University of Nevada, Reno THE ROLE OF URBAN CLIMATE AND
Transcription
University of Nevada, Reno THE ROLE OF URBAN CLIMATE AND
University of Nevada, Reno THE ROLE OF URBAN CLIMATE AND LAND COVER IN PHENOLOGY, NEST SUCCESS, AND HABITAT USE A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Ecology, Evolution, and Conservation Biology by Miles Becker Dr. Peter J. Weisberg/Dissertation Advisor August 2013 UMI Number: 3595627 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. UMI 3595627 Published by ProQuest LLC (2013). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, MI 48106 - 1346 UNIVERSITY OF NEVADA, RENO THE GRADUATE SCHOOL We recommend that the dissertation prepared under our supervision by MILES BECKER entitled The Role of Urban Climate and Land Cover in Phenology, Nest Success, and Habitat Use be accepted in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Peter J. Weisberg, PhD, Advisor Scott Bassett, PhD, Committee Member Matthew Forister, PhD, Committee Member James Sedinger, PhD, Committee Member Laurel Saito, PhD, Graduate School Representative Marsha H. Read, PhD, Dean, Graduate School August 2013 i ACKNOWLEDGEMENTS S. Bassett, M. Forister, L. Saito, and J. Sedinger gave invaluable insights and comments on all aspects of the project that greatly improved its quality and scope. P. Weisberg simultaneously buoyed and grounded the project with optimism and thoughtful criticism, skillfully adjusting focus between the trees and the forest. Numerous EECB students reviewed and commented on the work at varying stages. Special thanks deservedly goes to D. Nonne for patient and thorough assistance with the mechanics of nest survival analysis. C. Taylor and M. Johnson assisted with youthful energy and attention to detail in the field. Land cover and parcel data sets were courteously provided by the City of Reno and Washoe County. Access to restoration sites was granted by The Nature Conservancy. Small grants and research support were provided by the Ecology, Evolution, and Conservation Biology Graduate Program at University of Nevada, Reno. ii SUMMARY ABSTRACT Urban development is the fastest growing global land use and it increasingly brings drastic environmental change to natural systems. Climatic variability and spatial habitat heterogeneity within a city could influence species in habitat remnants. The objectives of this project were to investigate the relative importance of temperature variation and urban land cover for nest survival, the function of thermal refuges for habitat use by breeding birds, and the relationship between urban climate and phenology. Temperatures, leaf phenology, invertebrate abundance, nest habitat, and nest fate were monitored over a four-year study period at nine sites along an urban to rural gradient. The sites were located in a riparian corridor passing through the cities of Reno and Sparks, Nevada. Study area temperature anomalies from the 20-yr normal strongly impacted nest survival rates of American Robins (Turdus migratorius), Mourning Dove (Zenaida macroura), and Black-headed Grosbeaks (Pheucticus melanocephalus). Nest survival increased at warmer than normal temperatures and to a lesser extent at colder than normal temperatures. Canopy cover within a 20m radius of the nest and pedestrian traffic at a site were positively associated with nest survival. Microsites available to nesting birds that had more vegetation were cooler at night and warmer in the day when they had more bare ground, less impervious surfaces, and less canopy cover. The nests of doves and robins were constructed at cooler locations when air temperatures were warmer. Temperature varied between sites with the warmest temperatures recorded in rural sites, followed by the more urban sites, and with coolest temperatures in suburban sites. The timing of nesting by robins occurred earlier in the warmer sites and at sites with earlier bud break by Populus spp. Peaks in foliar invertebrate abundance were also weakly iii associated with leaf out, while the sequence of peaks by aerial invertebrates was related more to the level of urbanization at a site. The results from these studies demonstrate the ability of urban development to differentially affect phenology across trophic levels, have an additive effect with temperature variation on nest survival, and offer opportunities to provide thermal refuges to buffer the impacts of regional climate change. iv TABLE OF CONTENTS i ACKNOWLEDGEMENTS iii SUMMARY iv TABLE OF CONTENTS v LIST of TABLES vii LIST of FIGURES 1 PROJECT BACKGROUND 12 CHAPTER 1: THE INFLUENCE OF URBAN CLIMATE ON TREE, INVERTEBRATE, AND NESTING PHENOLOGY 52 CHAPTER 2: SYNERGISTIC EFFECTS OF SPRING TEMPERATURES AND LAND COVER ON NEST SURVIVAL OF URBAN BIRDS 88 CHAPTER 3: USING NEST HABITAT FOR SUITABLE MICROCLIMATES AT WARMER TEMPERATURES v LIST OF TABLES p.45 TABLE 1.1 A summary of logistic exposure nest survival models for three species along an urban to rural gradient in Reno and Sparks, Nevada. Models include covariates and are compared to the null model of constant survival S(.). Individual covariates included are the temperature deviation from 20-yr normal at hatching with a quadratic term (hdev2), amount of canopy cover in a 20m radius (canopy 20), and rates of pedestrians at a study site (peds). An interaction term for species and nest stage (stage*species) was also included to represent three stages of nesting that differ in daily survival rates. The lowest AICc value of all candidate models considered was 1444.39. p.81 TABLE 2.1 Categorical threshold values of temperatures and the timing of breeding used to analyze nest habitat use by three bird species breeding in urban habitat in northern Nevada. Percentages of four land cover types within a 20m radius of nests and nest height are also reported. All values are means (µ) ± SD for the associated sample size (N). p.118 TABLE 3.1 Qualitative indicators of sequential stages of bud break used to measure leaf phenology in Salix sp. and Populus sp. at nine sites on an urban to rural gradient. p.119 TABLE 3.2 Site level differences in environmental factors at nine sites on an urban to rural gradient. The urban index is a PCA score derived from canopy vi cover, impervious surface, bare ground, and road density. Estimates of relative robin densities are in the unit of birds per hectare. p.120 TABLE 3.3 Abundance by taxon of invertebrates captured with adhesive traps from May to July 2011 at nine sites in northern Nevada. vii LIST OF FIGURES p.46 FIGURE 1.1 Study area location (star) in the western United States. Enlarged area of Reno and Sparks, Nevada with study sites (black) connected by the Truckee River (black line). Location of the National Weather Station is marked by the +. Enlarged images of two study sites demonstrate the within-site variation in land cover and among-site variation along the urban to rural gradient. p.47 FIGURE 1.2 Weekly averages of daily maximum temperatures in the study area during the breeding season for 2009 – 2012. Upper horizontal dashed line marks a temperature threshold of 27 °C, the upper boundary of physiological zero. Lower horizontal dashed line marks the temperature threshold for 15 °C, below which nestlings are likely to be negatively impacted by cold. p.48 FIGURE 1.3 Temperature deviation from the 20-yr norm at the time of clutch initiation for three bird species nesting in urban habitat over a four-year study period. p.49 FIGURE 1.4A-B Brood parasitism of grosbeak nests by Brown-headed Cowbirds in A) 36 grosbeak nests that either fledged young or failed, and B) the number of cowbird eggs deposited at a range of temperature deviations from the 20-yr normal for the five days during egg laying. p.50 viii FIGURE 1.5 Fitted curve of nestling survival probability estimates (± SE) for temperature deviations from normal after hatching for three species breeding in urban habitat over a four-year study period. p.51 FIGURE 1.6A-B Differences in the amount of canopy cover within a 20m radius, and rates of pedestrian traffic, between the nests of three species that fledged young or failed in urban habitat over a four-year study period. p.82 FIGURE 2.1A - F Hypothesized effects of temperature on nest site selection. Nest habitat on the Y axis represents % cover of one of the land cover variables measured (bare ground, canopy cover, impervious surfaces, or understory vegetation). Relative timing of nesting in the breeding season on the X axis represents seasonal variation of any one of several time varying factors that have the potential to influence nest site selection (e.g. predation risk, human activity). Lines represent different levels to air temperature at the time of nesting (cold = dashed grey, mild = solid grey, warm = solid black). Interpretation of the graphical illustration is provided by the presence or absence of a significant effect. (after Gotelli and Ellison 2004) p.83 FIGURE 2.2 Daily variation of maximum air temperatures during spring and summer in Reno, Nevada from 2009 to 2012. p.84 ix FIGURE 2.3A - D The influence of land cover within 20m on Tmax (left) and Tmin (right) recorded at 48 microsites in urban reserves from April 21 to July 13, 2011. Relative levels of land cover types are categorized into three groups of low, mid, and high. The upper percentage threshold value of low, mid, and high levels were 30, 55, and 88 for understory vegetation (A); 6, 24, and 75 for impervious surfaces (B); 10, 30, and 92 for canopy cover (C); and 8, 19, and 48 for bare ground (D). p.86 FIGURE 2.4A - D Microhabitat used at 278 robin and dove nest sites for different levels of the timing of breeding (early, middle, late) and the relative daily temperature maximum in the five days before nesting (cold, mild, warm). Temperature levels differed by 7.5 °C and timing levels differed by an average of 21 days. The four land cover variables at a nest that could have varied with time or temperature were understory vegetation (A), impervious surfaces (B), bare ground (C), and canopy cover (D). p.87 FIGURE 2.5 Counts of corvids observed on transects at eight sites in a riparian corridor of northern Nevada. Relative maximum temperature on the day of a survey (cold, mild, warm) are compared within a site over eight survey intervals. Each temperature level spanned a 7.5 °C increment. p.121 FIGURE 3.1 Location of nine study sites (black) along the Truckee River (black line) on an urban to rural gradient across the cities of Reno and Sparks, Nevada. x p.122 FIGURE 3.2 Hypothesized pathways between urban development and phenology. Predictor variables are labeled in boxes and response variables are open text. Arrows indicate the direction of the relationship between predictor and response variable. p.123 FIGURE 3.3 Average daily temperature maximum at study sites in April and June 2011. Sites are arranged left to right according to their respective level of urban development. p.124 FIGURE 3.4 The timing of bud burst in Populus sp. and Salix sp., peak abundance for aerial, aquatic, and foliar invertebrates, and clutch initiation in American Robin at sites across an urban to rural gradient. Sites are arranged left to right according to their respective level of urban development. p.125 FIGURE 3.5 Path analysis modeling hypothesized relationships between predictor variables (in boxes) and the phenology of leaf out, peak invertebrate abundance, and robin lay date. The relative strength of associations between variables are indicated by line weight, based on standardized Beta coefficients. The residual variance (error) for each response variable are given in small boxes. Darker lines indicate a stronger relationship and dashed lines represent a negative relationship. xi p.126 FIGURE 3.6 Maximum abundance of invertebrates reached at the timing of peak abundance in sites on a rural to urban gradient. Invertebrate guilds were aerial, foliar, or aquatic. 1 PROJECT BACKGROUND The world reached a symbolic landmark in 2008 when the number of people living in cities crept past fifty percent. As a species, more humans now live in urban habitat than ever before. Issues of food supply, air and water quality, waste management, energy, and livability are practical concerns for urban dwellers. Urban sustainability is intimately tied to ecological processes – nutrient cycling, energy flow, organisms, hydrology - and achieving sustainability will require in-depth knowledge of urban ecosystem function. It will also be important to understand what changes will come about in natural systems as they become part of the matrix of land uses in the urban landscape. Natural areas are being converted to urban development faster than urban population growth, and the sprawl is engulfing the habitat of many species. Although urban and residential areas cover a small percentage of the land area, they often overlap with hotspots of biodiversity and critical habitat (Hansen et al. 2005). As a result, many species are exposed to a novel ecosystem that has the potential to unravel the integrity of ecological communities and alter regional biodiversity. Cities offer a unique opportunity to study basic questions in ecology. Many cities are at least partly planned and different designs provide opportunities for natural experiments in environmental change. Pickett et al. (2011) recently reviewed the major environmental changes introduced by urban development to natural areas. Most obvious is the removal of vegetation in forested biomes and the addition of canopy cover in prairie and desert biomes. Structural modification of habitat can greatly affect local 2 climate and the interface between urban climate and organisms is the focus of this dissertation. The biophysical properties of different land cover types within a city create the complex urban climate. At the landscape level, the net effect of the urban matrix may be warmer temperatures in the city than surrounding natural areas. The urban heat island effect exists for some cities, like Baltimore, in temperate forested regions of the United States where canopy cover is replaced by impervious surfaces. However, cities in arid regions, like Phoenix, may actually be cooler than surrounding natural areas from increased irrigation and canopy cover. The patchy mosaic of fine-scale heterogeneity in land cover types may be more important in determining microclimate within a city than the city-wide elevation in temperature (Peterson 2003). Organisms also interact with the thermal environment at finer spatial scales and temperature changes at the landscape or patch scale can have profound impacts on organisms. The phenology of many organisms is temperature sensitive and may shift along an urban thermal gradient. Plants have already been observed to do so (Zhang et al. 2004). The timing of breeding in Florida Scrub-jays (Aphelocoma coerolescens) advanced in cities compared to natural areas, but the shift was attributed to food resources, not temperatures (Schoech and Bowman 2003). The role of urban climate in the phenology of other taxa remains largely unexplored. Fitness of an organism can also be temperature sensitive, and only recently have the effects of urban climate on population dynamics been studied (Meineke et al. 2012). Physiological limitations may increase the risk of heat stress for some species in urban habitat, and reproduction can also be affected by microclimate. Eggs laid by long-tailed 3 skinks (Euthopis longicaudata) in concrete walls in urban areas were incubated at warmer temperatures and produced larger hatchlings than eggs laid under rocks in natural areas (Huang and Pike 2011). The fitness benefit for lizards nesting in urban habitat was lost after urban temperatures increased disproportionately from climate change and became too warm for egg development. Measures of fitness in urban habitat are necessary for establishing habitat quality. A diverse community of native species can inhabit cities and suburbs, yet presence does not signify persistence. Measurements related to reproduction and survival may be more reliable indicators (Aldridge and Boyce 2008, Johnson 2007) for helping to identify source and sink populations. The influence of urban development on productivity in birds does not show a consistent pattern across studies (Chamberlain et al. 2009), despite there being plenty of anthropogenic sources for lower survival and reproduction. House sparrows (Passer domesticus) at more urbanized sites were smaller and in poorer condition (Liker et al. 2008), suggesting that urban habitat was lower quality despite higher bird abundance. In contrast, mockingbird survival and productivity increased in urban sites, indicating a source population (Stracey and Robinson 2012). Even seemingly advantageous resources in urban habitat may undermine nest success. Organisms may have evolved to respond to specific cues that increase fitness in natural habitat, but those cues can become decoupled from fitness in urban habitat. For example, earlier nesters are often the more successful breeders and produce more young than nests started later in the breeding season. Earlier leaf out by Lonicera sp. at urban sites initiated earlier nesting by higher quality Northern Cardinals (Cardinalis cardinalis), yet those nests had lower estimated nest survival rates (Leston and Rodewald 4 2006, Rodewald et al. 2011). Such ecological traps could also result from urban climate influences on nest phenology and nest success. Habitat modifications from urban development are localized and not independent of regional changes occurring simultaneously. An emerging area of research is the synergistic effects of land use change and climate change on distributions and population dynamics (Mantyka-Pringle et al. 2012). The same predictions of species distributions under climate change scenarios can be predicted by land use change, potentially masking the signal of one or the other (Clavero et al. 2011). The rapid rate of land use change may have more of an impact than climate change effects for some sensitive species (Wilson et al. 2013). Land use change may also provide opportunities to buffer populations against negative consequences of climate change (Clausen et al. 2013). Many cities will be affected by climate change without a current mitigation plan in place (Stone et al. 2012, Tyler and Moench 2012) and improving resiliency should be a priority. The objectives of this project were to better understand the relative importance of temperature and land cover on nest survival, the function of thermal refuges for nest sites, and the influence of urban climatic variability on phenology. The three studies touch on global change processes with the intention of improving urban habitat resiliency while addressing questions relevant to ecology. Monitoring of temperatures, trees, invertebrates, and breeding birds occurred in a riparian corridor along an urban to rural gradient. Riparian habitat provides a disproportionate level of ecosystem services in arid biomes and urban development is the most common new land use in US catchments 5 (Jones et al. 2010). Birds received most of the attention in these studies as they are readily observable indicators of biodiversity and ecosystem health. The impacts of these studies extend well beyond conservation of native species. Maintaining biodiversity in and around cities may help to preserve the ecosystem functions on which people depend. Examples include the pollination of landscaped plants and urban gardens, sewage treatment, rainwater drainage, noise reduction, air filtration, energy savings, and recreational opportunities (Alberti 2005, Bolund and Hunhammer 1999, Taha 1996). Mental health and well-being are also at risk from a decline in native species richness. Cognitive abilities and childhood learning may improve in the presence of trees and other vegetation (Wells 2000). Experiencing nature or biodiversity on a frequent basis may strengthen family and neighborhood relationships (Dustin et al. 2010) or improve mental health (Dean et al. 2012). Perhaps most importantly, childhood experiences in nature, even in ways as simple as exploring a backyard, influence adult attitudes towards environmental issues (Wells and Lekies 2006). The significance of urban biodiversity may be lost to future generations if it is not conserved now. Fortunately, cities have the potential to incorporate thoughtful ecological design (Niemala et al. 1999), including microclimate. State and municipal governments are the most effective at implementing successful conservation strategies, and some cities are planning for biodiversity (Connery 2009, de Oliveira et al. 2011). Unlike natural systems in the United States, much of the land surface in urban habitat is privately owned and managed. Backyards and residential gardens offer substantial resources for organisms (Davies et al. 2009, Doody et al. 2010, French et al. 2005, Larson et al. 2009) and they 6 may be as important for urban biodiversity as public open spaces. Using those resources to the fullest advantage will require coordinated efforts (Hostetler and Drake 2009, Lerman et al. 2012). Citizen participation in urban ecology research is growing and provides an opportunity to educate the public about conservation (Cooper et al. 2007, Ryder et al. 2010). Close partnerships between urban dwellers and urban ecologists will hopefully increase the pace of learning and achieve biodiversity conservation. 7 LITERATURE CITED Alberti, M. 2005. The effects of urban patterns on ecosystem function. International Regional Science Review 28: 168–192. Aldridge, C.L. and Boyce, M.S. 2008. Accounting for fitness: combining survival and selection when assessing wildlife-habitat relationships. Israel Journal of Ecology & Evolution 54: 389–419. Bolund, P. and Hunhammar, S. 1999. Ecosystem services in urban areas. Ecological Economics 29: 293–301. Chamberlain, D.E., Cannon, A.R., Toms, M.P., Leech, D.I., Hatchwell, B.J., Gaston, K.J. 2009. Avian productivity in urban landscapes: a review and meta-analysis. Ibis 151: 1–18. Clausen, K.K., Stjernholm, M., Clausen, P., 2013. Grazing management can counteract the impacts of climate change-induced sea level rise on salt marsh-dependent waterbirds. Journal of Applied Ecology 50: 528–537. Clavero, M., Villero, D., Brotons, L., 2011. Climate change or land use dynamics: do we know what climate change indicators indicate? PLoS ONE 6. Connery, K. 2009. Biodiversity and urban design: seeking an integrated solution. Journal of Green Building 4: 23–38. Cooper, C.B., Dickinson, J., Phillips, T., Bonney, R. 2007. Citizen science as a tool for conservation in residential ecosystems. Ecology and Society 12. Davies, Z.G., Fuller, R.A., Loram, A., Irvine, K.N., Sims, V., Gaston, K.J. 2009. A national scale inventory of resource provision for biodiversity within domestic gardens. Biological Conservation 142: 761–771. 8 De Oliveira, J.A.P., Balaban, O., Doll, C.N.H., Moreno-Penaranda, R., Gasparatos, A., Iossifova, D., Suwa, A. 2011. Cities and biodiversity: Perspectives and governance challenges for implementing the convention on biological diversity (CBD) at the city level. Biological Conservation 144: 1302–1313. Dean, J., van Dooren, K., Weinstein, P. 2011. Does biodiversity improve mental health in urban settings? Medical Hypotheses 76: 877–880. Doody, B.J., Sullivan, J.J., Meurk, C.D., Stewart, G.H., Perkins, H.C. 2010. Urban realities: the contribution of residential gardens to the conservation of urban forest remnants. Biodiversity and Conservation 19: 1385–1400. Dustin, D.L., Bricker, K.S., Schwab, K.A. 2010. People and Nature: Toward an Ecological Model of Health Promotion. Leisure Sciences 32: 3–14. French, K., Major, R., Hely, K. 2005. Use of native and exotic garden plants by suburban nectarivorous birds. Biological Conservation 121: 545–559. Hansen, A.J., Knight, R.L., Marzluff, J.M., Powell, S., Brown, K., Gude, P.H., Jones, A. 2005. Effects of exurban development on biodiversity: Patterns, mechanisms, and research needs. Ecological Applications 15: 1893–1905. Hostetler, M. and Drake, D. 2009. Conservation subdivisions: A wildlife perspective. Landscape and Urban Planning 90: 95–101. Huang, W.-S. and Pike, D.A. 2011. Climate change impacts on fitness depend on nesting habitat in lizards. Functional Ecology 25: 1125–1136. Johnson, M.D. 2007. Measuring habitat quality: A review. Condor 109: 489–504. Jones, K.B., Slonecker, T.E., Nash, M.S., Neale, A.C., Wade, T.G., Hamann, S. 2010. Riparian habitat changes across the continental United States (1972–2003) and 9 potential implications for sustaining ecosystem services. Landscape Ecology 25: 1261-1275. Larson, K.L., Casagrande, D., Harlan, S.L., Yabiku, S.T. 2009. Residents’ yard choices and rationales in a desert city: social priorities, ecological impacts, and decision tradeoffs. Environmental Management 44: 921–937. Lerman, S.B., Turner, V.K., Bang, C. 2012. Homeowner associations as a vehicle for promoting native urban biodiversity. Ecology and Society 17. Leston, L.F.V. and Rodewald, A.D. 2006. Are urban forests ecological traps for understory birds? An examination using Northern cardinals. Biological Conservation 131: 566–574. Liker, A., Papp, Z., Bokony, V., Lendvai, A.Z. 2008. Lean birds in the city: body size and condition of house sparrows along the urbanization gradient. Journal of Animal Ecology 77: 789–795. Mantyka-Pringle, C.S., Martin, T.G., Rhodes, J.R. 2012. Interactions between climate and habitat loss effects on biodiversity: a systematic review and meta-analysis. Global Change Biology 18: 1239–1252. Meineke, E.K., Dunn, R.R., Sexton, J.O., Frank, S.D. 2013. Urban warming drives insect pest abundance on street trees. PLoS ONE 8: e59687. Niemela, J. 1999. Ecology and urban planning. Biodiversity and Conservation 8: 119– 131. Peterson, T. 2003. Assessment of urban versus rural in situ surface temperatures in the contiguous United States: No difference found. Journal of Climate 16: 2941–2959. 10 Pickett, S.T.A., Cadenasso, M.L., Grove, J.M., Boone, C.G., Groffman, P.M., Irwin, E., Kaushal, S.S., Marshall, V., McGrath, B.P., Nilon, C.H., Pouyat, R.V., Szlavecz, K., Troy, A., Warren, P. 2011. Urban ecological systems: Scientific foundations and a decade of progress. Journal of Environmental Management 92: 331–362. Rodewald, A.D., Shustack, D.P., Jones, T.M., 2011. Dynamic selective environments and evolutionary traps in human-dominated landscapes. Ecology 92: 1781–1788. Ryder, T.B., Reitsma, R., Evans, B., Marra, P.P. 2010. Quantifying avian nest survival along an urbanization gradient using citizen- and scientist-generated data. Ecological Applications 20: 419–426. Schoech, S. and Bowman, R. 2003. Does differential access to protein influence differences in timing of breeding of Florida scrub-jays (Aphelocoma coerulescens) in suburban and wildland habitats? Auk 120: 1114–1127. Stone, B., Vargo, J., Habeeb, D. 2012. Managing climate change in cities: Will climate action plans work? Landscape and Urban Planning 107: 263–271. Stracey, C.M. and Robinson, S.K. 2012. Are urban habitats ecological traps for a native songbird? Season-long productivity, apparent survival, and site fidelity in urban and rural habitats. Journal of Avian Biology 43: 50–60. Taha, H. 1996. Modeling impacts of increased urban vegetation on ozone air quality in the South Coast Air Basin. Atmospheric Environment 30: 3423–3430. Tyler, S. and Moench, M. 2012. A framework for urban climate resilience. Climate and Development 4: 311–326. Wells, N. 2000. At home with nature - Effects of “greenness’s on children’s cognitive functioning. Environment and Behavior 32: 775–795. 11 Wells, N.M. and Lekies, K.S. 2006. Nature and the life course: pathways from childhood nature experiences to adult environmentalism. Children, Youth and Environments 16: 1-25. Wilson, J.N., Bekessy, S., Parris, K.M., Gordon, A., Heard, G.W., Wintle, B.A. 2013. Impacts of climate change and urban development on the spotted marsh frog (Limnodynastes tasmaniensis). Austral Ecology 38: 11–22. Zhang, X.Y., Friedl, M.A., Schaaf, C.B., Strahler, A.H., Schneider, A. 2004. The footprint of urban climates on vegetation phenology. Geographical Research Letters 31. 12 SYNERGISTIC EFFECTS OF SPRING TEMPERATURES AND LAND COVER ON NEST SURVIVAL OF URBAN BIRDS MILES E. BECKER AND PETER J. WEISBERG ABSTRACT Climate change has the potential to influence population dynamics through a response in vital rates to temperature. Avian nest fate may be vulnerable to extreme weather events or directional shifts in spring temperatures during the breeding season. Nest fate varies across spatially heterogeneous habitat as well, and changing land uses may introduce additional stressors that contribute to reproductive outcome. We studied the nest fate of three sympatric species breeding in urban habitat in an arid region forecasted to become warmer. Nests of American Robin (Turdus migratorius), Mourning Dove (Zenaida macroura), and Black-headed Grosbeak (Pheucticus melanocephalus) were monitored over four years at seven sites on an urban to rural gradient. Nest fate in relation to temperatures at the time of egg laying and after eggs hatched was analyzed with logistic exposure models. Microhabitat measurements at the nest included nest height and the amount of canopy cover and impervious surfaces at three spatial scales of 20m, 100m, and 500m radii from the nest. Models with temperature deviation after hatching fit the data best when considering a quadratic relationship with temperature, nest stage, finescale canopy cover, and pedestrian traffic. Nest survival increased mainly at warmer temperature anomalies during the nestling stage. The nests of all three species were more likely to survive at locations with more canopy cover in a 20m radius and at sites with more pedestrian traffic. Warming spring temperatures may be beneficial for some breeding birds and increasing cover of native riparian canopy at fine scales in urban reserves may enhance habitat quality. For understanding vital rate and population responses to climate change, it will be important to simultaneously consider land-use influences on local habitat. 13 14 INTRODUCTION The impacts of global climate change on avian populations are widely recognized (reviewed in Møller et al. 2010), yet the underlying mechanisms are not well established. Several studies have demonstrated that the effects of changing temperatures or precipitation on the distribution or population dynamics of a species are often indirect. Examples include the contraction or expansion of current bioclimatic envelopes that can limit habitat availability (Virkkala et al. 2008) or help establish novel communities (Lurgi et al. 2012) where encounters with novel predators, parasites, or diseases could increase. Phenological shifts in the timing of breeding, or the absence of a shift, in response to temperature-related cues may lead to a temporal asynchrony between peak prey availability and peak nutrient demand by nestlings (the mismatch hypothesis: Visser et al. 1998). Other trophic interactions, such as predation rates, may be influenced by the amount of overlap in the activity of breeding birds and their predators (Both et al. 2009). Climate change also has the potential to negatively impact vital rates through direct abiotic effects. Heat waves are becoming more common and may increase mortality rates of adult birds in some regions (McKechnie and Wolf 2010). In addition to extreme stochastic events, slight increases in background temperatures could influence the reproductive success of birds breeding in temperate and high latitude areas. Egg viability is sensitive to temperatures above and below those needed to initiate embryonic development – physiological zero temperature (PZT) – and warmer ambient temperatures during incubation may decrease hatching success (Aldredge et al. 2012). Nestlings and chicks with low thermoregulatory ability are also vulnerable to heat stress. For example, 15 the proportion of fall harvest rates of juvenile sharptail grouse were negatively correlated with the number of days above chick heat stress tolerance during the breeding season (Flanders-Wanner et al. 2004). Alternatively, warming temperatures during the nesting period could benefit reproductive productivity. Temperature-related nest failure is more often associated with cold and wet weather and may be alleviated by regional warming. Hypothermia is a serious hazard for underdeveloped young. Oxidative stress in precocial chicks (Gallus gallus) maintained at low temperatures (20 °C) for 12 hours caused damage to heart and brain tissue (Mujahid and Furuse 2009). Increased nest survival at warmer temperatures during the breeding season has been documented in cavity nesters (Hollenbeck et al. 2011, Newlon and Saab 2011) and open-cup nesting species (Skagen and Adams 2012, Wright et al. 2009). Cases of the opposite response, increased nest survival at cooler temperatures, are reported (Dreitz et al. 2012) but rare, even for arid regions with extreme heat during the breeding season. Spring warming trends may therefore detract from or improve successful nesting attempts. It is also unclear if nest failure is more likely at extreme temperatures or simply at slight deviations from baseline temperatures. The stage of the nesting cycle, incubation or nestling, that is most sensitive to temperature variation needs elucidation as well. Wood Duck (Aix sponsa) eggs experimentally maintained at 0.9 – 2.3 °C cooler temperatures incubation produced females that were less likely to fledge, survive their first winter, and reproduce (Hepp and Kennamer 2012). In contrast, the daily minimum temperatures for the first 15 days after hatching in Great Tit (Parus major) nests explained most of the variation in post-fledging survival (Greno et al. 2008). Seasonal 16 variation in temperatures during the separate stages of the nest cycle may have independent effects on nest success. The relative sensitivity of nest survival to spring temperatures is poorly understood, especially for passerines breeding in temperate regions, and could explain population declines in addition to other climate change induced factors. Any temperature-related influences on nest survival are likely to be mediated by habitat. Different land covers have different biophysical effects on climate through a change in albedo, surface roughness, heat capacity, or water retention (Vargo et al. 2013). Concurrent with climate change, the ranges of many bird populations are undergoing rapid land use changes and land cover modification, with implications for local temperatures. The fastest growing global land use is urban development (Seto et al. 2012) that typically modifies the thermal environment in similar ways to forecasted regional temperatures for many natural areas, such as milder winters and warmer summers (Grimmond and Oke 1999, Oke 1988, Pickett et al. 2011). Thermodynamics in urban areas may influence breeding birds in similar ways as climate change would in less disturbed habitat, although the signal may be difficult to detect. Many other anthropogenic stressors exist in urban areas (Chamberlain et al. 2009) that may mask temperature effects on nest survival. Air and noise pollution, nestling food quality, and nest predation risk can change with urban development (Halfwerk et al. 2011, Peach et al. 2008, Robb et al. 2008, Ryder et al. 2010, Thorington and Bowman 2003). Dramatic and immediate land cover disturbances may be more important for nest survival than local temperature variation or long-term climate change, as it was for less developed systems (Clavero et al. 2011, Cox et al. 2013). Identifying the relative 17 importance of climate change and land use change, synergistic effects, and interactions is necessary for mitigating the impact of global change on native species (Mantyka-Pringle et al. 2012). We studied the influence of variation in spring temperatures during incubation and the nestling period on nest survival of three sympatric songbird species in northern Nevada, USA. Arid lands in the Southwest and Intermountain West regions of North America, including the Great Basin, are forecasted to become warmer (Diffenbaugh et al. 2008). The study area included urban habitat with substantial land cover heterogeneity within breeding territories. Our primary objectives were to determine 1) how much variation in nest survival was explained by temperature deviation from the norm, compared to more extreme temperatures above and below physiological tolerances of eggs and nestlings; 2) which stage of the nest cycle was most sensitive to temperature variation; and 3) the relative importance of habitat at the nest compared to temperatures during nesting. We expected more extreme heat or cold to influence nest survival more than slight deviations from background temperatures. We predicted that young nestlings would be more vulnerable than recently laid eggs to temperature stress for two reasons. First, temperatures during the incubation stage are closer in time to the favorable weather conditions used as a cue by a female to lay eggs. Second, food demand by nestlings puts more pressure on adults to leave the nest and forage, thereby leaving young chicks exposed to the weather. We expected that in a disturbed urban ecosystem, land cover and anthropogenic effects would be more important for nest survival than temperatures. METHODS 18 Study Area and Species Our study system was a series of municipal parks on a 15.25 km transect through the cities of Reno and Sparks in northern Nevada (Figure 1). The city boundaries have expanded outward rapidly in the past 40 years, establishing residential areas, suburbs, and industrial parks in a complex patchwork of land uses. Areas of natural vegetation, primarily sagebrush communities, and farmland remain where the boundaries of developed land parcels abruptly end. We selected seven public parks ranging in size from 3 ha to 18 ha that all contained at least some native riparian vegetation, were easy to access, and represented the composite land covers available in the urban matrix and reserves. The study area is on the border of the Great Basin and eastern Sierra Nevada where there is strong inter-annual, seasonal, and intra-seasonal temperature variation. Average winter temperatures are 7° C and summers reach highs of 33° C. The region is fairly arid and most precipitation falls as snow during winter. Temperatures during the breeding season in spring and summer can vary greatly over a short period of time (Figure 2). The three study species, American Robin (Turdus migratorius, hereafter robin), Black-headed Grosbeak (Pheucticus melanocephalus, grosbeak), and Mourning Dove (Zenaida macroura, dove), differ in length of the nesting season and nest habitat preferences. Doves and robins initiate nesting in early spring when temperatures are often cold and unpredictable and continue to nest through summer. Doves and robins are also synanthropic habitat generalists and construct nests in a wide range of substrates. 19 Grosbeaks arrive in the study area the very end of April and begin breeding in mid-May when temperatures are usually warmer. Grosbeaks typically nest in denser vegetation close to the river and they are less likely to be found in areas with high levels of human activity. Nest Monitoring We monitored 159 robin nests, 163 dove nests, and 49 grosbeak nests from 2009 to 2012. We began nest searching in late March or early April and located nests using behavioral cues of the adults or by visually scanning vegetation. All nests used in the analysis were lower than 8m because limited visibility of higher nests made it difficult to accurately assess nest age and fate. A mirror pole was used to examine nest contents unless parents were observed feeding chicks or fledges. Nests were visited every three to five days until the nest failed or nestlings fledged. Visits between study sites were systematically rotated to ensure consistent search effort across all parks. Females lay one egg a day and the day the first egg was laid could be estimated with a high level of accuracy for nests observed in the middle of the laying sequence. For all other nests, we estimated clutch initiation by counting back from hatch day or by aging nestlings. Nestlings of uncertain hatch date were aged using reference photographs taken of nests with the hatch day observed. A single observer aged all nests. Nest fate was determined from the nest contents at consecutive visits to the nest. Nests empty more than four days before expected fledge date were recorded as failed. Nests with live nestlings observed until at least within two days of the expected fledge date were designated successful if there was also supporting evidence of fledging from 20 auditory or visual observations, such as adults bringing food to fledges near the nest tree. Alternative sources of nest failure, such as abandonment or depredation, could not be distinguished for most nests. Predators may consume eggs or chicks after parents abandon a nest and before the next visit by an observer, making it difficult to know the primary source of nest failure. Since temperatures can affect nestling physiology and parental behavior in ways that influence rates of both abandonment and nest depredation, it was not possible to separate out sources of nest failure from the protocol we used. One potential source of nest failure that may have differed between the study species was nest parasitism by Brown-headed Cowbirds (Molothrus ater) in grosbeak nests. Cowbirds are one of the more common species in the study area during the breeding season (Lynn et al. 1998) and they may be an additional influence on grosbeak nest survival. Cowbird females deposit their eggs in the host nest during egg laying by the host, and parasitism is sometimes associated with increased rates of nest failure (Cox et al. 2012, Hannon et al. 2009). We recorded the presence (cow egg) and number of cowbird eggs (cow num) and nestlings in grosbeak nests. Temperatures Daily temperatures in the study area are recorded continuously by a National Weather Station at the Reno Tahoe International Airport located in close proximity to the study sites (Figure 1). We used five measures of temperature variability. The first was intended to detect subtle changes in background levels of ambient temperatures. Average daily temperature deviation from the 20-yr average (Tdev) was calculated for every day of the nesting cycle for all nests with known age. Tdev avoids confounding day of the 21 season with temperatures, because actual temperatures generally increase as the breeding season progresses. Tdev also provides a benchmark for biologically significant temperatures above or below the normal range of temperatures nesting birds may have experienced or adapted to in the past twenty years. We then calculated average Tdev for a moving window in the five days after clutch initiation (Cdev) and five days after hatching (Hdev) to isolate the relative importance of those two stages. A five day window was used because complete incubation and full nest attentiveness starts after the final egg is laid (Morton and Pereyra 2010, Wang and Weathers 2009) and songbird nestlings are least capable of thermoregulation the first few days after hatching (Pereyra and Morton 2001). The other two temperature metrics were intended to capture heat events beyond the physiological tolerances of eggs or nestlings. Physiological zero temperatures range from 24 – 27 °C and we used a conservative estimate of the upper limit to calculate the number of days with Tmax above 27 °C during the five days after clutch initiation for each nest (pzt). The temperature threshold for nestling survival is less specific, but reportedly 11 – 20 °C may be required for mortality or tissue damage (Mujahid and Furuse 2009, Rippon et al. 2011). We used an intermediate threshold value of 15 °C that may correspond to negative consequences for nestlings in this study. The number of days with Tmax below 15 °C in the first five days after eggs hatched was also calculated for each nest that survived to hatching (u15). Habitat 22 While we were primarily interested in the influence of temperature on nest success, nest site habitat could also mediate predation risk. Nest height was a significant factor for nest survival of Northern Cardinals (Cardinalis cardinalis; Smith-Castro and Rodewald 2010), silvereyes (Zosterops lateralis) and blackbirds (Turdus merula; van Heezik et al. 2008) breeding in urban habitat. Increased vegetation density surrounding the nest may also decrease detection by predators (Chalfoun and Martin 2009). Nest habitat variables considered in the analysis were nest height (nest ht), percentage of canopy cover within a 20m radius (canopy20), and percentages of impervious surface and all vegetation within a 100m radius (impev100, plants100) and 500m radius (imperv500, plants500). We measured nest height with a clinometer to the nearest meter. Quantities of land cover types within specified radii were calculated in ArcGIS (ArcMap 10, ESRI Software, Redlands, CA, USA) using layers provided by the City of Reno (AMEC 2012). The five land cover types were classified using a geographic object-based image analysis based on 1m resolution orthophotographs captured in summer 2010 by the National Agriculture Imagery Program that yielded an overall accuracy of 94%. Human and Predator Activity The seven study sites covered a gradient of urban development that likely influenced predator populations and disturbance from recreational park use. Some of the more active predators in urban areas are cats (Balogh et al. 2011) and corvids (Marzluff et al. 2007). Pedestrians also have the potential to disturb birds in urban areas (Schlesinger et al. 2008). Too few cats were detected to reliably estimate densities, so we used only corvid data for a predator activity index at a site. We estimated relative site 23 level densities of corvids, and pedestrians (corvid, peds) by distance sampling (Buckland et al. 2001, Gottschalk and Huettmann 2011) along transects through each park in the 2011 breeding season. Predator and pedestrian densities were only estimated in one year (2011) even though nesting data were collected over four years (2009-2012). However, resources that attract neighborhood predators and people to a park are fairly constant from year to year. Line transects followed a foot or bicycle path through the entire length of each site and generally ran parallel to the river. The perpendicular distance from an observer on the line to a detected bird or predator was measured with a laser rangefinder (Sport 450, Bushnell, Overland Park, Missouri, USA). Transects were started between 0630 and 1000 and sites were assigned to one of three groups based on geographic proximity. The sampling order within a group was rotated at every visit to avoid a temporal bias. Within a round, all groups were sampled over 3 days and repeated 8 times over the season every 5-8 days for a constant observer effort between sites. The direction of travel on transects from start to finish alternated between East and West on consecutive visits. ANALYSIS Predator and Pedestrian density Site level densities of potential corvid predators were estimated in Program DISTANCE (version 6.0 Release 2, Research Unit for Wildlife Population Assessment) that avoids biased estimates from decreasing detection probabilities at farther distances. Total density estimates were divided by the number of visits to the same site. 24 Observations were treated as clusters and distances were considered to be perpendicular. Data were stratified by site and estimated at the site level using the negative exponential key function and cosine series expansion. The functions were chosen because they best represented the distribution of the data and series expansions had the lowest AIC scores when compared to models with other combinations of functions. Pedestrian traffic was estimated with a belt transect estimate using a fixed distance of 100m on either side of the footpath. Belt transects were used for pedestrian estimates because we were able to observe everyone traveling on designated paths in relatively open habitat with a detection probability of approximately 100%. We estimated number of pedestrians per hour in Program DISTANCE using a uniform key function and cosine series expansion. Nest Success Nest survival was analyzed in Program MARK (White and Burnham 1999) for a series of logistic exposure models built with different combinations of individual covariates. Given that nests were clustered in time (years) and space (sites), we first examined site, year, and species level effects by placing nests into groups nested by those three factors. Species was the only group that had an effect on nest survival and data were pooled across years and sites. The two parameter columns for species groups in the design matrix were included in models built with the individual covariates for nest temperature (Tdev, Cdev, Hdev, pzt, u15) and habitat (nest ht, canopy20, imperv100, plant100, imperv500, plant500, corvid, peds). We standardized all covariates prior to the 25 survival analysis and employed the logit link function for maximum likelihood estimation. Tdev was a time varying covariate where every day of the nest cycle in the design matrix corresponded to a unique daily value for each nest in the input file. Temperatures during egg laying (Cdev, pzt) only contributed to variance in daily nest survival during incubation in the models, while temperatures in the first five days after hatching (Hdev, u15) were designed to influence only nestling survival. Two additional time varying covariates were built into models using the design matrix. The day of clutch initiation for a nest relative to the start of that year’s breeding season (season) was included in a single column parameter and increased for every consecutive day in the nesting cycle to represent the progression of dates over time. All days over the entire nest cycle were assigned to a nest stage of egg laying, incubation, or nestlings (stage) for a time varying trend in the design matrix. Habitat at a nest was unchanged for the duration of that nest cycle and those variables were consistent over time. The relative fit of all models in the candidate set was assessed with an Information Theoretic approach using Akaike’s Information Criterion adjusted for small sample sizes. Higher performing models were given support by smaller ΔAICc and larger AICc weights. Temperature and habitat covariates were first modeled independently and the variables in better performing models were selected for inclusion in additive models with other variables. We also modeled a quadratic term and species interaction with individual covariates prior to selecting the variable that best represented variance in daily survival rates. Important variables were included in the same number of 26 candidate models to provide an unbiased AICc weight. Only a summary of the top models are presented in the results for simplicity. Of 371 nests observed, 54 were of an unknown age and an additional 74 failed during incubation. Prior to z-transforming the data, missing values for nests with an unknown age were assigned the population average variables that were time dependent: Tdev, Cdev, Hdev, pzt, and u15. Nests that failed before eggs hatched could not have been influenced by Hdev or u15, and those empty cells were assigned the population average as well. Average values become zero after z-transformation and so did not influence estimates of nest survival. Daily survival estimates from individual parameters of the top model were back transformed from the logit scale. Estimates are often averaged among the top competing models within 4 ΔAICc to account for model uncertainty, but the top model in our candidate set differed by more than 4 ΔAICc from the next best model. Cumulative daily survival, or nest survival probability, was calculated by raising daily survival rate to the number of days in the nesting cycle of an individual species. We calculated 95% confidence intervals for beta estimates by bootstrapping back transformed standard errors at selected user-specified values. RESULTS All values are mean ± SD unless otherwise indicated. Temperatures 27 Average daily temperatures increased from 6.9 °C (± 3.9 °C) in March to 19.6 °C (± 4.5 °C) in June over the entire four year study period. Five-day windows of temperatures during incubation and hatching were more often below the 20-yr average, indicating the study period was colder than normal. Within the study period, 2009 and 2012 were warmer springs (Figure 2). Average daily temperatures between the first and second quartiles of clutch initiation were colder for doves and robins (11.7 °C and 11.8 °C) than for grosbeaks (15.0 °C), which is not surprising since the first quartile of grosbeak nests was started 21 to 37 days later than robins and doves, respectively. Most nests were initiated within ± 5 °C of the 20-yr norm across all years and appeared to be evenly distributed around ± 0 °C of the norm (Figure 3). Habitat Most nests we observed were constructed at mid- to low-level heights relative to the upper canopy available in mature trees in the study area (11.9m ± 7m). Doves and robins built significantly higher nests (4.2m ± 1.6m and 4.0m ± 1.6m) than grosbeaks (2.9m ± 1.5m; F2,370 = 11.941, P < 0.001). Canopy cover within 20m of a nest was also significantly greater for dove nests (31.1% ± 16.8%) and robin nests (28.1% ± 18.1%) than for grosbeak nests (22.2% ± 16.5%; F2,370 = 5.050, P = 0.007). Across all species, canopy and understory vegetation was 53.0% ± 13.4% and impervious surfaces were 7.4% ± 19.8% of the land cover within 100m of nests. Pedestrians, Corvids, and Cowbirds 28 The estimated number of pedestrians walking through a site every hour ranged from none to 136 (31.4 ± 30.3). Estimated corvid density at sites ranged from 2.7 to 56.4 birds per hectare (10.5 ± 11.3). Cowbird parasitism could be reliably confirmed or rejected in only 36 of the 49 grosbeak nests. Of those nests with known presence or absence of cowbird eggs and chicks, 75.7% were parasitized. The average number of eggs in parasitized nests was 1.7 (± 0.9) and did not differ significantly between failed and fledged nests (F1,36 = 0.242, P = 0.626; Figure 4A). The Cdev during egg laying did not differ significantly between nests with one, two, three, or no cowbird eggs (F3,35 = 0.190, P = 0.902; Figure 4B). Nest Success Several variables were important for variation in daily nest survival. Both dove and robin nests were more likely to fail during the nestling stage (doves: β = -0.932 ± 0.259 SE; grosbeaks: β = -1.049 ± 0.408 SE). Temperature during the first five days after eggs hatched (Hdev) was the most important temperature variable considered when fit with a quadratic term and included with species and nest stage (Table 1). Estimates of nestling survival probabilities from hatching to fledging were lowest at temperatures -2 °C from the 20-yr norm (Figure 5). Nestling survival increased by 12% for every +1°C above the 20-yr norm and also increased at temperature anomalies below -2 °C. Substituting Hdev with slight changes in deviation from normal temperatures on any given day in the nesting period (Tdev) or during egg laying (Cdev) did not improve the model (ΔAICc = 10.89, wi = 0.004; ΔAICc = 9.35, wi = 0.009). The models with the two variables for more severe temperatures, daily maximum temperatures over 27 °C (pzt) or 29 below 15 °C (u15), did not fit the data very well (ΔAICc = 22.92, wi = 0; ΔAICc = 22.36, wi = 0). For the habitat variables considered, canopy cover in a 20m radius best explained variance in nest survival (Table 1). The cumulative probability of nest survival increased by 3% for every increase of 10% canopy cover across all species (β = 0.159 ± 0.076 SE; Figure 6A). Sites with heavier pedestrian traffic (peds) had a positive influence on nest survival and 10 more people per hour translated into a 2% increase in cumulative nest survival probability (β = 0.164 ± 0.085 SE; Figure 6B). Neither variation in corvid predator densities (corvid) nor cowbird parasitism (cow egg, cow num) had much of an influence on nest survival (corvid: ΔAICc = 21.88, wi = 0; cow egg: ΔAICc = 22.11, wi = 0; cow num: ΔAICc = 22.96, wi = 0). DISCUSSION Temperatures Extreme weather has the potential to cause mass bird mortality (McKechnie and Wolf 2010, Newton 2007), and such events are expected to increase in frequency in association with climate change. Gradual directional changes in spring temperatures could also affect nest survival, although we predicted extreme temperatures would be more important. We did not find support for an effect of temperature extremes beyond the physiological tolerances of embryo development or nestling thermoregulatory ability on nest survival of robins, doves, or grosbeaks. Instead, a slight deviation of a few 30 degrees in ambient temperatures was enough to increase the probability of nest survival. Excessive heat and cold may not have influenced nest survival in this study because days with temperatures outside the observer-defined ‘optimal’ thermal window were rare events in the study period. Facultative behavioral adjustments, such as a shift in nesting phenology cued by favorable ambient temperatures, may also minimize the frequency of extreme temperatures during the nesting cycle. Extreme temperatures are difficult to define ecophysiologically. The temperature range we considered to be within comfortable limits for eggs or nestlings, between 15 27 °C, may not necessarily be biologically meaningful. Eggs need to reach a critical temperature threshold of 24 - 27 °C to initiate embryonic development (Webb 1987), which is usually reached in altricial songbirds once the female begins complete incubation after the penultimate egg is laid (Wang and Weathers 2009). Incubation maintains egg temperature within the narrow range of 34-36 °C required for proper development. A few days of warm weather during egg laying may prematurely initiate embryonic development and expose eggs to daily minimum temperatures below the optimal range, resulting in egg inviability (Beissinger et al. 2005). However, ambient air temperatures differ from egg temperatures since egg cooling rate is partly dependent on female nest attentiveness (Morton and Pereyra 2010, Wang and Weathers 2009). As such, there is no universal ambient temperature for when eggs are at risk of developmental arrest. Similarly, critical ambient temperature thresholds for passerine nestling survival are not well established. The sensitivity of nest survival for robins, doves, and grosbeaks to slight changes in ambient air temperatures well within the range of background levels is consistent with 31 previous studies. Nests of the cavity nesting white-headed woodpecker in central Oregon were more likely to fledge young at higher daily maximum temperatures (Hollenbeck et al. 2011), as were Lewis’ woodpeckers in riparian woodlands of Idaho (Newlon and Saab 2011). Two ground nesting species, Lark Bunting (Calamospiza melanocorys) and wood lark (Lullula arborea) also had higher nest survival at warmer average temperatures over the breeding season (Skagen and Adams 2012, Wright et al. 2009). Only mountain plovers (Charadrius montanus) breeding in eastern Colorado have been reported to increase nest survival at cooler temperatures (Drietz et al. 2012). Weather during the nesting cycle may also have only a weak effect or no effect on productivity (Bradbury et al. 2003), suggesting that the magnitude and direction of sensitivity may be related to species traits or habitat. A few degrees of temperature deviation made a significant difference in nest survival, but it is not clear why. Any negative consequences from cold or heat, such as nestling hypothermia or heat stress, would have resulted in higher nest survival at normal temperatures, the opposite of the quadratic trend we observed. Estimates of nest survival increased mostly at warmer anomalies, especially for robins, and there may be a greater benefit to nest survival at warmer actual temperatures with only a small benefit to nesting at colder temperatures in some circumstances. Additional data on sources of nest failure across a range of actual temperatures would be required to make an informed interpretation of the observed response. Results from these data suggest that gradual increasing spring temperatures associated with climate change may in fact benefit some species although it is possible that at some upper critical threshold, temperatures could become too warm. Increasing inter-annual temperature variability may also increase 32 productivity of some species. We were not able to observe the effects of extreme temperatures (e.g. heat wave, late spring blizzards) on nest survival, and such weather events may be detrimental to nest survival. Nest Stage Seasonal variation in temperatures during a single nest cycle may expose nest stages to different temperatures that could influence nest fate. We expected that the first five days after hatching would be a more vulnerable stage than egg laying. Nest survival was most sensitive to temperature deviation in the nestling period immediately after hatching when young nestlings are less capable of thermoregulation. Nestlings may be more vulnerable than developing embryos to temperature variability because adults with nestlings may face a trade-off between brooding young and leaving the nest to bring them food. Eggs are self-nourishing and incubating adults may be more capable of buffering eggs from suboptimal temperatures. Temperatures during egg laying could also be closer in time to the preferred short-term temperatures used as cues by females to initiate clutches (Wiebe and Gerstmar 2010). Temperatures during egg laying may still have influenced overall reproductive productivity even if we did not find evidence of an effect on survival. Deviation of egg temperatures from PZT may affect egg inviability and hatching success (Olsen et al. 2008) that reduce brood size but do not necessarily affect nest fate. Brood reduction from warmer temperatures can impact long-term population dynamics by reducing annual productivity or offspring fitness, but this is not as severe of an outcome as complete nest failure in response to temperatures at hatching. 33 Habitat Urban habitat has been shown to influence nest survival and reproduction in several ways. Air pollution can decrease nestling mass (Peach et al. 2008), traffic noise can decrease clutch size and fledging rate (Halfwerk et al. 2011), food resources may be of lower quality (Shawkey et al. 2004), and nest predator composition can change (Rodewald and Kearns 2011). For all these reasons, we expected habitat to be important for nest survival in this study, yet it was less important than temperature. The relatively low importance of habitat variables may be due to the scale at which the variables were measured. Ryder et al. (2010) found impervious surfaces and canopy cover at a scale of 1000m were more important than at finer scales down to 100m for nest survival of several species breeding within the matrix of Washington, D.C. In a study of riparian habitat adjacent to urban areas, land use within 400m of the nest was the most important factor in nest survival of Least Bell’s Vireo (Vireo bellii pusillus; Kus et al. 2008). When only one scale was considered for another riparian species, microhabitat within 5m of the nest of Yellow Warblers (Dendroica petechial) influenced nest survival (Latif et al. 2011). The importance of habitat is probably dependent on the scale and system studied. Inconsistencies in definitions of urban habitat may also contribute to mixed results on the influence of urban development on nest survival. Our study nests were located in parks that, although were surrounded by an urban matrix of land use and land cover, may have been less susceptible to disturbance than nests monitored in the matrix itself. Continuous nest monitoring at locations on an urban to rural gradient in North 34 America occurs in urban reserves similar to ours (e.g. Marzluff et al. 2007), in larger forested plots (e.g. Rodewald et al. 2011), and within backyards contained in the matrix (Ryder et al. 2010). Inferring generalizations about the response of reproductive output to urban development needs to consider the context in which the study sites are described. Urban habitat did influence nest survival in this study through variation in canopy cover and pedestrian traffic. Canopy cover within 20m of the nest was positively associated with a higher nest survival probability for all species. The benefit could be a function of a number of factors, including the moderation of nest site temperatures. Canopy cover, understory vegetation, and lawn generally cool air (Jenerette et al. 2007), and some species may benefit from cooler nest temperatures at warmer times of the season. Canopy cover can also increase nest survival by mitigating predation risk. Nests in denser woody vegetation require more search effort from a predator to find than a nest in an isolated tree or shrub (Chalfoun and Martin 2009). Some species may increase nest survival with more protective cover at the nest (Goddard and Dawson 2009) and some species select nest sites with more cover (Chalfoun and Martin 2007, Kolada et al. 2009). The other important habitat factor that probably had more to do with predation risk than temperature moderation was pedestrian traffic at a study site. Nest failure increased in towns with more pedestrians at multiple sites across Europe (Jokimaki et al. 2005), which is different from the results of this study that sampled pedestrian activity at much finer spatial scales. In general, predator abundance increases in urban areas even as predation rates decrease, posing a predation paradox (Fischer et al. 2012, Rodewald et al. 2011). Urban areas may be a safe refuge for bird species able to tolerant human 35 disturbance because of a disproportionate negative effect of pedestrians on predators (Møller 2012). The sites in our study with the most recreational users also had the least amount of woody vegetation, which may normally serve as refugia for mammalian mesopredators. The absence of suitable habitat and negative interactions with people may lead to predator interference in the most urban sites. Conclusions Small deviations from baseline ambient temperatures during the nestling stage were important for the nest survival of robins, doves, and grosbeaks, but it is not certain if the response is ecologically significant enough to impact the population dynamics of these or other species. Any true influence of temperature on population dynamics may be masked by facultative adjustments to compensate for warming springs, like phenological shifts in migration (Both et al. 2010, Møller et al. 2008) or breeding (Both et al. 2006), or by selecting nest sites with more favorable microclimates. The simultaneous modification of the local thermal and structural environment in urban and residential areas will likely influence species differently than populations responding to climate change in less developed habitats. The synergistic effects of land use change and climate change are gaining recognition for their potential to influence species distributions and extinction risk (Jongsomjit et al. 2013). Both contributed to estimates of nest success and our results were surprising in that air temperatures were more important for nest survival than metrics of the urban environment that we sampled. Low impact recreation by pedestrians did not appear to be negatively associated with nest survival for the more synanthropic species we studied. Other positive aspects of urban habitat may benefit species that are more sensitive to human disturbance, suggesting a high potential conservation value of urban parks and open spaces. 36 37 LITERATURE CITED Aldredge, R.A., LeClair, S.C., Bowman, R. 2012. Declining egg viability explains higher hatching failure in a suburban population of the threatened Florida scrub-jay Aphelocoma coerulescens. Journal of Avian Biology 43: 369–375. AMEC Environment and Infrastructure. 2012. Urban tree canopy assessment Truckee Meadows (Reno-Sparks, Nevada). 53pp. Balogh, A.L., Ryder, T.B., Marra, P.P. 2011. Population demography of Gray Catbirds in the suburban matrix: sources, sinks and domestic cats. Journal of Ornithology 152: 717–726. Beissinger, S., Cook, M., Arendt, W. 2005. The shelf life of bird eggs: Testing egg viability using a tropical climate gradient. Ecology 86: 2164–2175. Both, C., Bouwhuis, S., Lessells, C.M., Visser, M.E. 2006. Climate change and population declines in a long-distance migratory bird. Nature 441: 81–83. Both, C., van Asch, M., Bijlsma, R.G., van den Burg, A.B., Visser, M.E. 2009. Climate change and unequal phenological changes across four trophic levels: constraints or adaptations? Journal of Animal Ecology 78: 73–83. Both, C., Van Turnhout, C.A.M., Bijlsma, R.G., Siepel, H., Van Strien, A.J., Foppen, R.P.B. 2010. Avian population consequences of climate change are most severe for long-distance migrants in seasonal habitats. Proceedings of the Royal Society BBiological Sciences 277: 1259–1266. Bradbury, R., Wilson, J., Moorcroft, D., Morris, A., Perkins, A. 2003. Habitat and weather are weak correlates of nestling condition and growth rates of four UK farmland passerines. Ibis 145: 295–306. 38 Buckland, S.T. 2001. Introduction to distance sampling: estimating abundance of biological populations. Oxford University Press, 432 pp. Chalfoun, A.D. and Martin, T.E. 2007. Assessments of habitat preferences and quality depend on spatial scale and metrics of fitness. Journal of Applied Ecology 44: 983– 992. Chalfoun, A.D. and Martin, T.E. 2009. Habitat structure mediates predation risk for sedentary prey: experimental tests of alternative hypotheses. Journal of Animal Ecology 78: 497–503. Chamberlain, D.E., Cannon, A.R., Toms, M.P., Leech, D.I., Hatchwell, B.J., Gaston, K.J. 2009. Avian productivity in urban landscapes: a review and meta-analysis. Ibis 151: 1–18. Clavero, M., Villero, D., Brotons, L. 2011. Climate change or land use dynamics: do we know what climate change indicators indicate? PLoS ONE 6. Cox, W.A., Thompson, III, F.R., Reidy, J.L., Faaborg, J. 2013. Temperature can interact with landscape factors to affect songbird productivity. Global Change Biology 19: 1064–1074. Cox, W.A., Thompson, III, F.R., Root, B., Faaborg, J. 2012. Declining Brown-Headed Cowbird (Molothrus ater) populations are associated with landscape-specific reductions in brood parasitism and increases in songbird productivity. PLoS ONE 7. Diffenbaugh, N.S., Giorgi, F., Pal, J.S. 2008. Climate change hotspots in the United States. Geophysical Research Letters 35. 39 Dreitz, V.J., Conrey, R.Y., Skagen, S.K. 2012. Drought and cooler temperatures are associated with higher nest survival in Mountain Plovers. Avian Conservation and Ecology 7. Fischer, J.D., Cleeton, S.H., Lyons, T.P., Miller, J.R. 2012. Urbanization and the predation paradox: The role of trophic dynamics in structuring vertebrate communities. Bioscience 62: 809–818. Flanders-Wanner, B., White, G., McDaniel, L. 2004. Weather and prairie grouse: dealing with effects beyond our control. Wildlife Society Bulletin 32: 22–34. Goddard, A.D. and Dawson, R.D. 2009. Seasonal changes in habitat features influencing nest survival of sharp-tailed grouse in northeastern British Columbia, Canada. Ecoscience 16: 476–482. Gottschalk, T.K. and Huettmann, F. 2011. Comparison of distance sampling and territory mapping methods for birds in four different habitats. Journal of Ornithology 152: 421–429. Greno, J.L., Belda, E.J., Barba, E. 2008. Influence of temperatures during the nestling period on post-fledging survival of great tit Parus major in a Mediterranean habitat. Journal of Avian Biology 39: 41–49. Grimmond, C.S.B. and Oke, T.R. 1999. Heat storage in urban areas: Local-scale observations and evaluation of a simple model. Journal of Applied Meteorology 38: 922–940. Halfwerk, W., Holleman, L.J.M., Lessells, C.M., Slabbekoorn, H. 2011. Negative impact of traffic noise on avian reproductive success. Journal of Applied Ecology 48: 210– 219. 40 Hannon, S.J., Wilson, S., McCallum, C.A. 2009. Does cowbird parasitism increase predation risk to American redstart nests? Oikos118: 1035–1043. Hepp, G.R. and Kennamer, R.A. 2012. Warm is better: Incubation temperature influences apparent survival and recruitment of Wood Ducks (Aix sponsa). PLoS ONE 7. Hollenbeck, J.P., Saab, V.A., Frenzel, R.W. 2011. Habitat suitability and nest survival of White-Headed Woodpeckers in unburned forests of Oregon. Journal of Wildlife Management 75: 1061–1071. Jenerette, G.D., Harlan, S.L., Brazel, A., Jones, N., Larsen, L., Stefanov, W.L. 2007. Regional relationships between surface temperature, vegetation, and human settlement in a rapidly urbanizing ecosystem. Landscape Ecology 22: 353–365. Jokimaki, J., Kaisanlahti-Jokimaki, M., Sorace, A., Fernandez-Juricic, E., RodriguezPrieto, I., Jimenez, M. 2005. Evaluation of the “safe nesting zone” hypothesis across an urban gradient: a multi-scale study. Ecography 28: 59–70. Jongsomjit, D., Stralberg, D., Gardali, T., Salas, L., Wiens, J. 2013. Between a rock and a hard place: the impacts of climate change and housing development on breeding birds in California. Landscape Ecology 28: 187–200. Kolada, E.J., Sedinger, J.S., Casazza, M.L. 2009. Nest site selection by Greater SageGrouse in Mono County, California. Journal of Wildlife Management 73: 1333–1340. Kus, B.E., Peterson, B.L., Deutschman, D.H. 2008. A multiscale analysis of nest predation on Least Bell’s Vireos (Vireo bellii pusillus). Auk 125: 277–284. Latif, Q.S., Heath, S.K., Rotenberry, J.T. 2011. An `ecological trap’ for yellow warbler nest microhabitat selection. Oikos 120: 1139–1150. 41 Lurgi, M., Lopez, B.C., Montoya, J.M. 2012. Novel communities from climate change. Philosophical Transactions of the Royal Society B-Biological Sciences 367: 2913– 2922. Lynn, S., Morrison, M., Kuenzi, A., Neale, J., Sacks, B., Hamlin, R., Hall, L. 1998. Bird use of riparian vegetation along the Truckee River, California and Nevada. Great Basin Naturalist 58: 328–343. Mantyka-Pringle, C.S., Martin, T.G., Rhodes, J.R. 2012. Interactions between climate and habitat loss effects on biodiversity: a systematic review and meta-analysis. Global Change Biology 18: 1239–1252. Marzluff, J.M., Withey, J.C., Whittaker, K.A., Oleyar, M.D., Unfried, T.M., Rullman, S., DeLap, J. 2007. Consequences of habitat utilization by nest predators and breeding songbirds across multiple scales in an urbanizing landscape. Condor 109: 516–534. McKechnie, A.E. and Wolf, B.O. 2010. Climate change increases the likelihood of catastrophic avian mortality events during extreme heat waves. Biology Letters 6: 253–256. Møller, A.P., Fiedler, W., Berthold, P. eds. 2010. Effects of climate change on birds. Oxford University Press, New York, 321pp. Møller, A.P. 2012. Urban areas as refuges from predators and flight distance of prey. Behavioral Ecology 23: 1030–1035. Møller, A.P., Rubolini, D., Lehikoinen, E. 2008. Populations of migratory bird species that did not show a phenological response to climate change are declining. Proceedings of the National Academy of Sciences of the United States of America 105: 16195–16200. 42 Morton, M.L. and Pereyra, M.E. 2010. Development of incubation temperature and behavior in thrushes nesting at high altitude. Wilson Journal of Ornithology 122: 666–673. Mujahid, A. and Furuse, M. 2009. Oxidative damage in different tissues of neonatal chicks exposed to low environmental temperature. Comparative Biochemistry and Physiology A- Molecular and Integrative Physiology 152: 604–608. Newlon, K.R. and Saab, V.A. 2011. Nest-site selection and nest survival of Lewis’s woodpecker in aspen riparian woodlands. Condor 113: 183–193. Newton, I. 2007. Weather-related mass-mortality events in migrants. Ibis 149: 453–467. Oke, T. 1988. The urban energy-balance. Progress in Physical Geography 12: 471–508. Olsen, B.J., Felch, J.M., Greenberg, R., Walters, J.R. 2008. Causes of reduced clutch size in a tidal marsh endemic. Oecologia 158: 421–435. Peach, W.J., Vincent, K.E., Fowler, J.A., Grice, P.V. 2008. Reproductive success of house sparrows along an urban gradient. Animal Conservation 11: 493–503. Pereyra, M. and Morton, M. 2001. Nestling growth and thermoregulatory development in subalpine Dusky Flycatchers. Auk 118: 116–136. Pickett, S.T.A., Cadenasso, M.L., Grove, J.M., Boone, C.G., Groffman, P.M., Irwin, E., Kaushal, S.S., Marshall, V., McGrath, B.P., Nilon, C.H., Pouyat, R.V., Szlavecz, K., Troy, A., Warren, P. 2011. Urban ecological systems: Scientific foundations and a decade of progress. Journal of Environmental Management 92: 331–362. Rippon, R.J., Alley, M.R., Castro, I. 2011. Causes of mortality in a nestling population of free-living hihi (stitchbird-Notiomystis cincta). New Zealand Journal of Zoology 38: 207–222. 43 Robb, G.N., McDonald, R.A., Chamberlain, D.E., Bearhop, S. 2008. Food for thought: supplementary feeding as a driver of ecological change in avian populations. Frontiers in Ecology and the Environment 6: 476–484. Rodewald, A.D. and Kearns, L.J. 2011. Shifts in dominant nest predators along a rural-tourban landscape gradient. Condor 113: 899–906. Rodewald, A.D., Kearns, L.J., Shustack, D.P. 2011. Anthropogenic resource subsidies decouple predator-prey relationships. Ecological Applications 21: 936–943. Ryder, T.B., Reitsma, R., Evans, B., Marra, P.P. 2010. Quantifying avian nest survival along an urbanization gradient using citizen- and scientist-generated data. Ecological Applications 20: 419–426. Schlesinger, M.D., Manley, P.N., Holyoak, M. 2008. Distinguishing stressors acting on land bird communities in an urbanizing environment. Ecology 89: 2302–2314. Seto, K.C., Gueneralp, B., Hutyra, L.R. 2012. Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools. Proceedings of the National Academy of Sciences of the United States of America 109: 16083–16088. Shawkey, M.D., Bowman, R., Woolfenden, G.E. 2004. Why is brood reduction in Florida scrub-jays higher in suburban than in wildland habitats? Canadian Journal of Zoology-Revue Canadienne de Zoologie 82: 1427–1435. Skagen, S.K. and Adams, A.A.Y. 2012. Weather effects on avian breeding performance and implications of climate change. Ecological Applications 22: 1131–1145. Smith-Castro, J.R. and Rodewald, A.D. 2010. Effects of recreational trails on Northern Cardinals (Cardinalis cardinalis) in forested urban parks. Natural Areas Journal 30: 328–337. 44 Thorington, K.K. and Bowman, R. 2003. Predation rate on artificial nests increases with human housing density in suburban habitats. Ecography 26: 188–196. Van Heezik, Y., Ludwig, K., Whitwell, S., McLean, I.G. 2008. Nest survival of birds in an urban environment in New Zealand. New Zealand Journal of Ecology 32: 155– 165. Vargo, J., Habeeb, D., Stone, Jr., B. 2013. The importance of land cover change across urban-rural typologies for climate modeling. Journal of Environmental Management 114: 243–252. Virkkala, R., Heikkinen, R.K., Leikola, N., Luoto, M. 2008. Projected large-scale range reductions of northern-boreal land bird species due to climate change. Biological Conservation 141: 1343–1353. Visser, M.E., van Noordwijk, A.J., Tinbergen, J.M., Lessells, C.M. 1998. Warmer springs lead to mistimed reproduction in great tits (Parus major). Proceedings of the Royal Society of London Series B-Biological Sciences 265: 1867–1870. Wang, J.M. and Weathers, W.W. 2009. Egg laying, egg temperature, attentiveness, and incubation in the western bluebird. Wilson Journal of Ornithology 121: 512–520. Webb, D. 1987. Thermal tolerance of avian embryos – a review. Condor 89: 874–898. White, G. and Burnham, K. 1999. Program MARK: survival estimation from populations of marked animals. Bird Study 46: 120–139. Wright, L.J., Hoblyn, R.A., Green, R.E., Bowden, C.G.R., Mallord, J.W., Sutherland, W.J., Dolman, P.M. 2009. Importance of climatic and environmental change in the demography of a multi-brooded passerine, the woodlark Lullula arborea. Journal of Animal Ecology 78: 1191–1202. 45 TABLE 1. A summary of logistic exposure nest survival models for three species along an urban to rural gradient in Reno and Sparks, Nevada. Models include covariates and are compared to the null model of constant survival S(.). Individual covariates included are the temperature deviation from 20-yr normal at hatching with a quadratic term (hdev2), amount of canopy cover in a 20m radius (canopy 20), and rates of pedestrians at a study site (peds). An interaction term for species and nest stage (stage*species) was also included to represent three stages of nesting that differ in daily survival rates. The lowest AICc value of all candidate models considered was 1444.39. Model 2 species + stage*species + Hdev + Hdev + peds + canopy20 1 2 K 3 Deviance 0.923 10 1383.22 5.76 0.052 8 1392.99 9.12 0.010 8 1396.36 9.35 0.009 8 1396.58 10.89 0.004 8 1398.12 species + stage*species 15.23 0.000 6 1406.47 species + stage 21.00 0.000 4 1416.25 species 21.06 0.000 3 1418.32 S(.) 34.70 0.000 1 1435.95 year + site 41.14 0.000 10 1424.36 2 species + stage*species + peds + canopy20 species + stage*species + Cdev + Cdev species + stage*species + Tdev + Tdev 2 2 Measure of a model distance in AIC units to the best model Model weight 3 Number of parameters included in the model 4 Measure of model fit to the data 2 wi 0.00 species + stage*species + Hdev + Hdev 1 ΔAICc 4 46 FIGURE 1. Study area location (star) in the western United States. Enlarged area of Reno and Sparks, Nevada with study sites (black) connected by the Truckee River (black line). Location of the Reno/Cannon International Weather Station (Co-op ID: 266779) is marked by the +. Enlarged images of two study sites demonstrate the within-site variation in land cover and among-site variation along the urban to rural gradient. 47 FIGURE 2. Weekly averages of daily maximum temperatures in the study area during the breeding season for 2009 – 2012. Upper horizontal dashed line marks a temperature threshold of 27 °C, the upper boundary of physiological zero. Lower horizontal dashed line marks the temperature threshold for 15 °C, below which nestlings are likely to be negatively impacted by cold. FIGURE 3. Temperature deviation from the 20-yr norm at the time of clutch initiation for three bird species nesting in urban habitat over a four-year study period. 48 49 FIGURE 4A-B. Brood parasitism of grosbeak nests by Brown-headed Cowbirds in A) 36 grosbeak nests that either fledged young or failed, and B) the number of cowbird eggs deposited at a range of temperature deviations from the 20-yr normal for the five days during egg laying. 50 FIGURE 5. Fitted curve of nestling survival probability estimates (± SE) for temperature deviations from normal after hatching for three species breeding in urban habitat over a four-year study period. 51 FIGURE 6A-B. Differences in the amount of canopy cover within a 20m radius, and rates of pedestrian traffic, between the nests of three species that fledged young or failed in urban habitat over a four-year study period. 52 NEST HABITAT SHIFTS TO SUITABLE MICROCLIMATES IN RESPONSE TO WARMER TEMPERATURES MILES E. BECKER AND PETER J. WEISBERG ABSTRACT Bird populations at risk of losing access to their historic bioclimatic envelope may benefit from increased spatial heterogeneity of within-territory microclimates, provided those microclimates function to buffer air temperatures and they are used as thermal refuges. We monitored the nests of three species in an arid ecosystem to investigate the relationship between spatial microclimate heterogeneity, variation in air temperature, and shifts in nest locations. Microclimate was recorded in spring and summer at sites surrounded by four land cover types with different biophysical properties in urban habitat. The same four habitat variables were sampled at nests over a four year monitoring period. Microsites with more lawn and understory vegetation had cooler daily minimum temperatures. Lower levels of canopy cover and higher levels of bare ground increased daily maximum temperatures. High levels of impervious surface moderated temperatures by decreasing daily maximum and increasing daily minimum temperatures. Habitat within a 10m radius explained more of the variance in microclimate than at scales of 20m and 100m. Two generalist species shifted nest site microhabitat at warmer temperatures before clutch initiation to locations with less bare ground and more impervious surfaces. Nest habitat use was apparently sensitive to air temperatures, suggesting birds are able to compensate for unfavorable temperatures by using appropriate microclimates. More bird populations are likely to experience unfavorable regional temperatures due to climate change, increasing the importance of conserving spatial heterogeneity in habitat features that provide a broad range of microclimates. 53 54 INTRODUCTION Temperature is a defining characteristic of the fundamental niche of a species (Kearney and Porter 2004, Monahan 2009) and often a primary influence on local habitat use (Monasterio et al. 2009, Speed et al. 2012). At coarse regional scales, many species are projected to lose their preferred thermal envelope from their current range due to global climate change (Domisch et al. 2013, Keith et al. 2008, Thuiller et al. 2005), even within the boundaries of currently protected reserves (Ackerly et al. 2010, Araujo et al. 2011, Virkkala et al. 2013). Broadscale climate modeling overlooks microhabitat and microclimate that may be especially important for species with limited ability to shift range boundaries (Scherrer and Koerner 2010). As temperature resources shrink or shift, spatial thermal heterogeneity at fine scales of less than 100m may facilitate species persistence in current ranges by providing temporary in situ thermal refuges. Spatial heterogeneity is advantageous for population stability in theory (Oliver et al. 2010, Ruokolainen et al. 2011), but the extent to which at-risk populations make use of thermal refuges through facultative behavioral adjustments is not well known. Identifying fine-scale habitat features that correspond to microclimatic value is a challenge. A habitat patch or individual structure used as a thermal refuge must differ from surrounding temperatures and be selected by an organism for its thermal advantages (Keppel and Wardell-Johnson 2012). Thus, there are two steps to identifying thermal refuges. First, spatial variation in microclimate needs to be explained by habitat heterogeneity. Spatial heterogeneity in temperature can be related to distance from water, topography, wind exposure, vegetation structure, and habitat type (Ashcroft et al. 2009, 55 Holden et al. 2010, Suggitt et al. 2011). Relative spatial differences in temperature may be more pronounced in warmer seasons and during the day than at night (Suggitt et al. 2011), potentially creating a dynamic thermal value for static structural features. An equally necessary second step is to link thermal heterogeneity to resource selection by the species of interest. Many bird populations are declining in association with climate change (Both et al. 2010, Møller et al. 2008) and some are expected to lose their bioclimatic envelope from their current range (Virkkala et al. 2008). Birds may use behavioral thermoregulation to avoid heat stress, such as bobwhites resting at cooler microsites during the day in summer (Hiller and Guthery 2005), or to reduce the cost of thermoregulation in winter, as suggested by increased bird abundance in forests with more solar radiation penetrating the canopy (Huertas and Diaz 2001). Selecting habitat with favorable microclimate can have fitness consequences for survival of adults (Patten et al. 2005) that may ultimately impact population dynamics. Besides adult survival, another avian vital rate likely to be influenced by microclimate is reproduction. Hatching success and nest fate are associated with air temperatures in some species (Aldredge et al. 2012, Greno et al. 2008, Chapter 1). Nest site selection may be especially sensitive to microclimate because the location of a nest is immobile over a range of temperatures during the 30 to 80 day nest cycle. Studies on nest site microclimate are often limited to considerations of nest position and structure. Nest orientation can influence daytime temperatures at the nest (Lloyd and Martin 2004) and cavity nesters may select warmer cavities (Maziarz and Wesolowski 2013, Rhodes et al. 2009) or use more nest insulation at colder temperatures (Deeming et al. 2012). 56 Vegetation immediately adjacent to or in contact with the nest can shield it from wind or solar radiation (Kim and Monaghan 2005, Tieleman et al. 2008). Examples of preferences for warmer nest structures are evidence of microclimate management at the level of the nest itself. Open-cup nesters are additionally expected to control microclimate to some extent by adjusting nest location in the surrounding land cover, yet empirical evidence for shifts in nest site selection in response to temperature is surprisingly lacking from the extensive body of literature on bird-habitat associations. The aim of this study was to identify resource selection of thermal refuges for nest sites. We first generated a series of hypotheses and predictions about the potential influence of microclimate on nest habitat over the season that could be distinguished from the effects of other factors. We then established the relationship between microclimate and microsite with fine-scale temperature data loggers. To assess the use of microclimates over time, we sampled nest habitat for three songbird species over four breeding seasons in a region with high climatic variability that is forecasted to become warmer. METHODS Hypothesized Effects There are several ways in which temperature can influence nest habitat use independently or in addition to other factors driving nest site selection. We generated several hypothesized relationships (Figure 1) where a change in nest habitat with air temperature would be observed. To facilitate the isolation of temperature influences from other factors, the breeding season was divided into three time periods of early, mid-, 57 or late season nesters. Due to the highly variable spring temperatures in the study area, it was possible for late nesters to experience colder temperatures at the time of clutch initiation than nests started early in the season. As a result, we were effectively able to separate out temperature effects from other time varying influences on nest habitat selection, like predation risk or food availability. The null hypothesis was that nest habitat did not change over the season or with temperature (Figure 1A). Nest locations may have shifted over time but not a temperature range (Figure 1B), or the opposite could also occur (Figure 1C). Both time and temperature could contribute an additive effect to habitat use (Figure 1D). Temperature effects could be dependent on the timing of nesting and result in an interaction that obscures the individual effects of each (Figure 1E) or increases the strength of the association between temperature and nest habitat shifts (Figure 1F). We expected any constraints on using nest habitat for optimal microclimate from other factors, like low predation risk, to be manifested as an interaction illustrated in Figure 1E and 1F. Any temperature effects should stand out from the effects of other time varying factors, provided potentially confounding factors are uncorrelated with temperature. For example, canopy cover may prevent nests from overheating in hot weather while simultaneously reducing the risk of predators detecting the nest. If predator activity also increases when it is hot, then the effects of the two variables cannot be independently quantified. However, identical responses to multiple stressors are unlikely to occur across all land cover variables considered. Study area 58 The study area in northern Nevada is on the border of the Great Basin and eastern Sierra Nevada where there are four distinct seasons. The regional climate is characterized by cold winters (January high of 7° C) and hot summers (July high of 33° C) with most precipitation falling as snow in winter. Intra-seasonal temperatures during the breeding season in early spring and summer can be unpredictable over one to two weeks in a single year (Figure 2). The study sites pass through Reno and the adjoining city of Sparks, two mid-sized cities that supported a combined population of 315,485 people in 2010 (US Census 2010). Rapid outward growth from the city centers has led to the expansion of suburbs and housing developments into sagebrush habitat adjacent to riparian woodlands along the Truckee River that runs west to east through town. The urban ecosystem is particularly well suited to studies on microclimate use because the habitat is highly heterogeneous at fine-scales, increasing the opportunity for microclimate heterogeneity at the territory level. There are also strong barriers to dispersal from reserves in urban habitat for birds and other species (Delaney et al. 2010, Kennedy and Marra 2010, Tremblay and St Clair 2011) that prevent an individual from tracking its preferred thermal niche across the landscape and perhaps strengthen pressure on behavioral thermoregulation. Study sites were along a 15.7 km west to east transect that intersected rural, suburban, and urban zones. We selected eight municipal parks (mean size = 9.9 ha) with some native riparian vegetation next to the Truckee River and maintained by local municipalities for public use. The heterogeneity of fine-scale land cover types within the parks provided a wide selection of diverse microsites in a small geographic area. 59 Microclimate data We deployed 95 iButton temperature data loggers (Dallas Semiconductor/MAXIM, Dallas, TX, USA) in eight study sites in spring and summer 2011 to encompass the majority of the breeding season for my study species. The time period analyzed was 12 weeks from April 21 to July 13. Many of the deployed data loggers were missing or moved at the time of retrieval and only 48 were used for the microclimate analysis. All iButtons were placed 1.5m off the ground and under the bark on the NNW to NNE facing side of a tree. Data loggers oriented north on a tree are shielded from direct radiation (Holden et al. 2010). Air temperatures were recorded at 4hr intervals and used to calculate the temperature variables of daily temperature maximum (Tmax) and daily temperature minimum (Tmin), which are more likely than mean temperature to be influenced by local habitat (Suggitt et al. 2011). Tmax and Tmin are also more biologically meaningful metrics because organisms experience a full range of temperatures in a 24-hr period, not the mean. Only one year of microsite temperatures was recorded even though four years of microsite nest habitat data were sampled. There were no a-priori reasons to expect the relationship between microhabitat and microclimate in 2011 to change in other years, although the magnitude may have changed. Temperatures in the study area recorded by the local National Weather Station tower showed strong inter-annual temperature differences over the 4-yr study period (Figure 2). Two years, 2010 and 2011, were colder springs and microsite temperatures may have differed less but any thermal refuges should serve the same function from year to year. 60 The habitat variables we measured at each iButton location that may influence microclimate were land cover (bare ground, impervious surface, canopy cover, and understory vegetation) and distance to river. Land cover was calculated for 10m, 20m, and 100m radius buffer centered on the sampling point using ArcGIS software (ArcMap 10, ESRI, Redlands, California, USA). A base map of 1m resolution orthophotographs captured in summer 2010 by the National Agriculture Imagery Program (NAIP) was used to classify land cover types with a geographic object-based image analysis. A point-based accuracy assessment of 273 randomly selected points yielded an overall accuracy of 94% (AMEC 2012). Water surfaces were absent from almost all locations at the 10m and 20m scales, and we excluded this land cover type from the analyses. The west to east slope over the entire study area was 0.4% and every study site was essentially flat, so no topographical features were used as predictor variables. Nest site selection The nests of three native species, American Robin (Turdus migratorius, hereafter robin), Black-headed Grosbeak (Pheucticus melanocephalus), grosbeak), and Mourning Dove (Zenaida macroura, dove) were observed in seven study sites for four years from 2009 to 2012. Robins build heavy mud and grass nests in a variety of substrates, including human structures. Doves are also nest habitat generalists and construct nests made of a weak twig platform. Grosbeaks are riparian specialists and have the narrowest breadth of nest site substrates of the three study species. In our study sites, they invariably build nests in dense vegetation close to the river. Doves begin breeding the earliest in late March, robins in mid- to late-April, and grosbeaks in mid-May. 61 We started searching for nests in late March to early April at the start of the breeding season. Only active nests – indicated by an incubating female, the presence of eggs or nestlings, or parents delivering food to the nest – and nests with a known clutch initiation date were considered for analysis (N = 320). The same habitat variables sampled at iButton locations were calculated for each nest at the same three scales. An additional habitat variable, nest height, was included since vertical variation may represent a gradient of predation risk (van Heezik et al. 2008), human disturbance (Smith-Castro and Rodewald 2010), or air temperature (Suggitt et al. 2011). Predator Activity Birds are a major nest predator in urban areas (Marzluff et al. 2007, Thorington and Bowman 2003). We counted corvids along transects at each of the study sites for an index of predator activity at different air temperatures. The perpendicular distance from an observer on line transects through the entire length of each site to a detected bird or predator was measured with a laser rangefinder (Sport 450, Bushnell, Overland Park, Missouri, USA). Observations of corvids were made between 0630 – 1000 and the sampling order between sites was rotated at every visit to avoid a temporal bias. We repeated transects every five to eight days for eight replicates per site. The number of corvids observed at a site was divided by the area (ha) of the site for an activity index. We were interested in within-site variation in corvid counts over the eight survey intervals, which assumes that detection probability does not change with time but allows for different detection probabilities between sites. The transect survey period started after the breeding season when territories were established and ended 62 before most nests fledged, reducing the chance that differences in corvid counts were due to movement in or out of the study site or the appearance of juveniles. ANALYSES Microclimate We analyzed changes in temperatures at iButton locations over time and by cover type in a repeated measures generalized linear model with weekly intervals as the withinsubjects factor and land cover category or distance to river as the between subjects factor. Separate models were run for the four types of land cover at the three spatial scales and for two temperature response variables, Tmax and Tmin. Daily Tmax and Tmin were averaged over the seven-day time interval. We assigned treatment levels of a cover type to each iButton by binning values into three categorical groups of low, mid-level, and high using the 33rd, 66th, and 100th population percentiles as treatment threshold values. If at least 20% of the iButtons had a value of zero for that cover type, then a 4th group of “none” was added and the percentiles calculated from the remaining non-zero values. At least half of the iButtons had no bare ground or impervious surfaces at 10m and 20m, and only the non-zero values were included in the analyses to maintain a balanced design. Distance to river was similarly binned into three levels of near, mid-range, and far. Nest Site Selection The analysis of nest site selection treated temperature and time as predictor variables and microsite features at the nest as continuous response variables. Daily Tmax 63 values in the study area for five days before clutch initiation (Tmax5) were binned into three fixed treatment levels of temperature: cold, mild, and warm. Threshold values between temperature levels were species-specific and evenly divided over the temperature range experienced at the start of nesting over the entire study period (Table 1). Five to seven day temperature ranges have been shown to be good predictors of clutch initiation and nest structure in other species (Wiebe and Gerstmar 2010, Deeming et al. 2012). Ambient Tmax5 and Tmin5 in the study area were highly correlated (Pearson’s = 0.879, P < 0.001), so only Tmax5 was used as a predictor variable. Tmax temperatures were provided by data from the Reno/Tahoe National Weather Station, located close to the study sites. Three treatment levels of timing of nesting (early, mid, late) were calculated from percentiles by species and year as a relative measure of when a bird began nesting in that year compared to conspecifics (Table 1). Variation in the correlated multiple response variables at a nest was modeled in a 2 x 2 MANOVA. The assigned levels of temperature and timing of breeding were species-specific and effectively standardized treatment levels across species. An initial model with species as a fixed factor did not show any interaction between species and time (F20,963 = 1.361, P = 0.133) or species and Tmax5 (F20,963 = 1.152, P = 0.289). The similarity in robin and dove nest habitat and distinctly different grosbeak nest habitat (Table 1) was supported by a post-hoc Tukey’s HSD multiple comparisons test. Due to the smaller number of nests sampled for grosbeaks and their distinct difference from the other species, we modeled their nest selection separately and pooled robin and dove nests together in the same model. The statistical approach requires independence between samples in each of the treatments, and it is possible that nests initiated later in the season 64 were re-nesting attempts by the same unmarked female. We avoided this potential bias by not including multiple nesting attempts in the same territory. Predator Activity Corvid counts for a survey interval were paired with ambient Tmax in the study area on that day. Temperatures over the range of those experienced during predator monitoring transects were binned into cold, mild, and warm groups following the same threshold values as temperatures used for predicting nest site selection by the study species. Within-site differences in corvid counts at each interval and temperature were analyzed in a repeated measures ANOVA with the activity index as a response variable, temperature groups as the treatment level, and transect interval as the repeated measure. RESULTS All values are mean ± SD, unless otherwise noted. Microclimate Microsites with iButtons showed strong seasonal variation in Tmax and Tmin in 2011. Maximum daily temperatures increased from 13.8 ± 2.2 °C in the fourth week of April to 27.0 ± 5.0 °C in the second week of July. Tmin also increased over the same time period from 4.7 ± 1.4 °C to 13.8 ± 2.4 °C. There was substantial spatial variation in microclimate. Within a single study site in April, Tmax differed between iButtons by 4.3 ± 3.4 °C and by 3.3 ± 1.8 °C for Tmin. By July, spatial temperature differences within a 65 site had tripled to 14.0 ± 3.0 °C for Tmax and risen slightly to 5.7 ± 1.7 °C for Tmin. Overall, spatial variation in microclimate was amplified during the warmer part of the season and day. The propensity of microsites to gain or lose heat was related to surrounding habitat. Land cover types functioned differently in regulating microclimate. The most abundant land cover type was vegetation (understory shrubs or lawn). Within a 20m radius, vegetation did little to change daily Tmax (F2,45 = 1.999, P = 0.147), but microsites with vegetation cover above 30% were cooler at night (F2,45 = 3.776, P = 0.030; Figure 3A). Impervious surfaces was also associated with Tmin warming at night for surface areas covered by over 24% (F2,22 = 3.551, P = 0.046) but the relationship was not significant for Tmax (F2,22 = 2.742, P = 0.086; Figure 3B). Microsites with more than 10% canopy cover were cooler in the day (F2,45 = 4.305, P = 0.019), but canopy cover was not related to Tmin (F2,45 = 0.717, P = 0.494; Figure 3C). Within a 10m radius, bare ground was related to higher Tmax (F2,20 = 3.866, P = 0.038) but the same amounts of over 26% land cover did not have a significant relationship with Tmin (F2,20 = 1.641, P = 0.219; Figure 3D). Distance to river was not associated with Tmax or Tmin (Tmax: F2,45 = 1.04, P = 0.362; Tmin: F2,45 = 1.718, P = 0.191). The direction of thermal influence for each cover type did not change over time, but the relative magnitude of the variation did (Figures 3A-D). Microclimate variation between levels of any one land cover type was much more pronounced at air temperatures over 22 °C Tmax and 10 °C Tmin, starting by mid-June in 2011. Microhabitat at the 10m radius scale explained more variance in microclimate (10m: Wilk’s Λ = 0.20; 20m: Wilk’s Λ = 0.30; 100m; Wilk’s Λ = 0.28). The relative 66 importance of land cover differed slightly with each scale. At a 10m radius, Tmax significantly increased with bare ground and Tmin decreased with vegetation (F2,45 = 6.587, P = 0.003). The variables with significant relationships to Tmax and Tmin at a 20m radius were canopy, vegetation, and impervious surfaces. Only impervious surface significantly decreased Tmax at the broadest scale considered, a 100m radius (F2,45 = 3.25, P = 0.048). Nest site selection Over the four-year study period, we observed nests initiated from March 16 to July 2. Temperatures at the time of nesting ranged from 8.5 – 32.7 °C for Tmax and -2.2 – 14.5 °C for Tmin. Variation in microhabitat at robin and dove nests was better explained by lay date (Wilks Λ = 0.859) than by the temperatures at the time of nesting (Wilks Λ = 0.955) for all spatial scales. Late nesters were more likely to construct nests in areas with more vegetation (F2,268 = 9.690, P < 0.001; Figure 4A) and less impervious surface (F2,268 = 8.728, P < 0.001; Figure 4B) at 20m, and more bare ground (F2,268 = 5.657, P = 0.004; Figure 4C) in a 100m radius as predicted in Figure 1B. Canopy cover was consistent over time and the temperature range, matching the hypothesized relationship in Figure 1A (F2,268 = 0.312, P = 0.732; Figure 4D). For robins and doves, the variance explained by the three spatial scales of microhabitat were fairly similar (10m: Wilks Λ = 0.95; 20m: Wilks Λ = 0.94; 100m: Wilks Λ = 0.92). Although weak, the effect of temperature on nest site selection was significant for bare ground in a 100m radius (F2,268 = 5.717, P = 0.004). Independently of timing, nests started at warmer temperatures had less bare ground (Figure 4D), following the predicted 67 pattern of additive effects illustrated in Figure 1D. Nests constructed at colder temperatures were significantly more likely to have less impervious surfaces within 20m (F2,268 = 3.114, P = 0.046; Figure 4B) at least during the early and middle parts of the season, similar to the prediction of a change in the magnitude of a temperature effect illustrated in Figure 1F. Timing of breeding and temperature explained more variance in nest habitat use for grosbeaks than for robins and doves (Time: Wilks Λ = 0.660; Tmax5: Wilks Λ = 0.757). The only response variable that significantly changed over time for grosbeaks was nest height (F2,34 = 3.832, P = 0.032). Average height of nests initiated in the middle of the grosbeak season was 1.7m lower than earlier nests and 0.9m lower than nests constructed later. Predator Activity The most frequently observed corvids at the study sites were Western Scrub Jay (Aphelocoma californica), Steller’s Jay (Cyanocitta stelleri) and Black-billed Magpie (Pica hudsonia). Counts of corvids per hectare along transects in the sites were highest at cold temperatures below 19 °C (0.53 ± 0.48) and similar in mild temperatures (0.30 ± 0.21) and warm temperatures above 26 °C (0.31 ± 0.40; Figure 5). The differences were not statistically significant (F2,14 = 1.990, P = 0.174). DISCUSSION Microclimate 68 Thermal refuges were identified in this study as cooler habitat patches with less bare ground and more impervious surfaces and that were used as nest sites by doves and robins at warmer temperatures. Changes in microhabitat were associated with larger differences in microclimate at warmer temperatures. The opposite has been observed at a broader scale, where habitat heterogeneity was associated with warming of minimum temperatures and in the coldest season (Ashcroft et al. 2009). However, similar results to our study were found for open habitats with less canopy cover, and spatial variation in temperature was also more evident in warmer seasons (Suggitt et al. 2011). Bare ground and open habitat allow direct solar radiation to warm air and soil. Shading by canopy cover intercepts solar radiation and its effects are more likely to diminish at night or in regions with abundant cloud cover. Low growing vegetation, which often consisted of lawn and turf in this study, cannot effectively shade habitat patches at nest level and so did not influence Tmax. Most of the non-canopy vegetation in our study was irrigated and water retention in soils may have contributed to lower temperatures at night. Unexpectedly, the land cover type that moderated daily air temperature ranges the most was impervious surface. Light colored rooftops with high albedo have the ability to reflect light at sufficient levels to effectively cool surrounding areas (Taha 1997). Microsites with data loggers in this study were in parks and fine-scale impervious surfaces were limited to paved roads and small utility buildings. Therefore, the thermal properties of concrete and asphalt were likely responsible for the observed microclimate variation. Thermal inertia of man-made materials in cities is evident from a peak in the Urban Heat Island effect at sunset (Pickett et al. 2011), not at the time of peak air temperatures. The latent heating effect from impervious surfaces suggests they absorb 69 heat in the day, potentially creating a heat sink, and release stored heat later in the day, serving as a heat source. The amount of heat absorbed by paved surfaces before releasing heat is probably limited and the temporal dynamics of heat transfer may change at the height of summer. Land cover within a 10m radius of iButton locations explained most of the variance in microclimate. Other microclimate studies across a heterogeneous landscape did not test explicitly for scale effects, but found spatial variation in temperatures when iButtons were deployed at distances of approximately 40m (Suggitt et al. 2011) to over a kilometer (Ashcroft et al. 2009, Holden et al. 2010). The spatial scale important for microclimate at a nest will probably differ with habitat and should be evaluated prior to identifying thermal refuges. Nest sites The timing of breeding relative to other conspecific nesters explained more variance in nest site selection than did air temperature, possibly due to other time-varying influences on nesting that were not accounted for in this study. My study design was not able to rule out the possibility that nest site selection by the three species was influenced by other factors besides temperature. Nest site selection may be a response to perceived predation risk or actual nest fate (Chalfoun and Martin 2010). Nesting close to food resource rich patches may also pay off to meet the high nutrient demand by nestlings while minimizing adult energy expenditure on longer-distance foraging flights. Phainopepla (Phainopepla nitens) nested in patches with greater abundance of their preferred food, berries of the desert mistletoe (Phoradendron californicum), within a 5m 70 radius and a 11.3m radius (Crampton and Sedinger 2011). Competition for high quality habitat can also force individuals with a lower competitive ability, like juveniles or latearriving migrant breeders, into low quality habitat (Loehle 2012). Air temperature has the potential to influence predator activity and prey availability for some species, which could confound temperature effects on nest microhabitat use. Ectothermic predators, like snakes, may be more active and more of a threat to bird nests in warmer weather (Morrison and Bolger 2002). Corvids, a common nest predator in urban areas (Marzluff et al. 2007, Thorington and Bowman 2003), are locally abundant in the study area (Trammell and Bassett 2012). Detection rates of corvids in the study sites did not increase on warmer days, making it less likely that temperature influenced nest predator activity. Robins primarily forage for soil macroinvertebrates which could be less accessible at colder temperatures, but lawns and short grass are the dominant land cover in almost all robin territories, possibly sustaining a high prey abundance even at low temperatures. Doves consume seeds and other plant materials, which are available independently of short-term temperature variability, making it unlikely that food availability was confounded with temperature in nest habitat use by doves and robins. The effect of air temperature on nest site selection may have been stronger in a warmer series of years; temperatures at the time of clutch initiation during the four-year study period were cooler than the 20-yr normal (Chapter 1). Air temperatures in the study area also peak in late June and July, after nest searching ended but before the breeding season ended. In an extreme desert environment where maximum air temperatures later in the breeding season reach 47.6 °C, Hoopoe Larks (Alaemon 71 alaudipes) consistently changed nest site locations as the season progressed (Tieleman et al. 2008). Given that nest site selection for favorable microclimates may have intensified later in the summer at warmer temperatures, it would be useful to study nest microclimate over a wider range of air temperatures. Despite the effects being relatively weak, air temperatures were related to nest site selection for robins and doves. At warmer temperatures, both species were more likely to use nest sites with less bare ground land cover. At cooler temperatures, they nested at sites with less impervious surface land cover. The iButton data show that microsites with more bare ground are warmer and sites with more impervious surfaces are cooler. The results suggest robins and doves adjusted nest microclimate in response to air temperatures in addition to other factors, as predicted in the hypothesized relationships illustrated in Figure 1. Grosbeaks changed only the height of their nests in the middle of the season, the biological significance of which is difficult to determine. However, it is unlikely that grosbeaks are completely unresponsive to air temperatures. Instead, robins and doves may be able to follow their preferred thermal niche across multiple habitat structures and vegetation compositions because they are habitat generalists. Grosbeaks are habitat specialists that may be confined to one habitat type, able to capitalize on spatial variation in temperatures only from the presence of topographic heterogeneity, which did not exist in my study sites. Conclusions 72 Changing air temperatures will likely become an increasingly important constraint for birds nesting in the Great Basin, the southwestern United States, and other rapidly warming parts of the world. One solution to shrinking bioclimatic envelopes is the preservation of large areas of contiguous habitat or corridors connecting a network of smaller reserves to facilitate species dispersal and migration to preferred climatic habitat (Bernazzani et al. 2012, Hannah 2011, Nunez et al. 2013). However, such broad scale management actions may not be economically or logistically feasible for many reserves in disturbed landscapes with a matrix of other land uses, including urban habitat (Rudd et al. 2002). Instead, fine-scale in situ habitat heterogeneity and thermal refuges may be a better strategy for species persistence and ecosystem resilience (Ackerly et al. 2010). Temporary behavioral thermoregulation is apparent in ectotherms, such as reef sharks swimming to warmer waters to maintain higher metabolism (Speed et al. 2012) or ungulates like bighorn sheep that may seek out shade under tree canopies or in caves at warmer times of the day in desert environments (Cain et al. 2008). Such short-term temporary thermal refuges may influence fitness and habitat use, but microclimate may be more important for fixed nest locations that could benefit from being buffered by daily and weekly temperature fluctuations. Despite the apparent need to understand the influence of microclimate on nest habitat use, few studies have done so in the context of seasonal shifts. Two habitat generalists in this study shifted to more favorable nest microclimates at warmer temperatures, supporting the value of fine-scale habitat heterogeneity. It should be recognized that loss of suitable microclimates can result from either regional long-term climate change or localized vegetation structural changes from land use change. In this study, the maximum temperature range within a single urban 73 reserve (14 °C) was much greater than the predicted temperature change forecasted for the study area of a few degrees by the end of the century (Diffenbaugh et al. 2008). Since spatial thermal variance is so much greater than long-term annual variance, modification of land cover by land use change may be a more important ecosystem stressor than climate change. However, land use change, including urban development, also has the potential to create novel microclimates and buffer organisms against any negative impacts of climate change. Mitigation strategies that use management of local habitat to counteract climate change effects warrant further study (Clausen et al. 2013). 74 LITERATURE CITED Ackerly, D.D., Loarie, S.R., Cornwell, W.K., Weiss, S.B., Hamilton, H., Branciforte, R., Kraft, N.J.B. 2010. The geography of climate change: implications for conservation biogeography. Diversity and Distributions 16: 476–487. Aldredge, R.A., LeClair, S.C., Bowman, R. 2012. Declining egg viability explains higher hatching failure in a suburban population of the threatened Florida scrub-jay Aphelocoma coerulescens. Journal of Avian Biology 43: 369–375. AMEC Environment and Infrastructure. 2012. Urban tree canopy assessment Truckee Meadows (Reno-Sparks, Nevada). 53pp. Araujo, M.B., Alagador, D., Cabeza, M., Nogues-Bravo, D., Thuiller, W. 2011. Climate change threatens European conservation areas. Ecology Letters 14: 484–492. Ashcroft, M.B., Chisholm, L.A., French, K.O. 2009. Climate change at the landscape scale: predicting fine-grained spatial heterogeneity in warming and potential refugia for vegetation. Global Change Biology 15: 656–667. Bernazzani, P., Bradley, B.A., Opperman, J.J. 2012. Integrating climate change into habitat conservation plans under the U.S. Endangered Species Act. Environmental Management 49: 1103–1114. Cain, J.W., Jansen, B.D., Wilson, R.R., Krausman, P.R. 2008. Potential thermoregulatory advantages of shade use by desert bighorn sheep. Journal of Arid Environments 72: 1518–1525. Both, C., Van Turnhout, C.A.M., Bijlsma, R.G., Siepel, H., Van Strien, A.J., Foppen, R.P.B. 2010. Avian population consequences of climate change are most severe for 75 long-distance migrants in seasonal habitats. Proceedings of the Royal Society BBiological Sciences 277: 1259–1266. Chalfoun, A.D. and Martin, T.E. 2010. Facultative nest patch shifts in response to nest predation risk in the Brewer’s sparrow: a “win-stay, lose-switch” strategy? Oecologia 163: 885–892. Clausen, K.K., Stjernholm, M., Clausen, P. 2013. Grazing management can counteract the impacts of climate change-induced sea level rise on salt marsh-dependent waterbirds. Journal of Applied Ecology 50: 528–537. Crampton, L.H. and Sedinger, J.S. 2011. Nest-habitat selection by the phainopepla: congruence across spatial scales but not habitat types. Condor 113: 209–222. Deeming, D.C., Mainwaring, M.C., Hartley, I.R., Reynolds, S.J. 2012. Local temperature and not latitude determines the design of Blue Tit and Great Tit nests. Avian Biology Research 5: 203–208. Delaney KS, Riley SPD, Fisher RN. 2010. A rapid, strong, and convergent genetic response to urban habitat fragmentation in four divergent and widespread vertebrates. PLoS ONE 5: e12767. Diffenbaugh, N.S., Giorgi, F., Pal, J.S. 2008. Climate change hotspots in the United States. Geophysical Research Letters 35. Domisch, S., Araujo, M.B., Bonada, N., Pauls, S.U., Jaehnig, S.C., Haase, P. 2013. Modelling distribution in European stream macroinvertebrates under future climates. Global Change Biology 19: 752–762. 76 Greno, J.L., Belda, E.J., Barba, E. 2008. Influence of temperatures during the nestling period on post-fledging survival of great tit Parus major in a Mediterranean habitat. Journal of Avian Biology 39: 41–49. Hannah, L. 2011. Climate change, connectivity, and conservation success. Conservation Biology 25: 1139–1142. Hiller, T. and Guthery, F. 2005. Microclimate versus predation risk in roost and covert selection by bobwhites. Journal of Wildlife Management 69: 140–149. Holden, Z.A., Crimmins, M.A., Cushman, S.A., Littell, J.S. 2011. Empirical modeling of spatial and temporal variation in warm season nocturnal air temperatures in two North Idaho mountain ranges, USA. Agricultural and Forest Meteorology 151: 261–269. Huertas, D. and Diaz, J. 2001. Winter habitat selection by a montane forest bird assemblage: the effects of solar radiation. Canadian Journal of Zoology-Revue Canadienne de Zoologie 79: 279–284. Kearney, M. and Porter, W. 2009. Mechanistic niche modelling: combining physiological and spatial data to predict species’ ranges. Ecology Letters 12: 334–350. Keith, D.A., Akcakaya, H.R., Thuiller, W., Midgley, G.F., Pearson, R.G., Phillips, S.J., Regan, H.M., Araujo, M.B., Rebelo, T.G. 2008. Predicting extinction risks under climate change: coupling stochastic population models with dynamic bioclimatic habitat models. Biology Letters 4: 560–563. Kennedy, C.M. and Marra, P.P. 2010. Matrix mediates avian movements in tropical forested landscapes: Inference from experimental translocations. Biological Conservation 143: 2136–2145. 77 Keppel, G. and Wardell-Johnson, G.W. 2012. Refugia: keys to climate change management. Global Change Biology 18: 2389–2391. Kim, S. and Monaghan, P. 2005. Interacting effects of nest shelter and breeder quality on behaviour and breeding performance of herring gulls. Animal Behaviour 69: 301– 306. Lloyd, J. and Martin, T. 2004. Nest-site preference and maternal effects on offspring growth. Behavioral Ecology 15: 816–823. Loehle, C. 2012. A conditional choice model of habitat selection explains the source-sink paradox. Ecological Modelling 235: 59–66. Marzluff, J.M., Withey, J.C., Whittaker, K.A., Oleyar, M.D., Unfried, T.M., Rullman, S., DeLap, J. 2007. Consequences of habitat utilization by nest predators and breeding songbirds across multiple scales in an urbanizing landscape. Condor 109: 516–534. Maziarz, M. and Wesolowski, T. 2013. Microclimate of tree cavities used by Great Tits (Parus major) in a primeval forest. Avian Biology Research 6: 47–56. Møller, A.P., Rubolini, D., Lehikoinen, E. 2008. Populations of migratory bird species that did not show a phenological response to climate change are declining. Proceedings of the National Academy of Sciences of the Unites States of America 105, 16195–16200. Monahan, W.B. and Tingley, M.W. 2012. Niche tracking and rapid establishment of distributional equilibrium in the house sparrow show potential responsiveness of species to climate change. PLoS ONE 7: e42097. 78 Monasterio, C., Salvador, A., Iraeta, P., Diaz, J.A. 2009. The effects of thermal biology and refuge availability on the restricted distribution of an alpine lizard. Journal of Biogeography 36: 1673–1684. Morrison, S. and Bolger, D. 2002. Variation in a sparrow’s reproductive success with rainfall: food and predator-mediated processes. Oecologia 133: 315–324. Nunez, T.A., Lawler, J.J., Mcrae, B.H., Pierce, D.J., Krosby, M.B., Kavanagh, D.M., Singleton, P.H., Tewksbury, J.J. 2013. Connectivity planning to address climate change. Conservation Biology 27: 407–416. Oliver, T., Roy, D.B., Hill, J.K., Brereton, T., Thomas, C.D. 2010. Heterogeneous landscapes promote population stability. Ecology Letters 13: 473–484. Patten, M., Wolfe, D., Shochat, E., Sherrod, S. 2005. Effects of microhabitat and microclimate selection on adult survivorship of the lesser prairie-chicken. Journal of Wildlife Management 69: 1270–1278. Pickett, S.T.A., Cadenasso, M.L., Grove, J.M., Boone, C.G., Groffman, P.M., Irwin, E., Kaushal, S.S., Marshall, V., McGrath, B.P., Nilon, C.H., Pouyat, R.V., Szlavecz, K., Troy, A., Warren, P. 2011. Urban ecological systems: Scientific foundations and a decade of progress. Journal of Environmental Management 92: 331–362. Rhodes, B., O’Donnell, C., Jamieson, I. 2009. Microclimate of natural cavity nests and its implications for a threatened secondary-cavity-nesting passerine of New Zealand, the south island saddleback. Condor 111: 462–469. Rudd, H., Vala, J., Schaefer, V. 2002. Importance of backyard habitat in a comprehensive biodiversity conservation strategy: A connectivity analysis of urban green spaces. Restoration Ecology 10: 368–375. 79 Ruokolainen, L., Abrams, P.A., McCann, K.S., Shuter, B.J. 2011. The roles of spatial heterogeneity and adaptive movement in stabilizing (or destabilizing) simple metacommunities. Journal of Theoretical Biology 291: 76–87. Scherrer, D. and Koerner, C. 2010. Infra-red thermometry of alpine landscapes challenges climatic warming projections. Global Change Biology 16: 2602–2613. Smith-Castro, J.R. and Rodewald, A.D. 2010. Behavioral responses of nesting birds to human disturbance along recreational trails. Journal of Field Ornithology 81: 130– 138. Speed, C.W., Meekan, M.G., Field, I.C., McMahon, C.R., Bradshaw, C.J.A. 2012. Heatseeking sharks: support for behavioural thermoregulation in reef sharks. Marine Ecology Progress Series 463: 231–244. Suggitt, A.J., Gillingham, P.K., Hill, J.K., Huntley, B., Kunin, W.E., Roy, D.B., Thomas, C.D. 2011. Habitat microclimates drive fine-scale variation in extreme temperatures. Oikos 120: 1–8. Taha, H. 1997. Urban climates and heat islands: Albedo, evapotranspiration, and anthropogenic heat. Energy and Buildings 25: 99–103. Thorington, K.K. and Bowman, R. 2003. Predation rate on artificial nests increases with human housing density in suburban habitats. Ecography 26: 188–196. Thuiller, W., Lavorel, S., Araujo, M., 2005. Niche properties and geographical extent as predictors of species sensitivity to climate change. Global Ecology and Biogeography 14: 347–357. Tieleman, B.I., van Noordwijk, H.J., Williams, J.B. 2008. Nest site selection in a hot desert: Trade-off between microclimate and predation risk? Condor 110: 116–124. 80 Trammell, E.J. and Bassett, S. 2012. Impact of urban structure on avian diversity along the Truckee River, USA. Urban Ecosystems 15: 993–1013. Tremblay, M.A. and St Clair, C.C. 2011. Permeability of a heterogeneous urban landscape to the movements of forest songbirds. Journal of Applied Ecology 48: 679– 688. Van Heezik, Y., Ludwig, K., Whitwell, S., McLean, I.G. 2008. Nest survival of birds in an urban environment in New Zealand. New Zealand Journal of Ecology 32: 155– 165. Virkkala, R., Heikkinen, R.K., Fronzek, S., Kujala, H., Leikola, N. 2013. Does the protected area network preserve bird species of conservation concern in a rapidly changing climate? Biodiversity and Conservation 22: 459–482. Virkkala, R., Heikkinen, R.K., Leikola, N., Luoto, M. 2008. Projected large-scale range reductions of northern-boreal land bird species due to climate change. Biological Conservation 141: 1343–1353. Wiebe, K.L. and Gerstmar, H. 2010. Influence of spring temperatures and individual traits on reproductive timing and success in a migratory woodpecker. Auk 127: 917– 925. 81 TABLE 1. Categorical threshold values of temperatures and the timing of breeding used to analyze nest habitat use by three bird species breeding in urban habitat in northern Nevada. Percentages of four land cover types within a 20m radius of nests and nest height are also reported. All values are means (µ) ± SD for the associated sample size (N). Variable Tmax5 (°C) cold mild warm Timing (ordinal day) early middle late 20m Cover (%) bare ground impervious canopy vegetation Nest height (m) µ Robins SD Doves SD N µ 15.8 20.9 27.2 1.5 1.8 1.7 49 68 27 114 126 141 5 6 10 9 14 28 38 3.9 13 17 18 21 1.6 Grosbeaks µ SD N N 14.6 19.2 28.0 1.7 1.9 2.1 53 63 17 17.4 23.5 27.9 1.8 1.2 1.6 14 12 17 49 47 48 93 112 139 8 8 18 45 44 44 139 145 155 6 7 7 15 13 15 144 7 16 31 38 4.2 10 19 17 20 1.7 133 4 5 23 56 2.9 6 9 17 20 1.8 43 82 FIGURE 1A - F. Hypothesized effects of temperature on nest site selection. Nest habitat on the Y axis represents % cover of one of the land cover variables measured (bare ground, canopy cover, impervious surfaces, or understory vegetation). Relative timing of nesting in the breeding season on the X axis represents seasonal variation of any one of several time varying factors that have the potential to influence nest site selection (e.g. predation risk, human activity). Lines represent different levels to air temperature at the time of nesting (cold = dashed grey, mild = solid grey, warm = solid black). Interpretation of the graphical illustration is provided by the presence or absence of a significant effect. (after Gotelli and Ellison 2004) FIGURE 2. Daily variation of maximum air temperatures during spring and summer in Reno, Nevada from 2009 to 2012. 83 84 FIGURE 3A - D. The influence of land cover within 20m on Tmax (left) and Tmin (right) recorded at 48 microsites in urban reserves from April 21 to July 13, 2011. Relative levels of land cover types are categorized into three groups of low, mid, and high. The upper percentage threshold value of low, mid, and high levels were 30, 55, and 88 for understory vegetation (A); 6, 24, and 75 for impervious surfaces (B); 10, 30, and 92 for canopy cover (C); and 8, 19, and 48 for bare ground (D). FIGURE 3A – D, continued. 85 86 FIGURE 4A-D. Microhabitat used at 278 robin and dove nest sites for different levels of the timing of breeding (early, middle, late) and the relative daily temperature maximum in the five days before nesting (cold, mild, warm). Temperature levels differed by 7.5 °C and timing levels differed by an average of 21 days. The four land cover variables at a nest that could have varied with time or temperature were understory vegetation (A), impervious surfaces (B), bare ground (C), and canopy cover (D). 87 FIGURE 5. Counts of corvids observed on transects at eight sites in a riparian corridor of northern Nevada. Relative maximum temperatures on the day of a survey (cold, mild, warm) are compared within a site over eight survey intervals. Each temperature level spanned a 7.5 °C increment. 88 THE INFLUENCE OF URBAN CLIMATE ON TREE, INVERTEBRATE, AND NESTING PHENOLOGY MILES E. BECKER AND PETER J. WEISBERG ABSTRACT Urban development modifies the thermal environment in addition to causing structural changes in habitat. The timing of life cycle events in many species is sensitive to temperature variation and may shift along an urban to rural temperature gradient. Rates of temperature-related phenological shifts are often species-specific and there is the potential for uneven responses in timing between trophic levels in urban sites. We studied the phenology of leaf out, invertebrate abundance, and nesting of American Robin (Turdus migratorius) on an urban to rural gradient in northern Nevada. We recorded air temperatures and the dates of bud burst in Salix spp. and Populus spp., peak abundance of aerial, foliar, and aquatic guilds of invertebrates, and robin clutch initiation at nine sites along a riparian corridor. Daily maximum temperatures were highest in the rural sites and next highest in the urban sites. The sequence in phenology of robins, aerial and foliar invertebrates all shifted across sites, but in different directions and at different rates. The responses of robins and foliar invertebrates to the temperature gradient were also associated with the timing of bud break. The association between urban development and phenology was primarily indirect through a delay in life cycle events at cooler sites in the suburbs. Phenological shifts in the urban landscape are uneven between trophic levels, similar to the effects of regional climate change on phenology. Future studies should investigate the population level consequences for a potential phenological mismatch between predators and herbivore prey in urban habitat. 89 INTRODUCTION Urban development can be a major factor altering the ecological processes important for native species persistence. The primary manifestation of urban development is the loss or degradation of natural systems by vegetation removal, paving of topsoil, and the addition of air and water pollutants (Alberti 2005, Grimm et al. 2008). Direct disturbance of habitat structure can have immediate impacts on local populations of animals, like vehicle collisions on roads (Taylor and Goldingay 2009), utility line electrocutions and window collisions (Hager 2009), or the destruction of burrows (Trulio 1995). Habitat modification associated with development may also impact native species in less obvious ways, such as a shift in temperature regimes. The Urban Heat Island (UHI) typically creates warmer nights, milder winters, and hotter summers in a city than the surrounding natural habitat (Brazel et al. 2007, Imhoff et al. 2010). The expression of the UHI is similar to some climate change forecasts but it is strongly linked to local land cover. The replacement of vegetation by impervious surfaces absorbs heat in the day and radiates stored heat at night (Grimmond and Oke 1999, Oke 1988). The magnitude of the UHI may be counteracted by an increase in canopy cover and irrigated landscaping that provides daytime shade and evapotranspiration (Cao et al. 2010, Jenerette et al. 2007, Zhou and Wang 2011). In addition, the heterogeneity of land cover types between neighborhoods and individual properties within the umbrella of the UHI is likely to generate fine-scale thermal heterogeneity, which may have biological significance. The ecological relevance of changing temperature regimes is frequently studied in 90 the context of regional, long-term climate change. There is ever increasing documentation of a variety of subtle and dramatic responses to climate change across multiple taxa in natural systems (Chen et al. 2011, Parmesan and Yohe 2003). One welldocumented behavioral adjustment to changing temperatures is a shift in the timing of temperature sensitive life-cycle events such as reproduction and migration (Jonzen et al. 2006, Møller et al. 2008, Stefanescu et al. 2003). Since phenology has consequences for fitness (Chuine 2010), understanding the relationship between phenology, temperature, and habitat is of great value for conservation planning. Urban landscapes are especially under-represented in the phenological literature even though they are the fastest growing land use (Seto et al. 2012) and will be encountered by many more plant and animal populations in the future. Urban habitat may have consequences for a target population by altering its phenology or the phenology of other closely associated organisms. Plant bud burst and flowering, invertebrate emergence, and the onset of breeding in birds are all temperature sensitive and their fitness is often interdependent. Milder cold nights during winter in urban areas may advance the timing of green up through the acceleration of accumulated growing degree days (Zhang et al. 2004). Invertebrate emergence and activity can be related to temperature (Pearce-Higgins et al. 2010, Tobin et al. 2008) and may also advance in warmer urban habitat. Warmer temperatures from local weather stations were a good predictor of Collared Flycatcher (Ficedula albicollis) laying date (Brommer et al. 2005), and lay date was most closely associated with weather from local stations in the 5day period before laying for Northern Flickers (Colaptes auratus; Wiebe and Gerstmar 2010). The phenology of trees, invertebrates, and birds could advance in areas with 91 urban warming, though perhaps at different rates. Other cues besides temperature may also be used to initiate phenological events. Herbivorous invertebrates dependent on the foliage of a host plant may benefit from timing their life cycle events to leaf out instead of temperature. Similarly, many birds use foliage to shield nests from solar radiation or conceal nests from predators. Blue Tits (Cyanistes caeruleus) in Corsica timed breeding to coincide with leaf out of the dominant vegetation type, not spring temperatures (Bourgault et al. 2010). Altered plant community composition along the urban to rural gradient may produce a gradient of green up that is independent of temperature variation (Shustack et al. 2009). It is therefore possible for invertebrates and birds to track leaf phenology in addition to or instead of urban warming. Similarly, other factors unique to urban habitat, such as increased irrigation, pesticide applications, and increased densities of breeding birds may exert a direct influence on phenology. Phenology could be influenced by urban development through multiple pathways including temperatures, phenology of other taxa, or environmental drivers unique to urban areas. Trees, invertebrate communities, and birds have been studied independently in urban habitats to some extent, yet the relative phenological response of all three trophic levels is unknown because they have rarely been studied in the context of a single survey. We observed the timing of bud burst in two native tree genera, peak abundance of foliar, aerial, and aquatic invertebrates, and nesting by American Robin (Turdus migratorius) at sites along a gradient of urban development in northern Nevada. Many studies have successfully demonstrated phenological shifts in response to regional climate over long time periods (Both et al. 2006, Pereyra 2011). However, we used 92 spatial variation in temperatures within a single season rather than temporal variation between years to examine phenological variability. We expected all taxa to advance phenology in the more urban sites, but at different rates, and that phenological shifts would be mainly attributed to urban climatic variability and not other urban features such as irrigation, pesticides, or densities. METHODS Study Area The study area, a series of parks on a 28.4 km transect, provided a cross-section of urban, suburban, industrial, and rural land uses in Reno and Sparks, Nevada (Figure 1). Rapid outward growth from the city center has led to the expansion of suburbs and housing developments into sagebrush habitat adjacent to riparian woodlands along the Truckee River that runs west to east through town. Much of the study area is on the historic floodplain of the Truckee Meadows and is bounded in three directions by hills and mountain ranges, including the eastern Sierra Nevada. We selected eight public parks and one privately owned restoration site using the criteria of park size, accessibility, and the presence of riparian habitat that required all sites be located adjacent to the Truckee River. Remnant patches of riparian woodland were dominated by cottonwood (Populus spp.) and willow (Salix spp.). The less abundant non-native Russian olive (Elaeagnus angustifolia) and Siberian elm (Ulmus pumila) are naturalized along the river. Average study site size was 9.9 ha and neighborhoods adjacent to the parks include apartment complexes, single family 93 residences, industrial complexes, pasture, and rangeland. Urban metrics Urban development can be characterized in various ways. The level of urbanization of patches of native vegetation in a matrix of other land uses has been described by the dominant land use in the surrounding area (Kennedy et al. 2010) or a combination of land uses and land cover, such as agriculture to represent non-urban habitat, road length, and mowed surfaces (Sundell-Turner and Rodewald 2008). Ryder et al. (2010) used 30m resolution land cover data to quantify impervious surface and canopy density within 100 – 1000m radii of nests as a measure of ecologically relevant urban development in Washington, D.C. We used three land cover classifications – percent impervious surface, canopy cover, and bare ground – and road density to represent an urban to rural gradient. Land cover classifications were derived from 1m resolution orthophotographs collected in 2010 by the National Agriculture Imagery Program (NAIP) using a geographic object-based image analysis (GEOBIA) with an overall accuracy of 94% (AMEC 2012). Road data were publicly available from Washoe County GIS (Reno, Nevada, USA). Road density was calculated from sum of the length of all roads within each study site and a surrounding 100m buffer divided by the area of the study site in hectares. Land cover and land use percentages were calculated for a 100m buffer around each park that did not include the park surface itself using ArcGIS software (ArcMap 10, ESRI, Redlands, California, USA). The 100m scale represents the distance of effective heat transfer from surrounding neighborhoods across the park boundary (Jansson et al. 94 2007, Bowler et al. 2010). The four site variables were entered into a Principal Components Analysis (PCA) and the first two principal components explained 87% of the variation between sites. A PCA for five land cover and land use variables had a similar explanatory power (80%) for variation between forest stands on an urban to rural gradient in Ohio (Shustack and Rodewald 2010). We used a weighted average of the first two principal components scores for each site as an index of urban development using the eigenvalues for the respective axes. Rural sites scored negative and urban sites scored positive values. The PCA weighted average is referred to throughout the text as the ‘urban index’ and used for all site level analyses. The nine sites were clustered in three groups in ordination space, which we labeled as urban, suburban, and rural zones to facilitate interpretation of site level differences related to urbanization. Urban sites had more canopy cover and impervious surfaces, less bare ground, and higher road densities than rural and suburban sites. Rural sites had the most bare ground and lowest road densities. The zonal groups were used only in the analysis of invertebrate abundance and otherwise functioned as a descriptive tool to assist visualization of the relative urban development of a site. Our definitions of zones based on the habitat factors used in the PCA are intended to be biologically meaningful and they may differ from zonal definitions employed by city planners that use other criteria for designating land uses. Temperatures Five to nine sampling stations were established at each site to record air temperature, invertebrate abundance, and tree phenology in 2011. Stations within a site 95 were placed at evenly spaced intervals of approximately 150m. Thermocron iButton data loggers (N = 52; DS1921G, Dallas Semiconductors/MAXIM, Dallas, Texas, USA) were deployed in mid-March, programmed to record temperature every 4 hours, and retrieved in mid-July. Data analysis included temperature measurements from March 17 after all iButtons had time to adjust to field temperatures and ended on July 6 before they were retrieved. All iButtons were placed out of direct sunlight under the bark on the north side of a tree and at a height of 1.5m. Tree Phenology Leaf phenology in deciduous trees can be measured with a qualitative scale of bud break and flowering that is relevant to invertebrates and songbirds (McGrath et al. 2009). Bud break was measured in April and May 2011 at a sub-set of temperature stations that had a Salix (N = 37) or Populus (N = 30) tree within 20m of the station. Trees sampled for bud break had secondary growth taller than 2m. Bud break was measured by scoring the apical bud on four branches closest to chest height in each of the cardinal directions. The stem directly before the apical bud was marked by a dot with a permanent marker to allow resampling the same bud on returning visits. Developmental stages of buds were given a score of 0 – 5 following Pellis et al. (2004; Table 1). Reference photos were used to maintain observer consistency in scoring bud break. A tree was considered to have progressed to bud burst when the average of four apical buds on a tree had reached a score of 4. Site level bud break date was the average date for all trees sampled within the site. 96 Invertebrate Phenology Several methods exist for sampling invertebrates. We used a trapping method that allows continuous sampling over several days and temperature ranges. Pitfall traps are capable of multi-day trapping but are biased towards ground dwelling invertebrates (Rohr et al. 2007). Instead, we used adhesive traps that catch a combination of aerial, foliar, and adult aquatic invertebrates. Even so, relative abundance between taxa does not represent true relative abundance because the sampling method was biased to capture aerial invertebrates. Adhesive traps were yellow 7.6 x 12.7 cm cards coated in a clear adhesive (Stiky Strip Traps, Olson Products Inc.). Traps were located at 64 stations (5-8 per site) and hung with a short wire from a branch on the western side of the tree crown at a height of approximately 2m. Traps were retrieved and replaced at weekly intervals from May 1 – July 17, 2011. An index of the timing of peak abundance at a site for each taxon was identified by the sampling date with the highest count of individual invertebrates. A second measure of invertebrate activity, peak magnitude, was the value from the sampling date with peak abundance. The unit of measurement in all invertebrate analyses was number of invertebrates captured per day. Invertebrates were identified to order or sub-order and placed into guilds of aerial, aquatic, or foliar. Aerial insects primarily moved by flight as adults, were often captured in air by hawking birds, and generally acquired nutrition by means other than herbivory. Invertebrates whose larval stage is underwater in ponds or rivers were classified as aquatic. Foliar invertebrates traveled more often by walking, were herbivorous or strongly associated with plants (i.e. herbivore predators), and often found themselves the 97 target of foliage gleaning by foraging birds. Nest Phenology Robins are present year-round in the study area and begin breeding by mid-April. Robins are foraging and nest site generalists, but despite their apparent lack of niche specificity, they have been documented to adjust arrival at breeding habitat to spring temperatures (Inouye et al. 2008). Nests were located starting in March or April from 2009 to 2012 (N = 155). Nests higher than 8m were excluded because of less accurate ageing of nests. Nest contents were checked every three to four days with a mirror pole to age nests. Nests were aged using laying sequence during egg laying, by counting back from hatch day, or by ageing nestlings. Individuals in this study were not uniquely marked, and so first nesting attempts could not be reliably segregated from sequential nest attempts. Instead, we estimated site level clutch initiation dates using the 25th percentile of nesting date. Other studies have used median or mean nesting date for marked populations (Both et al. 2005, Bourgault et al. 2010, Pereyra 2011) and the 25th percentile is a more conservative estimate. While our predictor variables of interest were related to habitat and temperature, clutch initiation can be influenced by the density of breeding conspecifics (Evans et al. 2009) and breeding bird density often increases at more urban sites (Rodewald and Shustack 2008). We estimated the density of robins at a site in 2011 using distance sampling (Buckland et al. 2001, Gottschalk and Huettmann 2011) on line transects through each site. Observations on transects were between 0630 and 1000 with eight 98 replicates per site at five to eight day intervals. Distance sampling was used to calculate density estimates as individuals per ha. Estimates of robins per ha in each site were calculated in Program DISTANCE (version 6.0 Release 2, Research Unit for Wildlife Population Assessment). ANALYSIS Temperature, urbanization, and phenology We followed two analytical approaches to addressing the questions of how urban development influences temperatures and how phenology across taxa responds to both predictor variables. Individual relationships between sites and temperatures were first analyzed in a univariate ANOVA for average daily temperature maximum (Tmax) in April. Most of the phenological events occurred during May and June, and they were most likely influenced by prior temperatures in April. Tmax was used because it is related to the accumulation of growing degree days that influences leaf phenology and also invertebrate emergence and nest phenology. Easting was used as a covariate in the analysis because the sites follow a west to east transect at increasing distances from the Sierra Nevada. Cold air masses, hydrological characteristics, or other landscape scale thermal dynamics may change across the gradient. The second approach was to collectively examine the relationships between urban development, temperature, and phenology in a path analysis. Path analysis uses a series of multiple regressions to quantify hypothesized relationships between exogenous 99 predictor variables and endogenous response variables. The main advantage of a path analysis approach is to isolate the direct and indirect relationship of a predictor variable, urban development, with a response variable, phenology. Model fit for individual taxa phenologies were assessed with the coefficient of determination (R2) and the residual error associated with the response variable. Path coefficients displayed in the results are standardized Beta coefficients that show how much a one standard deviation change in the predictor variable is estimated to change the response variable. We used site level averages for the date of peak invertebrate abundance for three guilds, and clutch initiation as the response variables and urban index, Easting, bud break, and temperature as the predictor variables. In the models, urban index scores could influence invertebrate and robin phenology directly or indirectly through temperatures or date of leaf out by Populus (Figure 2). Populus phenology was used over Salix because a priori univariate testing showed a strong relationship between robin nest phenology and Populus bud break (Pearson’s = 0.817, P = 0.013) but not with Salix (Pearson’s = 0.175, P = 0.824). Populus was used for the other models to maintain consistency. Prior examination of temperature differences between sites revealed that there was a quadratic effect of urban development on daily Tmax. We included a quadratic term for urban index in the path analysis (Scheiner et al. 2000) that was modeled to influence only temperature (Figure 2). Four years of nest phenology data were sampled at the sites even though temperatures and bud break were only sampled in 2011. We included all four years of nest data for a larger sample size to better estimate the sequence of nest phenology across sites. Phenological sequences cued by urban development or geographic location of a site should be consistent from year-to-year since 100 neither variable changed during the study period. Breeding density of birds can influence nest phenology, but there was almost no relationship between 25th percentile lay date and the density of robins at a site (P = 0.525, r2 = 0.09), so robin density was also omitted from the model. RESULTS Site Features Individual sites differed in size, geography, and proportion of land cover types in the matrix outside their boundaries (Table 3). The urban gradient captured by the Urban Index represented a shift from rural sites with lower road densities and more bare ground to urban sites with more canopy cover and impervious surfaces. Impervious surface changed substantially between sites, up to 45%, while canopy cover changed the least amount. The sites were located on a west to east transect where Easting was highly correlated with elevation (Pearson’s = -0.982, P <0.001, N = 9). The most western site was 110m higher in elevation than the most eastern site, for a slope of 0.4% over the entire study area. Urban influence on Temperatures Maximum daily temperatures in April differed between sites by up to 4.6°C and by 5°C in June. Temperature differed significantly between sites in April (F6,289 = 4.003, P < 0.001) and June (F7,379 = 3.467, P = 0.001). Tmax was highest in the rural sites in both April and June (Figure 3). The most urban site had the next highest Tmax in April 101 and, on average, suburban sites had the lowest observed Tmax in April (Figure 3). Phenology Salix were the first to reach their phenological event across all individuals in the study population (4 April ± 10 days), followed by Populus bud break (1 May ± 7 days), nesting by robins (1 May ± 5 days), peak abundance of foliar (31 May ± 24 days), aerial (15 June ± 20 days), and aquatic invertebrates (26 June ± 18 days). The intensity of the shift in phenology between sites differed between taxa (Figure 4). Salix bud break advanced by up to 14 days between sites and Populus by up to 16 days. Robins nested up to 11 days earlier at some sites on the 28.4 km transect. Changes in the timing of peak invertebrate abundance were the most dramatic: up to 48 days for foliar invertebrates and up to 55 days for aerial invertebrates. The sequence of phenology for trees, insects, and birds also differed across sites (Figure 4). The sequence of bud break by Populus paralleled the site level temperatures with earlier phenology at warmer sites. Path analysis supported a strong negative association between temperature and bud break and a positive association between easting and bud break (Figure 5). The timing of nesting by robins at sites was similar to bud break by Populus (Figure 4) and the two phenologies were closely associated in the path analysis (Figure 5). Bud break was somewhat associated with peak abundance of foliar invertebrates (Figure 5) but the sequence of invertebrate peak abundances was fairly idiosyncratic for all three guilds (Figure 4). The path model was not an equally good fit for all the taxa. Robin clutch initiation was closest (R2 = 0.981) and aerial and foliar invertebrates were reasonable (aerial: R2 = 0.701; foliar: R2 = 0.764). The paths for 102 Populus bud break were a slightly better fit than invertebrates (R2 = 0.893). The model of site level temperatures in April also fit the data moderately well when explained by the urban index at sites (R2 = 0.746). More aerial invertebrates were captured than the two other guilds, although aerial invertebrates had the fewest number of orders, sub-orders, or families (Table 4). Too few aquatic invertebrates were captured to be able to demonstrate any pattern (Figure 6). The lowest observed abundances of foliar and aerial invertebrates occurred in the urban sites (Figure 6), but did not significantly differ between zoning groups (foliar: F3,8 = 4.23, P = 0.099; aerial: F3,8 = 0.49, P = 0.732). DISCUSSION Phenology has the potential to shift along an urban to rural gradient in response to temperatures, biological interactions, or other factors present in the urban environment. We found that the sequence of timing of bud break, peak invertebrate abundance, and nesting by robins differed between sites on an urban to rural gradient. The indirect effect of the surrounding matrix on site level temperatures was associated with advancing phenology of Populus, robins, and aerial invertebrates. Robin nest phenology and peak foliar abundance was also associated with leaf phenology. Urban sites tended to have the lowest abundances of invertebrates. Urban Development and Temperatures 103 The relationship between temperature and urban development in the study area was not typical of the Urban Heat Island effect described for cities in temperate and tropical regions. Usually, temperatures within the city boundaries are warmer than adjacent land used for other purposes. Rural sites in this study reached the warmest temperatures in spring and summer, most likely because of a limited 3-5 % canopy cover and abundant bare ground or low growing sparse vegetation characteristic of sagebrush communities in the western Great Basin. Solar radiation rapidly heats surfaces in the absence of shade (Suggitt et al. 2011, Vargo et al. 2013), promoting warmer temperatures at rural sites. The subrural, suburban, and urban sites were surrounded by 2 to 3 times more canopy cover, which could effectively cool air by shading or evapotranspiration (Cao et al. 2010). The urban sites were the next warmest, with the coolest spring temperatures generally in the suburban sites. This “U-shaped” spatial temperature trend has been described as uniquely characteristic to cities in arid and desert regions, where the Urban Heat Island is less pronounced (Imhoff et al. 2010). The relationship between urban development in the matrix surrounding parks and site temperatures may have been even greater if our study sites were located farther from the river because water can dampen temperature amplitudes. Phenology Temperature variation between sites was strongly associated with differences in the sequence of phenological events. The relationship between tree leaf out and growing degree days is well established (Bronson et al. 2009, Richardson et al. 2006) and it was expected that Populus would leaf out earlier at sites with higher Tmax, representing faster 104 accumulation of growing degree days. Populus bud break did occur earlier at warmer sites, but Salix did not follow the same pattern. Salix may have greater sensitivity to other urban factors, like irrigation levels, nutrient cycling and soil biochemistry, or soil compaction and composition. There are seven Salix species, including at least one naturalized species, and only two Populus species in the Truckee Meadows (pers. comm. Otis Bay Ecological Consultants, Mogul, Nevada). Genetic variation within a single genus can lead to variation in leaf out under the same environmental conditions (Pellis et al. 2004). Salix composition may have differed between sites and masked any unified response to temperature. The timing of bud break by Populus was also strongly associated with the position of sites on a West to East gradient. Temperatures and tree species composition could have changed along the geographic gradient, which may have accounted for the significant relationship. Easting and Tmax were related in the path model, suggesting that temperatures may have been influenced by geography instead of urban development. However, the variation in temperatures between sites did not follow a linear trend with Easting, allowing for another factor, likely urban development, to contribute to temperature differences between sites. Robins varied the timing of clutch initiation between sites by up to 11 days. The phenological shift was partly associated with the direct influence of temperatures at a site, but the relationship with leaf phenology of Populus was stronger. Temperature can be a secondary cue, after photoperiod, for birds to initiate nesting (Brommer et al. 2005, Coppack and Pulido 2004). Birds may also use leaf out as a cue to begin breeding. Great tits and blue tits initiated nesting in association with the timing of leaf out in different 105 habitats, not temperatures (Nilsson and Kallander 2006). Previous studies have found that other bird species synchronize the timing of nesting with the timing of foliar invertebrates, the preferred prey item in their nestling diets (Both et al. 2006, Burger et al. 2012). Closely synchronized phenology between robins and Populus in this study likely had little to do with food resources. Robins generally eat soil macroinvertebrates that are not associated with plant herbivory and the functional relationship between robins and trees is not linked by food. Instead, robins use trees primarily for nesting and the strong relationship between leaf out of Populus and not Salix suggests the former is more widely used by nesting robins. Indeed, over four years of observing robin nests, 8% were constructed in Salix and 43% in Populus even though the relative availability of each tree species was almost exactly one to one in the study area (unpublished data). Foliage provides nest concealment from predators (Chalfoun and Martin 2009) and some species select nest sites with more cover across spatially heterogeneous habitat (Chalfoun and Martin 2010, Kolada et al. 2009). Our results suggest that temporal variation in the availability of less risky nest sites may drive the timing of clutch initiation. The early timing of leaf out in the exotic shrub Lonicera promoted early nesting by Northern Cardinals, even though those nests were more likely to fail (Rodewald et al. 2010). The benefit of waiting for Populus to leaf out could be tested in future studies by comparing nest survival of nests constructed in trees before and after leaf out. Phenology of the invertebrate guilds responded differently to urban development, temperatures, and Populus leaf phenology. Peak abundance in aerial invertebrates was the most sensitive to the urban index and Tmax at a site, peaking earliest at the most rural 106 site that was warmest. Invertebrate abundance of foliar arthropods on potted plants in Phoenix, Arizona also peaked earlier at rural sites (Bang et al. 2012). Foliar insects in our study were expected to peak earlier at sites where their host plants also leaf out earlier (Mussey and Potter 1997). In this case, foliar invertebrates were expected to peak earlier at sites where Populus peaked earlier. The path analysis showed foliar invertebrates had the strongest association with bud break of the three guilds. However, foliar invertebrates in this study did not peak earlier in both the warmer urban and rural sites, perhaps because of the phenology of other plants besides Populus. Native plant richness may decrease in urban areas (Godefroid and Koedam 2007) where there may be fewer resources or narrower windows of leaf and flower phenology for foliar invertebrates. Urban areas had the lowest abundance of aerial and foliar invertebrates, perhaps from limited food resources. Further research on separating out the temperature effects from plant species richness effects on invertebrate phenology would be useful. Not surprisingly, there was no influence of any of the three terrestrial metrics sampled on the phenology of aquatic invertebrates. The sample size for that guild was relatively low compared to the other two guilds because adult aquatic invertebrates were mainly bycatch in the trapping method that was biased towards terrestrial invertebrates. Even so, the 1262 aquatic invertebrates collected across nine sites showed that the site level factors we measured were not significantly related to their phenology. However, aquatic invertebrates can be an important part of nutrient cycling between aquatic and terrestrial food webs (Nakano and Murakami 2001) and the relative timing of their peak abundances can respond to different factors (Dineen et al. 2007). Urbanization can affect aquatic systems from the addition of nutrients and chemicals into urban streams and 107 rivers or channelization (Nilsson et al. 2003), but at the scale of the entire city or watershed and not at individual parks that we sampled. Temperatures can also affect the timing of aquatic invertebrate emergence (Harper and Peckarsky 2006), but water temperatures in the river at a single park are likely more influenced by stream morphology, snow-runoff, and upstream factors that we did not sample. Urban warming and Phenology We found an indirect association of suburban cooling on phenology, but phenology was mediated by interactions with other species, other urban factors, and geography. The taxa spanning three trophic levels in this study differed in the magnitude of their response to the same predictor variable and some were more sensitive than others to different predictor variables. Even though there was not a direct link between all three trophic levels, changes in the sequence of phenology across the rural to urban gradient could result in a phenological mismatch similar to observations of natural systems undergoing long-term regional climate change. Lay date in some urban birds may be constrained by the availability of canopy cover at nest sites, a potential trade-off with synchronizing nesting to invertebrate prey availability. For example, shorter and lower magnitude peaks in invertebrate prey in urban habitat may make it less productive for some foraging guilds. Acadian Flycatchers that primarily capture aerial insects nested later and had higher turnover in urban forests, suggesting that it was marginal habitat (Shustack and Rodewald 2010). Species richness and abundance of insectivorous birds often decline in urban habitat (Dunford and Freemark 2005, Schlesinger et al. 2008), perhaps for the same reason. 108 Conclusions Urban development modifies the thermal environment in ways that may affect biological processes. Phenology is an important factor in fitness and can have consequences for population dynamics (Chuine 2010). Phenology is known to shift in response to climate change (Chen et al. 2011, Parmesan and Yohe 2003) and also across habitats and land uses (Altermatt 2012, Burger et al. 2012). Since the rates of phenological shifts are often species-specific (Donnelly et al. 2011), predicting community level responses to global change requires simultaneous observation of multiple species (Ibanez et al. 2010). This is the first study to our knowledge that follows the phenology of three trophic levels across the urban to rural gradient. While other major features of urban development, such as habitat fragmentation or habitat loss, may impact plant, invertebrate, and bird populations, the results of this study suggest indirect effects of temperature variation can also influence life cycles. The mistiming of phenology between trophic levels can have serious implications for population viability of a single species (Visser 1998) or multiple species interactions (Yang and Rudolf 2010). Whether or not phenological shifts in the urban landscape are sufficient to affect population dynamics remains an open question for future studies. 109 LITERATURE CITED Alberti, M. 2005. The effects of urban patterns on ecosystem function. International Regional Science Review 28: 168–192. Altermatt, F. 2012. Temperature-related shifts in butterfly phenology depend on the habitat. Global Change Biology 18: 2429–2438. AMEC Environment and Infrastructure. 2012. Urban tree canopy assessment Truckee Meadows (Reno-Sparks, Nevada). 53pp. Bang, C., Faeth, S.H., Sabo, J.L. 2012. Control of arthropod abundance, richness, and composition in a heterogeneous desert city. Ecological Monographs 82: 85–100. Both, C., Bijlsma, R.G., Visser, M.E. 2005. Climatic effects on timing of spring migration and breeding in a long-distance migrant, the pied flycatcher Ficedula hypoleuca. Journal of Avian Biology 36: 368–373. Both, C., Bouwhuis, S., Lessells, C.M., Visser, M.E. 2006. Climate change and population declines in a long-distance migratory bird. Nature 441: 81–83. Bourgault, P., Thomas, D., Perret, P., Blondel, J. 2010. Spring vegetation phenology is a robust predictor of breeding date across broad landscapes: a multi-site approach using the Corsican blue tit (Cyanistes caeruleus). Oecologia 162: 885–892. Bowler, D.E., Buyung-Ali, L., Knight, T.M., Pullin, A.S. 2010. Urban greening to cool towns and cities: A systematic review of the empirical evidence. Landscape and Urban Planning 97: 147–155. Brazel, A., Gober, P., Lee, S.-J., Grossman-Clarke, S., Zehnder, J., Hedquist, B., Comparri, E. 2007. Determinants of changes in the regional urban heat island in 110 metropolitan Phoenix (Arizona, USA) between 1990 and 2004. Climate Research 33: 171–182. Brommer, J.E., Merila, J., Sheldon, B.C., Gustafsson, L. 2005. Natural selection and genetic variation for reproductive reaction norms in a wild bird population. Evolution 59: 1362–1371. Bronson, D.R., Gower, S.T., Tanner, M., Van Herk, I. 2009. Effect of ecosystem warming on boreal black spruce bud burst and shoot growth. Global Change Biology 15: 1534–1543. Buckland, S.T. 2001. Introduction to distance sampling: estimating abundance of biological populations. Oxford University Press, 432pp. Burger, C., Belskii, E., Eeva, T., Laaksonen, T., Maegi, M., Maend, R., Qvarnstrom, A., Slagsvold, T., Veen, T., Visser, M.E., Wiebe, K.L., Wiley, C., Wright, J., Both, C. 2012. Climate change, breeding date and nestling diet: how temperature differentially affects seasonal changes in pied flycatcher diet depending on habitat variation. Journal of Animal Ecology 81: 926–936. Cao, X., Onishi, A., Chen, J., Imura, H. 2010. Quantifying the cool island intensity of urban parks using ASTER and IKONOS data. Landscape and Urban Planning 96: 224–231. Chalfoun, A.D. and Martin, T.E. 2009. Habitat structure mediates predation risk for sedentary prey: experimental tests of alternative hypotheses. Journal of Animal Ecology 78: 497–503. 111 Chalfoun, A.D. and Martin, T.E. 2010. Facultative nest patch shifts in response to nest predation risk in the Brewer’s sparrow: a “win-stay, lose-switch” strategy? Oecologia 163: 885–892. Chen, I.-C., Hill, J.K., Ohlemueller, R., Roy, D.B., Thomas, C.D. 2011. Rapid range shifts of species associated with high levels of climate warming. Science 333: 1024– 1026. Chuine, I. 2010. Why does phenology drive species distribution? Philosophical Transactions of the Royal Society B: Biological Sciences 365: 3149 –3160. Coppack, T. and Pulido, F. 2004. Photoperiodic response and the adaptability of avian life cycles to environmental change, in: Møller, A.P., Fielder, W., Berthold, P. (eds.). Birds and Climate Change, Advances in Ecological Research. pp. 131–150. Dineen, G., Harrison, S.S.C., Giller, P.S. 2007. Seasonal analysis of aquatic and terrestrial invertebrate supply to streams with grassland and deciduous riparian vegetation. Biology and Environment – Proceedings of the Royal Irish Academy 107B: 167–182. Donnelly, A., Caffarra, A., O’Neill, B.F. 2011. A review of climate-driven mismatches between interdependent phenophases in terrestrial and aquatic ecosystems. International Journal of Biometeorology 55: 805–817. Dunford, W. and Freemark, K. 2005. Matrix matters: Effects of surrounding land uses on forest birds near Ottawa, Canada. Landscape Ecology 20: 497–511. Evans, L.E., Ardia, D.R., Flux, J.E.C. 2009. Breeding synchrony through social stimulation in a spatially segregated population of European starlings. Animal Behaviour 78: 671–675. 112 Godefroid, S. and Koedam, N. 2007. Urban plant species patterns are highly driven by density and function of built-up areas. Landscape Ecology 22: 1227–1239. Gottschalk, T.K. and Huettmann, F. 2011. Comparison of distance sampling and territory mapping methods for birds in four different habitats. Journal of Ornithology 152: 421–429. Grimm, N.B., Foster, D., Groffman, P., Grove, J.M., Hopkinson, C.S., Nadelhoffer, K.J., Pataki, D.E., Peters, D.P.C. 2008. The changing landscape: ecosystem responses to urbanization and pollution across climatic and societal gradients. Frontiers in Ecology and the Environment 6: 264–272. Grimmond, C.S.B. and Oke, T.R. 1999. Heat storage in urban areas: Local-scale observations and evaluation of a simple model. Journal of Applied Meteorology 38: 922–940. Hager, S.B. 2009. Human-related threats to urban raptors. Journal of Raptor Research 43: 210–226. Harper, M. and Peckarsky, B. 2006. Emergence cues of a mayfly in a high-altitude stream ecosystem: Potential response to climate change. Ecological Applications 16: 612–621. Ibanez, I., Primack, R.B., Miller-Rushing, A.J., Ellwood, E., Higuchi, H., Lee, S.D., Kobori, H., Silander, J.A. 2010. Forecasting phenology under global warming. Philosophical Transactions of the Royal Society B: Biological Sciences 365: 3247 – 3260. 113 Imhoff, M.L., Zhang, P., Wolfe, R.E., Bounoua, L. 2010. Remote sensing of the urban heat island effect across biomes in the continental USA. Remote Sensing of Environment 114: 504–513. Inouye, D.W., Barr, B., Armitage, K.B., Inouye, B.D. 2000. Climate change is affecting altitudinal migrants and hibernating species. Proceedings of the National Academy of Sciences of the United States of America 97: 1630–1633. Jansson, C., Jansson, P.-E., Gustafsson, D. 2007. Near surface climate in an urban vegetated park and its surroundings. Theoretical and Applied Climatology 89: 185– 193. Jenerette, G.D., Harlan, S.L., Brazel, A., Jones, N., Larsen, L., Stefanov, W.L. 2007. Regional relationships between surface temperature, vegetation, and human settlement in a rapidly urbanizing ecosystem. Landscape Ecology 22: 353–365. Jonzen, N., Linden, A., Ergon, T., Knudsen, E., Vik, J.O., Rubolini, D., Piacentini, D., Brinch, C., Spina, F., Karlsson, L., Stervander, M., Andersson, A., Waldenstrom, J., Lehikoinen, A., Edvardsen, E., Solvang, R., Stenseth, N.C. 2006. Rapid advance of spring arrival dates in long-distance migratory birds. Science 312: 1959–1961. Kennedy, C.M. and Marra, P.P. 2010. Matrix mediates avian movements in tropical forested landscapes: Inference from experimental translocations. Biological Conservation 143: 2136–2145. Kolada, E.J., Sedinger, J.S., Casazza, M.L. 2009. Nest site selection by Greater SageGrouse in Mono County, California. Journal of Wildlife Management 73: 1333– 1340. 114 McGrath, L.J., van Riper, III, C., Fontaine, J.J. 2009. Flower power: tree flowering phenology as a settlement cue for migrating birds. Journal of Animal Ecology 78: 22–30. Meineke, E.K., Dunn, R.R., Sexton, J.O., Frank, S.D. 2013. Urban warming drives insect pest abundance on street trees. PLoS ONE 8: e59687. Møller, A.P., Rubolini, D., Lehikoinen, E. 2008. Populations of migratory bird species that did not show a phenological response to climate change are declining. Proceedings of the National Academy of Sciences of the United States of America 105: 16195–16200. Mussey, G. and Potter, D. 1997. Phenological correlations between flowering plants and activity of urban landscape pests in Kentucky. Journal of Economic Entomology 90: 1615–1627. Nakano, S. and Murakami, M. 2001. Reciprocal subsidies: Dynamic interdependence between terrestrial and aquatic food webs. Proceedings of the National Academy of Sciences of the United States of America 98: 166–170. Nilsson, C., Pizzuto, J., Moglen, G., Palmer, M., Stanley, E., Bockstael, N., Thompson, L. 2003. Ecological forecasting and the urbanization of stream ecosystems: Challenges for economists, hydrologists, geomorphologists, and ecologists. Ecosystems 6: 659–674. Nilsson, J.A. and Kallander, H. 2006. Leafing phenology and timing of egg laying in great tits Parus major and blue tits P-caeruleus. Journal of Avian Biology 37: 357– 363. Oke, T. 1988. The urban energy-balance. Progress in Physical Geography 12: 471–508. 115 Parmesan, C. and Yohe, G. 2003. A globally coherent fingerprint of climate change impacts across natural systems. Nature 421: 37–42. Pearce-Higgins, J.W., Dennis, P., Whittingham, M.J., Yalden, D.W. 2010. Impacts of climate on prey abundance account for fluctuations in a population of a northern wader at the southern edge of its range. Global Change Biology 16: 12–23. Pellis, A., Laureysens, I., Ceulemans, R. 2004. Genetic variation of the bud and leaf phenology of seventeen poplar clones in a short rotation coppice culture. Plant Biology 6: 38–46. Pereyra, M.E. 2011. Effects of snow-related environmental variation on breeding schedules and productivity of a high-altitutde population of dusky flycatchers (Empidonax oberholseri). Auk 128: 746–758. Richardson, A., Bailey, A., Denny, E., Martin, C., O’Keefe, J. 2006. Phenology of a northern hardwood forest canopy. Global Change Biology 12: 1174–1188. Rodewald, A.D. and Shustack, D.P. 2008. Consumer resource matching in urbanizing landscapes: Are synanthropic species over-matching? Ecology 89: 515–521. Rodewald, A.D., Shustack, D.P., Hitchcock, L.E. 2010. Exotic shrubs as ephemeral ecological traps for nesting birds. Biological Invasions 12: 33–39. Rohr, J.R., Mahan, C.G., Kim, K.C. 2007. Developing a monitoring program for invertebrates: Guidelines and a case study. Conservation Biology 21: 422–433. Ryder, T.B., Reitsma, R., Evans, B., Marra, P.P. 2010. Quantifying avian nest survival along an urbanization gradient using citizen- and scientist-generated data. Ecological Applications 20: 419–426. 116 Schlesinger, M.D., Manley, P.N., Holyoak, M. 2008. Distinguishing stressors acting on land bird communities in an urbanizing environment. Ecology 89: 2302–2314. Seto, K.C., Gueneralp, B., Hutyra, L.R. 2012. Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools. Proceedings of the National Academy of Sciences of the United States of America 109: 16083–16088. Shustack, D.P. and Rodewald, A.D. 2010. Attenuated nesting season of the Acadian Flycatcher (Empidonax virescens) in urban forests. Auk 127: 421–429. Shustack, D.P., Rodewald, A.D., Waite, T.A. 2009. Springtime in the city: exotic shrubs promote earlier greenup in urban forests. Biological Invasions 11: 1357–1371. Stefanescu, C., Penuelas, J., Filella, I. 2003. Effects of climatic change on the phenology of butterflies in the northwest Mediterranean Basin. Global Change Biology 9: 1494–1506. Suggitt, A.J., Gillingham, P.K., Hill, J.K., Huntley, B., Kunin, W.E., Roy, D.B., Thomas, C.D. 2011. Habitat microclimates drive fine-scale variation in extreme temperatures. Oikos 120: 1–8. Sundell-Turner, N.M. and Rodewald, A.D. 2008. A comparison of landscape metrics for conservation planning. Landscape and Urban Planning 86: 219–225. Taylor, B.D. and Goldingay, R.L. 2009. Can road-crossing structures improve population viability of an urban gliding mammal? Ecology and Society 14. Tobin, P.C., Nagarkatti, S., Loeb, G., Saunders, M.C. 2008. Historical and projected interactions between climate change and insect voltinism in a multivoltine species. Global Change Biology 14: 951–957. 117 Trulio, L. 1995. Passive relocation – a method to preserve burrowing owls on disturbed sites. Journal of Field Ornithology 66: 99–106. Vargo, J., Habeeb, D., Stone, Jr., B. 2013. The importance of land cover change across urban-rural typologies for climate modeling. Journal of Environmental Management 114: 243–252. Visser, M.E., van Noordwijk, A.J., Tinbergen, J.M., Lessells, C.M. 1998. Warmer springs lead to mistimed reproduction in great tits (Parus major). Proceedings of the Royal Society of London Series B-Biological Sciences 265: 1867–1870. Wiebe, K.L. and Gerstmar, H. 2010. Influence of spring temperatures and individual traits on reproductive timing and success in a migratory woodpecker. Auk 127: 917– 925. Yang, L.H. and Rudolf, V.H.W. 2010. Phenology, ontogeny and the effects of climate change on the timing of species interactions. Ecology Letters 13: 1–10. Zhang, X.Y., Friedl, M.A., Schaaf, C.B., Strahler, A.H., Schneider, A. 2004. The footprint of urban climates on vegetation phenology. Geophysical Research Letters 31. Zhou, X. and Wang, Y.-C. 2011. Dynamics of land surface temperature in response to land-use/cover change. Geographical Research 49: 23–36. 118 TABLE 1. Qualitative indicators of sequential stages of bud break used to measure leaf phenology in Salix spp. and Populus spp. at nine sites on an urban to rural gradient. Pellis Score 0 Indicator dormant bud, no sign of any activity, completely dry 1 buds slightly swollen, color starting to change on bud scales 2 buds fully swollen and rounded, no sign of breakage of buds 3 buds started breaking, tips of shoots visible 4 bud burst and shoots green and very young leaves visible 5 green leaves growing and leaf venation visible 119 TABLE 2. Site level differences in environmental factors at nine sites on an urban to rural gradient. The urban index is a PCA score derived from canopy cover, impervious surface, bare ground, and road density. Estimates of relative robin densities are in the unit of birds per hectare. 120 TABLE 3. Abundance by taxon of invertebrates captured with adhesive traps from May to July 2011 at nine sites in northern Nevada. Guild Aerial Taxon Diptera Hymenoptera Lepidoptera Guild total Total Abundance 64320 4328 2 68650 Aquatic Ephemeroptera Neuroptera Odonota Plecoptera Raphidiidae Trichoptera Guild total 108 449 3 53 90 559 1262 Foliar Hemiptera Arachnid Chrysopidae Cicadellidae Coleoptera Formicidae Larvae-unknown Guild total 4963 114 57 2691 2062 87 3 9977 121 FIGURE 1. Location of nine study sites (black) along the Truckee River (black line) on an urban to rural gradient across the cities of Reno and Sparks, Nevada. FIGURE 2. Hypothesized pathways between urban development and phenology. Predictor variables are labeled in boxes and response variables are open text. Arrows indicate the direction of the relationship between predictor and response variable. 122 123 FIGURE 3. Average daily temperature maximum at study sites in April and June 2011. Sites are arranged left to right according to their respective level of urban development. 124 FIGURE 4. The timing of bud burst in Populus sp. and Salix sp., peak abundance for aerial, aquatic, and foliar invertebrates, and clutch initiation in American Robin at sites across an urban to rural gradient. Sites are arranged left to right according to their respective level of urban development. 125 FIGURE 5. Path analysis modeling hypothesized relationships between predictor variables (in boxes) and the phenology of leaf out, peak invertebrate abundance, and robin lay date. The relative strength of associations between variables are indicated by line weight, based on standardized Beta coefficients. The residual variance (error) for each response variable are given in small boxes. Darker lines indicate a stronger relationship and dashed lines represent a negative relationship. 126 FIGURE 6. Maximum abundance of invertebrates reached at the timing of peak abundance in sites on a rural to urban gradient. Invertebrate guilds were aerial, foliar, or aquatic.