BEE COMMUNITY RESPONSE TO LOCAL AND LANDSCAPE FACTORS ALONG AN URBAN RURAL GRADIENT By KRISTEN R. BIRDSHIRE B.S., University of Missouri, Columbia, 2011 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in parti al fulfillment of the requirements for the degree of Master of Science Environmental Sciences Program 2018
ii This thesis for the Master of Science degree by Kristen R. Birdshire has been approved for the Environmental Sciences P rogram by Christy E. Briles, Chair Adrian Carper Peter Anthamatten Date: December 1 5 , 2018
iii Birdshire, Kristen R. (M.S., Environmental Sciences Program) Bee Community Response to Local and Landscape Factors Along an Urban Rural Gradient Thesis directed by Assistant Profess or Christy E. Briles ABSTRACT including fruits, vegetables, nuts, spices, and oilseed require insect pollination. Reliance on the pollination services that promote these food products continues to rise due to increasin g demands from growing human population s ; therefore, it is imperative to understand the ecology of insect pollinators. While extensive research describes pollination services in both agricultural and urban settings, few studies examine pollinator activity along urban rural gradient s . Even fewer address how bees with different bee life history characteristic s respon d to urbaniz ation, and to my knowledge , never from a high elevation semi arid environment. This study document ed different pollinator assemblag es from 12 sites along an urban rural gradient in Denver, Colorado, USA . Percent impervious surface was used define the extent of urbanization. Wild bees were collected in the summer of 2017 and local and landscape characteristics were estimated using fie ld assessments and geospatial analysis . Bee species life history characteristics were assigned to ecological guilds, and the relationships between urbanization and bee communities were explored using linear modeling. Given findings from a similar study , I hypothesize d that bee communities in urban landsca pe s would be dominated by smaller bees with more ecologically generalized characteristics (e.g., polylectic and eusocial) than bee s in rural landscape s, which were hypothesized to contain bee communities with more specialized traits (e.g., oligolectic). Overall, bee abundance and diversity decreased with increasing urbanization , suggest ing that urban areas negatively impact bee communities.
iv Moreover, all bee guilds responded positively to local floral r ichness, and negatively to the degree of urbanization in the landscape, suggesting that urbanization in the Denver metropolitan area is detrimental to entire bee communities . For the Denver metro politan area, the findings suggest incorporating a greater d iversity of floral species and increasing pollinator resource corridors throughout the city of Denver. The form and content of this abstract are approved. I recommend its publication. Approved: Christy Briles
v DEDICATION I ded icate this work to my bel oved spouse, Micah Birdshire. May your rivers flow without end, meandering through pastoral valleys tinkling with bells, past rs into a dark primeval forest where tigers belch and monkeys howl, through miasmal and mysterious swamps and down into a desert of rock, blue mesas, domes and pinnacles and grottos of endless stone, and down again into a deep vast ancient unknown chasm wh ere bars of sunlight blaze on profiled cliffs, where deer walk across the white sand beaches, where storms come and go as lightning clangs upon the high crags, where something strange and more beautiful and more full of w onder than your deepest dreams wait ~Isabelle Allende
vi ACKNOWLEDGEMENTS I would like to thank my adviser, Christy Briles, for her time and commitment to the successful development and completion of this research. I would also like to thank my committee members, Adrian Carper and Peter Anthamatten for their guidance, particularly through challenging statistical analyses. I adamantly thank Virginia Scott and Adrian Carper at the Univers ity of Colorado Boulder for hours of bee identific ations. I am grateful to The Garden Club of America Centennial Pollinator Fellowship, the American Association of Geographers Biogeography Specialty Group Graduate Research Award , the University of Colorad o Denver Graduate School and the College of Libera l Arts and Sciences for funding this research. I also thank the Auraria Higher Education Center, Duane McClanahan, Denver Urban Gardens, Christy Shires, Ellen Haroutunian, the Jay family, St. Andrews Unite d Methodist Church, Barr Lake State Park, and Gwen survey on their properties . Finally, I would like to thank Micah Birdshire, Julie Borja, Maida Pearce, Becca Jay and family, Audrey Yanos and family, Ellen Haroutunian, Paige Alexander, and Sarah Lambert for helping with fiel d collections.
vii TABLE OF CONTENTS CHAPTER I. INTRODUCTION ................................ ................................ ................................ ................. 1 II. BACKGROUND INFORMATION ................................ ................................ ..................... 5 Pollinator Value ................................ ................................ ................................ .................. 5 Reasons for Pollinator Declines ................................ ................................ ........................ 11 Rural Pollination ................................ ................................ ................................ ............... 15 Urban Pollination ................................ ................................ ................................ .............. 18 Bees along an Urban Rural Gradient ................................ ................................ ................ 22 Common Methodologies ................................ ................................ ................................ ... 23 III. METHODS AND DATA ANALYSIS ................................ ................................ ............. 25 Ethics Statement ................................ ................................ ................................ ................ 25 Site Description ................................ ................................ ................................ ................. 25 Field Methods ................................ ................................ ................................ ................... 28 Bee Sampling ................................ ................................ ................................ .............. 28 Local and Landscape Habitat Characteristics ................................ ............................. 30 Data Analysis ................................ ................................ ................................ .................... 31 Bee Community Composition ................................ ................................ ..................... 31 Bee Community Respons e to Urbanization ................................ ................................ 33 IV. RESULTS ................................ ................................ ................................ ......................... 35 Bee Community Composition ................................ ................................ ........................... 35 Be e Community Response to Urbanization ................................ ................................ ...... 38 V. DISCUSSION ................................ ................................ ................................ .................... 42
viii Bee Community Composition ................................ ................................ ........................... 42 Responses to Urbanization ................................ ................................ ................................ 44 Urban Bees ................................ ................................ ................................ .................. 44 Suburban Bees ................................ ................................ ................................ ............ 48 Rural Bees ................................ ................................ ................................ ................... 51 Limitations ................................ ................................ ................................ ........................ 53 Implications ................................ ................................ ................................ ....................... 55 Management Recommendations ................................ ................................ ................. 56 Future Research ................................ ................................ ................................ ................ 57 VI. CONCLUSIONS ................................ ................................ ................................ .............. 60 REFERENCE S ................................ ................................ ................................ ............................. 63 APPENDIX A. .. 82 B. Bee Abundance and Richnes s per Community ................................ ................................ .. 83 C. Genus Level Bee Rank Abundance Plot ................................ ................................ ............ 84 D. Floral Richness and Percent Imperviousness per Community ................................ ........... 85 E. Bee Species Abundances and Guild Classifications ................................ .......................... 86 F. Supplement for Significance of Response Variables to Bee Guild Explanatory Variables ................................ ................................ ................................ ............................. 88
ix LIST OF TABLES TABLE 1. Significance of response variables to bee guild explanatory variables for all sites. ............... 40 0
x LIST OF FIGURES FIGURE 1. Study sites. ................................ ................................ ................................ .............................. 28 8 2. Bee abundance, richness, and diversity across urban, suburban, and rural communities. ...... 36 6 3. Species accumulation curve combining captures from all study sites ................................ .... 36 6 4. Non metric multidimension al scaling (NMDS) using Bray Curtis dissimilarities for urban, suburban, and rural classes ................................ ................................ ................................ ..... 37 7 5. Total abundance, species richness, and diversity of collected bees ................................ ........ 38 8 6. Scatter plots from stepwise regression of significant bee guild responses to urbanization .... 41 1
1 CHAPTER I INTRODUCTION -including fruits, vegeta bles, nuts, spices, and oilseed -require insect pollination (Klein et al., 2007) , and our r eliance on pollination services t o promote these crops continues to rise due to increasing demands from growing human population s . As of 2010, the global value of i nsect pollination was estimated at US$212 billio n, representing 9.5% of the total value of agricultural production (van Engelsdorp and Meixner, 2010). In the US, bee pollination is responsible for $14 billion of agricultural production (Morse and Calderon e, 2000). Thus, understanding the biogeography of insect pollinators in a variety of environments ( e.g. , urban, rural, and agricultural) will become increasingly important as demand for pollination services continues to rise. In recent years bees have r eceived substantial attention from the media and scholarly researchers due to population declines in many parts of the world. These declines are well documented for honey bees in North America and Europe, with 49.5% managed honey bee colony losses in Nor th America and 25.5% colony losses in Europe between 1961 2007 (FAO, 2009). Wild bees have suffer ed similar population declines . A ccording to the Xerces Society (2017), 57 different wild bee species have been identified as endangered, threatened, or at ri sk in North America, with over 30 bee species classified as critically imperiled or possibly extinct. These losses may have serious implications for plant reproduction, animal survivability, and world food security (e.g. , Majewski, 2016 and Novais et al., 2016). With a 300% increase in deman d for pollinator dependent crops during the last half century the trends in pollinator population losses are alarming (Aizen and Harder, 2009) .
2 Bee declines are occurring due to a range of factors, and no single fact or is entirely responsible. The most commonly described risk factors relate to anthropogenic pressures, including land use change and intensification, climate change, pesticide applications, lack of food and nesting resources, and the spread of exotic spec ies and diseas es ( Vanbergen and the Insect Pollinators Initiative, 2013; Potts et al., 2010 b ; Potts et al., 2016; Grunewald, 2010) . U rban spaces are particularly notorious for creating these conditions . Maintaining bees in urban environments is not only important for the pollination of urban gardens and residential green spaces , but also for mitigating local extirpations of bee species that may otherwise be threatened , particularly specialist bees that collect pollen from just a few flower families or gen era . B ees and other pollinators are significantly impacted by anthropogenic disturbances such as land use modification (Ahrne et al., 2009; Leong et al., 2014; Theodorou et al., 2017) . H owever, less i s known about the degree to which bees are affec ted b y these changes. R esearch examining pollinator assemblages along spatial continua, such as urban rural gradients, is limited and highly variable . Some scholars have found that certain pollinators, such as bees, are positively associated with urban ization (Martins et al., 2017; Theodorou et al., 2017). Many o thers have found that bee communities decline with increasing urbanization (Geslin et al., 2016; Lagucki et al., 2017; Ahrne et al., 2009; Bates et al., 2011; Choate et al., 2018; Verboven et al., 201 4). Variation in these findings may be due to differences in pollinator assemblages. For example, specific bee guilds may respond differently to increasing urbanization (Banaszak Cibicka and Zmihorski, 2012; Fortel et al., 2014). Very few studies have e xplicitly attempted to explain how different bee guilds respond to increasing urbanization, and to my knowledge never from a high elevation sem i arid environment. A bee guild is defined as a species subgroup that
3 achieves similar functions in the communit y and has similar resource requirements (de M. Santos et al., 2013). This research assesses bee communities in a semi arid montane environment where resources are limited, and where the human population is rapidly growing resulting in a lack of ruderal sp aces which are key for pollinator success in urban environments . The objectives of this study are to understand how pollinators are influenced by urbanization by examining their abundance and diversity along an urban rural gradient and to determine if bee species with different life history characteristics are more resilient in the context of urbanization than others . The questions addressed in the research include : 1) how do bee communities respond to urbanization, and 2) how do specific bee guilds respo nd to urbanization? I hypothesized that bee diversity and abundance in the Denver region would be greatest in rural settings, intermediate in suburban areas, and lowest in urban landscapes. This hypothesis was based on research suggesting that while urban areas have the capacity to support certain bee species, bee diversity and abundance diminish with increasing urban intensity (Bates et al., 2011). It was unclear whether this remains true for rural semi arid settings at higher elevations (>5000 ft) due to availability of food resources with a shorter growing season and moisture restrictions. Additional ly , I hypothesized that bee communities in areas with greater urbanization would be dominated by smaller bees with more ecologically generalized characterist ics (e.g., polylectic and eusocial) than rural areas which likely harbor bee communities with more specialized traits (e.g., oligolectic). For example, studies indicate that small bodied species that begin their activity later in the season and are eusoci al are more capable of persisting in city centers (Banaszak Cibicka and Zmihorski, 2012). However, due to limited research on bee guilds in the
4 urban environment it is unclear whether this can be generalized to urban contexts with different climates and en vironments . This study comprises six chap ters. Chapter T wo addresses the relevant background of pollinator conservation. I discuss the value of pollination services and why bees are important . I address the mult ifaceted and complex issues that are likel y leading to pollinator population losses . Then I cover the current knowledge focusing on pollination in rural settings, in urban se ttings, and across urban rural gradients , and conclude with a summary of the common methodologies applied to bee studies. In C hapter T hree , I discuss in detail the methods employed to answer the research questions . C hapter F our discusses the results from data collection and stepwise linear regression modeling. Chapter F ive examines the implications of and discusses how the knowledge gained from this research can inform management strategies to improve pollinator conservation in Denver, CO. Chapter S ix concludes with a summary of this work .
5 CHAPTER II BACKGROUND INFORMATION Pollinator Value The poll ination services provided by bees and other insects is fundamental to the production and re generation of wild plants as well as agricultural crops. It is estimated that about 35% of the human diet benefits from animal pollination (Klein et al., 2007 ; Ghaz oul and pollination increased by 50% in developing nations and by 62% in developed countries (Aizen et al., 2009). As of 2010, the global value of insect pollination w as e stimated at US$212 billion, representing 9.5% of the total value of agricultural production (van Engelsdorp and Meixner, 2010). In the US, bee pollination is responsible for $14 billion of agricultural production (Morse and Calderone, 2000), and is im port ant for production of alfalfa, hay, apples, plums, cherries, pears, almonds, oilseed, canola, blueberries, cranberries, raspberries, squashes, melons, pumpkins, cucumbers, tomatoes, carrots, onion, among others (Cane, 2005). In Mexico, crop species di vers ity is exceptionally high, and nearly 85% of crops produced depend on insect pollination services (Ashworth et al., 2009). Farmers in Kakamega, Kenya experience improved crop yield and quality of produce from bee pollination of pollinator dependent cr ops, and the net benefit to farmers from bee pollination was approximately US$3.2 million, nearly 40% of the annual market value of the region's major food crops (Kasina et al., 2009). In Brazil, 60% of crops depend on pollination to some degree , with one thi rd of these crops having an essential or high dependence on pollinators (Giannini et al., 2015). T he economic value of pollinators in Brazil is estimated to be about 30% (or about US$12 billion) of the total annual agricultural income of pollinator de pend ent crops (almost US$45 billion). However, the actual costs are
6 difficult to estimate because analyses include soybeans , which may have varying degrees of pollinator dependence , contingent on the variety grown. Alternatively, Novais et al. (2016) fou nd t hat 68% of the 53 major food crops in Brazil depend to some extent on animal pollination, and while pollinator dependent crops produce a smaller volume of food compared to non dependent crops, the cultivated area and the monetary value of pollinator de pend ent crops are higher at 59% and 68%, respectively. Specifically, h oney bees play a vital role in plant pollination services, pollinating 90% of commercially insect pollinated crops (Steffan Dewenter et al., 2005). Other reports show that 52 out of 115 o either fruit o r seed set . M any of these crops, including almond, melons , apples, pears, cherries plums , prunes , pumpkins, and squash , would experience over a 90% decrease in yields without honey bee pollination services ( McGregor, 1976; Southwick and Southwick, 1992; Klein et al., 2007). Managed honey bee colonies are ideal for the pollination of large monocrops because hives can be transported and relocated as necessary, co lonies maintain a solid workforce of pollinators throughout the growing season, honey bees can travel longer distances (4.5 km average) to forage (Seel e y, 1985), and they are generalists that will forage from many different flowers. Additi onally, they pro duce important international economic commodities like honey, which had a global production value of about US$1.25 billion in 2007 (FAO, 2009). It is generally assumed that honey bees provide most of the crop pollination, with wild bees c ontributing betw een one sixth and one half of crop production value (Losey and Vaughan, 2006). However, several studies have found that certain wild bees are more efficient pollinators , particularly specialty crops, than honey bees (e.g. , Thomson and Goodell, 2001; Gardn er and Ascher, 2006; Park et al., 2010; Mackenzie and Averill, 1995; Cane, 1996; Stubbs and
7 Drummond, 1996; Stanghellini et al., 1998; Winfree et al., 2008; Morandin et al., 2001; Velthuis and van Doorn, 2006). Research by Adamson et al. (2012) found that non Apis bees contributed to the majority of crop visitation in Virginia, USA, and probably accounted for the greatest pollination services of apple, blueberry, caneberry, and cucurbit s . Specifically, wild bees made up 68 % and 83% of all floral visits to caneberries a nd cucurbits , respectively . The most abundant bees included Andrena , Bombus , and Osmia on apples and blueberries, Lasioglossum on caneberries, and Peponapis pruinosa and Bombus on cucurbits. Their research demonstrates that wild bees may pl ay a greater role in crop pollination than previously thought (Adamson et al., 2012). Bee conservation efforts ha ve been shown to have additional positive effects. Hipolito et al. (2016) e xamined the effectiveness of pollinator friendly practices (metho ds designed to enhance pollinator abundance and diversity) in coffee plantations of Mucuge and Ibicoara, Brazil. They found that increasing pollinator friendly practices resulted in improvem ent of flower visitor richness as well as increased coffee yield. Their results suggest that management decisions oriented towards natural assets such as pollination services can enhance financial assets in the form of crop yields (Hipolito et al., 2016). Other biological assets may be improved as well. Wratten et al . (2012) found that establishment of habitat designed to strengthen ecological fitness of pollinator populations in intensively managed agricultural landscapes in Europe and North America als o has secondary benefits. For example, such practices can increa se overall biodiversity around a farm, protect soil and water quality by reducing runoff potential and mitigating soil erosion, and improve rural aesthetics (Wratten et al., 2012). These co nsiderations can be important for assessing trade offs in decisio n making processes concerning local and landscape scale land management.
8 Bees and the pollination services they provide are also important for the restoration of native prairies in the Great Plains . Prairie restoration is highly dependent on the establis hment of bee p opulations to promote regeneration of native plants because most prairie plants require insect pollination for seed set (Reed, 1993) . However, restoration efforts rarely emphasize the reestablishment of native pollinators , which may hinder p rairie restoration if the site is isolated. For example, Reed (1995) found reduced pollinator richness in isolated restored prairies compared to prairie remnant s in eastern Minnesota . Furthermore, a study conducted in a highly fragmented prairie site in Iowa found reduced seed set of purple prairie clover ( Dalea purpurea ) and lead plant ( Amorpha canescens ) in small remnant prairie patches due to reduced pollinator visitation rates ( Hendrix, 1994 ). These findings suggest that prairie restoration efforts s hould prioritize native pollinator conservation in addition to plant reestablishment. B ees are also important in urban spaces for the develop ment and maintenance of managed vegetation , such as in parks and community gardens . S tudies have shown that be e visitation to purple coneflower ( Echinacea purpurea ) increased in urban areas with higher human population densities compared to natural sites in cities such as Chicago (Lowenstein et al., 2014) . Similarly, u rban community gardens, which conserve biodiv ersity, ecological footprint, and improve overall quality of life (Potter and LeBuhn, 2015), have been shown to benefit from bee pollination. It is estimate d that in New York City bumble bee p ollination is important for the production of numerous common crops grown in urban community gardens (Matteson and Langellotto, 2009) . B ee pollination of tomato plants has been shown to exhibit increased fruit set, fruit mass, yield, and seed set compared to artificial self, artificial cross, and co ntrol treatments (Potter and LeBuhn, 2015). Similarly, Ksiazek et al. ( 2012) found that seed set was significantly reduced when pollination of Chicago green roof
9 plants was hindered , compared to plants subject to open and hand pollinated treatments . Thes e results highligh t the importance of maintaining functional bee communities to promote pollinator dependent plant reproduction in urban spaces . In recent years , bees have received substantial media attention due to population declines around the world. These declines are well documented for honey bees in North America and Europe with 49.5 % 59% honey bee colony losses in US between 1947 2005 (Natural Res earch Council, 2006; van Englesdorp et al., 2008) and 25% 25.5% honey bee colony losses in centra l Europe between 1985 2005 (Potts et al., 2010a ; FAO, 2009 ). According to the Xerces Society, 57 different wild bee species have been identified as endanger ed, threatened, or at risk in North America, with over 30 bee species classified as critically impe riled or possibly extinct (The Xerces Society, 2017). According to van and/or failure of pollinator popu lations to increase at the rate of pollinator dependent crop expansion could have serious effects o loss of pollination services for 29 major food crops in Brazil would decrease food production b y between 16.55 and 51 million tons, worth US$4.86 to $14.56 billion , per year. In areas where rai n forest conversion and fragmentation are expected to reduce pollination services, coffee yields are predicted to decrease by up to 18% and net revenues per hectare may be reduced by up to 14% by 2020 (Priess et al., 2007). Pollinator losses will be spat ially variable along with corresponding reductions in crop production. The importance of pollination services in vulnerable countries is largely driven by poverty level, population density, and level of pollinator dependence for food production (Ashworth et al., 2009). Particularly vulnerable countries may include Guinea Bissau, Benin, -
10 dependent crops compose a considerable proportion of the agricultural area (12 58% ), and over two thirds of native forest growth has been lost (Ghazoul and Koh, 2010). Lautenbach et al. management decisions. Hot spots include areas in the Midwest, south, and west coast of the US, areas in Italy, France, Germany, Poland, Spain, and regions in southwest (e.g. , Turkey and Syria) and e ast Asia (e.g. , China and Japan). In Brazil, pollinator losses would primarily affect family farmers which co mprise pollination services and it is important to maintain free pollination services by protecting and restoring plant resources and conserving pollinators (Ashworth et al., 2009). The rain forests in Central Sulawesi, Indonesia are estimated to provide pollination services that are worth US$53 per hectare. In t hese rain forests, issues of f orest conversion to crop land may be mitigated if remnant forest patches are maintained in the agricultural landscape to support pollinators (Priess et al., 2007). However, an analysis by Ghazoul and Koh (2010) demonstrated no global trend between the di fferent measures of land use intensification (e.g. , forest cover, crop diversity, fertilizer use, rural population growth rates, tractor densities, pollution) and yields of pollinator dependent crops. Furthermore , they found no differences in yields betwe en pollinator dependent and non dependent crops in response to land use intensification. Thus, the authors reported limited evidence of global vulnerability of pollinator dependent crops. They concede that resul ts should be interpreted carefully as they are influenced by several factors, such as smaller scale interactions, farmland management practices, and pollinator efficiency. One reason for the discrepancy may be that pollinators have not yet declined to a c ritical threshold at which crop
11 production is adversely affected (Ghazoul and Koh, 2010). Nonetheless, others have shown that while agricultural dependency on pollination has remained stable from 1993 to 2009, producer prices for pollinator dependent crop s have increased, possibly indicating an ea rly warning sign for conflict between necessity for pollination services and other different land uses globally (Lautenbach et al., 2012). Whether or not the impacts of pollinator declines have begun, pollinator conservation and management remain imperat ive to mitigate population losses and enhance pollination benefits. Reasons for Pollinator Declines Many entomologists and other researchers agree that pollinator populations have declined in recent years. S pec ialist species may be particularly vulnerab le (Kleijn and Raemakers, 2008) as the quality and quantity of available food resources changes through time, but generalist pollinators are still vulnerable. This has important implications for reliable ecologic al functioning and filling specialist niche s and functional roles (Larsen et al., 2005). P ollinators in developed countries have reduced their ranges or have been extirpated in recent years with declines in local abundance and community diversity and composition (Potts et al., 2016 ) ; yet , less is known about the community structure of pollinators in countries like Africa, Asia, and Latin America (Gill et al., 2016). The challenge lies in the fact that there is simply not enough pollinato r abundance data for many regions of the globe, and it is di fficult to predict the complex biotic and abiotic interactions among generalist and specialist bees in response to environmental changes ( Keil et al., 2011 ; Warren et al., 2001; Bommarco et al., 2011; Cameron et al., 2011). Despite these obstacles, some r esearchers have provided evidence to d ocument these pollinator declines. Pauw and Hawkins (2011) , produced a longitudinal model of the relationship between the oil collecting bee Rediviva peringueyi (Melittidae) and a
12 guild of oil secreting orchid (Coryci inae) that are dependent on R. peringueyi for pollination in the Cape Region of South Africa . Their analyses show recent local decl ines in pollination of the bee dependent orchid species. They also documented a recent shift in orchid guild composition in urban areas as bee dependent orchid spec ie s decline d in response to R. peringueyi extirpation. Conversely, the composition of other orchid species that could reproduce asexually remain ed unchanged . Bee declines are occurring due to a range of factors , an d no single factor is entire ly responsible. T hose most commonly cited reasons for bee population decline include anthropogenic pressures, including land use change and intensification, climate change, pesticide applications , a lack of food and nesting reso urces , and the spread of exotic species and diseases ( Vanbergen and the Insect Pollinators Initiative, 2013; Potts et al., 2010 b ; Potts et al., 2016; Grunewald, 2010) . The interactions between these pressures and other naturally occurring biological proce sses are responsible for the decline in pollinators (V anbergen and the Insect Pollinators Initiative, 2013). For example, exposure to pesticides has been shown to result in declining nutrition , which may increase susceptibility to pathogens, reduc ing its lifespan. D eclining nutrition can reduce a be ability to detoxify pesticides (Vanbergen and the Insect Pollinators Initiative, 2013) , making them more susceptible to sub lethal effects such as d ecreased foraging activity (Gill et al., 2012; Fel tham et al., 2014; Gill et al., 2014), reduced homing capabilities (Henry et al., 2012; Fischer et al., 2014), and crippled reproductive efforts (Whitehorn et al, 2012; Rundolf et al., 2015). The combined impacts of pesticides and pathogens may also have additive and synergistic interactions which have been shown to increase energetic stress and mortality in honey bees (Vanbergen and the Insect Pollinators
13 Initiative, 2013; ex. Alaux et al., 2010 and Vidau et al., 2011). Furthermor e, these adverse effects may cascade from the individual to the entire colony ( Pettis et al., 2012) . The negative impacts from pesticides have received attention from the media and scholars in recent years. Stanley et al. (2015) demonstrated the effects of pesticides on bumblebe e pollination services of apple trees in Reading, UK. They found that bumblebees exposed to neonicotinoid pesticides experienced reduced flower visitation and decre ased pollen collection from apple trees. Consequently , pesticide bearing apple trees produ ced fruit with fewer seeds, indicating reduced bumblebee pollination services. The research has important implications for the reliable delivery of quality crops as well as for natural ecosystem functioning in crops where pesticides are applied. David et al. (2016) analyzed pollen grains for concentrations of neonicotinoids from insecticides and fungicides administered to oilseed rape fields and adjacent wildflower fields in East Sussex, UK. They compared these concentrations with pollen collected from h oney bee and bumble bee colonies located in near by farmland , as well as from bumble bee colonies placed in proximate urban areas. They found that pollen from oilse ed rape and adjacent wildflower patches , as well as b ee collected pollen , was contaminated with copious amounts of pesticides. Furthermore, they found that exposure to pesticides among urban bumble bees was substantially lower compared to rural bees. Ther e is c onsiderable interest in predict ing the impacts of climate change on future pollinator populations. For example, Giannini et al. (2017) modeled the effects of climate change on the geographic distribution of 95 pollinator species that facilitate the production of 13 crops in Brazil. The ir models predicted that pollinator occurrence probab ilities will decrease by almost 0.13 by 2050 , with nearly 90% of the analyzed districts experiencing pollinator reductions. Declines in pollinator visitation probabilities were crop dependent (e.g. , 0.08 for
14 persimmon and 0.25 for tomato), and so district s growing different crop varieties are predicted to be variously impacted. However, their results show that the decreases in pollinator occurrence probabilities will primarily affect areas with lower GDP or where crop production or population density is h igher (e.g. , more than 6 million people). Imbach et al. (2017) demonstrate that bee richness in certain Brazilian coffee plantations will decline 8 18% by 2050 due to climate change . Other research shows that many pollinators are shifting their ranges in response to climate change (Klein et al., 2007). These results have important implications for the quality and quantity of floral resources, and negatively impacts human food production (Klein et al., 20 07). O thers have examine d alternative causes for pollinator declines. Jha (2015), in a st udy of bumble bee gene flow, found that gene flow was more limited by oceanic barriers and human driven landscape changes in the southwest coast of the US than by physical terrestrial barriers like mountains. This research emphasizes the importance of creating a nd maintaining geographic resource corridors to facilitate the distribution of bumble bees and other native pollinators . Alternatively, infection experi ments and landscape wide field d ata conducted by F Ã¼ rst et al. (2014) show that honey bee colonies infec ted with deformed wing virus (DWV) share d the same strain of DWV as that found in adjacent infected bumble bee colonies, indicating that honey bees, which are susceptible to a wide variety of pathogens and di seases (Ratnieks and Carreck, 2010; Vanbergen an d the Insect Pollinator Initiative, 2013), are a source of at least one emerging infectious disease in native bees . Therefore, wild pollinators, including bumble bees , may be declining at least in part due t o pathogen spillover from honey bees (Evison et al., 2012; Genersch et al., 2006) or other managed bumble bee populations (Meeus et al., 2011 ). This has critical implications in terms of the risk of spread of emerging infectious disease s between managed a nd wild pollinators .
15 About 80% of wil d plant species are directly dependent on insect pollination for fruit and seed set (Ashman et al., 2004; Aguilar et al., 2006; Klein et al., 2007), while 75% of human food crops require insect pollination (Klein et al. , 2007). Many different floral speci es have declined in abundance and quality in conjunction with pollinator declines (Biesmeijer et al., 2006) , and food crops are becoming increasingly vulnerable to declines in honey bees and other wild pollinators . Hum an well being is at risk and i t is imperative to make significant strides to mitigate or reverse pollinator declines. This will require interdisciplinary collaboration among scientists, policy makers, and the community. Practical actions can include: (a ) increasing and improving upo n the exchange of knowledge pertaining to pollinators, (b) improving landscape management that encourages pollinator conservation, (c) minimizing pesticide risks by including a greater range of potentially impacted pollinator taxa and increasing understand ing of pesticide interactions with other stressors, and (d) enhancing understanding of disease epidemiology and improving management of bee diseases and pathogens; for example, reducing agricultural reliance on honey bee polli nation to reduce disease outbr eaks ( Vanbergen and the Insect Pollinators Initiative, 2013). Other approaches to enhance pollinator co nservation in agro ecosystems include placing a greate r emphasis on ecological intensification, improving ecological infra structure, and reinforc ing div ersifi ed farming management practices ( Potts et al. , 2016) . Rural Pollination Extensive research describ es the role of pollinators, and plant pollinator networks in rural settings. Because bees play such a significant role i n food production, numerous studies focus on agricultural pollination, specifically highlighting the means by which agricultural landscapes may drive or hinder pollination. For example, Cole et al. (2017) found that habitat heterogeneity
16 in agricultural l andscapes including riparian buffer strips, road verges, and open scrub is central to maintain seasonal pollinat or resources and enhance pollinator conservation. Cusser et al. (2016) found comparable results in cotton plantations grown in Texas, USA, while SamnegÃ¥rd et al. (2011) showed that pollinators benefit from residential gardens near intensively managed agri cultural areas. Others have found that bee abundance and richness measures were generally negatively correlated with distance from forest e dges in French oilseed rape fields (Bailey et al., 2014), indicating that natural forest buffers may afford sufficie nt foraging resources for wild bees. Alternatively, one study found that forest edges supported few bee species in New Jersey and Eastern P ennsylvania, USA. Rather, bees were more likely to transition from fallow areas early in the season, to agricultura l crops during mid season, to old field habitats late in the growing season (Mandelik et al. 2012), showing that bees may have a greater rel iance on intensively managed farmland than previously described. In response to agricultural intensification, the UK developed agri environmental schemes (AES) designed to promote biodiversity and protect ecological constituents such as air quality, wate r quality, soil quality, and climate change mitigation (Science for Environment Policy, 2017). The novelty of AES h as resulted in a critical review of the possible benefits for agricultural pollinators. For example, Carvell et al. (2011) examined the eff icacy of a targeted AES approach on bumble bees ( Bombus spp.) in the UK, and found that AES can provide greater bene fits to support abundance and diversity of different bumble bee species compared to other heterogeneous landscapes containing different flor al resources. Pywell et al. (2011) recommend improving AES guidelines by removing competitive grass species, and in cluding more mid and late season floral resources to field edges and fallow areas to enhance pollinator abundance and
17 diversity. Additiona lly, they recommend altering summer cutting to prolong the flowering season. Some studies have distinguished betwe en the impacts of conventional and organic farming on pollination networks and bee abundance measures. Specifically, one study found that i mproving habitat quality in agricultural landscapes enhances pollination in organic farms, more than reducing pestic ide use, which may have significant implications for organic and conventional farming (Chateil and Porcher 2015). Similarly, Munyuli et al. (201 3 ) and KovÃ¡cs HostyÃ¡nszki et al. (2016) found that landscape wide natural or semi natural habitats intermixed w ith traditional low intensity agricultural landscapes have a positive effect on pollinators. However, these results may be region specific, as demonstrated by Kehinde et al. (2018) , who found that bee abundance is higher in organically managed vineyards i n Italy, but not in similarly managed vineyards in South Africa. Other studies have defined the ways in which different pollinator network s enhance agricultural production. For example, Vergara and Badano (2009) found that higher pollinator diversity, i nfluenced by less impacted human systems, enhanced fruit production in Mexican coffee plantations. However, Tepedino et al. (2007) note a p otential risk of competition between native bees and honey bees for fruit orchard floral resources in Capitol Reef N ational Park, Utah , due to a substantially greater abundance of honey bees. They suggest a gradual withdrawal of managed honey bee hives to highlight the economic and ecological importance of native pollinators. To do this, it is important to increase local knowledge and i mprove farmer perceptions of wild bees and other important native pollinators (Hanes et al., 2015; Marques et al., 2017).
18 Alternatively, other studies have analyzed the presence and interaction webs of pollinators in managed rural landscapes other than a gricultural land. Romey et al. (2007) investigated the effects of logging on the diversity of native bees in the Adirond ack Mountains of New York, USA, and found that small scale clear cuttings had great potential to temporarily increase the diversity and abundance of wild bees due to a provisional influx of a flowering understory. Another study analyzed the differences in pollinator plant diversity between organic and conventionally managed grasslands in Ireland and found that organic grassland managemen t increases pollinator plant richness in field centers, while landscape complexity plays a stronger role in pollinator pl ant richness in both organic and conventionally managed grasslands (Power et al., 2012). Conversely, Cole et al. (2015) found that rip arian margins contained richer floral assemblages and supported a greater abundance of pollinators than did surrounding g rasslands in Scotland. Urban Pollination While some researchers consider the urban environment to be detrimental to the survival of b ee populations ( e.g., Carter, 2015), many others have found that cities have great potential to support pollinators (e.g. , McIntyre and Hostetler, 2001; Fetridge et al., 2008; Matteson et al., 2008; Fischer et al., 2016; Matteson et al., 2013; Hinners et a l., 2012; Tonietto et al., 2011; Ksiazek et al., 2012; Carper et al., 2014; Lowenstein et al., 2014). Research has shown that local factors are more important for flower visiting insects than landscape characteristics (Matteson and Langellotto, 20 10 ; Will iams and Winfree, 2013). Thus, many have concluded that local provision of diverse, abundant floral resources (Martins e t al., 2017; Matteson et al., 2013; Parmentier et al., 2014; Wojcik and McBride, 2011; Threlfall et al., 2015; McIntyre and Hostetler, 2001; Davis et al., 2017) with adequate sunlight (Matteson and Langellotto, 20 10 ;
19 Matteson et al., 2013; Williams and Win free, 2013) in urban garden spaces is key for attract ing more abundant and diverse pollinator communities. Furthermore, studies have demonstrated that bees appear to prefer native plants over exotic plants (Frankie et al., 2005; Threlfall et al., 2015; Mc Intyre and Hostetler, 2001), though these findings are m ore variable and tend to be context dependent. Frankie et al. (2005) found that pollinator attraction to host plants varied with native and exotic flowering plants in Albany and Berkeley, California, USA, largely due to the location of plants within resid ential areas. For example, yards with a higher proportion of diverse pollinator friendly plants were more attractive and sustained higher bee diversity compared to yards with a lower proportion of div erse pollinator friendly plants. Alternatively, pollen analyses by Hinners and Hjelmroos Koski (2009) showed that wild bees were unopposed to forage from exotic plants in Denver, Colorado, USA, with 45% of sampled pollen originating from exotic floral spe cies. Their findings demonstrated that bees are apparen tly more inclined to forage locally from non native floral species than from larger scale exotic landscapes such as residential yards. A few studies have explored the utility of green roofs to urban pollinators. Tonietto et al. (2011) compared pollinato r assemblages on green roofs in Chicago, Illinois, USA with those found in native tallgrass prairies and in proximate city parks. They found that bees use green roofs, but are present at lower abunda nces and species diversity compared to the reference hab itats. Similarly, Ksiazek et al. (2012) found lower abundance and diversity measures of bees on green roofs compared to ground level reference sites in Chicago. Both studies concluded that even thou gh abundance and diversity measures are lower on green r oofs compared to ground level sites, the presence of bees indicates that these spaces can be managed to provide functional habitat patches that sustain certain bee populations in a city. Moreover, Ma tteson and
20 Langellotto (2009) concluded that because sun light and floral area are the major limiting factors for local pollinator richness, these results indicate that rooftop garden units may have the capacity to support a rich assemblage of pollinators, even if the area seems relatively isolated from other po llinator habitat areas. Lowenstein et al. (2014) found that bee abundance and richness increased in areas with higher human population densities, which may be due to higher diversities in ornamental f lower plantings. However, others have shown that differ ent bee species are associated with different levels of urbanization; i.e., certain bee species are only found in urban spaces while other species are only recorded in rural settings. These trends ma y indicate a general shift to greater community homogeni zation in city centers. Deguines et al. (2016) demonstrate that urbanization is correlated with lower richness in flower visitors, showing a conversion toward more generalist flower visitor communiti es in Europe. These results indicate that urbanization may lead to large scale homogenization of flower visitor communities in urban landscapes, and they conclude that urban landscape management should place greater emphasis on conservation of flower visi tor communities to include more specialist flower visito rs. Studies from Fetridge et al. (2008) and from Jedrzejewska Szmek and Zych (2013) found that al though urban bee communities resembled natural reference sites, specialist bees were minimal or absent in species count data, and sites were dominated by gene ralist bee species. Similarly, Threlfall et al. (2015) found that residential gardens in Melbourne, Australia are dominated by Apis mellifera , a generalist exotic bee that may compete with native pol linators for floral resources (Tepedino et al., 2007). T he consequences of urban sprawl for pollinator population management are numerous. For example, problems may arise from the displacement or removal of native plants and the corresponding addition of exotic plants (Cane, 2005). Replacement of native with n on native
21 plants may have a negative impact on bee populations , as they generally prefer native floral resources, especially communities of pollen specialists (Frankie et al., 2002). Urban spaces als o attract pests such as rabbits that graze on and defoli ate important pollinator resources. Furthermore, urban sprawl disrupts ecological processes important for pollinator communities, (e.g. , ging the landscape; Cane, 2005). Efforts to encourage c ity residents to incorporate more pollinator friendly practices may require a fundamental paradigm shift , such as allowing weedy flowers to grow in lawns . Larson et al. (2014) showed that bees and ho verflies commonly visit weedy dandelion and white clover in residential turf grass in central Kentucky, USA. Results showed that species richness of bees visiting white clover was equivalent across urban, suburban, and peri urban lawns; however, honey bee s became more abundant and bumble bees proportiona lly less abundant with increasing landscape scale imperviousness , possibly due to an increase in managed honey bee hives in urban ar eas . Results like these suggest that awareness of and adaptation to bette r management practices will be critical for pollin ator conservation in cities. by 2050 (United Nations, 2014). To create urban spaces that effectively promote pollinator conservation, it will be important to advance urban management practices beyond the traditional concept of public makers will need to address the ecological complexity of city landscapes to incorporate efforts that benefit conservation should be a cooperative and integrative endeavor that creates more direct connections with the voting public ( Hall et al., 2016).
22 Bees along an Urban Rural Gradient Human activities such as land use modification strongly influence pollinator communities (Ahrne et al., 2009; Leong et al., 2014; Theodorou et al., 2017); however, research examining pollinator assembl ages along spatial continua, such as urban rural gradie nts, is somewhat limited. The research that does exist demonstrates significant variability. Some have found that some pollinators, such as bees, do well in the urban context (Martins et al., 2017; T heodorou et al., 2017). Others have found that bee com munities generally decline with increasing urbanization (Geslin et al., 2016; Lagucki et al., 2017; Ahrne et al., 2009; Bates et al., 2011; Choate et al., 2018; Verboven et al., 2014). Still others ha ve found that different bee guilds exhibit variable res ponses to increasing urbanization depending on their life history characteristics (Banaszak Cibicka and Zmihorski, 2012; Fortel et al., 2014). While several studies have sought to explain the effec ts of different land use types on pollinator population s, few have attempted to explain how different bee guilds respond to increasing urbanization. One study (Banaszak Cibicka and Zmihorski, 2012) examined the differences in bee species richness and comm unity composition along an urban periurban gradient in guilds according to their ecological traits, showing that small bodied eusocial bees that begin activity later in the season are more capable of persisting in city centers. Another study from Grand Lyon, France found that cavity nesting bees and long tongued bees were found to occur in areas with intermediate to high proportions of impervious surface s . Alternativel y, they found no strong effects of urbanizatio n on the occurrence of species according to body size or social behavior (Fortel et al., 2014). Th e current project builds on these previous studies by analyzing the effects of urbanization on bee communities in the United States, an area known for
23 reduct ions in pollinator populations. Specifically, this research will assess bee communities in a semi arid montane environment where resources are limited , and where the human population is expanding rapidly. Co mmon Methodologies Numerous studies analyzing insect pollinator communities seek to adhere to standardized field method s Plot website accessible from http://o nline.sfsu.edu/beeplot/ . T hese methods include setting pan traps filled with water and a surfactant and sweep netting for a specified amount of time in study sites (Banaszak Cibicka and Zmihorski, 2012; Winfree et al., 200 7 ; Bates et al., 2011; Carper et al., 2014; Fetridge et al., 2008; Hinners and Hjelmroos Koski, 2009; Tonietto et al., 2011; Hinners et al., 2012; Matteson et al., 2008; Matteson and Langellotto, 20 10 ; Quistberg et al., 2016). According to Westphal et al. (2008), pan trapping i s the most efficient means of collecting bees across a broad geographical range. However, pan trapping can lead to biased estimates of bee community compositions as it disproportionately attracts certain genera like Halictus (Can e et al., 2000), and is a poor collection method for other bees such as honey bees, bumble bees, and Colletes (Roulston et al., 2007; Wilson et al., 2008). Despite its shortcomings , pan trapping remains one of the most commonly applied sampling methods fo r bee studies. Sweep nett ing and/or hand searching are often implemented in conjunction with pan trapping. Some researchers have chosen to use vane traps with pan traps, instead of sweep netting (Choate et al., 2018), while others have used only pan trap s (Ksiazek et al., 2012; M cIntyre and Hostetler, 2001; Theodorou et al., 2017), or only sweep netting and visual observations (Larson et al., 2014; Martins et al., 2017; Matteson et al., 2013; Verboven et al., 2014).
24 Pan traps are often composed of small plastic Solo brand cups, colored fluorescent blue, fluorescent yellow, and white as the visual acuity of most bees tested are limited to these colors, in addition to ultraviolet (Grant, 1950; Butler, 1951; Proctor et al., 1996). A drop of liquid soap suc h as laundry detergent or dish detergent creates the surfactant and should be added to the water in each pan trap. Pan traps are often set in the field for an average of 24 hours (Lagucki et al., 2017; Hinners et al., 2012; Hinners and Hjelmroos Koski, 20 09; Threlfall et al., 2015 ; Ksiazek et al., 2012; Tonietto et al., 2011; Geslin et al., 2016; Fortel et al., 2014), though some have set traps for as few as 8 hours (Winfree et al., 2007) and others have set traps for as long as 3 weeks (Samneg Ã¥ rd et al., 2011). Similarly, sweep n etting times are somewhat variable in the literature, varying from 10 minutes (Hinners et al., 2012; Hinners and Hjelmroos Kosi, 2009; Lowenstein et al., 2014; Matteson and Langellotto, 20 10 ; Matteson et al., 2008) to 40 minutes ( Henning and Ghazoul, 2011), with 30 minutes most commonly dedicated to sweep netting (Tepedino et al., 2007; Winfree et al., 2007; Bates et al., 2011; Leong et al., 2014; Martins et al., 2017; Banaszak Cibicka and Zm ihorski, 2012). Sampling frequency shou ld be every two to four weeks during the growing season ( The Bee Inventory Plot , n.d. ). Furthermore, the Bee Inventory Plot suggests taking vegetation surveys within study sites and noting environmental conditions s uch as cloud cover and prevailing winds during each sampling period.
25 CHAPTER III METHODS AND DATA ANALYSIS Ethics Statement I received landowner permission prior to commencement of field work. Permits to collect pollinators were only required for Barr Lake State Park. Efforts were made at a ll sites to only removing specimens that could not be identified in the field and by administering lower intensity sampling methods that retained a small sub set of the bee popula tions at each study site (Bates et al., 2011). Site Description Located 1,610 meters ab ove sea level in a semi arid montane climate 19 km east of the Rocky Mountains, the Denver metropolitan area (39.7392Â° N, 104.9903Â° W) has an estimated population of o ver 2.7 million residents as of 2015 (World Population Review, 2016), making it the 19th most populous metropolitan area in the United States (U.S. Census Bureau Population Division, 2017). With a 2.8% growth rate , Denver is the fastest growing city in Col orado, adding nearly 15,000 people between 2011 and 2012 (World Population Review, 2016). The study area is a typical mosaic of paved surfaces, residential, commercial, and industrial areas, parks, and gardens containing diverse ornamental and native plan arid ecosystems transition from short grass steppe east of Denver (Sims et al., 1978) where tree growth is generally limited to riparian zones and mesic spaces (Hinners and Hjelmroos Koski, 2009) to a gamble oak juniper woodla nd up the eas tern foothills of the Rocky Mountains (Decker and the Colorado Natural Heritage Program, 2007). Average annual rainfall measures between 35 and 40 cm, with most precipitation occurring between May and July. Mean daily
26 temperatures reach lows of 0 Â° C in Januar y to highs of 23 Â° C in July ( National Oceanic and Atmospheric Administration , n.d. ). Approximately half of the study sites consisted of exotic landscaping, including turf grass, shade trees, and ornamental flower plantings, often requiring supplementary wa ter input during the summer months (Hinners and Hjelmroos Koski, 2009). Few bee se spaces. Howev er, for a gradient analysis it is critical to conceptually define each space along the gradient. Comparable studies have opted to use land use/land cover classification systems to identify urban, agricultural, natural, or other land use v ariables within t he study system (Leong et al., 2014; Verboven et al., 2014). Others have chosen to use impervious surface data as a proxy for increasing urban intensity (Fortel et al., 2014; Geslin et al., 2013; Geslin et al., 2016; McCluney et al., 2017 ; Lagucki et al., 2017; Choate et al., 2018). Similarly, the current study aims to use impervious surface information to identify areas of urban, suburban, and rural land use classifications because it is a highly standardized metric for defining urban, s uburban, and rural spaces . I sampled at sites for which I was granted permission , resulting in 12 study sites . Sites were then classified into urban, suburban, and rural categories using percent impervious surface raster data with 30 m resolution , downl oaded from the National Land Cover Database (Xian et al., 2011). S ites containing an average of 50% or more of imperviousness were classified as "urban," sites with an average of 25% to <50% impermeability as "suburban," and sites with an average of <25% impermeability as "rural" (Fig ure 1 ). This classification work was performed in ArcGIS, version 10.5 (Environmental Systems Research Institute, Redlands, CA, USA). Four sites within each category for the Denver metro politan area were identified. Many bee s pecie s
27 forage locally and seldom travel beyond a 2 km radius from their nest in urban setting s (Garbuzov et al., 201 5 ); therefore, each study site was located approximately 4 km or more from the other sites . Urban sites include AUR, TCG, ECG, and IVI. AUR is a university campus located near downtown Denver. TCG and ECG are urban community gardens, and IVI incorporates a small plot behind a commercial building. Suburban sites include IAA, EHH, SAU, and CSH. IAA is located near the intersection of two major r oads and is near the center of a commercial district. EHH and CSH are both residential, and the SAU site is in the lawn of a large suburban church. The rural sites are GCG, BJH, BLP, and GKH. GCG and BJ H are both located on the outskirts of suburban Denv er; the first is a community garden and the second is in a residential subdivision. BLP is a state park north of the city, and GKH is a privately owned horse ranch outside of town.
28 Figure 1 S tudy sites in the greater Denver, C O metropolitan area . Urban sites include IVI, TCG, AUR, and ECG. Suburban sites include EHH, CSH, IAA, and SAU. Rural sites include BLP, GCG, BJH, and GKH . Field Methods Bee Sampling Bees were captured using pan traps and sweep netting (The Bee Invent ory Plot). Bee surveys were conducted approximately once every 4 weeks from May to August on sunny days with light wind and no precipitation. I haphazardly suspended two feet from the gr ound one white, one yellow, and one blue pollinator pan trap, compos ed of painted plastic soufflÃ© cups (Solo brand), via modified wire coat hangers, in each of the sites for 24 hours once a month. The cups were half filled with a solution of 90% water and 10% All Free Clear liquid laundry detergent (to break the water surf ace tension). To minimize disturbances to the pan traps, cups
29 Bee s were captured within 25 m of each site using a long handled sweep net for 30 minutes once a mon th to apprehend the full extent of pollinator activity within each s ite, and to collect bees not attracted to traps (e.g. , pollen specialists) . Sites were visited in varying order each month so that sites were sampled both in the morning and in the afternoon. All honey bees, except the first captured specimen, were recor ded and released after the 30 minutes of sweep netting as they are easily identifiable. All recorded specimens were included in analyses. Specimens were strained, rinsed in water, a nd temporarily stored in denatured isopropyl alcohol to be pinned for iden tification. I pinned a ll bees that were hand netted or collected in pan traps , and A drian C arper and V irginia Scott at the University of Colorado Museum of Natural History assisted with species identification s . Because Agapostemon angelicus and A. texanu s females are morphologically identical (Roberts 1972), these species were analyzed together as a single species aggregate (McIntyre and Hostetler, 2001). All specimens were labeled and sorted to species or morphospecies and are housed in the University o f Colorado Museum of Natural History Entomology Section . Ecological characteristics for each species, such as mean body length, sociality (eusocial, solitary, or parasitic), preferr ed nesting substrate (soil or cavity), pollen specificity (oligolectic or polylectic), and phenology (i.e. , time of first activity) were determined from primary literature, catalogs, and other references (Wilson and Carril, 2016; Michener, 2000; Quistberg et al., 2016; Ascher and Pickering, 2012; Gibbs, 2011; Packer, 199 3 ; McGin ley, 1986; McGinley, 2003; LaBerge and Ribble, 1975; Bouseman and LaBerge, 1978; Scott et al., 2011; Brooks, 1988; Wright et al., 2017; Cockerell and Sumner, 1931; Cockerell, 1903). Species origin (whether a species is native or exotic to North America) w as also identified per Scott et al. (2011). Oligolectic species were identified as those that forage on a
30 single plant family or a subset of flowering genera; no attempt was made to evaluate the extent of oligolecty. Body size was quantified according to the full body length of each specimen, and collected bees were classified as small (<7 mm), medium (7 12 mm), and large (.>12 mm) (Quistberg et al., 2016). C haracteristics were lef t as undetermined when species specific ecological data were incomplete . There were fourteen species with undetermined sociality, and 11 species with undetermined phenology (Appendix E) ; these were excluded from analysis . Local and Landscape Habitat Characteristics For each of the twelve sites, local and landscape habitat cha racteristics were determined to identify the human environmental factors that drive changes in pollinator assemblages along the gradient. Local habitat characteristics mea sured or identified were based on Bates et al. (2011). Richness of flowering forbs a nd woody plants and trees, were sampled once in May and once in July, using 25 m radius circles centered around each study site. Flowers were identified in the field whene ver possible. When flowers could not be identified in the field, floral samples or p hotographs were taken for identification by Christy Briles in the Geography and Environmental Sciences Department at the University of Colorado Denver or by using the Wild flowers of Colorado database ( http:/ /wildflowersofcolorado.com/ ) . Floral abundance was composed of a single flowe r, umbel, head, spike, or capitulum (Carvell et al., 2007). Percent canopy cover was also estimated within 25 m of study site centers (Matteson et al., 2013), and I assessed wind exposure by estimating the percent of land within a 50 m radius not protecte d by tall shrubs, fences, hedgerows, trees, or buildings (Bates et al., 2011). Flora l abundance, percent canopy cover, and wind exposure were maintained as site descriptors and floral richness was maintained as a predictor variable for additional analyses .
31 Landscape metrics were digitized into maps using ArcGIS version 10.5 (Environment al Systems Research Institute, Redlands, CA, USA) to determine the degree of urbanization across the study area. Land use data (Homer et al., 2015) and impervious surface data (Xian et al., 2011) with 30 m resolution were downloaded from the National Land Cover Database. Adhering to the methods prescribed by Jha and Kremen (2013) and Quistberg et al. (2016), landcover types were divided into seven categories according to levels of resistance to bee nesting. Categories included 1) low intensity urban (<50 % impervious surface; NLCD values 21 and 22), 2) medium intensity urban (50 79% impervious surface area; NLCD 23), 3) high intensity urban (80% impervious surface area; NL CD 24), 4) forest (evergreen and deciduous; NLCD 41 and 42); 5) undeveloped open spac e (NLCD 31, 52 and 71), 6) agriculture (NLCD 81 and 82), and 7) wetland and open water (NLCD 11, 90, and 95). Data were quantified within 2 km buffers around each study site to summarize the values of all landcover features contained within the 2 km buff er. Data Analysis Bee Community Composition All analyses were conducted in R version 3.5.1 (R Core Team, 2013) . Bee abundance was calculated as the total number of recorded bees, and species richness was calculated as the total number of bee species doc umented across all study sites. Species diversity was calculated because species richness may increase with greater diversity of land uses within e ach study site (Matteson and Langellotto, 20 10 ; Verboven, 2014). common measure of diversity in ecology, is a measure of the degree of uncertainty that a randomly selected individual belongs to a certain species. Uncertainty increases as the number of species increases and as the distribution of individuals among species beco mes even
32 (Meerman, 2004). Bee abundance data were log transformed to generate a normal distribution. I used one way Analysis of Variance ( ANOVA ) t o determine whether bee abundance , richness, and diversity differed between community types . To determine how effectively I sampled the bee community , I combined all captures across sites and explored regional species richness by fitting an accumulation curve and estimat ing asymptotic species richness . Using the vegan package (Oksanen et al., 2018) for R ( R Deve lopment Core Team, 2015) I generated a species accumulation curve using the ra ndom method with 100 permutations to estimate the total number of unseen species in the study area. Total expected species richness was assessed using the Chao estimator because bee abundance data contains many singletons and it is a relatively unbiased e stimator for species rich assemblages (Fortel et al., 2014; Walther and Morand, 1998). I then ran an asymptotic model on function in vegan to predict the number of sites needed to be within 80% (as a reasonable goal for estimated community coverage) of the estimated Chao. linear species accumulation models such as in asymptoti c models. Differences in l ocal floral richness among sites was tested with a one way ANOVA. Landscape variables were found to be highly correlated with one another . Therefore, I used a principle components analysis (PCA) to incorporate the variables in to a single variable (PC1, hereafter referred to as that expla ins most of the variation in the original variables. The PCA yielded both positive and negative values and explained 56% of the variance. I then removed the negative values to explain 61% of the variance, creating a single
33 A table describing the landscape characteristics and the component urbanization value for each site may be found in Appendix A. To visua lize dissimilarities in bee community composition between urban, suburban, and rural sites, non metric multidimensional scaling (NMDS) was used ( vegan package ) based on Bray Curtis dissimilarities, which accounts for the incidence of unique species between two sites relative to the total of a ll species sampled (Carper et al., 2014). This measure has been shown to have a strong relationship with natural ecological distance when tested with simulated data (Faith et al., 1987). To evaluate the relative fit o f the NMDS , final stress measurements of the ordination were recorded , with values <0.05 indicating robust support for the ordination (Carper et al., 2014). In addition, community ellipses representing the 95% confidence interval around each community cen troid were plotted, and permutational analysis of variance ( Adonis function, with 999 permutations ) was used to test the significance of whether t he centroids differed from one another in the NMDS. To visualize how bee communities respond to different env ironmental factors, I variables as vectors on tests on vector directions to determine if a variable is signifi cantly driving differences between co mmunit ies in the ordination space . Bee Community Response to Urbanization Linear regression was use d to explore the relationships between all three community met ric s (total bee abundance, species richness, and diversity) and urbanization. Percent imperv iousness was used as the predictor variable . Correlation coefficients were used to test the strength of relationships, and one way ANOVA was used to test for model significance.
34 To explore guild specific responses within the bee community to urbanization , separate s tepwise linear regression s (step function in R) were used with the abundance of bees with different life history characteristic s as responses. Stepwise linear regression is a method of fitting regression models by adding or subtracting predict ive variables to find the most parsimonious model that reports the most predictive factors driving bee community dynamics (Efroymson, 1960; Hocking, 1976; Draper and Smith, 1981; SAS Institute Inc., 1989). Thus, in each step a variable is considered for a ddition into or subtraction from the set of explanatory variables based on A kaike Information Criterio n ( AIC ) . in the landscape and local floral richness as fixed effects and conducted forward and backward steps in model selection. Categories containing bees with indeterminate characteristics were excluded from the analysis. If a guild type was recorded at half the sites or less, it was also removed from the analysis.
35 CHAPTER VI RESULTS Bee Community Composition In all, 1 , 611 bees belonging to 5 families (Andrenidae, Apidae, Colletidae, Halictidae, and Megachilidae), 23 genera, and 81 species, were recorded throughout the field season. Bee communities varied between site types , but not always in the hypothesize d direction of dec li nes with increasing urbanization . Bee richness , according to expectations, was highest in rural s ite s with 52 % more species than suburban s ites and 55 % more than urban sites (F 1,2 = 3.4 7 , p = 0. 077 ) . However, bee abundance did not con form to the hypotheses , being highest in rural sites , intermediate in urban sites , and lowest in sub urban s etting s ( F 1,2 = 3.61 , p = 0. 071 ; Fig ure 2 , Appendix B ). According to the Chao estimate (Fig ure 3 ) the study area harbored approximately 129 (Â± 20) s pecies ; thus, I recorded 63% of the estimated number of species present. To obtain 80% (103 species) of the total estimated species abundance, I would have needed to sample approximately 40 sites (Fig ure 3 ). Most bees collected were native. Exotic speci es accounted for 9% of collected bees and included Apis mellifera , Megachile rotundata , and Hyleaus leptocephalus . The most abundant species, accounting for 52% of all specimens collected, was Halictus tripartitus ( Appendix C ) : a small, eusocial flower ge neralist that emerges in early spring. Other abundant species included A. mellifera , Lasioglossum semicaeurelum , Agapostemon texanus/angelicus , and L. sisymbrii . All species are polylectic, collecting pollen from a variety of f lowers. The most represent ed families included Halictidae (85% of individuals, 41% of species) and Apidae (7% of individuals, 20% of species). Thirty seven species were represented by a single individual.
36 Figure 2 Log b ee abu ndance (a) , richness (b) , a nd diversity (c) across urban, suburban, and rural communities. Figure 3 S pecies accumulation curve combining captures from all study sites. Suggested regional species richness was approximately 37 % higher than sample species r ichness. The d otted line represents the asymptotic model on the species accumulation curve , and t he v ertical hashed lin e identifies the approximate n umber of sites needed to sample within 80% of the Chao estimate.
37 NMDS results indicate d very little overl ap between urban and suburban communities, and no overlap with rural bee communities (Fig ure 4 ). The final stress of the two dimensional NMDS was 0.176, indicating acceptable confidence in the ordination distances of each community. The environmental fac tors urbanization and flower richness were significant in driving site community metrics ( p = 0.004), with decreasing urbanization, and increasing floral diversity a cross the gradient towards rural sites (Fig ure 4 ). Rural sites had the greatest floral ric hness with 113 recorded species. Suburban sites had the lowest floral richness with 44 species, and urban sites had intermediate floral richness with 64 species (F 1 ,2 = 1.46 , p = 0.283 ) . Conversely, rural sites had the lowest percent imperviousness with a mean of 7.7%, suburban sites had an intermediate amount of imperviousness with a mean of 35.7%, and urban sites had the highest amount of imperviousness with a mean of 62.0% (F 1,2 = 38 .45 , p = 0.000 ; Appendix D ) . Figure 4 N on metric multidimensional scaling (NMDS) using Bray Curtis dissimilarities for urban (dark gray), suburban (gray), and rural (light gray) classes . Polygons represent 95% confidence ellipses around centroids ( P =0.0 04 ).
38 Bee Community Response to Urbanizati on There was a weak negative relationship between all three community measures ( bee abundance R 2 = 0.1239 , richness R 2 = 0.2926 , and diversity R 2 = 0.2935 ) with increasing percent imperviousness. However , bee richness (p = 0.040) and diversity (p = 0.039) decline s were statistically significant (Fig ure 5 ) . Figure 5 Total abundance (p = 0.141) , species richness (p = 0.040) , and diversity (p = 0.039 ) of collected bees plotted on a scale of percent imperviousness in a 2 km radius around each study site.
39 Most bees across the study area were small (75% of individuals, 52% of species), social (66% of individuals, 14% of species), ground nesting (94% of individuals, 69% of species), pollen generalists (94%, 73% of species). Par asitic bees composed a very small fraction of the bee community for the study area (1% of individuals, 5% of species), and pollen specialists were also uncommon (5% of individuals, 22% of species). Collected species and their associated ecological charact eristic s are outlined in Appendix E . In cases where bee guild specific models showed responses to the explanatory variables , responses were always positive for floral richness and always negative for urbanization. No bee guild was positively associated with urb anization (Figure 6) . However, these results may be driven by total abundance , with substantially higher total abundance in rural spaces with low percent imperviousness compared to suburban and urban spaces with higher percent imperviou sness . I found gui ld specific responses to local and landscape factors for most bee guilds (Table 1) . Body S ize : W h ile small bees declined with increasing urbanization, medium and large bees were not affected by urbanization, though large bees did increase with increasin g local floral diversity. Nesting S ubstrate : Ground nesting bees responded to urbanization, though not significantly. H ive nesting bees were unaffected by urbanization but responded significantly to floral richness. Cavity nesting bees did not respond to either explanatory variable. Pollen Specificity : P olylectic bees declined significantly with increasing urbaniz ation (Fig ure 6) . Alternatively, oligolectic bees did not respond to either explanatory variable. Sociality : Both social and solitary bees responded to urbanization, though in both cases results were not significant.
40 Phenology : All types classified under phenology (early , mid , and late spring) responded to urbanization , though results wer e statistically significant only for early spring bees (Fig ure 6) . Species origin : Native bees declined with increasing urbanization, though not significantly. Alternatively, exotic bees were unaffected by urbanization but responded significantly to floral richness. T hese patterns were driven chiefly by high abundance s of bees in two sites; the rural BJH site and the florally rich urban ECG site. After removing these sites from analyses most results list statistical insignificance , suggesting that bee community response to urbanization is likely driv en by abundant bee species such as Halictus tripartitus . Appendix F provides additional detail . Table 1 Significance of response variables to bee guild explanatory variables for all sites according to stepwise linear regressions . Explanatory Response p Small Urbanization 0.0 47 Medium Flower rich ness 0.102 Large Flower richness + Urbanization 0.0 21 (FR) / 0. 163 (U) Ground Urbanization 0. 064 Cavity None -Hive Flower rich ness 0.0 08 Social Flower richness + Urbanization 0. 0 63 (FR ) / 0.082 (U) Solitary Urbanization 0.190 Polylectic Urbanization 0.0 40 Oligolectic None -Native Urbanization 0. 065 Exotic Flower rich ness 0.0 07 Early spring Urbanization 0. 0 49 Mid spring Urbanization 0.111 Late spring Urbanization 0.152
41 Figure 6 S catter plots from stepwise regression of significant bee guild responses to urbanization. Stepwise regression revealed a) small bees, b) po lylectic bees, and c ) early spring bees all decreased with increasing measures of urbanization (p<0.05) .
42 CHAPTER V DISCUSSION Bee Community Composition Urbanization has a dynamic impact on local flora and fauna; some species do well in a human dominated landscape, while others tend to disappear (Grimm et al., 2008). This research demonstrated that while wild bees exist along the entire urban rural gradie nt, total abundance and species diversity decrease with increasing urbanization. G eneral ly , rural areas had more abundant and diverse bee communities than suburban or urban sites. Moreover, the results from this study suggest that bee community response s to urbanization is guild specific, and that no bee guild s were positively associated with urbanization. Even the most dominant species, including H. tripartitus , A. mellifera , L. semicaeurelum , A. texanus/angelicus , and L. sisymbrii , were negatively asso ciated with urbanization. These species are polylectic, and many are small, native, ground nesters; despite their dominance these guilds responded negatively to urbanization. Thes e results suggest that urbanization has a negative impact on bee population s, and that no bee guild is better able than another to persist in urban spaces. I hypothesized that smaller bodied bees with ecologically generalized characteristics (e. g., polylectic and eusocial) would be more capable of persisting in urban spaces compared to other guilds with specialized traits (e.g., oligolectic). My results show that the most abundant urban bees had life history characteristics that include medium si zed, ground nesting, solitary, native, and with mid spring activity . Despite their abundance, none of these guilds responded positively to urbanization , suggesting a negative association between urbanization and all bee guilds . In addition, s ome of these characteristics , such as medium size and mid spring phenology , dif fer from those emphasized in the stepwise linear regressions , which showed that
43 small, early spring bees most strongly responded to urbanization. This discrepancy may be because the models included bee abundances across the full gradient. The most numeri cally dominant bees were also small and emerged in the early spring . Therefore, the mor e abundant bee guilds may be driving model outcomes. This is further demonstrated with the v ariation in model results when excluding the BJH and ECG outliers , which contained substantially more abundant bee communities . S ince overall abundances are likely driven by numerically dominant guilds (i.e., small, ground nesting, polylectic bees), overall respon se of the bee community to urbanization is likely driven by these most dominant bee guilds , suggesting that further analysis may be necessary to tease apart these relationships . One of the most dominant bee guilds collected in this study included small b ees. Small bees, which have short foraging distances gener ally between 100 and 500 m (Gathmann and Tscharntke, 2002; Greenleaf et al., 2007), were highly abundant in suburban and rural locations, but only a small p ercentage were collected from urban sites . Rather, I found a higher percentage of medium and large bees in urban sites compared to both suburban and rural areas. Small bees have lower flight mobility and may be more sensitive to highly dispersed floral resources and physical barriers, such as b usy roads an d tall buildings that can prevent movement between resource patches (Ahrne et al., 2009; Banaszak Cibicka and Zmihorski, 2012; Geslin et al., 2016). Instead, medium and large bees are likely less impacted by habitat fragmentation (Fortel et al ., 2014; Ste ffan Dewenter and Tscharntke, 1999; Westphal et al., 2006), and may show a higher performance in urban areas with high impervious surfaces because of their increased flight mobility .
44 Another highly dominant guild included ground nesting bees, which requi re bare, untreated ground as a nesting substrate. Consequently, ground nesting bees may be discouraged from existing in urban or suburban spaces where impervious surface density is significantly higher, and where mulch application is common , c ompared to s urrounding rural areas (McIntyre and Hostetler, 2001; Cane et al., 2006; Fortel et al., 2014). This study showed that cavity nesting bee abundances were greatest in urban settings, agreeing with findings from Fortel et al. (2014). However, ca vity nesters as well as other bee guilds, such as oligolectic and large bees, had very low abundances and may not be well enough represented in this study to detect meaningful relationships. M any researchers have reported decreasing abundances of ground n esting bees with increasing imperviousness (Cane et al. 2006; Geslin et al. 2016; Matteson et al. 2008; Winfree et al. 2009). Cavity nesting bees, on the other hand, may have a greater advantage with higher (manmade) nesting resource availability in an ur ban context (McFrederick and LeBuhn, 2006; Hernandez et al., 2009; Metteson et al, 2008). T hese results underscore the potential for functional homogenization of the bee community in urban spaces (Deguines et al., 2016; McKinney, 2006) , an undesirable yet common side effect of urbanization. Responses to Urbanization Urban Bees Overall, w ild bee abundance and diversity decrease with increasing urbanization. However, this work has demonstrated that floral resources have a positive effect on bee communitie s, particularly in urban spaces. The ECG site, an urban area with high floral richness, supported substantially more bee species compared to other urban and suburban sites, suggesting that local fl oral richness, more than the surrounding urban landscape, is important in sustaining a diverse bee community. If floral rich ECG is removed from the stepwise linear
45 regression models, urban bees showed higher response to urbanization than to floral richne ss in the models. These results suggest that local floral richness is more important than the landscape urban matrix in sustaining urban bee populations. These results are consistent with other research, such as Geslin et al. (2013) who found that plant pollinator interactions decline with increasing urbanizati on (also de fined by percent imperviousness) in Paris, France. McIntyre and Hostetler (2001) found a greater proportion of bees in natural desert landscapes compared to residential sites. Similarly , Ahrne et al. (2009), demonstrated that bumble bee abunda nce and species composition was positively associat ed with local floral abundance and negatively associated with increasing urbanization in Stockholm, Sweden. They concluded that landscape changes associated with urbanization negatively affect bumble bee diversity, while local factors, such as floral abundance, positively affect bumbl e bee abundance. Similar patterns were found for wild bees and other pollinators in Boulder, CO (Kearns and Oliveras , 2009), New York City, NY (Matteson et al., 2008), Paris, France (Geslin et al., 2016), Grand Lyon, France (Fortel et al., 2014), and Toledo, OH (Lagucki et al., 2017). Given these results , local factors may be more importa nt for managing bee populations than the surrounding urban landscape matrix. This study demonstrated a negative association between urbanization and all bee guilds that responded to urbanization , including small , polylectic, and early spring bees . However , others have found that some bee guilds may have a greater associated with u rbanizatio n. Specifically, generalist bees have been shown to maintain a stronger relationship with urbanization compared to specialists (Banaszak Cibicka and Zmihorski, 2012; Hernandez et al., 2009; Deguines et al., 2016; Frankie et al., 2005; Geslin et al., 2013; Jedrzejewska Szmek and Zych, 2013). For example, Hinners et al. (2012) found higher abun dances of small, social, and cavity nesting bees in grassland remnants around Denver and Boulde r suburbs. Alternatively,
46 Banaszak Cibicka and Zmihorski (2012) found a greater abundance of soil nesting species in Poland. Specifically, modeling efforts demonstrated that social bee species that began their activity later in the season had a greater c apacity to exist in the city center. Fortel et al. (2014) found that l ong tongued and cavity nesting bees had higher abundances in areas with a higher proportion of imperviousness. Differences between this study and other findings could indicate that bee community responses to urbanization may depend on the way in which the se spaces are managed. Th e current study occurred in a high elevation semi arid climate, which may be less supportive of wild bee populations due to the short length of the growing season and less floral diversity. Furthermore, Denver lawns are primarily composed of Kentucky blue grass and other non native grasses, which are exotic to the semi arid climate of Denver and do nothing to support wild bee populations (Davis et al., 2017). Therefore, inappropriate lawn management may be contributing to some of the contrasting findings to locatio ns with more native plantings. For example, Hinners et al. (2012) focused on remnant grassland spaces in the suburbs of Denver and Boulder, wh ich differed substantially from urban and suburban sites in this study, where ruderal vegetation was sparse. Sim ilarly, Banaszak Cibicka and Zmihorski (2012) selected meadows, for urban study sites. These sites might be tter represent suburban spaces i n the cur rent study. Alter natively, similar to the study design in this project , Fortel et al. (2014) focused on percent imperviousness for urban site selections. Thus, disparities between results may also be due to regional cli mate differences between the study areas . Analyses i n Fortel et al. (2014) took place in a Mediterranean climate which is known to be favorable to many wild bees (Michener, 1979). Given these region specific responses of bee guilds to urbanization, more studies, such as
47 this one, are needed to explore pote ntial mechanisms driving community responses to urbanization. The highly abundant and diverse bee community found at the ECG site suggests that community gardens may provide a diverse array of floral resources to support rich bee communities. Others agree that community gardens have great potential to support urban bees and other pollinators (e.g. , Ahrne et al., 2009; Gotlieb et al., 2011; Samneg Ã¥ rd et al., 2011; Hinners et al., 2012). Kaluza et al. (2016) found that nec tar forager abundances were higher in community gardens compared to plantations and adjacent natural forests in Queensland, Australia, likely due to a constant nectar availability t hroughout the season. Matteson and Langellotto (2009) found an abundance o f bumble bees in New York City community gardens to facilitate pollination of garden plants. Potter and LeBuhn (2015) found that urban spaces have the potential to support adequate pollination services by wild bees to facilitate urban agriculture. Altoge ther, these results suggest that community gardens can be highly beneficial to urban bee communities. On the other hand, w hile the ECG site contained a rich bee community, the TCG s ite, another small urban community garden, maintained substantially fewer i ndividuals and species, possibly because the garden is embedded in an impervious landscape. Given the greater p ercentage of medium bees captured at the TCG site, it was likely too ecologically remote for many smaller bee species to access. Similarly, Mat teson et al. (2008) found substantially reduced native bee richness in New York City community gardens compared to the surrounding natural landscape. Furthermore, Fetridge et al. ( 2008) showed that bee communities in Westchester County, NY suburban reside ntial gardens are ecologically like communities in a nearby research preserve, despite dissimilarities in plant communities, suggesting that suburban
48 gardens, more so than urban gar dens, may provide better support for taxonomically diverse bee communities. Therefore, urban gardens should be accessible and produce floral resources throughout the season to be most beneficial to bees. While the IVI site was not considered to be an ou tlier in this study, it also maintained somewhat higher abundances and grea ter diversity of bees relative to the TCG and AUR urban sites. The small lawn space at IVI was allowed to grow wild, and contained many different ruderal species compared to the su rrounding turf grass lawns. The IVI site was probably highly attractive to the local bee community in an area where resources were sparse. This has significant implications for the Denver metropolitan area where ruderal sites are disappearing with the in cursion of housing construction and infrastructural development. Similarly , Jedrzejewska Szmek and Zych (2013) found that even small patches of natural vegetation in urban spaces may support appreciable pollinator diversity. Additionally, Larson et al. (2014) found that weeds grown in turf lawns were an important food source, a nd may even act as resource corridors or stepping stones for certain bee species in Lexington, KY. The importance of maintaining diverse local resource patches for bees and other pollinators is well documented (e.g., Ahrne et al., 2009; Kearns and Olivera s, 2009; McFrederick and LeBuhn, 2006; Potts et al., 2003; Threlfall et al., 2015). Therefore, p atches of remnant vegetation, if they are accessible, are critical to maintain urb an bee populations. Suburban Bees While urbanization can be detrimental to some species, suburban spaces can offer a diversity of environmental characteristics , such as warmer winter microclimates, increased abundance of floral resources, and greater cavity nesting potential (Bates et al., 2011; Matteson et al., 2008; Cane et al. , 2006). My results show that suburban spaces maintain moderate levels
49 of bee abund ance and diversity relative to urban and rural sites. O thers have found that habitat s in suburban spaces either have no effect, or positively impact pollinators. For exam ple, Hinners et al. (2012) found no strong effect of suburban development on bee diversity, abundance, and community composition within patches of shortgrass steppe in the Front Range of Colorado . Similarly, Fortel et al. (2014) found that bee richness wa s hig hest in sites with intermediate imperviousness (i.e. , suburban sites). Others have obtained similar findings in Raleigh Durham, NC (Carper et al., 2014), Brentwood, CA (Leong et al., 2014), and Marzahn Hellersdorf, Germany (Fischer et al., 2016). D i ffer ences between my study and other findings may indicate that bees react differently to different landscape contexts. For example, the study by Hinners et al. (2012) was conducted in suburban areas associated with shortgrass steppe and grassland fragme nts, differing from the current study where suburban sites were in highly maintained commercial and residential plots with relatively low floral richness . If bees are sensitive to local factors (Bates et al., 2011; Davis et al., 2017; Matteson and Langell otto, 20 10 ; Williams and Winfree, 2013), then they may react positively to the more natural suburban sites in the Hinners et al. (2012) study, and negatively to the highly managed suburban sites in my study. Alternatively, climates in other studies, such as in Fortel et al. (2014), tend to be temperate or Mediterranean , which may have a greater capacity to support wild bees and their resources compared to the semi arid climate of Denver, CO. Another explanation may be that the sites in Denver, CO may be s o i ntensively managed that many of the more sensitive bee species have already declined, resulting in lower abundance and diversity in both urban and suburban sites than what may have existed previously . Habitat complexity , such as that seen in suburban dev elopment, has been cited as a habitat
50 suggests that introducing complexity to a landscape, such as through built structures, can benefit be e communities because the changes may supplement resource gains with the addition of new niches (MacArthur and Wilson, 1967) . However, too much complexity, like in a city center, negatively impacts pollinators as resources are removed or isolated to sm aller patches. Thus, the complexity in suburban development, rather than reducing habitat, is thought to increase habitat heterogeneity and alter the availability and distribution of resources increasing species coexistence (Hinners et al., 2012; McKinney , 2008). These areas of intermediate disturbance intensity often border the urban rural interface; thus, pollinator spillover from one community to the other may o ccur as resources become available throughout the season in the different landscape contexts (Pereira Peixoto et al., 2014). The current findings for suburban spaces fail to uphold the disturbance heterogeneity hypothesis with intermediate bee richness, an d even reduced bee abundances relative to urban spaces. This suggests that suburban spaces in Denver have sustained significant environmental shifts relative to natural conditions to reduce, rather than support bee populations. Or similarly, it may suggest that suburban s paces should be classified as something other than intermediate. Accordin g to the percent impervious surface definition, these areas are intermediate in that they co nsist of a transitional measure of imperviousness. However, given other definitions of land use or landscape composition , such as degree of anthropogenic modificati on or intensification , suburban areas may be classified with a higher intensity. In this case it would be necessary to explore the effects of such modifications as it pertains to suburban areas, as well as urban and rural spaces. Other researchers have f ound that bees and other pollinators have greater abundance or diversity in areas that have a moderate proportion of built space. Hinners et al. (2012) found that low level suburban development can have a positive effect on bee species
51 richness if a certa in amount of native habitat is retained. Zerbe et al. (2003) suggested that moderate intensities of disturbance increase habitat diversity with subsequent increases in species diversity. There was a greater abundance of exotic bees in suburban spaces re lative to urban and rural areas, possibly due to an increase in the amount of exotic and ornamental vegetation planted in these spaces (Thompson et al., 2003; Frankie et al., 2005; Matteson et al., 2008). Alternatively, exotic bee abund ance may be higher in the suburbs due to an influx of managed honey bee hives. There is increasing interest in maintaining honey bee ( Apis mellifera ) hives in Denver and other cities to help supplement pollination in community gardens (Cane, 2005; Peters, 2012). However, a ccording to abundance metrics, results suggest that wild bees may be just as or more relevant for sub urban pollination. Similarly, Lowenstein et al. (2015) found that most bees visiting experim ental crop plants in Chicago were wild. To my knowledge, ther e is no vegetation in Denver that relies exclusively o n honey bee pollination. I t may prove b eneficial to increase efforts to improve wild bee populations in the city to enhance urban pollination and concurrently reduce reliance on a single pollinator spe cies. Rural Bees Generally, rural sites contained bee abundances with over 100 specimens , a substantially greater number than in the suburban and urban bee communities (excluding the ECG site) . However, only 74 specimens were collected from the BLP site , a state park located north of Denver . This may be because at least 43% of the site was composed of water which substantially li mits food and nesting resources. If sampling had occurred further away from the water , it is possible I may have collected mo re individuals. Otherwise, relative to urban and
52 suburban communities, bee richness estimate s were similar across rural sites, ranging from 16 (at the GCG site) to 30 species (at the GKH site) . The most abundant site, and the second most diverse site wa s the rural BJH site, a residence on the exterior edge of suburban Denver with 10 percent impervious surface , which was ten times higher than the most remote rural sites. This site may sustain an appropriate level of habitat complexity to support highly a bundant and diverse bee populations. Specifically, w hen considering this site individu ally, these results may coincide with the habitat heterogeneity hypothesis sites that are more rural or urban, to support highly abundant and diverse bee communities. H owever, in this case xity The current study differs from many others in that none of the rural sites incorporate d agricultural land. Therefore, while the rural sites may experience local pesticide exposure, broadcast pesticide spraying was minimal. Furthermore, the rural sites in this study maintained higher floral richness relative to many agricultural areas (Tsc harntke et al., 2005; McCabe et al., 2017; Goulson et al., 2008; Carvell et al., 2006). The two most rural sites, BLP and GKH, are large tracts of land with a s ignificant proportion of the area dedicated to growth of ruderal vegetation. These sites are s urrounded by agricultural fields; however, both sites are large enough to sustain many different bee species, especially smaller bees, independent of the surroun ding agricultural landscape. This may explain why the rural areas in my study maintained signi ficantly higher amount s of bees compared to the urban and suburban sites, relative to other studies where rural bees had fewer resources and were likely exposed to agricultural pesticide sprays and landscape simplification (Verboven et al., 2014; Theodorou et al., 2017; Fortel et al., 2014; Sanchez Bayo and Koka, 2014). For example, Verboven et al.
53 (2014) found that bee abundance and diversity were highest in nat ural rural landscapes compared to urban spaces, like the current findings. But when they compa red findings to those from agricultural rural landscapes, they found that agricultural landscapes supported substantially fewer bees, likely due to low floral diversity and the application of chemicals. Altogether, these resu lts suggest that agricultural landscapes can negatively impact pollinator populations, and the fact that rural sites incorporated very little agricultural land may have positively influenced results such that I collected substantially more bee s from rural areas. Limitati ons The two most abundant genera found in the study included Halictus and Lasioglossum, which is a common finding (e.g., Marini et al. 2012; Morandin and Kremen 2013; Fortel et al. 2014; Geroff et al. 2014; Saunders and Luck 2 014; TornÃ© Noguera et al. 2014 ; Pisanty and Mandelik 2015; Geslin et al., 2016; Hinners and Hjelmroos Koski, 2009; Hinners et al., 2012; Cane et al., 2000). These bees are well represented in pan trap s (Rollin et al. 2015) , suggest ing that my findings are potentially biased towards th ese bee genera. Cane et al. (2000) found that pan trapping gave a poor estimate of the actual abundance and richness of native bee fauna existing on creosote bushes in the Silverbell bajada near Tucson, AZ. Fu rthermore, they found that traps failed to ca pture specialist bees. They conclude d that pan trapping should be used in conjunction with other passive trapping methods to obtain a more complete sample of species existing in the sampled area. Furthermore, the potential for pan trap bias exists in are as where resources are sparse and pan traps may be highly attractive to bees. An inverse relationship has been shown to exist between pan trap effectiveness and flower abundance in areas where flower abundance is low (Geslin et al., 2016; Roulston et al., 2007; Baum and Wallen, 2011). Sweep netting was supplemented to address this bias.
54 Relationships between urbanization and specific bee guilds may be highly correlated with one another (Williams et al., 2010). For example, nearly all small bees nest in t he ground (A.C., personal communication). Teasing apart thes e relationships to find the precise ecological mechanism that responds to urbanization may be difficult. For example, d o small ground nesting bees re spond negatively to urbanization because thei r size limits their flight mobility through a city, or because they are unable to find adequate nesting substrate in an impervious surface environment, or a combination of factors? Interactions among species li fe history characteristics may be critical in understanding a species response to urbanization and other disturbances (Williams et al., 2010). Understanding and incorporating these interactions into analys i s can give greater meaning and interpretability of results in similar investigations. A dditi onal research that teases apart relationships among bee guilds in urban and other disturbance environments would help identify confounding variables . Study sites were selected using a convenience based appro ach . T hus, s ampling wa s conducted where perm ission was granted , and consideration was given to the number of bees euthanized in the study, especially given the low abundance and diversity in the urban environment. Therefore, I was restricted to only twelve study sites for field collections. This re duced sample size confined the analyses, and additional study sites would help to validate current findings and increase the reliability of these research outcomes. For example, I estimated that approximately 129 bee species are present in the study area, 37% more than what I recorded. Scott et al. (2011) documents 156 species for Denver County alone . Furthermore, the 81 bee species recorded in this study comprise only 8% of the 946 species documented in the state of Colorado (Scott et al., 2011), indica ting that the city may be an unprofitable habitat to sustain a diverse bee community. However, year to year variation in bee abundance due to seasonality
55 differences may have impacted abundances recorded in this study (Cane and Payne, 1993; Choate et al., 2018). This underscores the need for large geographic studies an d over multiple seasons to explore spatio temporal patterns of community response s to urbanization and other anthropogenic disturbance regimes. F inally, it is important to underscore the ax iom that correlation does not prove causation. The fact that some bee guilds were negatively associa ted with urbanization (here defined by the percent of impervious surface in 2 km around each s tudy site) does not mean that urbanization is decidedly the c ause of their low abundances in the city. There may be other factors affecting communities that were not accounted for in this study , including socioeconomic, ecological, or historical factors. Implications Generally, bees require two main resources food and shelter to survive. Several bee species, such as H. tripartitus nest in the ground (Wilson and Carril, 2015). Others, like Hylaeus leptocephalus , prefer to create shelters out of wood cavities (Ascher and Pickering, 2012). Many be es are polylectic ; however, a few are oligolectic and require food resources from one family or genus of plants, such as Diadasia enavata that specializ es on sunflower pollen (Wilson and Carril, 2015). In natural settings bees are usually able to meet the se requirements. Alternatively, urban settings are often associated with habitat homogenizat ion via the removal of native plants and an associated increase in impermeable surfaces and physical barriers, resulting in the elimination of many important bee r esources and the incident al decline in local bee populations ( Decourtye et al. 2010; Winfree 2010; Roulston and Goodell 2011) . Increasing urban habitat heterogeneity by incorporating higher vegetation diversity and providing resource corridors may signifi cantly benefit urban pollinator populations (Geslin et al., 2016).
56 Management Reco mmendations Urbanization has had a significant impact on animal diversity and species loss, and studies such as this highlight the extent of these losses (Lagucki et al., 2 017) . T his study highlights s ome simple recommendations for improving bee community structure in the urban environment . Because bees responded positively to floral richness, planting a wide variety of flowers should be an effective strategy to better man age bee communities along the urban rural gradient. Thi s was especially evident in the ECG site , a florally rich urban site. Furthermore, because bees responded negatively to urbanization , it should prove beneficial to incorporate more natural park space in citywide management. For example, improvements may include plant ing native shortgrass steppe grassland species in parking lot islands and along road verges instead of mulching and exotic shrub plantings. A more visionary goal may include managing for native grassland vegetation growth aro in tegrate pollinator corridors throughout the metropolitan area . Increasing habitat heterogeneity by planting a variety of pollinator friendly plants as well as leaving patches o f fallen wood, leaves, and bare ground for nesting substrate would significantly improve bee communities around Denver. Finally, actively involving Denver residents in pollinator habitat restoration would advance local knowledge on the importance of maint aining urban pollinators and encourage residents to manage lawns more sustainably. This could include volunteer opportunities in citywide pollinator habitat management, and making classes focus ing on pollinator value and habitat requirements freely availa ble to reside nts. Others have demonstrated the effectiveness of these approach es. I mproving the quality and quantity of local floral resources has a positive impact on nearby pollinator communities (Davis et al., 2017) , and m any have shown that increasi ng floral richness in an urban space can
57 increase bee species richness (Kearns and Oliveras 2009; Hennig and Ghazoul 2011; Wojcik and McBride 2011; Geslin et al., 2016). Specifically, modeling efforts from Davis et al. ( 2017 ) demonstrated that co nversion from turf grass to more florally rich lawns would help increase pollinator populations throughout Chicago, IL. Furthermore, it has been shown that wild bees prefer native vegetation over exotic plant species (Frankie et al., 2005) . T herefore, ot her simpl e adjustments could include planting more native pollinator open spaces. Fischer et al. (2016) have explicitly shown that habitat restoration efforts in urban spaces such as th ose recommended here can positively impact bee species abundances. Furthermore, Marques et al. (2017) showed that integrating educational activities such as visual aids (e.g. , posters, videos , museum specimens , and bee nests) and games involving pollinat ion and pollinator co ncepts (e.g. , memory g ames, puzzles, board games) into rural elementary school curricula increase d knowledge and awareness of bee morphotypes . They also showed that educational efforts improved negative perceptions about bee human interactions . In Denver, city plann ers and subject matter experts may consider integrating similar activities to educate and involve the public in pollinator conservation. These unelaborate and cost effective changes have the potential to invite bees back into Denver and significantly miti gate the declining trends in the conservation scient ists, and the community to effectively move forward with urban pollinator conservation (Hall et al., 2016). Future R esearch The number of sites sampled in this study was a limiting factor in analys es. If sample size was not a limitation, this study coul d be expanded to integrate a more in depth analysis of the interactions and processes of bee communities along the urb an rural interface. Increasing the
58 sample size could allow for the capture and subsequent investigation of other bee guilds not collected or recorded in this study. It could build on the current analysis by evaluating local and landscape variables more t horoughly, for example, by comparing sites with more variability in percent imperviousness , sites with different proportions of native vegetation, or sites with greater variability in ground cover composition. Furthermore, future work could incorporate ot her potentially important factors such as socioeconomic inequalities between sites, other environmental variables such as differences in regional precipitation, and historical data that may have influenced the development of certain areas around the city. Furthermore, f uture efforts should focus on developing new methods of identifying bees i n the field to reduce the need for destructive measures of identification. Systemati cally k illing bees for research is a common and currently necessary method of obt aining species level identifications. While various trapping and netting methods remove only a n insignificant subset of bees from the overall population , it would be prudent, and may become necessary in the future, to develop less destructive methods for bee and other insect studies. Doing so c ould marginally mitigate pollinator popul ation reductions and aid research in terms of acceptance and funding. Other possible methods could incorporate the use of high resolution cameras or even artificial intellig ence identification capabilities. Finally, a n applied study examining the efficacy of the preceding management recommendations for Denver would facilitate meaningful action towards urban pollinator conservation. At this time , pollinator research has aff orded enough of an understanding of bee habitat requirements to begin implementing the strategies that many researchers focus on , such as incorporating more floral diversity and creating a greater network of habitat corridors . It is time to discover th ose strategies that truly make a difference for urban bee communities. To be
59 effective, t his will require cooperation between subject matter experts, volunteers, organizational leaders, city planners , and policy makers .
60 CHAPTER VI CONCLUSIONS This study sought to answer two questions: how do bee communities respond to urbanization, and how do specific bee guilds respond to urbanization? To answer the se questions , I classified the study area into urban, suburban, and rural spaces and collected data on be es from each of twelve study sites existing within the urban rural gradient around Denver, CO . I then modeled their guild responses to landscape scale urbanization and local floral richness. Bees responded negatively to urbanization , showing declines in both abundance and diversity with increasing imperviousness. These results indicate that currently the city of Denver is unsuitable to sustain diverse and abundant bee communities. Several factors may be driving these results, inclu ding the higher perce ntage of impervious surfaces in urban and suburban spaces relative to rural areas. Impervious surfaces prohibit ground nesting bees from finding adequate nest substrate in cities, which significantly reduced the number of bees collecte d in urban spaces. S imilarly, small bees are often prohibited by large structural barriers common to city centers , which also diminished the number of urban bees collected. Furthermore, the semi arid climate of Denver may further limit pollinator resourc es throughout the cit y. Contrary to other studies , n o bee guilds were positively associated with urbanization . Guilds that responded to urbanization declined with increasing urbanization, indicating that semi arid , high altitude urban spaces negatively impact many bee guil d s . Life history characteristics analyzed included pollen specificity, nesting substrate, sociality, phenology, body size, and species origin . C avity nesting medium sized bees were relatively more abundant in urban sites compared to suburban or rural site s. However, even these guilds failed to show a
61 positive association with urbanization. Altogether, t hese results may be biased towards the more dominant small, ground nesting species which we re highly abundant in rural sites . Alternatively, I found that bees responded positively to flower richness, and that floral richness, more so than urbanization, may have a higher importance in managing bee populations in a city. T he positive response to floral richness indicates that while the city of Denver is cur rently unsuitable to support diverse and abundant bee communities, increasing floral diversity throughout Denver has the potential to benefit bee populations. I therefore recommend that city p lanners and policy makers consider incorporating greater floral diversity in the managed vegetation throughout the city. This may even include the development of new pollinator habitat corridors through preexisting thoroughfares. I also suggest a greater inclusion of native vegetation and habitat heterogeneity given findings from other studies. This research has improved the current understanding of the effects of increasing urban intensity on bee pollinators in Denver, CO. It is crucial to think criti cally about how management decisions in urban spaces affect bee and other pollinator populations because of deep reliance on pollination services . Every third bite of food we take is dependent to some degree on pollination ( Ghazoul and Koh, 2010 ) . Therefore, declining pollinator populations has serious im plications for human kind . In Denver, r educed urban bee populations suggest inhibited pollinatio n services and reduced seed set of urban plants, which will negatively impact urban landscape managers as well as community and residential gardeners. Furtherm ore, extirpations of local bee communities have negative implications for regional and global po pulations. It will become critical for the residents of Denver to alter current management rove urban bee populations.
62 Pollination is a natural global service that humanity cannot afford to lose. The more we understand pollinators, the more we can do individually and as a community to protect pollinator populations. Any opportunity to promote pollinator education and awareness will benefit both pollinators and the people who rely on the m.
63 REFERENCES Adamson, N. L., Roulston, T. H., Fell, R. D. & Mullins, D. E. (2012). From April to August Wild bees pollinating crops through the growing s eason in Virginia, USA. Environmental Entomology, 41 (4), 813 821. doi: 10.1603/EN12073 Aguilar, R. Ashworth, L., Galetto, L., & Aizen, M. A. (2006) Plant reproductive susceptibility to habitat fragmentation: Review and synthesis through a meta analysis. Ecology Letters, 9( 8), 968 980. doi: 10.1111/j.1461 0248.2006.00927.x Ahrne, K., Bengtss on, J., Elmqvist, T., & Iantbruksuniversitet, S. (2009). Bumble bees (Bombus spp) along a gradient of increasing urbanization. PLoS One, 4 (5), e5574. doi: 10.1371/j ournal.pone.0005574 Aizen, M. A. & Harder, L. D. (2009) The global stock of domesticated honey bees is growing slower than agricultural demand for pollination. Current Biology, 19 (11), 915 918. doi: 10.1016/j.cub.2009.03.071 Alaux, C., Brunet, J. L., Duss aubat, C., Mondet, F., Tchamitchan, S., Cousin, M., Brillard, J., Baldy, A., Belzunces, L . P., & Le Conte, Y. (2010). Interactions between Nosema microspores and a neonicotinoid weaken honeybees (Apis mellifera). Environmental Microbiology, 12 (3), 774 782. doi: 10.1111/j.1462 2920.2009.02123.x Ascher, J. S. & Pickering, J. (2012). Discover Li fe bee species guide and world checklist (Hymenoptera: Apoidea: Anthophila). (http://www.discoverlife.org/mp/20q?guide= Apoidea_species). Ashman, T. L., Knight, T. M., Steets, J. A., Amarasekare, P., Burd, M., Campbell, D. R., Dudash, M. R., Johnston, M. O ., Mazer, S. J., Mitchell, R. J., Morgan, M. T., & Wilson, W. G. (2004). Pollen limitation of plant reproduction: Ecological and evolutionary causes and consequences. Ecology, 85 (9), 2408 2421. doi: 10.1890/03 8024 Ashworth, L., Quesada, M., Casas, A., Agu ilar, R., & Oyama, K. (2009). Pollinator dependent food production in Mexico. Biological Conservation, 142 (5), 1050 1057. doi: 10.1016/j.biocon.2009.01.016 Bailey, S ., Requier, F., Nusillard, B., Roberts, S. P. M., Potts, S. G., & Bouget, C. (2014). Dis tance from forest edge affects bee pollinators in oilseed rape fields. Ecology and Evolution, 4 (4), 370 380. doi: 10.1002/ece3.924 Banaszak M. (2012). Wild bees along an urban gradient: Winners and losers. Journal of Insect Conse rvation, 16 (3), 331 343. doi: 10.1007/s10841 011 9419 2 Bates, A. J., Sadler, J. P., Fairbrass, A. J., Falk, S. J., Hale, J. D., & M atthews, T. J. (2011). Changing bee and hoverfly pollinator assemblages along an urban rural gradient. PLoS ONE, 6 (8), e23459. doi: 10.1371/journal.pone.0023459
64 Baum, K. A., & Wall en, K. E. (2011) Potential bias in pan trapping as a function of floral abun dance. Journal of the Kansas Entomological Society, 84 (2): 155 159. doi:10.2317/ JKES100629.1 Biesmeijer, J. C., Roberts, S. P. M., Reemer, M., Ohlemuller, R., Edwards, M., Peeters , T., Schaffers, A. P., Potts, S. G., Kleukers, R., Thomas, C. D., Settele, J., & Kunin, W. E. (2006) Parallel declines in pollinators and insect pollinated plants in Britain and the Netherlands. Science, 313 (5785), 351 354. doi: 10.1126/science.1127863 B ommarco, R., Lundin, O., Smith, H. G., & Rundlof, M. (2011). Drastic histori c shifts in bumble bee community composition in Sweden. Proceedings: Biological Sciences, 279 (1727), 309 15. Bouseman, J. K., & LaBerge, W. E. (1978). A revision of the bees of the genus Andrena of the western hemisphere. Part IX. Subgenus Melandrena . Transactions of the American Entomological Scoiety, 104 , 275 389. Brooks, R.W. (1988). Systematics and phylogeny of the Anthophorine bees (Hymenoptera: Anthophoridae; Anthophor ini). The University of Kansas Science Bulletin, 53 (9), 436 575. Butler, C. G. (1951) The importance of perfume in the discovery of food by the worker honeybee (Apis mellifera L.). Proceedings of the Royal Society of London. Series B, Biological Science s, 138 (892), 403 413. Cameron, S. A., Lozier, J. D., Strange, J. P., Koch, J . B., Cordes, N., Solter, L. F., Griswold, T. L., & Robinson, G. E. (2011). Patterns of widespread decline in North American bumble bees. Proceedings of the National Academy of Sci ences of the United States of America, 108 (2), 662 667. doi: 10.1073/pnas.10 14743108 Cane, J. H. (1996). Lifetime monetary value of individual pollinators: the bee Habropoda laboriosa at rabbiteye blueberry ( Vaccinium ashei Reade), pp. 67 70. In D. E. Yarb orough & J. M. Smagula (eds.), 6th International Symposium on Vaccinium Cult ure, 12 17 August 1996, Orono, ME. Cane, J. H. (2005). Bees, pollination, and the challenges of sprawl. In E. Johnson & M. Klemens (eds.), Nature in Fragments: The Legacy of Sprawl . New York: Columbia Univ. Press. Cane J. H., & Payne, J. A. (1993) Regiona l, annual, and seasonal variation in pollinator guilds: Intrinsic traits of bees (Hymenoptera: Apoidea) underlie their patterns of abundance at Vaccinium ashei (Ericaceae). Annals of the Entomological Society of America, 86 (5), 577 588. doi: 10.1093/aesa/ 86.5.577 Cane, J. H., Minckley, R. L., & Kervin, L. J. (2000) Sampling Bees (Hymenoptera: Apiformes) for pollinator community studies: Pitfalls of pan trapping. Journal of t he Kansas Entomological Society, 73 (4), 225 231.
65 Cane, J. H., Minckley, R. L., Ker vin, L. J., Rouston, T. H., & Williams, N. M. (2006) Complex responses within a desert bee guild (Hymenoptera: Apiformes) to urban habitat fragmentation. Ecological Applicat ions, 16 (2), 632 644. doi: 10.1890/1051 0761(2006)016[0632:CRWADB]2.0CO;2 Carper, A . L., Adler, L. S., Warren, P. S., & Irwin, R. E. (2014) Effects of Suburbanization on Forest Bee Communities. Environmental Entomology, 43 (2), 253 262. doi: 10.1603/EN13078 Carter, H. (2015). Urban life can be a danger to bees: Bees living in the city ha ve a lower immune system and are three times more likely to contract deadly diseases than rural bees. The Telegraph Online . Accessed June 14, 2018 from http://go.galegroup.com.aurarialibrary.idm.oclc.org/ps/i.do?p=WHIC&u=auraria_main&i d=GALE%7CA433857922&v=2.1&it=r&sid=summon . Carv ell, C., Roy, D. B., Smart, S. M., Pywell, R. F., Preston, C. D., & Goulson, D. (20 06) Declines in forage availability for bumblebees at a national scale. Biological Conservation, 132 (4), 481 489. doi: 10.1016/j.biocon.2006.05.008 Carvell, C., Meek, W. R., Pywell, R. F., Goulson, D., & Nowakowski, M. (2007). Comparing the efficacy of ag ri environment schemes to enhance bumble bee abundance and diversity on arable field margins. Journal of Applied Ecology, 44 (1), 29 40. Carvell, C. Osborne, J. L., Bourke , A. F. G. Freeman, S. N., Pywell, R. F., & Heard, M. S. (2011). Bumble bee specie landscape context and habitat quality. Ecological Applications, 21 (5), 1760 1771. doi: 10.1890/10 0677.1 Chateil, C., & Porcher, E. (2015). Landscape features are a better correlate of wild plant pollination than agricultural practices in an intensive cropping system. Agriculture, Ecosystems & Environment, 201 (1), 51 57. doi: 10.1016/j.agee.2014.12.008 Choate, B. A. , Hickman, P. L, & Moretti, E. A. (2018). Wild bee species abundance and richness across an urban rural gradient. Journal of Insect Conservation , 1 13. doi: 10.1007/s10841 018 0068 6 Cockerell, T. D. A. (1903). Bees of the genus Nomada from California. Proceedings of the Academy of Natural Sciences of Philadelphia , 559 614. Cockere ll, T. D. A., & Sumner, R. (1931). Rocky Mountain Bees III: The Genus Hylaeus (Prosopis). The American Museum of Natural History . American Museum Novitates Number 490. Cole, L. J., Brocklehurst, S., Robertson, D., Harrison, W., & McCracken, D. I. (20 15). Riparian buffer strips: Their role in the conservation of insect pollinators in intensive grassland systems. Agriculture, Ecosystems & Environment, 211 , 207 220. doi : 10.1016/j.agee.2015.06.012
66 Cole, L. J., Brocklehurst, S., Robertson, D., Harrison , W., & McCracken, D. I. (2017). Exploring the interactions between resource availability and the utilization of semi natural habitats by insect pollinators in an intensive agricultural landscape. Agriculture, Ecosystems & Environment, 246 , 157 167. doi: 10.1016/j.agee.2017.05.007 Cusser, S., Neff, J. L., & Jha, S. (2016). Natural land cover drives pollinator abundance and richness, leading to reductions in pollen limitati on in cotton agroecosystems. Agriculture, Ecosystems & Environment, 226 , 33 42. do i: 10.1016/j.agee.2016.04.020 David, A., Botias, C., Abdul Sada, A., Nicholls, E., Rotheray, E. L., Hill, E. M., & Goulson, D. (2016). Widespread contamination of wildflowe r and bee collected pollen with complex mixtures of neonicotinoids and fungicides c ommonly applied to crops. Environment International, 88 , 169 178. doi: 10.1016/j.envint.2015.12.011 Davis, A. Y., Lonsdorf, E. V., Shierk, C. R., Matteson, K. C., Taylor, J . R., Lovell, S. T., & Minor, E. S. (2017). Enhancing pollination supply in an urb an ecosystem through landscape modifications. Landscape and Urban Planning, 162 , 157 166. doi: 10.1016/j.landurbplan.2017.02.011 Decker, K., & Colorado Natural Heritage Program (2007). Rocky mountain lower montane foothill shrubland ecological system: E cological integrity assessment . Ft. Collins, CO: Colorado State University, Colorado Nautral Heritage Program. Decourtye, A., Ma honey bees in agro ecosystems. Apidologie 41(3), 264 277. doi: 10.1051/apido/2010024 Deguines, N., Julliard, R., de Flores, M., & Fontaine, C. (2016). Functional homogenization of flower visitor communities with urbanization. Ecology and Evolution, 6 (7), 1967 1976. doi: 10.1002/ece3.2009 de M. Santos, G. M., de C arvalho, C. A. L., Aguiar, C. M. L., Macedo, L. S. S. R., & Mello, M. A. R. (2013). Overlap in trophic and temporal ni ches in the flower visiting bee guild (Hymenoptera, Apoidea) of a tropical dry forest. Apidologie, 44 (1), 64 74. doi: 10.1007/s13592 012 0155 8 Draper, N., & Smith, H. (1981). Applied Regression Analysis (2 nd ed.). New York: John Wiley & Sons, Inc. E froymson, M. A. (1960). Multiple regression Analysis. In A. Ralston and H. S. Wilf (eds.) Mathematical Methods for Digital Computers . Wiley, New York Evison, S. E. F., Roberts, K. E., Laurenson, L., Pietravalle, S., Hui, J., Biesmeijer, J. C., Smith, J. E., Budge, G., & Hughes, W. O. H. (2012). Pervasiveness of parasites in pollinators. PLoS One, 7 (1), e30641. doi: 10.1371/journal.pone.0030641 Faith, D. P., Minchin, P. R., & Belbin, L. (1987). Compositional dissimilarity as a robust measure of ecologic al distance. Vegetatio 69 (1/3). 57 68. doi: 10.1007/BF00038687
67 Feltham, H., Park, K., & Goulson, D. (2014). Field realistic doses of pesticide imidacloprid reduce bumblebee pollen foraging efficiency. Ecotoxicology, 23 (3), 317 323. doi: 10.1007/s10646 01 4 1189 7 Fetridge, E. D., Ascher, J. S., & Langellotto, G. A. (2008). The Bee fauna of residential gardens in a suburb of New York City (Hymenoptera: Apoidea). Annals of the Entomological Society of America, 101 (6), 1067 1077. doi: 10.1603/0013 8746 101 .6.1067 Fischer, J., Muller, T., Spatz, A., Greggers, U., Grunewald, B., & Menzel, R. (2014). Neonicotinoids interfere with specific components of navigation in honeybees. PLoS One, 9 (3), e91364. doi: 10.137/journal.pone.0091364 Fischer, L. K., Eichfeld, J., Kowarik, I., & Buchholz, S. (2016). Disentangling urban habitat and matrix effects on wild bee species. PeerJ, 4 , e2729. doi: 10.7717/peerj.2729 Food and Agriculture Organization of the United Nations (FAO), 2009. FAOSTAT. Fortel, L., Henry, M., Gui lbaud, L., Guirao, A. L., Kuhlmann, M., Mouret, H., Rollin, O., & Vaissiere, B. E. (2014). Decreasing abundance, increasing diversity and changing structure of the wild bee community (Hymenoptera: Anthophila) along an urbanization gradient. PLoS One, 9 ( 8), e104679. doi: 10.1371/journal.pone.0104679 Frankie, G. W., Thorp, R. W., Schindler, M. H., Ertter, B., & Przybylski, M. (2002). Bees in Berkeley? Fremontia, 30 , 50 58. Frankie, G. W., Thorp, R. W., Schindler, M., Hernandez, J., Ertter, B., & Rizza r di, M. (2005). Ecological patterns of bees and their host ornamental flowers in two northern California cities. Journal of the Kansas Entomological Society, 78 (3), 227 246. doi: 10.2317/0407.08.1 FÃ¼rst, M. A., McMahon, D. P., Osborne, J. L., Paxton, R. J ., & Brown, M. J. F. (2014). Disease associations between honeybees and bumblebees as a threat to wild pollinators. Nature, 506 (7488), 364 366. doi: 10.1038/nature12977 Garbuzov, M., SchÃ¼rch, R., & Ratnieks, F. L. W. (2015). Eating locally: Dance decodi n g demonstrates that urban honey bees in Brighton, UK, forage mainly in the surrounding urban area. Urban Ecosystems, 18 (2), 411 418. doi: 10.1007/s11252 014 0403 y Gardner, K. E., & Ascher, J. S. (2006). Notes on the native bee pollinators in New York app l e orchards. Journal of the New York Entomological Society, 114 (1/2), 86 91. doi: 10.1664/0028 7199(2006)114[86:NOTNBP]2.0.CO;2 Gathmann, A., & Tscharntke, T. (2002). Foraging ranges of solitary bees. Journal of Animal Ecology, 71 (5), 757 764. doi: 10.10 4 6/j.1365 2656.2002.00641.x
68 Genersch, E., Yue, C., Fries, I., & de Miranda, J. R. (2006). Detection of deformed wing virus, a honey bee viral pathogen, in bumble bees ( Bombus terrestris and Bombus pascorum ) with wing deformities. Journal of Invertebrate Pathology, 91 (1), 61 63. doi: 10.1016/j.jip.2005.10.002 Geroff, R. K., Gibbs, J., & McCravy, K. W. (2014). Assessing bee (Hymenoptera: Apoidea) diversity of an Illinois restored tallgrass prairie: Methodology and conservation considerations. Journal of In sect Conservation, 18 (5), 951 964. doi: 10.1007/s10841 0 14 9703 z. Geslin, B., Gauzens, B., Thebault, E., & Dajoz, I. (2013). Plant pollinator networks along a gradient of urbanisation. PLoS One, 8 (5), e63421. doi: 10.1371/journal.pone.0063421 Geslin, B. , Feon, V. L., Folschweiller, M., Flacher, F., Carmignac , D., Motard, E., Perret, S., & Dajoz, I. (2016). The proportion of impervious surfaces at the landscape scale structures wild bee assemblages in a densely populated region. Ecology and Evolution, 6 (18), 6599 6515. doi: 10.1002/ece3.2374 Ghazoul, J., & K oh, L. P. (2010). Food security not (yet) threatened by declining pollination. Frontiers in Ecology and the Environment, 8 (1), 9 10. doi: 10.1890/10.WB.001 Giannini, T. C., Cordeiro, G. D., Freitas, B. M., Saraiva, A. M., & Imperatriz Fonseca, V. L. (201 5). The dependence of crops for pollinators and the economic value of pollination in Brazil. Journal of Economic Entomology, 108 (3), 849 857. doi: 10.1093/jee/tov093 Giannini, T. C., Costa, W. F., C ordeiro, G. D., Imperatriz Fonseca, L., Saraiva, A. M., Biesmeijer, J., & Garibaldi, A. J. (2017). Projected climate change threatens pollinators and crop production in Brazil. PLoS One, 12 (8), e0182274. doi: 10.1371/journal.pone.0182274 Gibbs, J. (2011) . Revision of the metallic Lasioglossum ( Dialictus ) of eastern North America (Hymenoptera: Halictidae: Halictini). Zootaxa, 3073 , 1 216. Gill R. J., Ramos Rodriguez, O., & Raine, N. E. (2012). Combined pesticide exposure severely affects individual and colony level traits in bees. Nature, 491 (7422), 10 5 108. doi: 10.1038/nature11585 Gill, R .J., & Raine, N. E. (2014). Chronic impairment of bumblebee natural foraging behavior induced by sublethal pesticide exposure. Functional Ecology, 28 (6), 145 9 1471. doi: 10.1111/1365 2435.12292 Gill, R. J., Baldoc k, K. C. R., Brown, M. J. F., Cresswell, J. E., Dicks, L. V., Fountain, M. T., Garratt, M. P. D., Gough, L. A., Heard, M. S., Holland, J. M., Ollerton, J., Stone, G. N., Tang, C. Q., Vanbergen, A. J., Vogler, A. P., Woodward, G., Arce, A. N., Boatman, N. D ., Brand Osborne, J. L., Phillips, J., Sutton, P. B., & Potts, S. G. (2016). Protecting an ecosystem service: Approaches to unde rstanding and mitigating threats to wild insect pollinat ors. Advances in Ecological Research, 54 , 135 206. doi: 10.1016/bs.aecr.2015.10.007
69 Gotlieb, A., Hollender, Y., & Mandelik, Y. (2011). Gardening in the desert changes bee communities and pollination n etwork characteristics. Basic and Applied Ecology, 12 (4) , 310 320. doi: 10.1016/j.baae.2010.12.003 Goulson, D., Lye, G. C., & Darvill, B. (2008) Decline and conservation of bumble bees. Annual Review of Entomology, 53 (1) , 191 208. doi: 10.1146/annurev.ento .53.103106.093454 Grant, V. (1950) The flower constancy of bees. Botanical Review , 16 (7), 379 398. Greenleaf, S. S., Williams, N. M., Winfree, R., & Kremen, C. (2007). Bee foraging ranges and their relationship to body size. Oecologia, 153 (3), 589 596. d oi: 10.1007/s00442 007 0752 9 Grimm, N. B., Faeth, S. H., Golubiewski, N. E., Redman, C. L., Wu, J. , Bai, X., & Briggs, J. M. (2008). Global Change and the Ecology of Cities. American Association for the Advancement of Science, 319 (5864), 756 760. doi: 1 0.1126/science.1150195 Grunewald, B. (2010). Is Pollination at Risk? Current Threats to and Conse rvation of Bees. GAIA Ecological Perspectives for Science and Society, 19 (1), 61 67. doi: 10.14512/gaia.19.1.13 Hall, D. M., Camilo, G. R., Tonietto, R. K., Ollerton, J., Ahrne, K., Arduser, M., Ascher, J. S., Baldock, K. C. R., Fowler, R., Frankie, G ., Goulson, D., Gunnarsson, B., Hanley, M. E., Jackson, J. I., Langellotto, G., Lowenstein, D., Minor, E. S., Philpott, S. M., Potts, S. G., Sirohi, M. H., S pevak, E. M., Stone, G. N., & Threlfall, C. G. (2016). The city as a refuge for insect pollinators . Conservation Biology, 31 (1), 24 29. doi: 10.1111/cobi.12840 Hanes, S. P., Collum, K. K., Hoshide, A. K., & Asare, E. (2015). Grower perceptions of nativ e pollinators and pollination strategies in the lowbush blueberry industry. Renewable Agriculture and Food Systems, 30 (2), 124 131. doi: 10.1017/S1742170513000331 Hendrix, S. D. (1994). Effects of population size on fertilization, seed production, and se ed predation in two prairie legumes. North American Prairie Conference Proceedings, 13 , 115 119. Henni n g, E. I., & Ghazoul, J. (2011). Pollinating animals in the urban environment. Urban Ecosystems, 15 (1),149 166. doi: 10.1007/s11252 011 0202 7 Henry, M ., Beguin, M., Reguier, F., Rollin, O., Odoux, J., Aupinel, P., Aptel, J., Tchamitchian, S., & Deco urtye, A. (2012). A common pesticide decreases foraging success and survival in honey bees. Science, 336 (6079), 348 350. doi: 10.1126/science.1215039 Hernan dez, J. L., Frankie, G. W., & Thorp, R. W. (2009). Ecology of urban bees: A review of current knowl edge and directions for future study. Cities and the Environment, 2 (1), Article 3 15 pp.
70 Hinners, S. J., & Hjelmroos Koski, M. K. (2009). Receptiveness of Foraging Wild Bees to Exotic Landscape Elements. The American Midland Naturalist, 162 (2), 253 265. doi: 10.1674/0003 0031 162.2.253 Hinners, S. J., Keaans, C. A., & Wessman, C. A. (2012). Roles of scale, matrix, and native habitat in supporting a diverse suburban pollinator assemblage. Ecological Applications, 22 (7), 1923 1935. doi: 10.1890/11 1590.1 Hipolito, J., Viana, B. F., & Garibaldi, L. A. (2016). The value of pollinator friendly practices: Synergies between natural and anthropogenic assets. Ba sic and Applied Ecology, 17 (8), 659 667. doi: 10.1016/j.baae.2016.09.003 Hocking, R. R. (1976). Th e analysis and selection of variables in linear regression. Biometrics, 32 (1), 1 49. d oi: 10.2307/25 29336 Homer, C. G., Dewitz, J. A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N. D., Wickham, J. D., & Megown, K. (2015). Completion of the 2011 National Land Cover Database for the conterminous United States Representing a decade of land cover change information . Photogrammetric Engineering and Remote Sensing, 81 (5), 345 354. Imbach, P., Fung, E., Hannah, L., Navarro Racines, C. E., Roubik, D. W., Ricketts, T. H., Harvey, C. A., Donatti, C. I., Laderach, P., Locatelli, B., & Roehrdanz, P . R. (2017). Coupling of pollination services and coffee suitability under climate change. Proceedings of the National Academy of Sciences, 114 (39), 10438 10442. doi: 10.1073/pnas.1617940114 Jedrzejewska Szmek, K., & Marcin, Z. (2013). Flower visitor an d pollen transport networks in a large city: Structure and properties. Arthropod Plant Int eractions 7 (5), 503 516. doi: 10.1007/s11829 013 9274 z Jha, S. (2015). Contemporary human altered landscapes and oceanic barriers reduce bumble bee gene flow. Mo lecular Ecology, 24 (5), 993 1006. doi: 10.1111/mec.13090 Jha, S., & Kremen, C. (2013). Urban land use limits regional bumble bee gene flow. Molecular Ecology, 22 (9), 2483 2495. doi: 10.1111/mec.12275 Kaluza, B. F., Wallace, H., Heard, T. A., Klein, A. M., & Leonhardt, S. D. (2016). Urban gardens promote bee foraging over natural habitats and pla ntations. Ecology and Evolution, 6 (5), 1304 1316. doi: 10.1002/ece3.1941 Kasina, J. M., Mburu, J., Kraemer, M., & Holm Mueller, K. (2009). Economic benefit of cro p pollination by bees: A case of Kakamega small holder farming in Western Kenya. Journal o f Economic Entomology, 102 (2), 467 473. doi: 10.1603/029.102.0201 Kearns, C. A., & Oliveras, D. M. (2009) Environmental factors affecting bee diversity in urban and remote grassland plots in Boulder, Colorado. Journal of Insect Conservation, 13 (6), 655 665 . doi: 10.1007/s10841 009 9215 4
71 Keil, P., Biesmeijer, J. C., Barendregt, A., Reemer, M., & Kunin, W. E. (2011). Biodiversity change is scale dependent: An example from Dutch and UK hoverflies (Diptera, Syrphidae). Ecography, 34 (3), 392 401. doi: 10.1111/j .1600 0587.2010.06554.x Kehinde, T., von Wehrden, H., Samways, M., Klein, A., & Brittain, C. (2018). Organic farming promotes bee abundance in vineyards in Italy b ut not in South Africa. Journal of Insect Conservation, 22 (1), 61 67. doi: 10.1007/s10841 0 17 0038 4 Kleijn, D., & Raemakers, I. (2008). A retrospective analysis of pollen host plant use by stable and declining bumblebee species. Ecology 89 (7), 1811 1823. doi: 10.1890/07 1275.1 Klein, A. M., VaissiÃ¨re, B. E., Cane, J. H., Steffan Dewenter, I., Cunningham, S. A., Kremen, C., & Tscharntke, T. (2007). Importance of pollinators in changing landscapes for world crops. Proceedings of the Royal Society B: Biologi cal Sciences, 274 (1608), 303 313. doi: 10.1098/rspb.2006.3721 KovÃ¡cs HostyÃ¡nszki, A., FÃ¶ldesi R., MÃ³zes, E., SzirÃ¡k, A., Fischer, J., Hanspach, J., & BÃ¡ldi, A. (2016). Conservation of pollinators in traditional agricultural landscapes New challenges in Transylvania (Romani) posed by EU accession and recommendations for f uture research. PLoS One, 11 (6), e0151650. doi: 10.131/journal.pone.0151650 Ksiazek, K., Fant, J., and Skogen, K. (2012). An assessment of pollen limitation on Chicago green roofs. Lan dscape and Urban Planning 107 (4), 401 408. doi: 10.1016/j.landurbplan .2012.07.008 LaBerge, W. E., & Ribble, D. W. (1975). A revision of the bees of the genus Andrena of the western hemisphere. Part VII. Subgenus Euandrena. Transactions of the American Entomological Society, 101 (3), 371 446. Lagucki, E., Burdine, J. D. , & McCluney, K. E. (2017). Urbanization alters communities of flying arthropods in parks and gardens of a medium sized city. PeerJ, 5 , e3620. doi: 10.7717/peerj.3620 Larsen, T. H., Wil liams, N. M., & Kremen, C. (2005) Extinction order and altered commun ity structure rapidly disrupt ecosystem functioning. Ecology Letters 8 (5), 538 547. doi: 10.1111/j.1461 0248.2005.00749.x Larson, J. L., Kesheimer, A. J., & Potter, D. A. (2014). Pollina tor assemblages on dandelions and white clover in urban and suburban lawns. Journal of Insect Conservation, 18 (5), 863 873. doi: 10.1007/s10841 014 9694 9 Lautenbach, S., Seppelt, R., Liebscher, J., & Dormann, C. F. (2012). Spatial and Temporal Trends of Global Pollination Benefit. PLoS One, 7 (4), e35954. doi: 10.1371/jo urnal.pone.0035954 Leong, M., Kremen, C., & Roderick, G. K. (2014). Pollinator interactions with yellow starthistle ( Centaurea solstitialis ) across urban, agricultural, and natural lands capes. PLoS One, 9 (1), e86357. doi: 10.1371/journal.pone.0086357
72 Los ey, J. E., & Vaughan, M. (2006). The economic value of ecological services provided by insects. Bioscience, 56 (4), 311 323. doi: 10.1641/0006 3568(2006)56[311:TEVOES]2.0.CO;2 Lowenstein, D. M., Matteson, K. C., Xiao, I., Silva, A. M., & Minor, E. S. (2014) . Humans, bees, and pollination services in the city: The case of Chicago, IL (USA). Biodiversity Conservation, 23 (11), 2857 2874. doi: 10.1007/s10531 014 0752 0 Lowenstein, D. M., Mat teson, K. C., & Minor, E. S. (2015). Diversity of wild bees supports pollination services in an urbanized landscape. Oecologia, 179 (3), 811 821. doi: 10.1007/s00442 015 3389 0 MacArthur, R., & Wilson, O. E. (1967). The theory of island biogeography. Prin ceton: Princeton University Press. Mackenzie, K. E., & Averill, A. L . (1995). Bee (Hymenoptera: Apoidea) diversity and abundance on cranberry in southeastern Massachusetts. Annals of the Entomological Society of America 88 (3): 334 341. doi: 10.1093/aesa/8 8.3.334 Majewski, J. (2016). The importance of insect pollinators fo r Polish food security. Journal of Agribusiness and Rural Development, 3 (41), 329 336. Mandelik, Y., Winfre, R., Neeson, T., & Kremen, C. (2012). Complementary habitat use by wild bees in agro natural landscapes. Ecological Applications, 22 (5), 1535 15 46. doi: 10.1890/1051 0761 22.5.1535 Marini, L., Quaranta, M., Fontana, P., Biesmeijer, J. C., & Bommarco, R. (2012). Landscape context and elevation affect pollinator communities in int ensive apple orchards. Basic and Applied Ecology, 13 (8), 681 689. doi : 10.1016/j.baae.2012.09.003 Marques, M. F., Hautequestt, A. P., Oliveira, U. B., de Freitas ManhÃ£es Tavares, V., Perkles, O. R., Antunes, C., & Gaglianone, M. C. (2017). Local knowledge on native bees and their role as pollinators in agricultural communities. Journal of Insect Cons ervation, 21 (2), 345 356. doi: 10.1007/s10841 017 9981 3 Martins, K. T., Gonzalez, A., & Lechowicz, M. J. (2017). Patterns of pollinator turnover and increas ing diversity associated with urban habitats. Urban Ecosystems, 20 (6), 1359 1371. doi: 10.1007/s1 1252 017 0688 8 Matteson, K. C., Ascher, J. S., & Langellotto, G. A. (2008). Bee Richness and Abundance in New York City Urban Gardens. Annals of the Entomolo gical Society of America, 101 (1), 140 150. doi: 10.1603/0013 8746(2008)101[140:BRAAIN]2.0CO;2 Matt eson, K. C., & Langellotto, G. A. (2009). Bumble bee abundance in New York City community gardens: Implications for urban agriculture. Cities and the Envir onment, 2 (1), 5 12.
73 Matteson, K. C., & Langellotto, G. A. (20 10 ) . Determinates of inner city butterfly and bee species richness. Urban Ecosystems, 13 (3), 333 347. doi: 10.1007/s11252 010 0122 y Matteson, K. C., Grace, J. B., & Minor, E. S. (2013). Direc t and indirect effects of land use on floral resources and flower v isiting insects across an urban landscape. Oikos, 122 (5), 682 694. doi: 10.1111/j.1600 0706.2012.20229.x McCabe, E., Loeb, G., & Grab, H. (2017). Responses of Crop Pests and Natural Enemi es to Wildflower Borders Depends on Functional Group. Insects, 8 (3 ), 73. doi: 10.3390/insects8030073 McFrederick, Q. S., & LeBuhn, G. (2006). Are urban parks refuges for bumble bees Bombus spp. (Hymenoptera: Apidae)? Biological Conservation, 129 (3), 372 382. doi: 10.1016/j.biocon.2005.11.004 McGinley, R. J. (1986). St udies of Halictinae (Apoidea: Halictidae), I: Revision of New World Lasioglossum Curtis. Smithsonian Contributions to Zoology no. 429 . Smithsonian Institution Press, Washington D.C., US A. McGinley, R. J. (2003). Studies of Halictinae (Apoidea: Hali ctidae), II: Revision of Sphecodogastra Ashmead, Floral Specialists of Onagraceae. Smithsonian Contributions to Zoology no. 610 . Smithsonian Institution Press, Washington D.C., USA. Mc Gregor, S. E. (1976). Insect Pollination of Cultivated Crop Plants . Washington DC: U.S. Department of Agriculture, Agricultural Handbook No. 496. McIntyre, N. E.; & Hostetler, M. E. (2001). Effects of urban land use on pollinator (Hymenoptera: Apoid ea) communities in a desert metropolis. Basic and Applied Ecology, 2 (3), 209 218. doi: 10.1078/1439 1791 00051 McKinney, M. L. (2006) Urbanization as a major cause of biotic homogenization. Biological Conservation, 127 (3), 247 260. doi:10.1016/j.biocon.20 05.09.005 McKinney, M. L. (2008). Effects of urbanization on speci es richness: A review of plants and animals. Urban Ecosystems, 11 (2), 161 176. doi: 10.1007/s11252 007 0045 4 Meerman, J. (2004). Rapid ecological assessment Columbia River Forest Reserv e past Hurricane Iris. Accessed October 31, 2018 from http://biological diversity.info/Downloads/CRFR_REA_report.pdf . Meeus, I., Brown, M. J. F., DeGraaf, D. C., & Smagghe, G. (2011). Effects of invasive parasites on bumble bee declines. Conservation Biology, 25 (4), 662 667. doi: 10.1111/j.1523 1739.2011.01707.x Michener, C. D. (1979). Biogeography of the bees. Annals of the Missouri Botanical Garden, 66 (3), 277 347. doi:10. 2307/2398 833
74 Michener, C. D. (2000). The Bees of the World. Baltimore, MD: Johns Hopkins University Press. Multi Resolution Land Characteristics Consortium (2017). Morandin, L. A., Laverty, T. M., & Kevan, P. G. (2001). Effect of bumble bee (Hymenoptera: Apidae) p ollination intensity on the quality of greenhouse tomatoes. Journal of Economic Entomology, 94 (1), 172 179. doi: 10.1603/0022 0493(2001)094[0172:EOBBHA]2.0.CO;2 Morandin, L. A., & Kremen, C. (2013). Hedgerow restoration promotes pollinator populat ions and Ecological Applications, 23 (4), 829 839. doi: 10.1890/12 1051.1 Morse, R. A., & Calderone, N. W. (2000). The value of honey bees as pollinators of U.S. crops in 2000. Bee Culture, 128 , 1 15. Munyuli, M. B., Nyeko, P., Potts, S., Atkinson, P., Pomeroy, D., & Vickery, J. (2013). Patterns of bee diversity in mosaic agricultural landscapes of central Uganda: Implication of pollination services conservation for food security. Journal of Insect Conservation, 17 (1), 79 93. doi: 10.1007/s10841 012 9488 x National Oceanic and Atmospheric Administration (n.d.). Accessed May 15, 2018 from https://www.esrl.noaa.gov/psd/boulder/dailyrecords/ Natural Research Council (2006). Status of Pollinators in North America. National Ac ademic Press. Novais, S. M. A., Nunes, C. A., Santos, N. B., Amico, A. R. D., Fernandes, G. W., Quesada, M., Braga, R. F., & Neves, A. C. O. (2016). Effects of a Possible Pollinator Crisis on Food Crop Production in Brazil. PLoS One, 11 (11), e0167292. do i: 10.1371/journal.pone.0167292 Oksanen, J., Blanchet, F. G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, R. R., , Stevens, M. H. H., Szoecs, E., & Wagner, H. (2018). Vegan: Community Ecology Pac kage. R package version 2.5 2. https://CRAN.R project.org/package=vegan . Packer, L. (1993). Lasioglossum (Di alictus) tenax (Sandhouse) (Hymenoptera: Halictidae) as a solitary sweat bee. Inse ctes Sociaux, 41 (3), 309 313. doi: 10.1007/BF01242302 Park, M. G., Orr, M. C., & Danforth, B. N. (2010). The role of native bees in apple pollination. New York Fruit Quarte rly, 18 (1), 21 25. Parmentier, L., Meeus, I., Cheroutre, L., Mommaerts, V., Louwye, S., & Smagghe, G. (2014). Commercial bumblebee hives to assess an anthropogenic environment for pollinator support: A case study in the region of Ghent (Belgium). Enviro nmental Monitoring and Assessment, 186 (4), 2357 2367. doi: 10.1007/s10661 013 3543 2
75 Pauw, A., & Hawkins, J. A. (2011). Reconstruction of historical pollination rates reveals linked declines of pollinators and plants. Oikos, 120 (3), 344 349. doi: 10.1111 /j.1600 0706.2010.19039.x Pereira Peixoto, M. H., Pufal, G., Martins, C. F., & Klein , A. M. (2014). Spillover of trap nesting bees and wasps in an urban rural interface. Journal of Insect Conservation, 18 (5), 815 826. doi: 10.1007/s10841 014 9688 7 Peters, K. A. (2012). Keeping bees in the city? Disappearing bees and the explosion of u rban agriculture inspire urbanities to keep honeybees: Why city leaders should care and what they should do about it. Drake Journal of Agricultural Law, 17 (3), 597 644. Pettis, J. S., van Engelsdor p, D., Johnson, J., & Dively, J. G. (2012). Pesticide expo sure in honey bees results in increased levels of the gut pathogen Nosema. Naturwissenschaften, 99 (2), 153 158. doi: 10.1007/s00114 011 0881 1 nators: life history traits predict habitat use and crop v isitation by Mediterranean wild bees. Ecological Applications, 25 (3), 742 752. doi: 10.1890/14 0910.1 Potter, A., & LeBuhn, G. (2015). Pollination service to urban agriculture in San Francisco, CA. Urban Ecosystems 18 (3), 885 893. doi: 10.1007/s11252 015 0435 y flowers: How do floral communities structure pollinator communities? Ecology, 84 (10), 2628 2642. do i: 10.1890/02 0136 Potts, S. G., Roberts, S. P. M., Dean, R., Marris, G., Brown, M. A., Jones, R., Neumann, P., & Settele, J. (2010)a. Declines of managed honeybees and beekeepers in Europe? Journal of Apicultural Research, 49 (1), 15 22. doi: 10.3896/IBRA. 1.49.1.02 Potts, S. G., Biesmeijer, J. C., Kremen, C., Neu mann, P., Schweiger, O., & Kunin, W. E. (2010)b. Global pollinator declines: Trends, impacts and drivers. Trends in Ecology and Evolution, 25 (6), 345 353. doi: 10.1016/j.tree.2010.01.007 Potts, S . G., Imperatriz Fonseca, V., Ngo, H. T., Aizen, M. A., Bi esmeijer, J. C., Breeze, T. D., Dicks, L. V., Garibaldi, L. A., Hill, R., Settele, J., & Vanbergen A. J. (2016). Safeguarding pollinators and their values to human well being. Nature, 540 (7632), 2 20 229. doi: 10.1038/nature20588 Power, E. F., Kelly, D. L ., & Stout, J. C. (2012). Organic farming and landscape structure: Effects on insect pollinated plant diversity in intensively managed grasslands. PLoS One 7 (5), e38073. doi: 10.1371/journal.pone .0038073 Priess, J. A., Mimler, M., Klein, A. M., Schwarze , S., Tscharntke, T., & Steffan Dewenter, I. (2007). Linking deforestation scenarios to pollination services and economic returns in coffee agroforestry systems. Ecological Applications, 17 (2), 40 7 417. doi: 10.1890/05 1795
76 Proctor M., Yeo P., Lack, A. ( 1996). The Natural History of Pollination . Timber Press, Inc., Portland, OR. Pywell, R. F., Meek, W. R., Hulmes, L., Hulmes, S., James, K. L., Nowakowski, M., & Carvell, C. (2011). Management to en hance pollen and nectar resources for bumblebees and butte rflies within intensively farmed landscapes. Journal of Insect Conservation, 15 (6), 853 864. doi: 10.1007/s10841 011 9383 x Quistberg, R. D., Bichier, P., & Philpott, S. M. (2016). Landscape and l ocal correlates of bee abundance and species richness in u rban gradients. Environmental Entomology, 45 (3), 592 601. doi: 10.1093/ee/nvw025 Ratnieks, F. L. W., & Carreck, N. L. (2010). Clarity on honey bee collapse? Science, 327 (5962), 152 153. doi: 10.1 126/science.1185563 R Core Team (2013). R: A language an d environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R project.org/. Reed, C. C. (1993). Reconstruction of pollinator communities on res tored prairies in eastern Minnesota. Final report to the M innesota Department of Natural Resources Nongame Wildlife Program. Reed, C. C. (1995). Species richness of insects on prairie flowers in Southeastern Minnesota. Pages 103 115 in Proceedings of the 14th Annual North American Prairie Conference . D.C. Hartn ett editor. Kansas State University Press. Manhattan, KS. Roberts , R. R. (1972). Revision of the bee genus Agapostemon (Hymenoptera: Halictidae). University of Kansas Science Bulletin, 49 , 437 590. Rollin, O., Bretagnolle, V., Fortel, L., Guilbaud, L ., & Henry, M. (2015). Habitat, spatial and temporal drivers of div ersity patterns in a wild bee assemblage. Biodiversity and Conservation, 24 (5), 1195 1214. doi: 10.1007/s10531 014 0852 x Romey, W. L., Ascher, J. S., Powell, D. A., & Yanek, M. (2007). Im pacts of logging on midsummer diversity of native bees (Apoidea) in a northern hardwood forest. Journal of the Kansas Entomological Society, 80 (4), 327 338. doi: 10.2317/0022 8567(2007)80[ 327:IOLOMD]2.0.CO;2 Roulston, T. H., Smith, S. A., & Brewster, A. L . (2007) A comparison of pan trap and intensive net sampling techniques for documenting a bee (Hymenoptera: Apiformes) fauna. Journal of the Kansas Entomological Society, 80 (2), 179 181. do i: 10.2317/0022 8567(2007)80[179:ACOPTA]2.0.CO;2 Roulston, T. A. H. , & Goodell, K. (2011). The role of resources and risks in regulating wild bee populations. Annual Review of Entomology, 56 (1), 293 312. doi: 10.1146/annurev ento 120709 144802
77 Rundlof, M., Andersson, G. K. S., Bommarco, R., Fries, I., Hederstrom, V., Herb ertsson, L., Jonsson, O., Klatt, B. K., Pedersen, T. R., Yourstone, J., & Smith, H. G. (2015). Seed coating with a neonicotinoid insecticide negatively affects wild bees. Nature, 521 (7550) , 77 80. doi: 10.1038/nature14420 SamnegÃ¥rd, U., Persson, A.S., & S mith, H.G. (2011). Gardens benefit bees and enhance pollination in intensively managed farmland. Biological Conservation, 144 (11), 2602 2606. doi: 10.1016/j.biocon.2011.07.008 Samways, M. J., McGeoch, M. A., & New, T. R. (2010). Insect conservation: A handbook of approaches and methods. Sutherland, W.J., editor. Oxford: Oxford University Press. Sanchez Bayo, F., & Goka, K. (2014). Pesticide residues and bees A risk assessment. PLoS ONE 9 (4), e94482. doi: 10.1371/journal.pone.0094482 SAS Institute Inc. (1989). (4 th ed., Vol. 2). Cary, NC: SAS Institute Inc. Saunders, M. E., & Luck, G. W. (2014). Spatial and temporal variation in pollinator communi ty structure relative to a woodland almond plantation edge. Agricul tural and Forest Entomology, 16 (4), 369 381. doi: 10.1111/afe.12067 Science for Environment Policy (2017) Agri environmental schemes: how to enhance the agriculture environment relationship . Thematic Issue 57. Issue produced for the European Commission DG Environment by the Science Communication Unit, UWE, Bristol. Available at: http://ec.europa.eu/science environmentpolicy . Scott, V. L., Ascher, J. S., Griswold, T., & Nufio, C. R. (2011). The Bees of Colorado (Hymenoptera: Apoidea: Anthophila) . Natural History Inventory of Colorado 23: i vi + 100pp. http://www.colorado.edu/ cumuseum/sites/default/files/attached files/the _bees_of_colorado.pdf . Seeley, T. D. (1985). Honeybee Ecology . Prince ton University Press, Princeton, NJ. Sims, P. L., Singh, J. S., & Lauenroth, W. K. (1978). The structure and function of ten western Nor th American grasslands. Journal of Ecology, 66 (3), 983 1009. doi: 10.2307/2259309 Southwick, E., & Southwick, L. (1992). Estimating the economic value of honey bees (Hymenoptera: Apidae) as agricultural pollinators in the United States. Journal of Econo mic Entomology, 85 (3), 621 633. doi: 10.1093/jee/85.3.621 Stanghellini, M. S., Ambrose, J. T., & Schultheis, J. R. (1998 ). Seed production in watermelon: A comparison between two commercially available pollinators. HortScience, 33 (1), 28 30. Stanley, D. A., Garratt, M. P. D., Wickens, J. B., Wickens, V. J., Potts, S. G., & Raine, N. E. (2015). Neonicotinoid pesticide exp osure impairs crop pollination services provided by bumblebees. Nature, 528 (7583), 548 556. doi: 10.1038/nature16167
78 Steffan Dewenter I ., Potts, S. G., & Packer, L. (2005) Pollinator diversity and crop pollination services are at risk. Trends in Ecology a nd Evolution, 20 (12), 651 652. doi: 10.1016/j.tree.2005.09.004 Steffan Dewenter, I., & Tscharntke, T., (1999). Effects of habitat isola tion on pollinator communities and seed set. Oecologia, 121 (3), 432 440. doi: 10.1007/s004420050949 Stubbs, C. S., & Drummond, F. A. (1996). Blueberry and cranberry ( Vaccinium spp) p ollination: A comparison of managed and native bee foraging behavior, pp. 341 344. In K.W. Richards (ed.), 7 th International Symposium on Pollination, Pollination From Theory to Practice . 23 28 June 1996, Lethbridge, Canada. Tepedino, V. J., Alton, D. G., Bradley, B. A., Toler, T. R., & Griswold, T. L. (2007). Orchard pollina tion in Capitol Reef National Park, Utah, USA. Honey bees or native bees? Biodiversity Conservation, 16 (11), 3083 3094. doi: 10.1007/s10531 007 9164 8 The Bee Inventory Plot (n.d.) . Accessed March 31, 2018 from http://onlinesfsu.edu/beeplot/ . Theodorou, P., Albig, K., Radzeviciute, R., Settele, J., Schweiger, O., Murray, T. E., & Paxton, R. J. (2017). The structure of flower visitor networks in relation to pollination across an agricultural to urban gra dient. Functional Ecology, 31 (4), 838 847. doi: 10.1111/1365 2435.12803 The Xerces Society. Red Lists. Accessed April 2, 2017 from http://www.xerces.org/red lists/ . Thomson, J. D., & Goodell, K. (2001). Pollen removal and deposition by honeybee and bumblebee visitors to apple and almond flowers. Journal of Applied Ecology, 38 (5), 1032 1044. Thompson, K ., Austin, K. C., Smith, R. M., Warren, P. H., Angold, P. G., & Gaston, K. J. (2003). Urban domestic ga rdens (I): Putting small scale plant diversity in context. Journal of Vegetation Science 14 (1), 71 78. doi: 10.1111/j.1654 1103.2003.tb2129.x Threlfall, C. G., Walker, K., Williams, N. S. G., Hahs, A. K., Mata, L., Stork, N., & Livesley, S. J. (2015). The conservation value of urban green space habitats for Australian native bee communities. Biological Conservation, 187 , 240 248. doi: 10.1016/ j.biocon.2015.05.003 Tonietto, R., Fant, J., Ascher, J., Ellis, K., & Larkin, D. (2011). A comparison of bee commu nities of Chicago green roofs, parks and prairies. Landscape and Urban Planning, 103 (1), 102 108. doi: 10.1016/j.landurbplan.2011.07.004 TornÃ© Noguera, A., Rodrigo, A., Arnan, X., Osorio, S., Barril Graells, H., da Rocha Filho, L. C., & Bosch, J. (2014). D eterminants of spatial distribution in a bee community: nesting PLoS ONE, 9 (5), e97255. doi: 10.137 1/journal.pone.0097255
79 Tscharntke, T., Klein, A. M., Kruess, A., Steffan Dewenter, I., & Thies, C. (2005). Landsc ape perspectives on agricultural intensification and biodiversity Ecosystem service management. Ecology Letters, 8 (8), 857 874. doi: 10.111 1/j.1461 0248.2005.00782.x United Nations (2014). Our urbanizing world. Populations facts. UN Department f Eco nomic and Social Affairs, New York, New York. Accessed May 24, 2018 from http://www.un.org/en/development/desa/population/publications/pdf/pop facts/PopFacts_ 2014 3.pdf . U.S. Census Bureau Population Division (2017). Annual Estimates of the Resident Population: April 1, 2010 to Jul y 1, 2017. Accessed May 24, 2018 from https://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid= PEP_2 017_PEPANNRES&prodType=table . Vanbergen, A., & Insect Pollinators Initiative. (2013). Threats to an ecosyst em service: Pressures on pollinators. Frontiers in Ecology and the Environment, 11 (5), 251 259. Accessed June 12, 2018 from http://www.jstor.o rg.aurarialibrary.idm.oclc.org/stable/23470505 v an Engelsdorp, D., Hays, J., Underwood, R. M., & Pettis, J. (2008) A survey of honey bee colony losses in the U.S., Fall 2007 to Spring 2008. PLoS One, 3 (12), e4071. doi:10.1371/ journal.pone.0004071 v an Enge lsdorp, D., & Meixner, M. D. (2010). A historical review of managed honey bee populations in Europe and the Unite d States and the factors that may affect them. Journal of Invertebrate Pathology, 103 , 580 595. doi: 10.1016/j.jip.2009.06.011 Velthuis, H. H . W., & van Doorn, A. (2006). A century of advances in bumblebee domestication and the economic and environmental aspects of its commercialization for pollination. Apidologie, 37 (4), 421 451. doi: 10.1051/apido:2006019 Verboven, H. A. F., Uyttenbroeck, R., Brys, R., & Hermy, M. (2014). Different responses of bees and hoverflies to land use in an urban rural gradient show the importance of the nature of the rural land use. Landscape and Urban Planning, 126 , 31 41. doi: 10.1016/j.landurbplan.2014.02.017 Vergara, C. H., & Badano, E. I. (2009). Pollinator diversity increases fruit production in Mexican coffee plantation s: The importance of rustic management systems. Agriculture, Ecosystems & Environment, 129 (1), 117 123. doi: 10.1016/j.agee.2008.08.001 Vidau, C., Diogon, M., Aufauvre, J., Fontbonne, R., Vigues, B., Brunet, J., Texier, C., Biron, D. G., Blot, N., Alaoui , H., Belzunces, L. P., & Delbac, F. (2011). Exposure to sublethal doses of fipronil and thiacloprid high ly increases mortality of honeybees previously infected by Nosema ceranae. PLoS One, 6 (6), e21550. doi: 10.1371/journal.pone.0021550
80 Walther, B. A., & Morand, S. (1998). Comparative performance of species richness estimation methods. Parasitology, 116 (4), 395 405. Warren, M. S., Hill, J. K., Thomas, J. A., Asher, J., Fox, R., Huntley, B., Roy, D. B., Telfer, M. G., Jeffcoate, S., Harding, P., Jeffcoat e, G., Willis, S. G., Greatorex Davies, J. N., Moss, D., & Thomas, C. D. (2001). Rapid responses of Briti sh butterflies to opposing forces of climate and habitat change. Nature, 414 (6859), 65 69. doi: 10.1038/35102054 Westphal, C., Steffan Dewenter, I., & Tscharntke, T. (2006). Bumblebees experience landscapes at different spatial scales: Possible implicati ons for coexistence. Oecologia, 149 (2), 289 300. doi: 10.1007/s00442 006 0448 6 Westphal, C., Bommarco, R., Carre, G., Lamborn, E., Morison, N., Petani dou, T., Potts, S. G., Roberts, S. P. M., Szentgyorgyi, H., Tscheulin, T., Vaissiere, B. E., Woyciechowsk i, M., Biesmeijer, J. C., Kunin, W. E., Settele, J., Steffan Dewenter, I. (2008). Measuring bee diversity in different European habitats and biogeogra phical regions. Ecological Monographs, 78 (4), 653 671. doi: 10.1890/07 1292.1 S., Wackers, F. L., & Goulson, D. (2012). Neonicotinoid pesticide reduces bumble bee colony growth and queen production. Science, 336 (6079), 351 352. doi: 10.1126/science.1215025 Williams, N. M., & Winfree, R. (2013). Local habitat characteristics but n ot landscape urbanization drive pollinator visitation and native plant pollination in forest remnants. Biological Conservation, 160 , 10 18. doi: 10.10 16/j.biocon.2012.12.035 Wilson, J. S., Griswold, T., & Messinger, O. L. (2008). Sampling bee communitie s (Hymenoptera: Apiformes) in a desert landscape: Are pan traps sufficient? Journal of the Kansas Entomological Society, 81 (3), 288 300. doi: 10.231 7/JKES 802.06.1 Wilson, J. S. , & Carril, O. (2016). The Bees in Your Backyard: A Guide to North America's Bees . Princeton University Press. Winfree, R. (2010). The conservation and restoration of wild bees. Pp. 169 197 in R. S. Ostfeld, W. H. Schlesinger , eds. Year in ecology and conservation biology. (2010). WileyBlackwell, Malden. Winfree, R., Gr iswold, T., & Kremen, C. (2007). Effect of human disturbance on bee communities in a forested ecosystem. Conservation Biology, 21 (1), 213 223. doi: 10.1111/j.15 23 1739.2006.00574.x Winfree, R., Williams, N. M., Gaines, H., Ascher, J. S., & Kremen, C. (2008 ). Wild bee pollinators provide the majority of crop visitation across land use gradients in New Jersey and Pennsylvania, USA. Journal of Applied Ecology, 45 (3) , 793 802. doi: 10.1111/j.1365 2664.2007.01418.x
81 Winfree, R., Aguilar, R., Vazquez, D. P., LeBuh n, G., Aizen, M. A. (2009) A meta analysis of Ecology, 90 (8), 2068 2076. doi: 10.1890/08 1245.1 Wojcik, V. A., & M cBride, J. R. (2011) Common factors influence bee foraging in urban and wildland landscapes. Urb an Ecosystems, 15 (3), 581 598. doi: 10.1007/s11252 011 0211 6 World Population Review (2016). Denver Population 2017 . Accessed May 4, 2018 from http://worldpopu lationreview.com/us cities/denver population/. Wratten, S. D., Gillespie, M., Decourtye, A., Mader, E., & Desneux, N. (2012). Pollinator habitat enhancement: Benefits to other ecosystem services. Agriculture, Ecosystems & Environment, 159 , 112 122. doi: 10.1016/j.agee.2012.06.020 Wright, A., Boyd, C . L., Bowers, M. D., & Scott, V.L. (2017). The Bumble Bees of Colorado: A Pictorial Identification and Information Guide . The University of Colorado Boulder: 66pp. https://www.colorado.edu/cumuseum/sites/default/files/attached files/thebumblebeesofcolorado 2017.pdf . Xian, G., Homer, C., Dewitz, J., Fry, J., Hossain, N., & Wickham, J. (2011). The change of impervious surface area between 2001 and 2006 in the conterminous United States. Photogrammetric Engineering and Remote Sensing , 77(8), 758 762. Zerbe, S., Maurer, U., Schmi tz , S., & Sukopp, H. (2003) Biodiversity in Berlin and its potential for nature conservation. Landscape and Urban Planning, 62 (3), 139 148. doi: 10.1016/S0169 2046(02)00145 7
82 APPENDIX A S ite L andscape C haracteristics and the C as D e fined by PCA Site % impervious % low intensity % medium intensity % high intensity % water % forest % shrub/ grass % crops Component AUR 71.8 15.8 35.5 48.0 0.6 0.0 0.0 0.0 47.448 TCG 67.1 23.3 34.7 41.4 0.6 0.0 0.0 0.0 41.517 ECG 57.2 40 .1 28.5 28.9 2.0 0.0 0.5 0.0 28.062 IVI 52.1 30.8 28.4 27.9 1.8 0.0 9.3 1.8 22.179 IAA 46.3 49.5 29.7 15.6 1.2 0.0 3.9 0.0 15.199 EHH 34.9 67.2 20.0 3.7 4.9 0.2 3.7 0.4 0.248 SAU 32.6 54.2 23.0 3.9 2.8 0.0 16.1 0.0 2.832 CSH 29.0 85.5 9.7 2.4 2.1 0. 0 0.4 0.0 6.452 GCG 20.4 33.5 12.3 3.9 0.4 6.1 43.7 0.0 19.617 BJH 10.1 24.2 8.0 0.7 8.2 1.0 56.7 1.3 32.177 BLP 0.3 3.2 0.0 0.0 43.3 0.0 3.6 49.7 46.509 GKH 0.1 3.5 0.0 0.0 0.1 2.0 94.1 0.3 46.570
83 APPENDIX B Bee Abundance and Richness per Com munity Site Class (% imperv.) Abundance Richness AUR Urban (71.8%) 8 7 TCG Urban (67.1%) 21 9 IVI Urban (57.2%) 126 27 IAA Urban (52.1%) 99 15 Total Urban 254 32 ECG Suburban (46.3%) 19 8 EHH Suburban (34.9%) 42 14 CSH Suburban (32.6%) 90 23 SAU S uburban (29.0%) 40 10 Total Suburban 191 34 GCG Rural (20.4%) 181 16 BJH Rural (10.1%) 750 27 BLP Rural (0.3%) 74 30 GSH Rural (0.1%) 161 30 Total Rural 1166 62
84 APPENDIX C Genus Level Bee Rank Abundance Plot
85 APPENDIX D Floral Richness and Pe rcent Imperviousness per Community
86 APPENDIX E B ee S pecies Abundances and G uild C lassifications Species Abundance Sociality Pollen specificity Nest substrate Body size Phenology O rigin Colletes latitarsus 1 S P G M E N Hylaeus leptocephalus 1 S P C S M E Hylaeus mesillae cressoni 4 S P C S M N Hylaeus mesillae mesillae 3 S P C S M N Hylaeus wootoni 1 S P C S L N Augochlorella aurata 22 E P G S M N Agapostemon femoratus 24 S P G M M N Agapostemon texanus/angelicus 71 S P G M M N Agapostemon vir escens 51 S P G M L N Sphecodes morpho1 1 CP [P] [G] S L N Sphecodes morpho2 12 CP [P] [G] S L N Sphecodes morpho3 4 CP [P] [G] S L N Sphecodes morpho4 1 CP [P] [G] S L N Halictus ligatus 15 E P G M E N Halictus confuses 1 E P G M E N Halictus tripa rtitus 843 E P G S E N Halictus rubicundus 19 E P G M E N Lasioglossum albipenne 4 -P G S M N Lasioglossum anomalum 1 -P G S M N Lasioglossum hudsoniellum 12 -P G S -N Lasioglossum imitatum 2 E P G S M N Lasioglossum knereri 4 -P G S -N Lasioglossum occidentale 8 -P G S -N Lasioglossum pectoral 2 S P G S E N Lasioglossum pectoraloides 1 -P G S -N Lasioglossum pictum 1 -P G S M N Lasioglossum pruinosum 31 -P G S E N Lasioglossum semicaeurelum 111 -P G S M N Lasioglossum succinipenne 2 -P G S -N Lasioglossum synthyridis 1 -P G S -N Lasioglossum tegulare 10 E P G S M N Lasioglossum tenax 10 S P G S M N Lasioglossum trigeminum 7 -P G S -N Lasioglossum versatum 1 -P G S E N Lasioglossum desertum 7 S P G M L N Lasioglossum egregium 4 S P G M M N Lasioglossum sisymbrii 65 S P G M E N Lasioglossum lusoria 25 S O G M M N Lasioglossum morpho1 2 -P G S -N Andrena auricoma 1 S P G M M N Andrena commoda 2 S P G M M N Perdita morpho1 1 S O G S L N Perdi ta morpho2 1 S O G S L N Perdita morpho3 1 S O G S L N Perdita morpho4 1 S O G S L N Perdita morpho5 1 S O G S L N Perdita morpho6 1 S O G S L N Perdita morpho7 1 S O G S L N Perdita morpho8 1 S O G S L N Perdita morpho9 1 S O G S L N
87 Dianthidium p arvum 1 S P C S L N Ashmeadiella cactorum 1 S P C S L N Chelostoma philadelphi 1 S P C M E N Hoplitis pilosifrons 1 S P C M M N Hoplitis producta 5 S P C M M N Hoplitis aibifrons argentifrons 1 S P C M M N Osmia morpho1 2 S P C S M N Osmia morpho2 1 S P C S M N Osmia morpho3 1 S P C M M N Osmia morpho4 1 S P C S M N Megachile rotundata 4 S P C M L E Megachile inimica 1 S P C L L N Megachile frigida 2 S P C L L N Megachile perihirta 2 S P C L L N Megachile morpho1 1 S P C L L N Ceratina calcar ata 2 S P C S E N Ceratina neomexicana 3 S P C S -N Nomada rhodosoma 1 CP [P] [G] M M N Diadasia australis 1 S O G M -N Diadasia enavata 3 S O G M -N Melissodes agilis 10 S O G M M N Melissodes coreopsis 1 S O G M L N Melissodes tristis 1 S O G M M N Melissodes vernoniae 1 S O G M M N Melissodes bimaculatus 16 S O G M M N Melissodes morpho1 5 S O G M M N Peponapis pruinosa 1 S O G M L N Anthophora affabilis 1 S P G M E N Bombus grisecollis 2 E P G L L N Bombus huntii 4 E P G M M N Apis meliffera 136 E P H M M E Abundance is the total number of individuals collected across all sites. Sociality refers to whether a species is (eu)social (E), solitary (S), or cleptoparasitic (CP). Pollen specificity indicates whether a species is polylect ic (P collecting pollen fro m many flower families) or oligolectic (O collecting pollen from just a few families or genera); [P] indicates a parasitic species whose host is polylectic. Nest substrate indicates whether a species nests in the ground (G), in a cavity (C), or a hive (H) ; [G] indicates a parasitic species whose host nests in the ground. Body size is the total body length of a specimen, indicating whether it is small (S; <7 mm), medium (M; 7 12 mm), or large (L; >12 mm). Phenology refers to the time of first seasonal act ivity, with E indicating early spring (January through March), M indicating mid spring (April through May), and L indicating late spring (June through July). O rigin refers to whether a species is classified as native (N) or exot ic (E) to Colorado, USA. U ndetermined -
88 APPENDIX F S upplement for S ignificance of R esponse V ariables to B ee Guild E xplanatory V ariables According to Stepwise Linea r Regressions Table A All sites except the BJH site . Explanatory Response p Small Flower richness 0. 045 Medium Flower richness 0.111 Large Flower richness 0.025 Ground Flower richness 0. 080 Cavity None -Hive Flower richness + Urbanization 0.0 12 (FR ) / 0.203 (U) Social Flower richness + Urbanization 0. 0 14 (FR) / 0.187 (U) Solitary None -Polylectic Flower richness + Urbanization 0. 056 (FR) / 0.195 (U) Oligolectic None -Native Flower richness 0. 076 Exotic Flower richness 0.0 11 Early spring F lower richness + Urbanization 0. 0 40 (FR) / 0.240 (U) Mid spring Flower richness 0.125 Late spring Urbanization 0.170 T here was a distinct difference in bee abundance at the BJH site where I collected 658 Halictus tripartitus over the three trapping even ts. Excluding this site, I found that the only bee guild that responded to urbanization was late spring bees, though this relationship was not significant . Table B All sites except the ECG site . Explanatory Response p Small Urbanization 0.0 40 Medium Urbanization 0. 059 Large Flower richness + Urbanization 0.0 34 (FR) / 0.20 1 (U) Ground Urbanization 0. 0 49 Cavity None -Hive Flower richness + Urbanization 0.0 20 (FR) / 0.153 (U) Social Urbanization 0.0 23 Solitary Urbanization 0. 099 Polylectic Urban ization 0.0 26 Oligolectic None -Native Urbanization 0. 0 46 Exotic Flower richness + Urbanization 0.0 24 (FR) / 0.214 (U) Early spring Urbanization 0. 0 34 Mid spring Urbanization 0. 058 Late spring Urbanization 0.0 49
89 Bee abundance differed substantiall y at the ECG site, an urban site where bee abundance (126 specimens) was unusually high relative to other urban and suburban sites. Removing this site from the analysis revealed that small or medium b ees, ground nesting, social or solitary, polylectic, na tive, and early spring bees responded to urbanization. Table C All sites except both BJH and ECG sites . Explanatory Response p Small Flower richness 0. 062 Medium Urbanization 0. 083 Large Flower richness 0.0 41 Ground Urbanization 0. 099 Cavity None -Hive Flower richness + Urbaniz ation 0.0 24 (FR) / 0.111 (U) Social Flower richness + Urbanization 0.029 (FR) / 0.195 (U) Solitary Urbanization 0.150 Polylectic Urbanization 0. 052 Oligolectic None -Native Urbanization 0. 091 Exotic Flower richness + Urbanization 0.0 31 (FR) / 0.185 (U) Early spring Urbanization 0. 063 Mid spring Urbanization 0. 096 Late spring Urbanization 0. 063 Excluding both sites from the analysis showed that medium, ground nesting, solitary, polylectic, a nd native bees responded to urbanization, though these results were not significant. Like the model including all sites, all phenological types were responsi ve to urbanization, though insignificantly.