 2.3. (cont’d) AQU Species Scarlet tanager Four Letter Abbreviation SCTA Foraging 2010 d Guild R UCG Snow goose SNGO GG - SOSP Haworth Headquarters CCE c L - 2010 R L - - 2011 R L - - 2011 R L - - - - - - - - - LCF/GF 0.018 - - 0.014 - - - - SWTH GF/LCF - - - - - - - - TRES ASC - 0.014 - 0.017 - - - - WAVI UCG - - - - - - - - White-crowned sparrow a WCSP GF - - - - - - - 0.003 a WTSP GF - - - - - - - 0.077 WODU GG/FSG - - - - - - - - Yellow warbler YWAR LCG - - - - - - - - Yellow-bellied flycatcher YBFL ASA - - - - - - - - Yellow-rumped warbler YRWA LCG - - - - - - - - Species Richness 9 19 5 18 4 11 3 15 No. Survey Dates 4 a Song sparrow Swainson's thrush a Tree swallow Warbling vireo White-throated sparrow Wood duck 5 162 4 4 Table 2.3. (cont’d) CHA NAM Four Letter Abbreviation AMCR Foraging d Guild GF 2011 R - L - 2010 R - AMGO GF - - AMKE AH - AMRO GG ATSP MIA c 2011 R - L 0.035 2011 R L - 0.042 0.010 - - 0.007 - - - - - - - - 0.031 - 0.010 0.010 - 0.012 - - GF - - - - - - - - BAOR UCF - - - - - - - - Barn swallow BARS ASC - - - - - - - - Bay-breasted warbler BBWA UCG - - - - - 0.002 - - BEVI LCG/SG - - - 0.001 - 0.003 - - BEKI WP - - - - - 0.002 - - Blackburnian warbler Black-capped chickadee BLBW UCF - - - - - - - - BCCH LCG - - - 0.001 - 0.008 - - Black-crowned night heron BCNH WA - - - - - - - - Black-headed grosbeak BHGR UCF - - - - - - - - Blue jay BLJA GF/UCF - - - - - - - - Blue-winged teal BWTE FD - - - 0.001 - - - - Brown-headed cowbird BHCO GF - - - - - - - - CAGO WD - - 0.089 0.356 0.270 0.200 - - CEDW ASA/UCG - - - - - - Species American crow a a American goldfinch American kestrel American robin a American tree sparrow Baltimore oriole ab a Bell's vireo Belted kingfisher a a Canada goose Cedar waxwing 163 L 0.014 - - - Table 2.3. (cont’d) City Hall Species Chestnut-sided warbler Four Letter Abbreviation CSWA Foraging d Guild LCG 2011 R - Chimney swift CHSW ASC - CHSP NAM MIA c L - 2011 R - - 0.003 0.046 - - ASC - - COGR GF - COTE WP COYE L - 2011 R - L - - - - - - - 0.002 0.818 - - - - - - - - - 0.013 - 0.028 - - - - - 0.000 - - - - LCG - - - - - 0.008 - - DEJU GF - - - - - - - - DOWO BG/LCG 0.015 - - 0.002 - 0.003 - - EABL GG - - - - - - - - EAKI ASA - - - - - 0.008 - - EAWP ASA - - - - - - - - EUST GF 0.062 0.017 0.416 0.215 0.344 0.206 - - FISP GF - - - - - - - - GRCA GF/LCF - - - 0.000 - - - - Great blue heron GBHE WA - - - 0.000 - - - - Green heron GRHE WA - - - 0.001 - - - - Hermit thrush HETH GG - - - - - - - - House finch HOFI GG - - - - - - - - Chipping sparrow Cliff swallow Common grackle Common tern a Common yellowthroat Dark-eyed junco a Downy woodpecker Eastern bluebird a a Eastern kingbird Eastern wood-pewee European starling Field sparrow Gray catbird ab a a - GF CLSW a L - 2010 R - 164 Table 2.3. (cont’d) City Hall Four Letter Abbreviation HOSP Foraging 2011 d Guild R L 0.015 0.168 GG HOWR LCG KILL NAM MIA c 2010 R L 0.337 0.123 2011 R L 0.361 0.096 2011 R L - 0.229 0.092 - - - - - - - GG - - - - - - - - MAWA LCG - - - - - 0.002 - - MALL GG - - 0.040 0.112 - 0.099 - - MODO GG - - - 0.004 - - - - MUSW FD - - - - - - - - ab NAWA LCG 0.077 - - - - - - - a NOCA GF 0.015 - - 0.001 0.008 - - - Northern flicker Northern waterthrush NOFL GG - - - - - 0.002 - - NOWA BG/GF - - - - - 0.002 - - Palm warbler PAWA GG - - - 0.000 - - - - Prothonotary warbler PROW LCG/BG - - - - - - - - Purple finch PUFI UCG - - - - - - - - Red-bellied woodpecker RBWO BG/GF - - - - - - - - RWBL GF - - 0.059 0.082 0.016 0.097 - - RBGU GG - - - 0.025 - 0.007 - 0.021 ROPI GF - 0.702 0.040 0.008 - 0.050 - 0.708 RCKI LCG - - - - 0.010 - - Species a House sparrow House wren a a Killdeer Magnolia warbler Mallard a Mourning dove Mute swan a Nashville warbler Northern cardinal a Red-winged blackbird Ring-billed gull a a a Rock pigeon Ruby-crowned kinglet 165 - Table 2.3. (cont’d) City Hall Species Scarlet tanager Four Letter Abbreviation SCTA Foraging 2011 d Guild R UCG Snow goose SNGO GG SOSP NAM MIA c L - 2010 R L - - 2011 R L - - 2011 R - L - - - - 0.001 - 0.007 - - LCF/GF - - - - - - - - SWTH GF/LCF - - - 0.000 - - - - TRES ASC - - - - - - - - WAVI UCG - - - - - - - - White-crowned sparrow a WCSP GF 0.046 0.027 - - - 0.079 0.091 - a WTSP GF 0.600 0.086 - 0.008 - - 0.091 - WODU GG/FSG - - - 0.014 - 0.020 - - Yellow warbler YWAR LCG - - - 0.001 - - - - Yellow-bellied flycatcher YBFL ASA - - - - - - - - Yellow-rumped warbler YRWA LCG - - - - - 0.008 - - Species Richness 10 5 8 27 5 27 3 4 No. Survey Dates 4 a Song sparrow Swainson's thrush a Tree swallow Warbling vireo White-throated sparrow Wood duck 4 166 4 4 Table 2.3. (cont’d) SCH GCY c PSS 2011 R L - - 2010 R - Four Letter Abbreviation AMCR Foraging d Guild GF 2010 R - L 0.006 2011 R - AMGO GF - 0.022 0.048 0.038 - - 0.500 0.212 AMKE AH - - - - - - AMRO GG 0.400 0.076 0.810 0.180 - ATSP GF - - - - - - - - BAOR UCF - - - - - - - - Barn swallow BARS ASC - - - - - - - - Bay-breasted warbler BBWA UCG - - - - - - - - BEVI LCG/SG - - - - - - - - BEKI WP - - - - - - - - Blackburnian warbler Black-capped chickadee BLBW UCF - - - - - - - - BCCH LCG - - - 0.010 - - - 0.001 Black-crowned night heron BCNH WA - - - 0.003 - - - - Black-headed grosbeak BHGR UCF - - - - - - - - Blue jay BLJA GF/UCF - - - - - - - 0.001 Blue-winged teal BWTE FD - - - - - - - - Brown-headed cowbird BHCO GF - - - 0.018 - - - 0.001 CAGO WD - 0.155 - 0.083 - - - - CEDW ASA/UCG - - - - - - - 0.023 Species American crow a American goldfinch American kestrel American robin a a American tree sparrow Baltimore oriole ab a Bell's vireo Belted kingfisher a a Canada goose Cedar waxwing 167 L 0.008 - 0.216 L - 0.500 0.257 Table 2.3. (cont’d) SCH Species Chestnut-sided warbler Four Letter Abbreviation CSWA Foraging 2010 d Guild R LCG Chimney swift CHSW ASC CHSP GCY c PSS 2010 R L - - L - 2011 R - L - 2011 R L - - - - - - - - - - GF - - 0.095 0.003 - - - 0.039 CLSW ASC - - - - - - - - COGR GF 0.267 0.133 - 0.043 - - - 0.184 COTE WP - - - - - - - - COYE LCG - - - - - - - - DEJU GF - - - - - - - - DOWO BG/LCG - - - 0.003 - - - - EABL GG - - - - - - - - EAKI ASA - - - - - - - - EAWP ASA - - - - - - - - EUST GF 0.267 0.386 - 0.326 - 0.265 - 0.056 FISP GF - - - - - - - - GRCA GF/LCF - - - - - - - - Great blue heron GBHE WA - - - - - - - - Green heron GRHE WA - - - - - - - - Hermit thrush HETH GG - - - - - - - - House finch HOFI GG - - - - - - - 0.010 Chipping sparrow Cliff swallow Common grackle Common tern a a Common yellowthroat Dark-eyed junco a Downy woodpecker Eastern bluebird a a Eastern kingbird Eastern wood-pewee European starling Field sparrow Gray catbird ab a a 168 Table 2.3. (cont’d) SCH GCY Four Letter Abbreviation HOSP Foraging 2010 d Guild R L 0.067 0.026 GG HOWR LCG - KILL GG MAWA c PSS 2010 R L - 0.114 2011 R - L - 2011 R L 1.000 0.480 - - - - - - - - - - - - - - - LCG - - - - - - - - MALL GG - 0.018 - 0.033 - - - 0.008 MODO GG - - - - - - - 0.035 MUSW FD - - - - - - - - ab NAWA LCG - - - - - - - - a NOCA GF - - - 0.018 - 0.039 - 0.048 Northern flicker Northern waterthrush NOFL GG - 0.002 - - - - - - NOWA BG/GF - - - - - - - - Palm warbler PAWA GG - - - - - - - - Prothonotary warbler PROW LCG/BG - - - - - - - - Purple finch PUFI UCG - - - - - - - 0.001 Red-bellied woodpecker RBWO BG/GF - - - - - - - - RWBL GF - 0.070 - 0.058 - - - 0.007 RBGU GG - 0.080 - 0.083 - - - - ROPI GF - 0.026 0.048 0.078 - - - - RCKI LCG - - - - - - - Species a House sparrow House wren a a Killdeer Magnolia warbler Mallard a Mourning dove Mute swan a Nashville warbler Northern cardinal a Red-winged blackbird Ring-billed gull a a a Rock pigeon Ruby-crowned kinglet 169 - Table 2.3. (cont’d) SCH a Four Letter Abbreviation SCTA SNGO SOSP Foraging d Guild UCG GG LCF/GF 2010 R - a SWTH TRES GF/LCF ASC WAVI WCSP WTSP UCG GF GF GCY c PSS L - 2010 R - L 0.001 - L - 2011 R - L - 2011 R - - - - - - - - - - - - 0.003 0.013 - - - - WODU GG/FSG - - - 0.005 - - - - Yellow warbler YWAR LCG - - - - - - - - Yellow-bellied flycatcher Yellow-rumped warbler YBFL YRWA Species Richness No. Survey Dates ASA LCG 4 5 12 4 4 19 1 4 4 2 6 17 Species Scarlet tanager Snow goose Song sparrow Swainson's thrush Tree swallow Warbling vireo a White-crowned sparrow a White-throated sparrow Wood duck a b c Species observed on green roofs Species only observed on green roofs Study site for this year was not included in the model d Feeding guilds: AH, Air Hawker; ASC, Air Screener ; ASA, Air Sallier; BG, Bark Gleaner; FD, Freshwater Dabbler; FSG, Freshwater Surface Gleaner; GF, Ground Forager; GG, Ground Gleaner; LCF, Lower Canopy Forager; LCG, Lower Canopy Gleaner ; GF, Ground Forager; GG, Ground Gleaner; LCF, Lower Canopy Forager ; LCG, Lower Canopy Gleaner; SG, Shrub Gleaner; UCF, Upper Canopy Forager; UCG, Upper Canopy Gleaner; WA, Water Ambusher; WD, Water Dabbler; WP, Water Plunger e Bird survey areas: R, Green roof; L, Landscape 170 40 35 30 25 Estimated species richness Landscape Rooftop Figure 2.1. Estimated species richness of non-invasive, native bird species, excluding waterfowl and urban associated species, at the point-level for green roof sites in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Mean estimated species richness and quartiles shown are estimated using a multi-species hierarchical Bayes multi-scale model. 171 40 40 35 35 30 30 25 25 Estimated species richness Estimated species richness Landscape Landscape Rooftop Rooftop Figure 2.2. Estimated species richness of non-invasive, native bird species, excluding waterfowl and urban associated species, at the site-level for green roof sites in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Mean estimated species richness and quartiles shown are estimated using a multi-species hierarchical Bayes multi-scale model. 172 30 20 10 0 173 Estimated species richness 40 50 Rooftop Landscape PSS GCY SCH MIA NAM CHA CCE HAW AQU MCC FOR DCP Figure 2.3. Estimated species richness of non-invasive, native bird species, excluding waterfowl and urban associated species, for green roofs and surrounding landscapes in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Mean richness and 95% credible intervals shown are estimated using a multi-species hierarchical Bayes multi-scale model. 173 40 40 30 30 20 20 10 10 0 0 174 Species richness Estimated species richness Species richness 50 50 Estimated Estimated Observed Observed Smallest Largest PSS GCY SCH MIA NAM CHA CCE HAW AQU MCC FOR DCP MIA NAM CHA CCE HAW AQU MCC FOR DCP PSS GCY SCH Figure 2.4. Estimated and observed (uncorrected for detection probabilities) species richness of non-invasive, native bird species, excluding waterfowl and urban associated species, from the first year of sampling of green roofs and surrounding landscapes in Illinois and Michigan, U.S.A. in 2010 and 2011. Mean richness and 95% credible intervals shown are estimated using a multi-species hierarchical Bayes multi-scale model. 174 Rooftop 0.6 0.4 0.0 0.2 Occurrence probability 0.8 1.0 Landscape CAGO MALL Figure 2.5. Occurance probabilities of native waterfowl bird species, excluding urban-associated species, observed on three or more green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are Canada goose (Branta canadensis, CAGO) and mallard (Anas platyrhynchos, MALL). Mean occurrence and 95% credible intervals shown are estimated using a single-species hierarchical Bayes multi-scale occupancy model. 175 Rooftop 0.6 0.4 0.2 0.0 176 Occurrence probability 0.8 1.0 Landscape AMRO AMGO CHSP COGR NOCA RWBL WTSP Figure 2.6. Occurance probabilities of native bird species, excluding urban-associated and waterfowl species, observed on 3 or more green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are American robin (Turdus migratorius, AMRO), American goldfinch (CardueFfigurelis tristis, AMGO), chipping sparrow (Spizella passerina, CHSP), common grackle (Quizcalus quizcula, COGR), northern cardinal (Cardinalis cardinalis, NOCA), red-winged blackbird (Agelaius phoeniceus, RWBL), and white-throated sparrow (Zonotrichia albicollis, WTSP). Mean occurrence and 95% credible intervals shown are estimated using a multi-species hierarchical Bayes multi-scale model. 176 Rooftop 0.6 0.4 0.2 0.0 177 Occurrence probability 0.8 1.0 Landscape BEVI EAKI FISP NOFL Figure 2.7. Occurence probabilities of declining native bird speciesobserved on green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are Bell’s vireo (Vireo bellii, BEVI), eastern kingbird (Tyrannus tyrannus, EAKI), field sparrow (Spizella pusilla, FISP), and northern flicker (Colaptes auratus, NOFL). Mean occurrence and 95% credible intervals shown are estimated using a multi-species hierarchical Bayes multi-scale model. 177 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 Occurence probability on roofs Occurence probability on roofs 0.0 0.0 0.2 0.2 0.4 0.4 0.6 0.6 0.8 0.8 1.0 1.0 Occurrence probability in Occurrence probability landscapes in landscapes Figure 2.8. Mean occurrence probabilities on green roofs compared with mean occurrence probabilities in surrounding landscapes for all non-invasive, native bird species, excluding waterfowl and urban associated species, in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Mean occurrence probabilities shown are estimated using a multi-species hierarchical Bayes multi-scale model. 178 Rooftop 0.6 0.4 0.0 0.2 Use probability 0.8 1.0 Landscape CAGO MALL Figure 2.9. Use probabilities of native waterfowl bird species, excluding urban-associated species, observed on three or more green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are Canada goose (Branta canadensis, CAGO) and mallard (Anas platyrhynchos, MALL). Mean use and 95% credible intervals shown are estimated using a single-species hierarchical Bayes multi-scale occupancy model. 179 Rooftop 0.6 0.4 0.2 0.0 180 180 Use probability 0.8 1.0 Landscape AMRO AMGO CHSP COGR NOCA RWBL WTSP Figure 2.10. Use probabilities of native bird species, excluding urban-associated and waterfowl species, observed on 3 or more green Figure 2.10. Use probabilities of native bird species, excluding urban-associated and waterfowl species, observed on 3 ormigratorius, roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are American robin (Turdus more green roofs in Illinois and goldfinch (Carduelis tristis, AMGO), 2010 and sparrow (Spizella passerina,American robin (Turdus migratorius, AMRO), American Michigan, U.S.A. during sampling in chipping 2011. Speices included are CHSP), common grackle (Quizcalus AMRO), American goldfinch (Carduelis tristis, AMGO), chipping sparrow (Spizella passerina, CHSP), common RWBL),(Quizcalus quizcula, COGR), northern cardinal (Cardinalis cardinalis, NOCA), red-winged blackbird (Agelaius phoeniceus, grackle and whitequizcula, COGR), northern cardinal (Cardinalis cardinalis, and 95% credible intervals shown are estimated using a multi-species throated sparrow (Zonotrichia albicollis, WTSP). Mean useNOCA), red-winged blackbird (Agelaius phoeniceus, RWBL), and whitethroated sparrow (Zonotrichia model. hierarchical Bayes multi-scale albicollis, WTSP). Mean use and 95% credible intervals shown are estimated using a multi-species hierarchical Bayes multi-scale model. 180 180 Rooftop 0.6 0.4 0.2 0.0 181 Use probability 0.8 1.0 Landscape BEVI EAKI FISP NOFL Figure 2.11. Use probabilities of declining native bird speciesobserved on green roofs in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Speices included are Bell’s vireo (Vireo bellii, BEVI), eastern kingbird (Tyrannus tyrannus, EAKI), field sparrow (Spizella pusilla, FISP), and northern flicker (Colaptes auratus, NOFL). Mean use and 95% credible intervals shown are estimated using a multi-species hierarchical Bayes multi-scale model. 181 1.0 0.8 0.6 0.4 0.0 0.2 Use probability PSS MIA CCE HAW AQU FOR Extensive GCY SCH NAM CHA MCC DCP Intensive Figure 2.12. Mean use probabilities on extensive and intensive green roofs for non-invasive, native bird species, excluding waterfowl and urban associated species, that were observed on at least 25% of green roofs during sampling in Illinois and Michigan, U.S.A. in 2010 and 2011. Mean use probabilities shown are estimated using a multi-species hierarchical Bayes multi-scale model. 182 1.0 0.8 0.6 0.0 0.2 0.4 Use probability 1 3 10 2 100 Roof size (10 m ) Figure 2.13. Mean use probabilities on green roofs organized by size for non-invasive, native bird species, excluding waterfowl and urban associated species, that were observed on at least 25% of green roofs during sampling in Illinois and Michigan, U.S.A. in 2010 and 2011. Mean use probabilities shown are estimated using a multi-species hierarchical Bayes multi-scale model. 183 1.0 0.8 0.6 0.0 0.2 0.4 Use probability 0 25 50 75 100 Green space (%) Figure 2.14. Mean use probabilities on green roofs organized by green space at study sites for non-invasive, native bird species, excluding waterfowl and urban associated species, that were observed on at least 25% of green roofs during sampling in Illinois and Michigan, U.S.A. in 2010 and 2011. Mean use probabilities shown are estimated using a multi-species hierarchical Bayes multi-scale model. 184 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 Use probability on roofs on roofs Use probability 0.0 0.0 0.2 0.2 0.4 0.4 0.6 0.6 0.8 0.8 1.0 1.0 Use probability Use probability in landscapes in landscapes Figure 2.15. Mean use probabilities on green roofs compared with mean use probabilities in surrounding landscapes for all non-invasive, native bird species, excluding waterfowl and urban associated species, in Illinois and Michigan, U.S.A. during sampling in 2010 and 2011. Mean use probabilities shown are estimated using a multi-species hierarchical Bayes multi-scale model. 185 LITERATURE CITED 186 LITERATURE CITED Anderson, J. R., Hardy, E. E., Roach, J. T., Witmer, R. E., 1976. A land use and land cover classification system for use with remote sensor data. Geological Survey Professional Paper 964. United States Government Printing Office, Washington. 41pp. Banting, D., Doshi, H., Li, J., Missios, P., Au, A., Currie, B.A., Verrati, M., 2005. Report on the environmental benefits and costs of green roof technology for the city of Toronto. City of Toronto and Ontario Centres of Excellence–Earth and Environmental Technologies, 88pp. Basore, N.S., Best, L.B., Wooley, J.B., 1986. Bird nesting in Iowa: No-tillage and tilled cropland. The Journal of Wildlife Management. 50(1), 19-28. Baumann, N., 2006. Ground-nesting birds on green roofs in Switzerland: preliminary observations. Urban Habitats. 4(1), 37-50. Best, L.B., 1978. Field sparrow reproductive success and nesting ecology. The Auk. 95, 9-22. Best, L.B., Campa III, H., Kemp, K.E., Robel, R.J., Ryan, M.R., Savidge, J.A., Weeks Jr., H.P., Winterstein, S.R., 1997. Bird abundance and nesting in CRP fields and cropland in the Midwest: A regional approach. Wildlife Society Bulletin. 25(4), 864-877. Bibby, C. J., N. D. Burgess, D. A. Hill., 1992. Bird census techniques. Academic Press, New York. Bierwagen, B., 2007. Connectivity in urbanizing landscapes: The importance of habitat configuration, urban area size, and dispersal. Urban Ecosystems. 10, 29-42. Blair, R., 1996. Land use and avian species diversity along an urban gradient. Ecological Applications. 6(2), 506-519. Bollinger, E.K., Gavin, T.A., 1992. Eastern bobolink populations: ecology and conservation in an agricultural landscape. in Hagan, J.M., Johnston, D.W. Ecology and Conservation of Neotropical Migrant Landbirds, 497–506. Washington, DC: Smithsonian Institution Press. Brenneisen, S., 2006. Space for urban wildlife: Designing green roofs as habitats in Switzerland. Urban Habitats. 4(1), 27-36. Brewer, D., 2010. Wrens, dippers and thrashers. Christopher Helm Publishers, London, pp. 230. 187 Brewer, R., McPeek, G. A., Adams, R. J. Jr., 1991. The atlas of breeding birds of Michigan. Michigan State University Press, East Lansing, Michigan. 594pp. Canfield, R.H., 1941. Application of the Line Interception method in sampling range vegetation. Journal of Forestry. 39(4), 388-394. Christensen, N. L., 1997. Managing for heterogeneity and complexity on dynamic landscapes, in: Pickett, S.T., Ostfeld, R. S., Shachak, M., Likens, G. E., (Eds.), The ecological basis for conservation: Heterogeneity, ecosystems, and biodiversity. Chapman and Hall, New York: pp. 167-186. Cimprich, D., Moore, F,. Guilfoyle, M., 2000. The Birds of North America Online: Red-eyed Vireo (Vireo olivaceus). Accessed March 16, 2011 at http://bna.birds.cornell.edu/bna/species/527 . Clark, C., Adriaens, P., Talbot, B., 2008. Green roof valuation: A probabilistic economic analysis of environmental benefits. Environmental Science and Technology. 42, 21552161. Clements, F.E., 1905. Research methods in ecology. University Publishing Company, Lincoln, Nebraska. Cornell lab of Ornithology and the American Ornithologists Union, 2012. The Birds of North America Online. Accessed September 17, 2012 at http://bna.birds.cornell.edu/bna. Coupland, R. T., 1958. The effects of fluctuations in weather upon the grasslands of the Great Plains. Botanical Review. 24 (5): 273-317. De Graff, R.M., Tilghman, N.G., Anderson, S.H.,1985. Foraging guilds of North American birds. Environmental Management. 9(6),493-536. Delisle, J., Savidge, J., 1997. Avian use and vegetation characteristics of Conservation Reserve Program fields. Journal of Wildlife Management. 61(2), 318-325. Donnelly, R., Marzluff, J. M., 2004. Importance of reserve size and landscape context to urban bird conservation. Conservation Biology. 18, 733–745. Dorazio, R.M., Royle, J.A., Soderstrom, B., Glimskar, A., 2006. Estimating species richness and accumulation by modeling species occurrence and detectability. Ecology 87, 842–854. Dvorak, B., Volder, A., 2010. Green roof vegetation for North American ecoregions: A literature review. Landscape and Urban Planning. 96(4), 197–213. Eakin, C., Campa, H., Rowe, D., Westphal, W., Roloff, G., (In Review). Vegetation characteristics of green roofs for wildlife: Habitat contributions in Michigan and Illinois. Journal of Landscape and Urban Planning. 188 Fernandez-Canero, R., Gonzalez-Redondo, P., 2010. Green roofs as a habitat for birds: A review. Journal of Animal and Veterinary Advances. 9(15), 2041-2052. Fox, J., Weisberg, S., 2011. An R companion to applied regression, Second edition. Thousand Oaks, CA. Franzreb, K. E., 1989. Ecology and conservation of the endangered least Bell's vireo. U.S. Fish Wildlife Service, Biological Report. 89(1), 17 pp. Gedge, D., 2003. From rubble to redstarts. Paper presented at the Greening Rooftops for Sustainable Communities conference, Chicago. 9 pp. Gelman, A., Carlin, J.B., Stern, H.S., Rubin, D.B., 2003. Bayesian Data Analysis, second ed. Chapman & Hall, Boca Raton, Florida, USA. Getter, K.L., Rowe, D.B., 2006. The role of green roofs in sustainable development. HortScience. 41(5), 1276–1286. Getter, K.L., Rowe, D.B., Andresen, J.A., Wichman, I.S., 2011. Seasonal heat flux properties of an extensive green roof in a Midwestern U.S. climate. Energy and Buildings. 43, 35483557. Google Earth, 2011. Google Earth Version 6.1. Accessed November 2, 2011 at www.earth.google.com. Hanaburgh, C., 2001. Modeling the effects of management approaches on forest and wildlife resources in northern hardwood forests. Doctoral dissertation, Michigan State University. Heady, H.F., Gibbens, R.P., Powell, R.W., 1959. A comparison of the Charting, Line Intercept, and Line Point Methods of sampling shrub types of vegetation. Journal of Range Management. 12(4), 180-188. Herkert, J.A., Sample, D. W., Warner, R. E., 1996. Management of Midwestern grassland landscapes for the conservation of migratory birds, in: F. R. Thompson, III (Ed.), Managing Midwestern landscapes for the conservation of neotropical migratory birds. United States Department of Agriculture Forest Service, General Technical Report GTRNC-187, North Central Forest Experiment Station, St. Paul, Minnesota: pp. 89116.Herkert, J.A.,1995. An analysis of Midwestern breeding bird population trends: 1966-1993. American Midland Naturalist. 134(1), 41-50. Hohtola, E., 1978. Differential changes in bird community structure with urbanization: A study in central Finland. Ornis Scandinavica. 9(1), 94-100. Jackson, B., J. Jackson. 2000. Killdeer (Charadrius vociferus). Pp. 1-28 in A. Poole, F. Gill, eds. The Birds of North America, Vol. 517. Philadelphia, PA: The Birds of North America, Inc. 189 Gramza, A.F., 1967. Responses of brooding nighthawks to a disturbance stimulus. The Auk. 84(1), 72-86. Johnson, D.H., Igl, L.D., 2001. Area requirements of grassland birds: a regional perspective. The Auk. 118, 24-34. Keitt, T.H., Urban, D.L., Milne, B.T., 1997. Detecting critical scales in fragmented landscapes. Conservation Ecology. 1, 4. Accessed August 11, 2012 from www.ecologyandsociety.org /vol1/iss1/art4/. Lawrence, L., 1967. A comparative life-history study of four species of woodpeckers. Ornithological Monographs. 5, 1-156. Lin, S.C., 2006. The ecologically ideal road density for small islands: The case of Kinmen. Ecological Engineering. 27, 84–92. Lundholm, J.T., 2006. Green roofs and facades: A habitat template approach. Urban Habitats. 4(1), 87-101. Lunn, D.J., Thomas, A., Best, N., Spiegelhalter, D., 2000. WinBUGS, a Bayesian modeling framework: concepts, structure, and extensibility. Statistics and Computing, 10:325-337. McKinney, M.L., 2006. Urbanization as a major cause of biotic homogenization. Biological Conservation. 127, 247–260. 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). Ecological Applications. 13, 1041–1059. Monsma, J. L., 2011. A comparison of insect and spider species presence and composition across multiple vegetated roof systems. Master Thesis. Michigan State University. Mordecai, R.S., Mattsson, B.J., Tzilkowski, C.J., Cooper, R.J., 2011. Addressing challenges when studying mobile or episodic species: Hierarchical Bayes estimation of occupancy and use. Journal of Applied Ecology. 48, 56-66. Murphy, M.T., 1983. Nest success and nesting habits of eastern kingbirds and other flycatchers. The Condor. 85, 208-219. Oberndorfer, E., Lundholm, J., Bass, B., Coffman, R.R., Doshi, H., Dunnett, N., Gaffin, S., Kohler, M., Liu, K.K.Y., Rowe, D.B., 2007. Green roofs as urban ecosystems: Ecological structures, functions, and services. BioScience. 57(10), 823-833. 190 Peck, S., Callaghan, C., Kuhn, M., Grass, B., 1999. Greenbacks from green roofs: forging a new industry in Canada. Status Report prepared for Canada Mortgage and Housing Corporation. Peck and Associates. Pollet, Ingrid L., Dave Shutler, John Chardine and John P. Ryder. 2012. Ring-billed Gull (Larus delawarensis), The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology.Accessed September 15, 2012 at http://bna.birds.cornell.edu/bna/species/033. R Core Team, 2012. R: A language and environment for statistical computing: Version 2.14.2. R Foundation for Statistical Computing. Vienna, Austria. Robbins, C. S., 1981. Effect of time of day on bird activity. Studies in Avian Biology. 6, 275282. Rowe, D.B., 2011. Green roofs as a means of pollution abatement. Environmental Pollution. 159(8-9), 2100-2110. Royle, J.A., Dorazio, R.M., 2008. Hierarchical Modeling and Inference in Ecology. Academic Press, Boston, USA. Samson, F., Knopf, F., 1994. Prairie conservation in North America. BioScience. 44, 418-421. Sauer, J. R., Hines, J. E., Fallon, J. E., Pardieck, K. L., Ziolkowski, D. J. Jr., Link, W. A., 2011. The North American Breeding Bird Survey, results and analysis 1966 - 2010. Version 12.07.2011 USGS Patuxent Wildlife Research Center, Laurel, MD. Sauer, J. R., Peterjohn, B. G., Link, W. A., 1994. Observer differences in the North American Breeding Bird Survey. The Auk. 111, 50-62. Savard, J., Clergeau, P., Mennechez, G., 2000. Biodiversity concepts and urban ecosystems. Landscape and Urban Planning. 48(3-4), 131-142. Schroeder, R. L., 1982. Habitat suitability index models: Downy woodpecker. United States Department of the Interior, Fish Wildlife Service FWS/OBS-821l0.38. 10 pp. Short, H.L., 1985. Habitat suitability index models: Red-winged blackbird. U.S. Fish and Wildlife Service Biological Report 82(10.95). 20 pp. Sousa, P.J., 1983. Habitat suitability index models: Field sparrow. U.S. Fish and Wildlife Service Report FWS/OBS-82(10.62). 14 pp. Spiegelhalter, D.J., Thomas, A., Best, N.G., Lunn, D., 2003. WinBUGS version 1.4 user manual. in: MRC Biostatistics Unit, Cambridge, UK. 191 Sturtz, S., Ligges, U., Gelman, A., 2005. R2WinBUGS: A package for running WinBUGS from R. Journal of Statistical Software. 12(3), 1-16. Thompson, F.R. III., 2005. Landscape level effects on forest bird populations in eastern broadleaf forests: principles for conservation. USDA Forest Service General Technical Report PSW-GTR-191. Presented at the Third International Partners in Flight Conference, March 20-24, 2002, Asilomar Conference Grounds, California. U.S. Fish and Wildlife Service, 2010. Midwest birds of concern. Accessed July 11, 2012 at http://www.fws.gov/midwest/MidwestBird/birds_of_concern.htm. United States Fish and Wildlife Service, 2011. Accessed May 23, 2012 at Flyways.us. Villalobos, F. J., Ritchie, J. T., 1992. The effect of temperature on leaf emergence rates of sunflower genotypes. Field Crops Research. 29 (1), 37-46. Verbeek, N. A. and C. Caffrey. 2002. American Crow (Corvus brachyrhynchos), The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology. Accessed September 15, 2012 at http://bna.birds.cornell.edu/bna/species/647. Walkinshaw, L. H., 1991. Brown-headed cowbird, Molothrus ater. in: The atlas of breeding birds of Michigan. Brewer, R., McPeek, G. A., Adams, R. J. Jr. (Eds.) Michigan State University Press, East Lansing, Michigan. 594pp. Warnes, G.R., 2012. Gplots: Various R programming tools for plotting data. R package version 2.11.0. Accessed May 30, 2012 at http://CRAN.R-project.org/package=gplots. Wickham, H., 2007. Reshaping data with the reshape package. Journal of Statistical Software. 21(12). Wiens, J.A., 1974. Habitat heterogeneity and avian community structure in North American grasslands. American Midland Naturalist. 91(1), 195-213. Williams, E. D., 1983. Effects of temperature, light, nitrate and prechilling on seed germination of grassland plants. Annals of Applied Biology. 103 (1), 161–172. Zipkin, E.F., Royle, J.A., Dawson, D.K., Bates, S., 2010. Multi-species occurrence models to evaluate the effects of conservation and management actions. Biological Conservation. 143, 479-484. 192 CONCLUSIONS This project evaluated if green roofs could contribute to bird conservation by providing additional green space in urban landscapes. Associated objectives were addressed by collection and analysis of data on vegetation characteristics and bird communities on and surrounding green roofs. Analysis of vegetation on and surrounding green roofs demonstrated significant differences between intensive and extensive green roof vegetation characteristics. Intensive roofs had taller perennial and woody species, whereas extensive green roofs generally had lowgrowing, drought-tolerant perennial or shrubs species. Analyses of bird surveys demonstrated a tendency for birds to use intensive green roofs for more time than extensive roofs. In addition, intensive green roofs appear better suited to support a greater richness of bird species and successful nesting because of increased niche opportunities in vegetation. Ground foragers were observed on intensive and extensive roofs, and those in shrub and low canopy foraging and bark gleaning guilds (6 species) were only observed on intensive roofs with shrub and/or tree cover. However, successful nesting attempts have been observed on large extensive green roofs, indicating the ability of these roofs to also support nesting activities. Wildlife species that require shorter vegetation and less woody cover may be better supported on extensive roofs. Comparisons of vegetation characteristics provided on green roofs with those required for various native grassland bird species habitat requirements demonstrated the ability of green roofs to provide bird habitat. Green roofs may be able to support native bird species due to the ability of green roof vegetation to fulfill grassland bird species habitat requirements. In addition, 25 non-invasive, native bird species were observed on green roofs, and nearly all bird species observed in 193 landscapes and not on green roofs are estimated to have a > 70% probability to occur on green roofs. Among these bird species with high occurrence probabilities are those with populations in decline throughout the Midwest United States. These estimates coupled with the ability of vegetation on green roofs to support the habitat requirements of species in decline demonstrate the ability of green roofs to provide habitat for species of conservation concern. Green roofs were estimated to have higher median bird species richness than in surrounding landscapes. However, comparatively low use probabilities on green roofs indicate that birds primarily use landscapes and green roofs may function as complimentary bird habitat. The high bird species richness and low use probability for green roofs also suggest that birds may use green roofs as stepping stones to traverse urban areas. Bird species present in landscapes directly surrounding green roofs appear to influence which bird species frequently use on green roofs, as those on green roofs are generally also observed in landscapes. Future research is needed to examine the effect other green roof factors (roof height, human presence, non-native plant species, landscape matrix, lack of mesopredators) have on bird use and wildlife habitat connectivity. Green roofs with greater variability in roof size and vegetation structural diversity and/or roofs with similar structure could be studied to hone in on the bird community that uses a specific type of roof (e.g., intensive, native prairie vegetation with >60% cover, mean height of 1.1m) and the effect various green roof factors have on observed and predicted bird communities. Telemetry studies could be used to further understand how birds move through the landscape, when they are on green roofs, and where they are nesting. Pre- and post-construction bird surveys of green roofs would help explain how to achieve bird conservation goals (i.e., abundance, species richness, diversity). 194 Wildlife managers, land planners, environmental designers and policy makers who aim to improve ecosystem function and wildlife habitat quality in urban and developing landscapes may refer to this manuscript to better understand how wildlife communities may interact with green roofs, green roof vegetation, and surrounding landscapes. Presented information could be used to help select a wildlife group to target with conservation efforts through green roof installation. Results from our study could also demonstrate the ability of green roof installations to address conservation objectives at the landscape scale. Our research has demonstrated that green roofs have the ability to drastically increase (>300%) the amount of green space that may provide ecosystem functions (i.e., stormwater management, air pollution mitigation) and that is important for wildlife conservation. Wide-spread implementation of green roofs focused on creating bird habitat throughout urban areas could help minimize population declines of various grassland and neo-tropical migratory bird species, and promote biodiversity conservation in urban areas. 195