Nest Site Selection in Hudsonian Godwits: Effects of Habitat and Predation Risk

NEST SITE SELECTION IN HUDSONIAN GODWITS: EFFECTS OF HABITAT AND PREDATION RISK

A Thesis Presented to the Faculty of the Graduate School of Cornell University In Partial Fulfillment of the Requirements for the Degree of Master of Science

by Rose J. Swift February 2016

ABSTRACT

The choice of where to nest has enormous fitness consequences for individuals and, when scaled up, can even affect population demography. Nest site selection in birds is, therefore, guided by a wide variety of proximate and ultimate factors, including protection from predators, access to favorable microclimates, proximity to food resources, and costs/benefits stemming from interactions with neighboring con- and hetero-specifics. Nevertheless, the features associated with, cues used to identify, and availability of favorable nest sites may be changing rapidly with accelerating human development and climate change, especially in the Arctic and sub-Arctic. I investigated how habitat characteristics, predation risk, and proximity to human disturbance (i.e., roads) influenced nest site selection for the Hudsonian Godwit (Limosa haemastica). First, I identified habitat characteristics most associated with nests of Hudsonian Godwits in two breeding populations – Beluga River, Alaska and Churchill, Manitoba. I was particularly interested in habitat attributes that are (or will be) strongly impacted by climate change and human development. Second, I assessed the spatial pattern of nests in Beluga River, Alaska to evaluate if there was evidence of aggregation and then determined the extent to which nest location and fate were associated with predation risk and habitat features. Hudsonian Godwits displayed strong preferences for certain habitat features when selecting nest sites, with individuals preferring nest sites with less bare ground and more graminoid and tall shrubby cover than random locations. Nest site preferences were relatively similar in the two study populations, though birds in Churchill also preferred to nest near shallow water. Given the demonstrated preferences of nesting godwits, our study suggests that habitat degradation due to grubbing by overabundant Snow Geese (Chen caerulescens) and changing climate will reduce the suitability

and availability of habitats for breeding godwits. My study also provides evidence that, despite the importance of nesting habitat attributes, additional factors must be guiding nest site selection because nests of godwits were clustered within bogs. Because the spatial aggregation of nests was not explained by either underlying patchiness in habitat or by predation risk, I suggest that clustering may be related to social cues. Therefore, understanding behavioral processes can be as important as ecological requirements for habitat selection.

BIOGRAPHICAL SKETCH

Rose Swift was born in the Bay Area of California. Her parents, Brian and Karen, instilled a love of the outdoors in her from an early age. Camping when she was three months old and skiing when she was three years old, age was not a limitation to explore wilderness. She took that love for the outdoors with her as she began to explore universities. She went on to attend the University of California, Davis where she graduated with a B.S. in Wildlife, Fish, and

Conservation Biology with an emphasis on Wildlife Biology. Her first field position was camping at a remote lake in Alaska for three months. There she not only fell in love with the spectacular landscape, but also became engaged with the bird communities she was seeing.

Thanks to her field partner, Scarlett, an old pair of binoculars, and the trusty Sibley guide, she developed a passion for birds. At UC Davis, she engaged in many research projects and developed an honors thesis with Dr. John Wingfield investigating the phenological responses of arctic breeding birds to climate change. From these experiences launched a dream – to follow arctic breeding birds from the breeding grounds south and back north again.

After graduating from UC Davis, Rose went on to work many field positions where she gained a variety of ornithological skills. She worked for several seasons at Hastings

Natural History Reserve on Dr. Janis Dickinson’s long-term Western Bluebird (Sialia mexicana) behavioral ecology study. While there, she began to develop a project looking at the repeatability and heritability of egg size. From there, she flew to New Zealand to assist on a project looking at the breeding biology of North Island Brown Kiwi (Apteryx mantelli) where she had the opportunity to explore the biology of the endemic Kiwi Tick (Ixodes anatis) with Dr. Sarah Jamieson. From nest-searching in Montana to shorebird banding in California,

she began to explore the topics she would choose to study for her graduate work. Soon she would begin to engage in conversations with professors, ultimately finding her love of birds would send her heading to the Lab of Ornithology at Cornell University.

At Cornell, Rose has been involved in the Natural Resources and Lab of Ornithology communities. Under the guidance of Dr. Amanda Rodewald, she has been able to fulfill her dream to work with a long-distance migratory bird and travel to Alaska again. Rose will continue at Cornell for her Ph.D. continuing to work with Hudsonian Godwits on the breeding grounds, but she will spread her wings and begin to follow them from the non-breeding grounds in Chile north to Alaska.

For Garrett

For your blood, sweat, and encouragement on this project you deserve so much.

ACKNOWLEDGMENTS

First and foremost, I wish to thank Dr. Amanda Rodewald for her supervision, support and expeditious editing. This project has not been a small one, and her attention to details, strategic thinking, and support have been essential in continuing this project. Special thanks also go to Dr.

Nathan Senner for supervision, encouragement, and assistance from the beginning. I am indebted for his expertise, establishing this project, his years of data, and continued support. I also appreciate the contributions of my committee: Dr. John Fitzpatrick and Dr. Walter Koenig.

I am also grateful for the statistical advice of Dr. Patrick Sullivan and the Cornell

Statistical Consulting Unit. I am indebted to Brad Walker for his time and effort spent entering vegetation records and consulting with me on the floristics of the Churchill area. Additionally, I must thank Joe Barron and Katherine Hambury for their dedication to entering data.

I send thanks to my lab mates, who contributed in many ways: Gemara Gifford, Ruth

Bennett, Zephyr Mohr-Felson Züst, and Steven Sevillano Ríos for your input and assistance. And to all the scientists associated with the Bird Populations Studies and Conservation Sciences programs at the Lab of Ornithology for their expertise and advice.

This work would never have been possible without the many field assistants on the project. To William Abbott, Amy Alstad, Hope Batcheller, Shawn Billerman, Jon DeCoste,

Doug Gochfeld, Mike Harvey, Mike Hilchey, Tom Johnson, Andy Johnson, Julia Karagicheva,

Jess Marion, Madi McConnell, Jay McGowan, Brittany Schultz, Glenn Seeholzer, Hannah

Specht, Brad Walker, Ben Lagasse, Bret Davis, Kyle Parkinson, Justin Heseltine, James

Klarevas-Irby, and Garrett MacDonald. Thank you. Your hours of work, companionship, and good humor were not overlooked.

vii

To the administrative staff at the Cornell Lab of Ornithology, I owe you gratitude for your assistance. In Beluga, Judy and Larry Heilman continue to provide us a happy home; the staff at ConocoPhillips has made our work run smoothly.

Thanks to funding sources: the David and Lucile Packard Foundation, U.S. Fish and

Wildlife Service, Faucett Family Foundation, National Science Foundation, the Athena Fund at the Cornell Lab of Ornithology, the Atkinson Sustainable Biodiversity Fund at Cornell

University, Cornell Graduate School Mellon Fund, Cornell Chapter of Sigma Xi, American

Museum of Natural History, Ducks Unlimited Canada, Churchill Northern Studies Centre,

American Ornithologists’ Union, and Arctic Audubon Society.

Finally, I wish to thank my partner Garrett, for his patience and understanding. Without his constant and unwavering support this would not have been possible and he has poured many hours into data collection, data entry, editing, and brainstorming with me. And my family: Brian,

Karen, Lily, and Holly, for their support. My father deserves special thanks for reading through several edits. Thank you for instilling in me the confidence and curiosity that has gotten me here.

viii

TABLE OF CONTENTS

BIOGRAPHICAL SKETCH iv

ACKNOWLEDGEMENTS vii

TABLE OF CONTENTS ix

CHAPTER 1 1

LITERATURE CITED 6

CHAPTER 2 10

LITERATURE CITED 21

APPENDIX 31

CHAPTER 3 37

LITERATURE CITED 58

APPENDIX 93

CHAPTER ONE

INTRODUCTION

Breeding is a critical part of the avian annual cycle and can profoundly affect population dynamics. Choice of a nesting site, in particular, has consequences for both survival and reproductive success (Burger 1985, Martin 1992, Wiebe and Martin 1998), and nest preferences are generally assumed to have evolved to maximize fitness (Wiens 1989). For example, minimizing risk of nest predation, which is the primary cause of reproductive failure in birds, is an important driver of habitat selection, including nest sites (Martin 1993). Thus, patterns of nest placement are assumed to reflect the fitness benefits of nesting in preferred habitats (Cody 1981,

Wiens 1989, Martin 1998).

Species typically display strong nest site preferences because the success of a nest can be determined by a large suite of ecological factors, including availability of food (Lack 1968,

Martin 1987), nest microclimate (Walsberg 1985), and predation risk (Martin 1995). Across a wide range of species, studies have shown that nest microhabitats are selected non-randomly

(Colwell and Oring 1990, Rodrigues 1994, Clark and Shutler 1999, Smith et al. 2007). Although most studies of nest site selection examine the microhabitat surrounding nests (Martin 1998,

Clark and Shutler 1999), nest site selection clearly is also shaped by interactions with others species (Tarof and Ratcliffe 2004, Gascoigne and Lipcius 2004), including interactions with conspecific or heterospecific neighbors (Hildén 1965, Fretwell and Lucas 1970, Pitelka et al.

1974, Betts et al. 2008).

Birds may use a wide variety of cues to identify suitable and/or preferred nest sites.

Among cues that birds use in selecting nest sites are the presence of already settled individuals,

the configuration and quality of vegetation, and prior knowledge (Cody 1985). For example, public information on the number of fledged young or the presence of conspecifics in an area, may lead to increased rates of settlement in an area (Doligez et al. 2002). Understanding the role of each of these cues to habitat and nest site selection is critical for management and conservation given the importance of habitat selection on nest success.

One emergent property of nest site selection is the spatial patterning of nests. Spatial aggregations of nests can result from an underlying clustering of habitat features or in response to other species (Tarof and Ratcliffe 2004). In species where ecological resources influence settlement patterns, clusters could form where nesting sites or food resources were locally abundant. Individuals also may use the presence or reproductive performance of conspecifics to assess habitat quality when selecting a territory resulting in clustered territories (Kiester and

Slatkin 1974, Stamps 1988, Smith and Peacock 1990, Reed and Dobson 1993, Doligez et al.

2004a, Doligez et al. 2004b, Ward and Schlossberg 2004, Bourque and Desrochers 2006).

Further, aggregations could occur due to heterospecific attraction if other species are used as sources of information on habitat quality or if individuals gain protective benefits from nesting near other species (Bulmer 1984, Burger 1984, Goodale et al. 2010). Nesting in groups also can be a successful anti-predator defense strategy (i.e. through swamping effects due to densities of nests) (Hamilton 1971). As group size increases, the per capita predation risk to each individual declines, in part due to swamping but also because of earlier detection of predators and greater opportunities for group defense against approaching predators (Hamilton 1971, Martin 1988,

Martin 1993). Therefore, spatial aggregations that minimize predation risk can maximize an individual’s fitness. Ultimately, nest placement could be influenced by a variety of factors,

including the underlying spatial pattern of nests, and understanding what drives aggregations could have important conservation implications.

Though strong nest site preferences usually act to improve fitness, they can be a handicap in regions experiencing environmental changes. Such is the case for Arctic and sub-Arctic regions where climate change is predicted to severely impact the floristic and faunal communities in the Arctic and sub-Arctic (IPCC 2013). For example, higher summer temperatures and a longer frost-free season are predicted to increase permafrost thawing (Schurr et al. 2007), lengthen the growing season (Myneni et al. 1997, Jia et al. 2003, Goetz et al. 2005,

Bunn and Goetz 2006), and result in the northward expansion of shrubs (Sturm et al. 2001, Stow et al. 2004). Shrub encroachment and treeline advancement in the Churchill, Manitoba region has already eliminated some historic (1970s) breeding grounds for Whimbrel (Numenius phaeopus)

(Ballantyne and Nol 2015). Additionally, human disturbance associated with energy development introduces infrastructure, such as roads, wells, and buildings, that fragment habitat, limit the availability of nesting sites, and change the faunal composition of areas – such as subsidizing the population of Common Ravens (Corvus corax; Restani et al. 2001, Liebezeit et al. 2009). Roads in the Arctic are known to alter the vegetation communities in acidic soils, reducing species richness in half within 100 m of a dirt highway (Auerbach et al. 1997). Further, booming populations of Snow Geese (Chen caerulescens), supported by human activities in the mid-continental USA, degrade the vegetative communities around colonies for decades reducing the available graminoid habitat for nesting shorebirds (Peterson et al. 2013). Ultimately, climate change, development, and human-supported populations are changing habitats in the Arctic and sub-Arctic and could potentially alter the available breeding habitat for nesting birds.

One shorebird species that utilizes habitats in the Arctic and sub-Arctic is the Hudsonian

Godwit (Limosa haemastica, hereafter ‘godwit’). An extreme long-distance migrant, godwits breed in three disjunct regions across the Nearctic (Walker et al. 2011). A small population size and reliance upon threatened habitats have garnered godwits a status of high conservation concern (Senner 2010). Godwits have traditionally nested in sedge-dominated, Carex spp., meadows interspersed with trees, ponds, and upland areas, but the specific characteristics of nest sites are unknown (Walker et al. 2011). Occurrence and density of breeding individuals varies widely within and across suitable habitat as well as across their range, suggesting that non- habitat cues might factor prominently in the selection of breeding sites. The importance of non- habitat cues is further suggested by local absences in areas where sub-arctic/boreal wetlands are readily available and seemingly suitable (Walker et al. 2011, N. R. Senner personal observation).

Thesis Format:

I investigated how habitat features, predation risk, and proximity to human disturbance

(i.e. roads) influenced nest site selection for the Hudsonian Godwit. My thesis is organized in chapters that represent separate manuscripts for publication. In chapter two, I examined nest site selection of Hudsonian Godwits in Beluga River, Alaska and Churchill, Manitoba.

Understanding the microhabitat vegetation characteristics of godwit nests in two areas impacted differently by changing environments is critical for conservation recommendations for the species. In chapter three, I described the spatial patterning of godwit nests, and tested if habitat characteristics and predation risk explained spatial aggregations. Additionally, I investigated the relative influence of predation risk and proximity to roads on nest fate. Godwits are relatively uncommon across a broad range of presumably available suitable habitat. Understanding the

drivers of settlement decisions and habitat selection will inform conservation practitioners about whether unused habitats are due to the physical habitat or due to low population sizes of godwits.

LITERATURE CITED

Auerbach, N. A., M. D. Walker, and D. A. Walker (1997). Effects of roadside disturbance on substrate and vegetation properties in arctic tundra. Ecological Applications 7:218–235.

Ballantyne, K., and E. Nol (2015). Localized habitat change near Churchill, Manitoba and the decline of nesting Whimbrels (Numenius phaeopus). Polar Biology 38:529–537.

Betts, M. G., A. S. Hadley, N. Rodenhouse, and J. J. Nocera (2008). Social information trumps vegetation structure in breeding-site selection by a migrant songbird. Proceedings of the Royal Society of London B: Biological Sciences 275:2257–2263.

Bourque, J., and A. Desrochers (2006). Spatial Aggregation of Forest Songbird Territories and Possible Implications for Area Sensitivity. Avian Conservation and Ecology 1:3.

Bulmer, M. G. (1984). Risk avoidance and nesting strategies. Journal of Theoretical Biology 106:529–535.

Bunn, A. G., and S. J. Goetz (2006). Trends in satellite-observed circumpolar photosynthetic activity from 1982 to 2003: the influence of seasonality, cover type, and vegetation density. Earth Interactions 10:1–19.

Burger, J. (1984). Grebes nesting in gull colonies: Protective associations and early warning. The American Naturalist 123:327–337.

Burger, J. (1985). Habitat selection in temperate marsh-nesting birds. Habitat selection in birds (pp. 253–281). Academic Press

Clark, R. G., and D. Shutler (1999). Avian habitat selection: Pattern from process in nest-site use by ducks? Ecology 80:272–287.

Cody, M. L. (1981). Habitat selection in birds: the roles of vegetation structure, competitors, and productivity. BioScience 31:107–113.

Cody, M. L. (Ed.). (1985). Habitat selection in birds. Academic Press.

Colwell, M. A., and L. W. Oring (1990). Nest-site characteristics of prairie shorebirds. Canadian Journal of Zoology 68:297–302.

Doligez, B., E. Danchin, and J. Clobert (2002). Public information and breeding habitat selection in a wild bird population. Science 297:1168–1170.

Doligez, B., T. Pärt, E. Danchin, J. Clobert, and L. Gustafsson (2004a). Availability and use of public information and conspecific density for settlement decisions in the collared flycatcher. Journal of Animal Ecology 73:75–87. 6

Doligez, B., T. Pärt, and E. Danchin (2004b). Prospecting in the collared flycatcher: gathering public information for future breeding habitat selection? Animal Behaviour 67:457–466.

Fretwell, S. D., and H. L. Lucas (1970). On territorial behavior and other factors influencing habitat distribution in birds. Acta Biotheoretica 19:16–36.

Gascoigne, J. C., and R. N. Lipcius (2004). Allee effects driven by predation. Journal of Applied Ecology 41:801–810.

Goetz, S. J., A. G. Bunn, G. J. Fiske, and R. A. Houghton (2005). Satellite-observed photosynthetic trends across boreal North America associated with climate and fire disturbance. Proceedings of the National Academy of Sciences 102:13521–13525.

Goodale, E., G. Beauchamp, R. D. Magrath, J. C. Nieh, and G. D. Ruxton (2010). Interspecific information transfer influences animal community structure. Trends in Ecology & Evolution 25:354–361.

Hamilton, W. D. (1971). Geometry for the Selfish Herd. Journal of Theoretical Biology 31:295– 311.

Hildén, O. (1965). Habitat selection in birds: a review. Annales Zoologici Fennici 2:53–75.

IPCC Climate Change 2013: The Physical Science Basis. Available at: http://www.ipcc.ch/ipccreports/ar4-wg1.htm

Jia, G. J., H. E. Epstein, and D. A. Walker (2003). Greening of Arctic Alaska, 1981–2001. Geophysical Research Letters 30:2067.

Kiester, A. R., and M. Slatkin (1974). A strategy of movement and resource utilization. Theoretical Population Biology 6:1–20.

Lack, D. L. (1968). Ecological adaptations for breeding in birds. Meuthen. London.

Liebezeit, J. R., S. J. Kendall, S. Brown, C. B. Johnson, P. Martin, T. L. McDonald, D. C. Payer, C. L. Rea, B. Streever, A. M. Wildman, and S. Zack (2009). Influence of human development and predators on nest survival of tundra birds, Arctic Coastal Plain, Alaska. Ecological Applications 19:1628–1644.

Martin, T. E. (1987). Food as a limit on breeding birds: a life-history perspective. Annual review of ecology and systematics 18:453–487.

Martin, T. E. (1988). Processes organizing open-nesting bird assemblages: competition or nest predation? Evolutionary Ecology 2:37–50.

Martin, T. E. (1992). Interaction of nest predation and food limitation in reproductive strategies. Current Ornithology (pp. 163-197). Springer US.

Martin, T. E. (1993). Nest predation and nest sites: new perspectives on old patterns. BioScience 43:523–532.

Martin, T. E. (1995). Avian life history evolution in relation to nest sites, nest predation, and food. Ecological monographs 65:101–127.

Martin, T. E. (1998). Are microhabitat preferences of coexisting species under selection and adaptive? Ecology 79:656–670.

Myneni, R. B., C. D. Keeling, C. J. Tucker, G. Asrar, and R. R. Nemani (1997). Increased plant growth in the northern high latitudes from 1981 to 1991. Nature 386:698–702.

Peterson, S. L., R. F. Rockwell, C. R. Witte, and D. N. Koons (2013). The legacy of destructive Snow Goose foraging on supratidal marsh habitat in the Hudson Bay lowlands. Arctic, Antarctic, and Alpine Research 45:575–583.

Pitelka, F. A., R. T. Holmes, and S. F. MacLean (1974). Ecology and evolution of social organization in arctic sandpipers. American Zoologist 14:185–204.

Reed, J. M., and A. P. Dobson (1993). Behavioural constraints and conservation biology: conspecific attraction and recruitment. Trends in Ecology and Evolution 8:253–256.

Restani, M., J. M. Marzluff, and R. E. Yates (2001). Effects of anthropogenic food sources on movements, survivorship, and sociality of Common Ravens in the Arctic. The Condor 103:399–404.

Rodrigues, R. (1994). Microsite variables influencing nest-site selection by tundra birds. Ecological Applications 4:110–116.

Schuur, E. A., K. G. Crummer, J. G. Vogel, and M. C. Mack (2007). Plant species composition and productivity following permafrost thaw and thermokarst in Alaskan tundra. Ecosystems 10:280–292.

Senner, N. R. (2010). Conservation Plan for the Hudsonian Godwit. Version 1.1. Manomet Center for Conservation Science, Manomet, MA.

Smith, A. T., and M. M. Peacock (1990). Conspecific attraction and the determination of metapopulation colonization rates. Conservation Biology 4:320–323.

Smith, P. A., G. Gilchrist, H. G., and J. N. M. Smith (2007). Effects of nest habitat, food, and parental behavior on shorebird nest success. The Condor 109:15–31.

Stamps, J. A. (1988). Conspecific attraction and territorial aggregation: a field experiment. The American Naturalist 131:329–347.

Stow, D. A., A. Hope, D. McGuire, D. Verbyla, J. Gamon, F. Huemmrich, S. Houston, C. Racine, M. Sturm, K. Tape, L. Hinzman, K. Yoshikawa, C. Tweedie, B. Noyle, C. Silapaswan, D. Douglas, B. Griffith, G. Jia, H. Epstein, D. Walker, S. Daeschner, A. Petersen, L. Zhou, and R. Myneni (2004). Remote sensing of vegetation and land-cover change in Arctic tundra ecosystems. Remote Sensing of Environment 89:281–308.

Sturm, M., C. Racine, and K. Tape (2001). Climate change - increasing shrub abundance in the Arctic. Nature 411:546–547.

Tarof, S. A., and L. M. Ratcliffe (2004). Habitat characteristics and nest predation do not explain clustered breeding in least flycatchers (Empidonax minimus). The Auk 121:877–893.

Walker, B. M., N. R. Senner, C. S. Elphick, and J. Klima (2011). Hudsonian Godwit (Limosa haemastica), The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology; Retrieved from the Birds of North America Online: http://bna.birds.cornell.edu/bna/species/629

Walsberg, G. E. (1985). Physiological consequences of microhabitat selection. Habitat selection in birds (pp. 389–413). Academic Press.

Ward, M. P., and S. Schlossberg (2004). Conspecific attraction and the conservation of territorial songbirds. Conservation Biology 18:519–525.

Wiebe, K. L., and K. Martin (1998). Costs and benefits of nest cover for ptarmigan: changes within and between years. Animal Behaviour 56:1137–1144.

Wiens, J. A. (1989). Spatial scaling in ecology. Functional ecology 3:385–397.

CHAPTER TWO

NEST SITE SELECTION OF HUDSONIAN GODWITS IN A CHANGING ENVIRONMENT

Abstract:

Changes to arctic and sub-arctic environments due to climate change and human activities are expected to alter their faunal and floristic communities. Woody shrub encroachment into previously open habitats is projected with climate change, and booming populations of Snow Geese (Chen caerulescens) create vast areas of bare mud in the Eastern

Canadian Arctic. These changes are likely to influence availability of nest sites, cues utilized to select favorable sites, and attributes of nest sites that are most favorable. The Hudsonian Godwit

(Limosa haemastica) is one of a suite of arctic- and sub-arctic-nesting shorebirds for which changing environments may cause a reduction in breeding habitat. Nest site selection of godwits was studied at two study sites – one near Churchill, Manitoba, Canada, from 2008-2011 and the other near Beluga River, Alaska, USA, from 2009-2011 and 2014-2015 – to determine habitat preferences and the degree to which the species avoided woody vegetation and non-vegetated areas. We used a multiple analysis of variance (MANOVA) to test for significant microhabitat differences between nest sites (Alaska n = 107, Manitoba n = 75) and random sites (Alaska n =

670, Manitoba n = 310) within each study area. Godwits selected nest sites characterized by high amounts of graminoid and tall shrubby cover with little non-vegetated area. Worrisomely, the habitat attributes that godwits preferred at our study sites are expected to be less available as climate changes. Likewise, avoidance of non-vegetated areas suggests that high densities of

Snow Geese may further degrade breeding habitat for godwits. Our study suggests that habitat degradation due to grubbing by overabundant Snow Geese and changing climate are likely to

reduce the suitability and availability of habitats for breeding godwits.

Introduction:

In the Arctic and sub-Arctic, two major threats to habitat are climate change and large- scale human activities (Brown et al. 2001, Saalfeld et al. 2013). Climatic change can alter landscapes through a variety of processes and are likely to have profound impacts on physical and ecological variables (e.g., surface water, vegetation community, and insect community), which will in turn affect habitats in the Arctic and sub-Arctic (Post et al. 2009, IPCC 2013). For example, higher summer temperatures, changes in precipitation regimes, and a longer frost-free season are predicted to accelerate permafrost thawing (Schuur et al. 2007), lengthen the growing season (Myneni et al. 1997, Jia et al. 2003, Goetz et al. 2005, Bunn and Goetz 2006), dry wetlands (Kaplan and New 2006), and promote the northward expansion of shrubs (Sturm et al.

2001, Stow et al. 2004) and faunal and parasite communities (Piersma 1997, Tannerfeldt et al.

2002, Parmesan and Yohe 2003, Kutz et al. 2005, Hudson et al. 2006, Post et al. 2009).

Predictions of global climate change suggest that novel communities characterized by different species compositions and structure will dominate arctic and sub-arctic habitats.

In addition to a changing climate, habitats in the Arctic and sub-Arctic are influenced by human modifications and development. Landscapes can be modified by a wide variety of human activities, including direct development (e.g. oil pads, roads) and the presence of human- subsidized or facilitated species, such as Common Ravens (Corvus corax, Liebezeit et al. 2009,

Restani et al. 2001) and Snow Geese (Chen caerulescens, Abraham et al. 2005a). Changes in species interactions can also affect the habitat directly in some areas of the Arctic (Peterson et al.

2013). For example, Snow Goose grubbing can alter soil chemistry, and overgrazed areas may

take decades or longer to recover (Peterson et al. 2013). In this way, increased numbers of Snow

Geese may have lasting impacts on floristic communities and, especially, suitable habitat for ground-nesting Arctic birds.

Global climate change has been hypothesized to explain declining population trends for many species of shorebirds (Morrison et al. 2001, Brown et al. 2001, Piersma and Lindström

2004, Morrison et al. 2006, Bart et al. 2007), in part because it may limit available habitat for breeding (Zöckler and Lysenko 2000). Many shorebirds exhibit preferences to nest within or near specific habitats and vegetation characteristics, and those traits are generally thought to be associated with higher nest success (Colwell and Oring 1990, Smith et al. 2007, Skrade and

Dinsmore 2013). Lack of those preferred attributes may even contribute to population declines

(Kentie et al. 2015) or compromise population viability (Martin 1993, Newton 1998). Therefore, understanding nest site selection can help us to better interpret drivers of population trends, especially for species of conservation concern.

Habitat alterations caused by changing environments could, in turn, affect nest site selection and nest survival, and ultimately population trends, in three ways. First, changing climate and faunal communities could affect the type of nest that is most favorable. For instance, with longer ice-free seasons, polar bears (Ursus maritimus) depredate colonial nesting birds more often (Smith et al. 2010, Iverson et al. 2014, Prop et al. 2015). Thus, there may be a stronger selective pressure to nest on inaccessible cliffs or in less dense colonies to avoid polar bear predation. Second, changing climate could disrupt the cues used to identify nest sites. For example, in cold environments, such as the Arctic, microclimates that are more amenable to embryo development and that reduce energetic costs to the parents during incubation may be favored for nest sites (Tulp et al. 2012). As temperatures rise, however, warm sites may become

less advantageous and potentially create ecological traps if birds continue to rely upon the same cues (Martin 2001, Battin 2004). Third, changing landscapes could limit or eliminate the availability of particular characteristics individual species select for in nesting locations. Shrub encroachment since the 1970s has already eliminated some historic sub-Arctic breeding areas for

Whimbrel (Numenius phaeopus), reducing local breeding densities (Ballantyne and Nol 2015).

In short, changing environments may dramatically influence nest site selection and survival.

One shorebird species that utilizes habitats in the Arctic and sub-Arctic is the Hudsonian

Godwit (Limosa haemastica; hereafter ‘godwits’). An extreme long-distance migrant, godwits breed in three disjunct regions across the Nearctic (Walker et al. 2011). A small population size and reliance upon threatened habitats have garnered godwits a status of high conservation concern (Senner 2010). Additionally, there is some indication that the Eastern Canadian breeding population is declining (Bart et al. 2007), and previous research has shown that godwits breeding in at least some parts of this region have a much lower hatching success rate than do godwits breeding elsewhere in the species’ range (Senner 2013). Factors thought to contribute to low reproduction include alteration of breeding habitat due to climate change (Tape et al. 2006), creation of nesting structures for Common Ravens, and increased foraging by Snow Geese

(Sammler et al. 2008). Godwits have traditionally nested in sedge-dominated meadows at the interface of forest, but the specific characteristics of nest sites are unknown (Walker et al. 2011).

We studied the nest site selection of Hudsonian Godwits in two disjunct breeding populations: Beluga River, Alaska, and Churchill, Manitoba. The Churchill area is experiencing dramatic habitat changes due to woody shrub encroachment and overgrazing by Snow Geese. In contrast, Beluga River’s habitat has thus far remained relatively stable. Our study aimed to identify the habitat attributes within a one meter diameter circle around nests selected by

breeding godwits. We predicted that nesting godwits would choose to nest in areas dominated by non-woody vegetation and avoid non-vegetated areas. Because areas with woody vegetation as well as patches of mud are expected to increase in availability, the suitability of nesting habitat for godwits is expected to decline in the future as climate change continues.

Methods:

Study Areas:

We monitored the breeding effort of godwits at two study sites representing the geographic centers of the south-central Alaska and Hudson Bay breeding populations: Beluga

River, Alaska, USA (61.21°N, 151.03°W), and Churchill, Manitoba, Canada (58.93°N,

93.80°W). Research in Churchill was conducted in roughly a 15-km2 area from 2008-2011; the

Beluga River study area (8 km2) was surveyed from 2009-2011, and again from 2014-2015. Both areas support large breeding populations of godwits (5.0 breeding pairs per km2 at Beluga River, and 2.3 pairs per km2 at Churchill; N.R. Senner unpubl. data). Both study sites were dominated by sedges, Carex spp., and dwarf birch, Betula glandulosa/nana, although they appear different – the Churchill study site was located on a large, open fen, while the Beluga River site was situated in a black spruce, Picea mariana, dominated muskeg bog (Walker et al. 2011, Senner 2013).

Field Methods:

Nest Searching and Habitat Characteristics:

Field teams began nest searching after pair formation and continued through the onset of hatch. We located nests through knowledge of former territories, systematic walking surveys, and observing behavioral cues of nesting pairs.

After nests had fledged or failed, we measured microhabitat characteristics at the nest site

and associated random points. We defined the microhabitat (nest site) scale as the area within a 1 m diameter circle centered on the nest. In each territory, we placed ten 1 m diameter circular plots at randomly selected points within a 200 m radius of the nest.

From the center of each circular plot, we measured the distance to the nearest tree (≥ 2 m tall) and nearest water (≥ 2 cm deep). Following Viereck et al.’s (1992) classification, we also estimated percent cover within the plot by trees (≥ 2 m tall), saplings (1 m

(30 cm

Statistical Methods:

We used a multiple analysis of variance (MANOVA) criterion to test for significant microhabitat differences between nest sites and random sites. We reduced risk of multicolinearity by either combining or removing one pair of any highly correlated variables

(Pearson’s r>0.7; Sokal and Rohlf 1995). Additionally, we compared the characteristics of the nest cup and mound between study populations using an additional MANOVA test.

Results:

We found 75 godwit nests at Churchill and 107 in Beluga River. Overall, vegetation structure differed significantly between nest sites and random points (Beluga River: Wilk's λ =

0.775, p= 0.001; Churchill: Wilk's λ = 0.699, p= 0.001). Only 6.5% of located nests were associated with tall trees providing cover, and no nests were found within 0.5 m of a tree (>2 m tall).

In Beluga River, godwits selected nest sites where the closest water was shallower than random sites (Table 1). Nest sites had 14% more tall shrubby cover (between 30 cm and 1 m tall), 6% less bare ground, and 9% fewer graminoids (less than 30 cm tall) than available at random points. Nest sites were disproportionately comprised of sedges, grasses, and forbs, and fewer low-mat species (moss, fungi, and lichen) than random sites. Only one pair of godwits nested where >35% of the circular plot was bare ground comprised of mud, water or rocks.

Lastly, nest sites were more floristically diverse than available sites, with three more species, on average, found at nest sites (Table 1).

In Churchill, godwit nest sites were also closer to shallow water than expected, and the diameter at breast height of the closest tree was smaller than available. Nest sites had 6% more tall shrubby cover (between 30 cm and 1 m tall) and 8% less mud than was available at random points (Table 2). Nest areas had 2% more forbs and low-lying mat species, and nearly five more species than at random points (Table 2). No godwits nested in areas with >50% tall shrubby cover, and only two pairs nested where >20% of the circular plot was comprised of mud caused by Snow Goose grubbing.

Nest characteristics of godwits were similar in Beluga River and Churchill (Table 3).

The width of the mound was smaller by 20 cm and the nest cup was 2 cm narrower and 1.5 cm deeper in Churchill than in Beluga River.

Discussion:

Hudsonian Godwits selected specific vegetative features when choosing nest sites. In particular, godwits at our study sites selected species rich areas with high graminoid and tall shrubby cover and comparatively low amounts of open mud, water, or rocks. These results are consistent with the literature on known breeding habitats of godwits (Walker et al. 2011,

Harwood 2014) and are similar to reports of nonrandom nest site selection in many other species of Arctic and sub-Arctic breeding shorebirds, such as Whimbrel (Numenius phaeopus,

Ballantyne and Nol 2011), Terek Sandpiper (Xenus cinereus, Meissner et al. 2013),

Semipalmated Sandpiper (Calidris pusilla, Rodrigues 1994), White-rumped Sandpiper (C. fuscicollis, Smith et al. 2007), Ruddy Turnstone (Arenaria interpres, Smith et al. 2007), Red

Phalarope (Phalaropus fulicarius, Smith et al. 2007), Red-necked Phalarope (P. lobatus,

Rodrigues 1994, Walpole et al. 2008), Semipalmated Plover (Charadrius semipalmatus, Nguyen et al. 2003, 2013, Smith et al. 2007), and Black-bellied Plover (Pluvialis squatarola, Smith et al.

2007). For some of these species, shrub encroachment and treeline advancement have already limited available nest sites and dramatically decreased nesting densities in some areas

(Ballantyne and Nol 2011, Ballantyne and Nol 2015). However, because godwit nest sites already incorporate some woody vegetation, the consequences of future changes in sub-Arctic and Arctic habitats for godwit are unclear.

Projections of global climate change suggest that the availability of suitable nest habitats for Hudsonian Godwits will decline in the future; though, the interplay between factors introduces a level of uncertainty. Various scenarios of climate change predict the drying of

Arctic ponds (Yoshikawa and Hinzman 2003, Smith et al. 2005), a decline in graminoid habitats

(Chapin et al. 1995, Kaplan and New 2006), increased shrub cover (Chapin et al. 1995, Sturm et al. 2005a, Sturm et al. 2005b, Tape et al. 2006), and treeline advancement (Caccianiga and

Payette 2006, Kaplan and New 2006, Danby and Hik 2007). Thus, the habitat attributes that godwits preferred in our study areas, such as shallow bodies of water and graminoid habitats, are expected to decline under these scenarios, which may limit the amount of suitable nesting habitat. Such a reduction in suitable nesting habitat could increase competition for nest sites and limit populations via territoriality, potentially increase nest predation if habitat changes improve search efficiency by predators, or otherwise increase risk of predation (Martin 1993). However, future shrub encroachment caused by climate change may provide more suitable nest areas for godwits if the shrubs are between 30 cm and 1 m tall, which often comprised a large percentage of nest sites. Therefore, the effect on the availability of suitable nest sites is unclear due to the potential interplay between the effects of climate change.

Both populations of Hudsonian Godwits avoided non-vegetated or barren areas, especially those caused by artificially-supported Snow Geese populations in the Eastern

Canadian Arctic. Driven by agricultural modifications and associated nutrient subsidies on their wintering grounds and migratory stopover areas (Boyd et al. 1982, Abraham et al. 2005a), the mid-continent population of Snow Geese has soared at an annual rate of 5–14% since the 1970s

(Alisauskas et al. 2011). This human-subsidized Snow Goose population degrades arctic habitats through overgrazing and grubbing of vegetation, which ultimately eliminates plants and alters the soil chemistry, leading to large patches of barren ground/mud and extensive loss of graminoids and shrubs (Jefferies and Rockwell 2002, Abraham et al. 2005b). These habitat alterations can have lasting impacts on the vegetation community long after the initial vegetation removal occurs. For example, Peterson et al. (2013) documented a 46% reduction in graminoid cover and an 84% reduction in shrub cover 35 years after a Snow Goose colony stopped breeding in their study area. This habitat alteration has been shown to lead to smaller and more isolated

patches of the shrubs and graminoids upon which many shorebirds depend for nesting (Peterson et al. 2013, Flemming et al. 2015). Indeed, the low hatching success of the Churchill population of godwits, which largely avoided areas of bare mud caused by Snow Goose grubbing (Senner

2013), suggests that overabundant Snow Geese may have important demographic consequences for godwits.

Changes in habitat ultimately could reduce nest survival, especially by increasing risk of predation. Others have suggested that many shorebirds nest in exposed sites to facilitate early detection of predators (Maclean 1984, Burger 1987, Lauro and Nol 1995), as many adult shorebirds do not use cover in the face of predation but rather take flight. In this case, antipredator or cryptic behaviors of adults at the nest, rather than the characteristics of the nest site itself, may have an overriding effect on nest success (Cresswell 1997, Ghalambor and Martin

2002). Nesting where vegetation obstructs vision may hamper detection of predators, allowing the incubating adult to be caught by surprise and compromising their own survival (Götmark et al. 1995), and open habitats often are comprised of a greater percentage of aggressive species than habitats offering cover (Larsen et al. 1996). Thus, species that nest in open habitats prioritize detecting predators early, and many engage in aggressive mobbing behaviors (Metcalfe

1984, Lima 1995, Götmark et al. 1995, Larsen et al. 1996, Koivula and Rönkä 1998). However, other species may use camouflage to decrease the risk of predation, rather than taking flight and mobbing predators. Godwits selected sites with higher percent cover of graminoids and tall woody cover than at random sites, which could obstruct their detection to predators through disruptive camouflage. Shrubs are often used as overhead cover of nest sites (R. Swift personal obs.), which may serve to camouflage incubating adults from their main predator, Northern

Harriers (Circus cyaneus). Camouflage provided by shrubs may be especially important for

Hudsonian Godwits because they typically do not defend nests, and rather rely on crypsis to avoid detection (R. Swift personal obs.).

Although our findings suggest that anticipated changes in climate and habitat may reduce the suitability or availability of nest sites for godwits, there are important caveats. One, because we could not measure every possible attribute that might be selected by breeding godwits, there remains the possibility that godwits respond to habitat features or cues not included in our study.

Two, a wide variety of other factors, not necessarily related to vegetation, also can influence territory and nest site selection. Indeed, others have demonstrated that invertebrate and/or predator abundance, conspecific attraction, site fidelity, previous experience, individual specialization, and natal habitat preferences can shape habitat selection in birds (Van Horne

1983, Block and Brennan 1993, Ramsay et al. 1999, Bolnick et al. 2003, Battin 2004, Davis and

Stamps 2004). Three, as we did not study direct links among nest-site attributes, reproduction, survival, or site fidelity we can only infer the implications of our findings to godwits. Future work should address the relationship between nest placement and survival of nests or incubating adults to better understand potential fitness consequences.

Our study suggests that habitat degradation due to grubbing by overabundant Snow

Geese and changing climate are likely to reduce the suitability of habitats for breeding godwits.

Additional management for Snow Geese numbers or exclusions may provide more nesting habitat near Snow Goose colonies. The projected woody shrub encroachment, advancement in treeline, and drying of ponds due to climate change will alter the amount of suitable habitats for nesting godwits, but the interplay among these factors is unclear and warrants further study.

LITERATURE CITED

Abraham, K. F., R. L. Jefferies, and R. T. Alisauskas (2005a). The dynamics of landscape change and snow geese in mid-continent North America. Global Change Biology 11:841– 855.

Abraham, K. F., R. L. Jefferies, and R. F. Rockwell (2005b). Goose-induced changes in vegetation and land cover between 1976 and 1997 in an Arctic coastal marsh. Arctic, Antarctic, and Alpine Research 37:269–275.

Alisauskas, R. T., R. F. Rockwell, K. W. Dufour, E. G. Cooch, G. Zimmerman, K. L. Drake, J. O. Leafloor, T. J. Moser, and E. T. Reed (2011). Harvest, survival, and abundance of midcontinent lesser snow geese relative to population reduction efforts. Wildlife Monographs 179:1–42.

Ballantyne, K., and E. Nol (2011). Nesting Habitat Selection and Hatching Success of Whimbrels Near Churchill, Manitoba, Canada. Waterbirds 34:151–159.

Ballantyne, K., and E. Nol (2015). Localized habitat change near Churchill, Manitoba and the decline of nesting Whimbrels (Numenius phaeopus). Polar Biology 38:529–537.

Bart, J., S. Brown, B. Harrington, and R. I. G. Morrison (2007). Survey trends of North American shorebirds: population declines or shifting distributions? Journal of Avian Biology 38:73–82.

Battin, J. (2004). When good animals love bad habitats: ecological traps and the conservation of animal populations. Conservation Biology 18:1482–1491.

Block, W. M., and L. A. Brennan (1993). The habitat concept in ornithology: theory and applications. Pages 35-91 in D. M. Power, editor. Current Ornithology. Plenum Press, New York, N.Y.

Bolnick, D. I., R. S. Svanbäck, J. F. Fordyce, L. H. Yang, J. M. Davis, C. D. Hulsey, and M. L. Forister (2003). The ecology of individuals: incidence and implications of individual specialization. The American Naturalist 161:1–28.

Boyd, H., G. J. Smith, and F. G. Cooch (1982). The lesser snow goose of the eastern Canadian Arctic: their status during 1964–1979 and their management from 1982 to 1990. Canadian Wildlife Service Occasional Paper 46:1–21.

Brown, S., C. Hickey, B. Harrington, and R. E. Gill, Jr. (Eds.) (2001). United States Shorebird Conservation Plan, 2nd ed. Manomet Center for Conservation Sciences, Manomet, MA.

Burger, J. (1987). Physical and social determinants of nest-site selection in Piping Plover in New Jersey. The Condor 89:811–818.

Caccianiga, M., and S. Payette (2006). Recent advance of white spruce (Picea glauca) in the coastal tundra of the eastern shore of Hudson Bay (Québec, Canada). Journal of Biogeography 33:2120–2135.

Chapin, F. S., G. R. Shaver, A. E. Giblin, K. J. Nadelhoffer, and J. A. Laundre (1995). Responses of arctic tundra to experimental and observed changes in climate. Ecology 76:694–711.

Colwell, M. A., and L. W. Oring (1990). Nest-site characteristics of prairie shorebirds. Canadian Journal of Zoology 68:297–302.

Cresswell, W. (1997). Nest predation rates and nest detectability in different stages of breeding in Blackbirds Turdus merula. Journal of Avian Biology 28:296–302.

Danby, R. K., and D. S. Hik (2007). Variability, contingency and rapid change in recent subarctic alpine tree line dynamics. Journal of Ecology 95:352–363.

Davis, J. M., and J. A. Stamps (2004). The effect of natal experience on habitat preferences. Trends in Ecology and Evolution 19:411–416.

Flemming, S. A., P. A. Smith, and E. Nol (2015). Impacts of Goose Colonies on Bird Community Composition and Abundance. Conference Paper: 13th North American Arctic Goose Conference, at Winnipeg.

Ghalambor, C. K., and T. E. Martin (2002). Comparative manipulation of predation risk in incubating birds reveals variability in the plasticity of responses. Behavioral Ecology 13:101–108.

Götmark, F., D. Blomqvist, O. C. Johansson, J. Bergkvist (1995). Nest site selection: a trade-off between concealment and view of the surroundings? Journal of Avian Biology 26:305–312.

Harwood, C. M. (2014). Intraseason re-use of Numenius nest by Limosa. Wader Study Group Bulletin 121:199–200.

Hudson, P. J., I. M. Cattadori, B. Boag, and A. P. Dobson (2006). Climate disruption and parasite-host dynamics: patterns and processes associated with warming and the frequency of extreme climatic events. Journal of Helminthology 80:175–182.

IPCC Climate Change 2013: The Physical Science Basis. Available at: http://www.ipcc.ch/ipccreports/ar4-wg1.htm

Iverson, S. A., H. G. Gilchrist, P. A. Smith, A. J. Gaston, and M. R. Forbes (2014). Longer ice- free seasons increase the risk of nest depredation by polar bears for colonial breeding birds in the Canadian Arctic. Proceedings of the Royal Society of London B: Biological Sciences 281:20133128.

Jefferies, R. L., and R. F. Rockwell (2002). Foraging geese, vegetation loss and soil degradation in an Arctic salt marsh. Applied Vegetation Science 5:7–16.

Jia, G. J., H. E. Epstein, and D. A. Walker (2003). Greening of Arctic Alaska, 1981–2001. Geophysical Research Letters 30:2067.

Kaplan, J. O., and M. New (2006). Arctic climate change with a 2 °C global warming: timing, climate patterns and vegetation change. Climatic Change 79:213–241.

Kentie, R., C. Both, J. C. Hooijmeijer, and T. Piersma (2015). Management of modern agricultural landscapes increases nest predation rates in Black!tailed Godwits Limosa limosa. Ibis 157:614–625.

Koivula, K., and A. Rönkä (1998). Habitat deterioration and efficiency of antipredator strategy in a meadow-breeding wader, Temminck's stint (Calidris temminckii). Oecologia 116:348–355.

Kutz, S. J., E. P. Hoberg, L. Polley, and E. J. Jenkins (2005). Global warming is changing the dynamics of Arctic host-parasite systems. Proceedings of the Royal Society of London Series B: Biological Sciences 272:2571–2576.

Larsen, T., T. A. Sordahl, and I. Byrkjedal (1996). Factors related to aggressive nest protection behaviour: a comparative study of Holarctic waders. Biological Journal of the Linnean Society 58:409–439.

Lauro, B., and E. Nol (1995). Patterns of habitat use for pied and sooty oystercatchers nesting at the Furneaux Islands, Australia. The Condor 97:920–934.

Lima, S. L. (1995). Back to the basics of anti-predatory vigilance: the group-size effect. Animal Behaviour 49:11–20. 23

Maclean, G. L. (1984). Arid-zone adaptations of waders (Aves: Charadrii). South African Journal of Zoology 19:78–81.

Martin, T. E. (1993). Nest predation and nest sites: new perspectives on old patterns. BioScience 43:523–532.

Martin, T. E. (2001). Abiotic vs. biotic influences on habitat selection of coexisting species: climate change impacts? Ecology 82:175–188.

Meissner, W., M. Golovatin, and S. Paskhalny (2013). Geographical Differences in Nesting Habitats of Terek Sandpiper (Xenus cinereus). The Wilson Journal of Ornithology 125:811– 815.

Metcalfe, N. B. (1984). The effects of habitat on the vigilance of shorebirds: is visibility important? Animal Behaviour 32:981–985.

Morrison, R. I. G., R. E. Gill Jr., B. A. Harrington, S. K. Skagen, G. W. Page, C. L. Gratto- Trevor, and S. M. Haig (2001). Estimates of shorebird populations in North America (No. 104, pp. 0-64). Canadian Wildlife Service.

Morrison, R. I. G., B. J. McCaffery, R. E. Gill, S. K. Skagen, S. L. Jones, G. W. Page, C. L. Gratto-Trevor, and B. A. Andres (2006). Population estimates of North American shorebirds, 2006. Wader Study Group Bulletin 111:66–84.

Myneni, R. B., C. D. Keeling, C. J. Tucker, G. Asrar, and R. R. Nemani (1997). Increased plant growth in the northern high latitudes from 1981 to 1991. Nature 386:698–702.

Newton, I. (1998). Population limitation in birds. Academic Press, San Diego, CA.

Nguyen, L. P., E. Nol, and K. F. Abraham (2003). Nest success and habitat selection of the Semipalmated Plover on Akimiski Island, Nunavut. The Wilson Bulletin 115:285–291.

Nguyen, L. P., E. Nol, K. F. Abraham, and C. Lishman (2013). Directional selection and repeatability in nest-site preferences of Semipalmated Plovers (Charadrius semipalmatus). Canadian Journal of Zoology 91:646–652.

Parmesan, C., and G. Yohe (2003). A globally coherent fingerprint of climate change impacts across natural systems. Nature 421:37–42.

Piersma, T. (1997). Do global patterns of habitat use and migration studies co-evolve with relative investments in immunocompetence due to spatial variation in parasite pressure? Oikos 90:623–631.

Piersma, T., and Å. Lindström (2004). Migrating shorebirds as integrative sentinels of global environmental change. Ibis 146:61–69.

Post, E., M. C. Forchhammer, M. S. Bret-Harte, T. V. Callaghan, T. R. Christensen, B. Elberling, A. D. Fox, O. Gilg, D. S. Hik, T. T. Høye, R. A. Ims, E. Jeppesen, D. R. Klein, J. Madsen, A. D. McGuire, S. Rysgaard, D. E. Schindler, I. Stirling, M. P. Tamstorf, N. J. C. Tyler, R. van der Wal, J. Welker, P. A. Wookey, N. M. Schmidt, and P. Aastrup (2009). Ecological dynamics across the Arctic associated with recent climate change. Science 325:1355–1358.

Prop, J., J. Aars, B. J. Bårdsen, S. A. Hanssen, C. Bech, S. Bourgeon, J. de Fouw, G. W. Gabrielsen, J. Lang, E. Noreen, T. Oudman, B. Sittler, L. Stempniewicz, I. Tombre, E. Wolters, and B. Moe (2015). Climate change and the increasing impact of polar bears on bird populations. Frontiers in Ecology and Evolution 3:33.

R Development Core Team (2015). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.

Ramsay, R. J., K. Otter, and L. M. Ratcliffe (1999). Nest-site selection by Black-capped Chickadees: settlement based on conspecific attraction. The Auk 116:604–617.

Restani, M., J. M. Marzluff, and R. E. Yates (2001). Effects of anthropogenic food sources on movements, survivorship, and sociality of common ravens in the arctic. The Condor 103:399–404.

Rodrigues, R. (1994). Microsite variables influencing nest-site selection by tundra birds. Ecological Applications 4:110–116.

Saalfeld, S. T., R. B. Lanctot, S. C. Brown, D. T. Saalfeld, J. A. Johnson, B. A. Andres, and J. R. Bart (2013). Predicting breeding shorebird distributions on the Arctic Coastal Plain of Alaska. Ecosphere 4:16.

Sammler, J. E., D. E. Andersen, and S. K. Skagen (2008). Population trends of tundra-nesting birds at Cape Churchill, Manitoba, in relation to increasing goose populations. The Condor 110:325–334.

Schuur, E. A., K. G. Crummer, J. G. Vogel, and M. C. Mack (2007). Plant species composition and productivity following permafrost thaw and thermokarst in Alaskan tundra. Ecosystems 10:280–292.

Senner, N. R. (2010). Conservation Plan for the Hudsonian Godwit. Version 1.1. Manomet Center for Conservation Science, Manomet, MA.

Senner, N. R. (2013). The Effects of Global Climate Change on Long-Distance Migratory Birds. PhD Dissertation. Cornell University.

Skrade, P. D. B., and S. J. Dinsmore (2013). Egg crypsis in a ground-nesting shorebird influences nest survival. Ecosphere 4:151.

Smith, L. C., Y. Sheng, G. M. MacDonald, and L. D. Hinzman (2005). Disappearing arctic lakes. Science 308:1429.

Smith, P. A., G. Gilchrist, H. G., and J. N. M. Smith (2007). Effects of nest habitat, food, and parental behavior on shorebird nest success. The Condor 109:15–31.

Smith, P. A., K. H. Elliott, A. J. Gaston, and H. G. Gilchrist (2010). Has early ice clearance increased predation on breeding birds by polar bears? Polar Biology 33:1149–1153.

Sokal, R. R., and F. J. Rohlf (1995). Biometry: the principles and practice of statistics in biological research. W.H. Freeman, New York.

Sturm, M., C. Racine, and K. Tape (2001). Climate change - increasing shrub abundance in the Arctic. Nature 411:546–547.

Sturm, M., T. Douglas, C. Racine, and G. E. Liston (2005a). Changing snow and shrub conditions affect albedo with global implications. Journal of Geophysical Research 110.

Sturm, M., J. Schimel, G. Michaelson, V. E. Romanovsky, J. M. Welker, S. F. Oberbauer, G. E. Liston, and J. Fahnestock (2005b). Winter biological processes could help convert arctic tundra to shrubland. BioScience 55:17–26.

Tannerfeldt, M., B. Elmhagen, and A. Angerbjörn (2002). Exclusion by interference competition? The relationship between red and arctic foxes. Oecologia 132:213–220.

Tape, K., M. Sturm, and C. Racine (2006). The evidence for shrub expansion in northern Alaska and the pan-Arctic. Global Change Biology 12:686–702.

Tulp, I., H. Schekkerman, and J. De Leeuw (2012). Eggs in the freezer: energetic consequences of nest site and nest design in Arctic breeding shorebirds. PLoS ONE 7:e38041.

Van Horne, B. (1983). Density as a misleading indicator of habitat quality. Journal of Wildlife Management 47:893–901

Viereck, L. A., C. T. Dyrness, A. R. Batten, and K. J. Wenzlick (1992). The Alaska vegetation classification.

Walpole, B., E. Nol, and V. Johnston (2008). Breeding habitat preference and nest success of Red-necked Phalaropes on Niglintgak Island, Northwest Territories. Canadian Journal of Zoology 86:1346–1357.

Yoshikawa, K., and L. D. Hinzman (2003). Shrinking thermokarst ponds and groundwater dynamics in discontinuous permafrost near Council, Alaska. Permafrost and Periglacial Processes 14:151–160.

Zöckler, C., and I. Lysenko (2000). Water Birds on the Edge: The First Circumpolar Assessment of Climate Change Impact on Arctic Breeding Water Birds. World Monitoring Centre, Cambridge, UK. World Conservation Press.

TABLES AND FIGURES

Table 1. Mean and standard error (SE) of Hudsonian Godwit nest sites and randomly-selected, available sites in the Beluga, AK region. Percent cover types were estimated within 1m diameter circles. P-values are derived from MANOVA tests.

Occupied (n=107) Unoccupied (n=670) Mean SE Mean SE F value p-value Distance to nearest water body (m) 9.24 (0.80) 11.54 (0.53) 1.74 0.19 Depth of water (cm) 31.25 (2.99) 48.49 (1.21) 16.34 0.0001 Distance to nearest tree (m) 12.03 (0.74) 11.55 (0.36) 0.001 0.99 Tree DBH (cm) 3.83 (0.15) 3.86 (0.06) 0.02 0.89 % >2m 0.2 (0.08) 0.37 (0.09) 0.56 0.46 % 1m

Table 2. Mean and standard error (SE) of Hudsonian Godwit nest sites and randomly-selected, available sites in the Churchill, MB region. Percent cover types were estimated within 1m diameter circles. P-values are derived from MANOVA tests.

Occupied (n=75) Unoccupied (n=310) Mean SE Mean SE F value p-value Distance to nearest water body (m) 3.46 (1.47) 12.97 (1.95) 5.56 0.01 Depth of water (cm) 4.50 (0.39) 6.22 (0.34) 3.97 0.04 Distance to nearest tree (m) 34.85 (3.67) 30.66 (1.79) 1.06 0.31 Tree DBH (cm) 2.81 (0.16) 3.26 (0.1) 3.96 0.04 % >2m 0.01 (0.01) 0.25 (0.17) 0.45 0.51 % 1m

Table 3. Mean, sample size of mean, and standard error (SE) of Hudsonian Godwit nest sites in Beluga, AK and Churchill, MB. Percent cover types were estimated within 10m diameter circles. P-values are derived from MANOVA tests with 120 complete observations.

Beluga, AK Churchill, MB

Mean n SE Mean n SE F value p-value % Cover over nest 15.64 107 (1.25) 12.51 75 (1.94) 0.085 0.77 10m: >2m 0.61 107 (0.18) 0.71 75 (0.21) - - 10m: 1m

APPENDIX

Table S1. Mean and standard error (SE) of Hudsonian Godwit nest sites and randomly-selected, available sites in the Beluga River, AK region. Species percentages were estimated within 1m diameter circles. P-values are derived from MANOVA tests.

Occupied Unoccupied

(n=107) (n=670) Species Category Mean SE Mean SE F value p-value Alnus rubra Shrub 0.02 (0.01) 0.02 (0.02) 0.004 0.948 Andromeda polifolia Forb 3.17 (0.29) 1.68 (0.09) 34.76 0.000 Aster spp. Forb 0 - 0.01 (0.01) 0.16 0.690 Betula glandulosa/nana Shrub 8.47 (0.61) 4.46 (0.27) 32.13 0.000 Betula papyrifera Tree 0.01 (0.01) 0.01 (0.01) 0.02 0.882 Carex spp. Sedge/grass 10.87 (1.3) 10.58 (0.55) 0.04 0.846 Chamaedaphne calyculata Shrub 2.28 (0.42) 0.66 (0.07) 41.88 0.000 Comarum palustre Forb 0.53 (0.11) 0.36 (0.04) 2.13 0.145 Coptis trifolia Forb 0.01 (0.01) 0.02 (0.01) 0.16 0.689 Cornus canadensis Forb 0.59 (0.24) 0.58 (0.12) 0 0.987 Dasiphora fruticosa Shrub 0.64 (0.23) 0.2 (0.06) 6.76 0.010 Drosera anglica Forb 0.38 (0.06) 0.25 (0.03) 2.47 0.116 Drosera rotundifolia Forb 0.95 (0.08) 0.73 (0.05) 2.69 0.101 Empetrum nigrum Shrub 1.43 (0.34) 5.36 (0.42) 13.44 0.000 Equisetum arvense Forb 0.08 (0.05) 0.07 (0.02) 0.05 0.816 Eriophorum spp. Sedge/grass 8.79 (0.88) 5.46 (0.43) 8.43 0.004 Fragaria virginiana Forb 0.17 (0.06) 0.08 (0.02) 2.14 0.144 Fungi spp. Mat-species 0 - 0 - 0.16 0.690 Gentiana douglasiana Forb 0.09 (0.04) 0.06 (0.01) 0.56 0.454 Geocaulon lividum Forb 0.08 (0.03) 0.23 (0.03) 3.93 0.048 Iris setosa Forb 0.09 (0.03) 0.01 (0) 27.39 0.000 Juncus alpinoarticulatus Sedge/grass 0 - 0.09 (0.05) 0.48 0.491 Juncus stygius Sedge/grass 0.16 (0.13) 0.13 (0.04) 0.07 0.788 Linnaea borealis Forb 0 - 0 - 0.32 0.572 Lichen spp. Mat-species 0.25 (0.1) 0.35 (0.08) 0.24 0.626 Menyanthes trifoliata Forb 0.66 (0.11) 0.99 (0.12) 1.26 0.262 Moss spp. Mat-species 27.67 (2.02) 35.69 (1.19) 6.75 0.010 Myrica gale Shrub 6.68 (0.65) 6.8 (0.35) 0.02 0.893 Nuphar polysepala Forb 0.06 (0.03) 0.12 (0.05) 0.23 0.635 Pedicularis labradorica Forb 0.17 (0.04) 0.09 (0.02) 2.82 0.093 Picea mariana Tree 1.62 (0.33) 2.32 (0.28) 0.98 0.322 Platanthera dilatata Forb 0 - 0.01 (0.01) 1.34 0.247 Platanthera obtusata Forb 0.02 (0.01) 0.01 (0) 0.86 0.355 Poa spp. Sedge/grass 1.72 (0.33) 0.75 (0.13) 7.41 0.007 Polemonium caeruleum Forb 0.04 (0.02) 0.01 (0.01) 0.86 0.354 Potamogeton natans Forb 0 - 0 - 0.32 0.572 Potentilla gracilis Forb 0.01 (0.01) 0.05 (0.03) 0.36 0.549 Puccinellia spp. Sedge/grass 0.05 (0.02) 0.01 (0.01) 3.51 0.061 Rhododendron tomentosum Shrub 1.86 (0.22) 3.34 (0.23) 6.32 0.012 32

Rosa acicularis Shrub 0.11 (0.04) 0.03 (0.02) 3.01 0.083 Rubus arcticus Forb 0.47 (0.1) 0.07 (0.02) 52.5 0.000 Rubus chamaemorus Forb 0.75 (0.12) 1.28 (0.11) 3.82 0.051 Salix spp. Shrub 0.01 (0.01) 0.01 (0.01) 0.003 0.957 Solidago multiradiata Forb 0.06 (0.02) 0.04 (0.02) 0.28 0.600 Sparganium angustifolium Forb 0 - 0.01 (0.01) 0.67 0.414 Spiraea stevenii Shrub 3.95 (0.61) 1.12 (0.12) 55.47 0.000 Spiranthes romanzoffiana Forb 0.01 (0.01) 0 - 6.31 0.012 Trichophorum cespitosum Sedge/grass 5.02 (0.75) 1.94 (0.26) 18.33 0.000 Trientalis europaea Forb 0.77 (0.07) 0.18 (0.02) 114.63 0.000 Triglochin maritima Sedge/grass 0.02 (0.01) 0.01 (0.01) 0.65 0.422 Utricularia intermedia Forb 0.05 (0.02) 0.05 (0.01) 0 0.986 Vaccinium oxycoccos Forb 1.36 (0.12) 0.81 (0.04) 22.71 0.000 Vaccinium uliginosum Shrub 0.27 (0.07) 0.94 (0.11) 5.6 0.018 Vaccinium vitis-idaea Forb 0.21 (0.04) 0.4 (0.04) 3.08 0.080 Valeriana capitata Forb 0.02 (0.01) 0 - 12.73 0.000 Viola epipsila Forb 0.56 (0.08) 0.23 (0.03) 17.24 0.000 Unknown spp. - 0.84 (0.35) 0.1 (0.06) 13.95 0.000

Table S2. Mean and standard error (SE) of Hudsonian Godwit nest sites and randomly-selected, available sites in the Churchill, MB region. Species percentages were estimated within 1m diameter circles. P-values are derived from MANOVA tests.

Occupied Unoccupied (n=75) (n=310) Species Category Mean SE Mean SE F value p- value Andromeda polifolia Forb 2.83 (0.23) 1.8 (0.12) 14.78 0.000 Arctous rubra Forb 0.36 (0.12) 0.31 (0.08) 0.11 0.74 Bartsia alpina Forb 0.25 (0.08) 0.07 (0.02) 7.85 0.005 Betula glandulosa/nana Shrub 8.4 (0.94) 3.89 (0.36) 27.27 0.000 Bistorta vivipara Forb 0.41 (0.06) 0.07 (0.02) 60.63 0.000 Carex aquatilis Sedge/grass 22.89 (2.6) 17.7 (0.98) 4.73 0.03 Carex capillaris Sedge/grass 2.01 (0.48) 1.95 (0.31) 0.01 0.91 Carex capitata Sedge/grass 0.93 (0.29) 0.09 (0.03) 29.22 0.000 Carex glacialis Sedge/grass 0.03 (0.02) 0 - 8.39 0.004 Carex gynocrates Sedge/grass 0.49 (0.16) 0.01 (0.01) 34.51 0.000 Carex limosa/rariflora Sedge/grass 1.67 (0.35) 5.33 (0.6) 9.02 0.003 Carex microglochin Sedge/grass 0.11 (0.05) 0.03 (0.01) 4.84 0.03 Carex saxatilis Sedge/grass 0.27 (0.11) 0.21 (0.06) 0.17 0.68 Carex scirpoidea Sedge/grass 0.19 (0.08) 0.33 (0.13) 0.29 0.59 Carex spp. Sedge/grass 7.77 (2.16) 2.44 (0.4) 15.86 0.000 Carex vaginata Sedge/grass 0.03 (0.03) 0.14 (0.04) 1.92 0.17 Chamerion angustifolium Forb 0.29 (0.09) 0.01 (0.01) 39.73 0.000 Dryas integrifolia Forb 0.15 (0.06) 0.41 (0.14) 0.88 0.35 Empetrum nigrum Shrub 0.05 (0.03) 0.9 (0.22) 3.65 0.06 Equisetum variegatum Forb 0 - 0 - 0.24 0.62 Eriophorum angustifolium Sedge/grass 0.28 (0.06) 0.16 (0.03) 2.50 0.11 Eriophorum vaginatum Sedge/grass 0.16 (0.04) 0.13 (0.04) 0.12 0.73 Fungi spp. Mat-species 0.01 (0.01) 0 - 4.14 0.04 Grass spp. Sedge/grass 0.05 (0.03) 0.33 (0.11) 1.44 0.23 Juncus arcticus Sedge/grass 0 - 0.02 (0.02) 0.24 0.62 Juncus balticus Sedge/grass 0 - 0.06 (0.06) 0.24 0.62 Larix laricina Tree 1.15 (0.33) 0.57 (0.13) 3.18 0.08 Leymus mollis Sedge/grass 0 - 0.01 (0.01) 0.24 0.62 Lichen spp. Mat-species 1.91 (0.46) 2.24 (0.37) 0.16 0.69 Menyanthes trifoliata Forb 0.01 (0.01) 0.03 (0.02) 0.11 0.74 Moss spp. Mat-species 5.13 (0.45) 2.35 (0.24) 25.96 0.000 Myrica gale Shrub 0.03 (0.02) 0.23 (0.17) 0.35 0.56 Orthilia secunda Forb 0.01 (0.01) 0 - 4.14 0.04 Parnassia palustris Forb 0 - 0.01 (0.01) 0.24 0.62 Pedicularis flammea Forb 0 - 0.01 (0.01) 0.73 0.39 Pedicularis lapponica Forb 0.03 (0.02) 0.01 (0.01) 0.54 0.46 Pedicularis sudetica Forb 0.31 (0.06) 0.12 (0.02) 10.19 0.002 Petasites frigidus Forb 0.01 (0.01) 0.05 (0.02) 0.49 0.49 Picea glauca Tree 0 - 0.14 (0.09) 0.53 0.47 35

Picea mariana Tree 0 - 0.02 (0.02) 0.24 0.62 Pinguicula villosa Forb 0.09 (0.03) 0 - 31.54 0.000 Pinguicula vulgaris Forb 0.41 (0.06) 0.26 (0.03) 6.66 0.01 Platanthera hyperborea Forb 0 - 0.01 (0.01) 0.98 0.32 Rhododendron groenlandicum Shrub 0.03 (0.02) 0.01 (0.01) 0.60 0.44 Rhododendron lapponicum Shrub 1.23 (0.2) 0.85 (0.11) 2.24 0.14 Rhododendron tomentosum Shrub 0.39 (0.09) 0.62 (0.15) 0.53 0.47 Rubus arcticus Forb 0.04 (0.04) 0.04 (0.03) 0.00 0.99 Rubus chamaemorus Forb 0.01 (0.01) 0.12 (0.04) 1.37 0.24 Salix arbusculoides Shrub 1.79 (0.22) 1.83 (0.23) 0.01 0.93 Salix candida Shrub 0 - 0.04 (0.02) 0.70 0.40 Salix glauca Shrub 0 - 0.17 (0.09) 0.84 0.36 Salix lanata Shrub 0 - 0.02 (0.01) 0.49 0.49 Salix planifolia Shrub 0 - 0.26 (0.16) 0.59 0.44 Salix reticulata Shrub 0.03 (0.02) 0.18 (0.07) 1.11 0.29 Salix spp. Shrub 0 - 0.15 (0.07) 0.99 0.32 Saxifraga aizoides Forb 0.07 (0.04) 0 - 11.78 0.001 Shepherdia canadensis Shrub 0 - 0.02 (0.01) 0.47 0.49 Spiranthes romanzoffiana Forb 0.04 (0.02) 0 - 8.01 0.005 Stellaria longipes Forb 0 - 0 - 0.24 0.62 Taraxacum ceratophorum Forb 0 - 0 - 0.24 0.62 Tofieldia pusilla Forb 0.43 (0.1) 0.23 (0.03) 7.56 0.006 Trichophorum cespitosum Sedge/grass 8.88 (1.09) 12.5 (0.86) 3.69 0.06 Triglochin maritima Sedge/grass 0.07 (0.03) 0.02 (0.01) 4.85 0.03 Trisetum spicatum Sedge/grass 0 - 0.01 (0.01) 0.24 0.62 Utricularia intermedia Forb 0.05 (0.03) 0.03 (0.02) 0.39 0.53 Vaccinium uliginosum Shrub 0.97 (0.15) 1.2 (0.19) 0.36 0.55 Vaccinium vitis-idaea Forb 0.09 (0.04) 0.32 (0.13) 0.76 0.38 Unknown spp. - 0.11 (0.07) 0 - 8.97 0.003

CHAPTER THREE

DO HABITAT CHARACTERISTICS AND RISK OF PREDATION EXPLAIN SPATIAL

PATTERNING OF NESTS?

Abstract

While nest placement may be guided by a wide variety of proximate abiotic and biotic cues, it often reflects preferences for particular habitats and interactions (or lack thereof) with other species, including predators. Examining spatial patterns of nest placement provides an opportunity to evaluate how various components of habitat and predation risk may define habitat suitability. We assessed the spatial distribution of 112 nests of a migratory shorebird, the

Hudsonian Godwit (Limosa haemastica), found in Beluga River, Alaska, between 2009 and

2012, and explicitly tested the relative influence of habitat features and predation risk on nest locations. We examined relationships between nest location, predation risk, and proximity to roads on nest fate using 64 nests that were monitored through nest completion between 2009 and

2011. Hudsonian Godwit nests were significantly clustered across the landscape, despite a lack of significant spatial autocorrelation (i.e. patchiness) in landscape-scale vegetation. At the micro- scale, vegetation patchiness within the bog also did not correlate with the density of godwit nests. Nest fate was not predicted by either the distance to the nearest conspecific neighbor or proximity to roads. Thus, neither habitat characteristics nor predation risk explained the clustering of godwit nests. These results suggest that Hudsonian Godwit nest locations may be based more on social cues than underlying heterogeneity in vegetation or predation risk.

Introduction

Selection of breeding habitat has fundamental consequences for the demography, ecology, and conservation of a species. One emergent property of nest site selection is the spatial patterning of nests, which may affect breeding success and survival (Cornell and Donovan

2010). As many as eight hypotheses have been proposed for clustering in birds and other animals

(Tarof and Ratcliffe 2004); however, the primary hypothesis to explain spatial clustering in a species is that their resources are also spatially clustered (Kiester and Slatkin 1974, Tarof and

Ratcliffe 2004). Nevertheless, aggregations may reflect not only underlying heterogeneity in environmental attributes and resource availability, but also interactions with other species (e.g., competitors, predators) (Hamilton 1971, Danchin et al. 2004, Armstrong and Nol 1993, Tarof and Ratcliffe 2004, Gascoigne and Lipcius 2004). Understanding the drivers of aggregations is difficult because patterns are ultimately the consequence of preferences of and decisions made by individuals (Melles et al. 2009, Bayard and Elphick 2010).

Spatial aggregations of animals are most often attributed to habitat heterogeneity and patchiness, where individuals settle in high quality sites or areas with the appropriate resources.

The material resources hypothesis (Kiester and Slatkin 1974) proposes that individuals cluster in response to discontinuities in the distribution of ecological resources, such as vegetation or food, where the resource itself shows clustering. In species where ecological resources influence settlement patterns, clusters could form where nesting sites or food resources were locally abundant. Individuals also may use the presence or reproductive performance of conspecifics to assess habitat quality when selecting a territory, resulting in clustered territories (Stamps 1988,

Smith and Peacock 1990, Reed and Dobson 1993, Bourque and Desrochers 2006). Although distribution of resources is expected to constrain and guide settlement decisions to some extent,

aggregation may result from non-habitat factors as well, though these alternatives are generally less understood (Melles et al. 2009).

Among the non-habitat cues thought to influence spatial distribution is predation risk

(Gascoigne and Lipcius 2004). Individuals are known to select breeding locations that minimize risk of predation (Hamilton 1971, Götmark and Andersson 1984, Møller 1987, Brown and

Brown 1996) or human-associated disturbances (Knight and Fitzner 1985, Klein et al. 1993, Frid and Dill 2002). Nesting in groups can reduce the per capita predation risk on each individual in the group, allows for earlier detection of predators, and provides opportunities for group defense against approaching predators (Hamilton 1971, Martin 1988, Martin 1993). For example, Least

Flycatchers (Empidonax minimus) respond more quickly with alarm calls to predators when inside a cluster of territories (Perry and Andersen 2003) and show reduced predation rates for nests inside clusters compared to the periphery (Perry et al. 2008). The extent to which predation risk influences nest placement depends, in part, on the predator community. For example, in systems where the predator community is diverse, there may be no predictably safe nest sites

(Sih et al. 1998, Forstmeier and Weiss 2004). Similarly, human-modified landscapes are known to influence nest locations and may even limit the availability of suitable nesting habitat by creating non-suitable buffers around roads and other development. For instance, Lesser Prairie-

Chickens (Tympanuchus pallidicinctus) nest farther from buildings, transmission lines, and roads than random sites (Pitman et al. 2005).

Determining whether spatial aggregations reflect patchiness in the physical habitat rather than the social environment is a crucial step in identifying the drivers of settlement decisions and the distributional patterns of individuals and species. In cases where a species is absent despite abundant resources or appropriate physical habitat, the influence of species interactions may be

apparent (Melles et al. 2009). Inter- and intra-specific interactions are known to influence nest locations and settlement patterns of many birds, including Barn Swallows (Hirundo rustica,

Brown and Brown 2000), Black-capped Chickadees (Poecile atricapillus, Ramsay et al. 1999), and Snowy Plovers (Charadrius nivosus, Nelson 2007) – all of which preferentially nest in areas with conspecifics. Further, many territorial species have clustered distributions of breeding individuals across the landscape, including many passerines (Etterson 2003, Tarof and Ratcliffe

2004, Bourque and Desrochers 2006, Melles et al. 2009) and shorebirds (Howe 1982, Armstrong and Nol 1993, Lengyel 2006, Saalfeld et al. 2012, Patrick 2013). Species may exhibit heterospecific attraction if they use other species as sources of information on habitat quality or if they gain protective benefits from nesting near other species (Bulmer 1984, Burger 1984,

Goodale et al. 2010). For example, Semipalmated Plovers (Charadrius semipalmatus) nesting near Arctic Terns (Sterna paradisaea) have higher rates of nest survival (Nguyen et al. 2006),

Brent Geese (Branta bernicla) that nest near Snowy Owls (Bubo scandiacus) have lower risk of predation by arctic foxes (Vulpes lagopus, Kharitonov et al. 2008), and Canada Geese (Branta canadensis) nesting near Snow Geese (Chen caerulescens) have lower predation rates depending on the size of the Snow Goose colony (Reiter and Andersen 2013). The extent to which social cues are used for habitat selection and drive species distributions has significant implications for both habitat selection theory and conservation practice (Stamps 1988, Ahlering and Faaborg

2006).

Elucidating the habitat selection process is especially important for species that are uncommon across landscapes comprised of seemingly suitable habitat. The migratory shorebird, the Hudsonian Godwit (Limosa haemastica, hereafter: ‘godwits’) is one such species of concern where conservation is limited by a poor understanding of the full suite of cues used to select

breeding habitat (Senner 2010, Walker et al. 2011). Broadly, the basics are well described: godwits breed in sedge bogs that are dominated by muskeg, interspersed with small ponds, spruce tree islands, and drier upland areas (Walker et al. 2011). However, occurrence and density of breeding individuals varies widely within and across bogs as well as across their range, suggesting that non-habitat cues might factor prominently in the selection of breeding sites. The importance of non-habitat cues is further suggested by local absences in areas where sub- arctic/boreal wetlands are readily available and seemingly suitable (Walker et al. 2011, N. R.

Senner personal obs.). Whether the disjunct breeding distribution and patchy occurrence within breeding areas of godwits is due to low population numbers or simply reflects our poor understanding of their habitat preferences is unknown.

Two possible cues that could influence locations and/or fate of godwit nests are predation risk and the proximity of roads. Along the Cook Inlet of Alaska, the main predators on godwit nests and incubating adults are Common Ravens (Corus corax), Northern Harriers

(Circus cyaneus), Sandhill Cranes (Grus canadensis), and red foxes (Vulpes vulpes). Ravens, harriers, and cranes rely on sight to locate nests; whereas, foxes primarily use olfactory cues.

These two different predator types presumably would have different selective pressures on efforts by individuals to protect their nests and could result in different nest placement patterns.

Furthermore, predator population sizes and distributions can be enhanced by supplemental food and land uses, including roads and buildings (Restani et al. 2001, Kristan and Boarman 2007,

Liebezeit et al. 2009, Hackney et al. 2013). Abundances of both ravens and foxes in our study area have increased with human activities and development, including oil-well pads, roads, and transmission lines (N. R. Senner unpubl. data). Further, predators may respond to nest density, or change searching efforts when they successfully find a nest (Ratti and Reese 1988, Neimuth and

Boyce 1995). Therefore, birds may avoid placing nests near roads, transmission lines, or conspecific neighbors due to their increased predation risk.

In this study, we asked if Hudsonian Godwits nested in social aggregations and, if so, whether spatial patterns were explained by habitat or predation risk. If vegetation parameters were key determinants of settlement decisions and if the distribution of vegetation were patchy, we expected that nest placement would correspond to patches of favorable habitat. Alternatively, if settlement patterns were primarily a function of predation risk or other social cues, then nest placement was expected to be independent of vegetation patches and, rather, related to proximity to conspecifics or predators. We employed spatial point-pattern analyses to quantify patterns of godwit nests and nest fate at two study plots in south-central Alaska. First, we tested the null hypothesis that godwit nests were randomly distributed. Second, we determined if the spatial structure of the bog vegetation influenced placement of godwit nests. Third, we examined the spatial pattern of depredated nests to address the hypothesis that predation risk influenced nest locations. Lastly, we tested the hypothesis that proximity to roads/transmission lines and/or conspecific nests influenced nest fate.

Methods:

Study species:

Hudsonian Godwits are shorebirds of high conservation concern, in part because their small population faces a wide variety of threats across its range (Senner 2010). Godwits migrate from southern South America to arctic and sub-arctic breeding grounds in Alaska and Canada, where there are three disjunct breeding populations – Hudson Bay, northern Northwest

Territories and northeastern Alaska, and south-central and western Alaska (Walker et al. 2011).

After arrival on the breeding grounds, most clutches are complete within 10 days (Senner 2013).

Incubation lasts ~22 days and chicks leave the nest <24 hours after hatching (Walker et al. 2011).

Previous work indicated that during the ≤30 day period when nests are sensitive to predators,

26% of nests are depredated (Senner 2013). Depredation of parents is relatively rare, although

Northern Harriers and red foxes occasionally kill incubating adults.

Study Area:

From 2009-2012, we monitored the breeding effort of godwits within an ~8 km2 area at

Beluga River, Alaska (61.21°N, 151.03°W). This area supports a large breeding population of godwits (5.0 breeding pairs per km2; N.R. Senner unpubl. data). The study site was dominated by sedges, Carex spp., and dwarf birch, Betula glandulosa/nana, in a black spruce, Picea mariana, dominated muskeg bog (Walker et al. 2011, Senner 2012). The study area was divided into two study plots of unequal sizes (North: 550 ha; South: 120 ha), and each plot was searched completely for godwit nests within the study plot boundaries. Plots were separated by ~7 km of boreal forest and muskeg bog not monitored for breeding godwits.

Nest Searching and Monitoring:

Field teams began searching for nests after pair formation and continued through the onset of hatch. We attempted to locate nests for all breeding pairs detected in the study area.

Nests were located through knowledge of former territories, systematic walking surveys, and behavioral cues such as flushing during incubation switchovers. Locations were recorded with a

Global Positioning System (GPS) and were not physically marked to avoid associative learning of predator species (Reynolds 1985). Nests were monitored every 2-3 days until signs of hatching, when nests were monitored daily. We typically checked nests by resighting incubating birds with binoculars from 20-30 m away. Adults were flushed at most weekly to minimize

disturbance that might increase the probability of nest failure, and field teams did not visit nests directly when predators were seen nearby. A nest was considered successful if ≥1 egg hatched and chicks successfully left the nest site. We used presence of small eggshell fragments (Mabee

1997), alarm-calling parents, and/or young at or near the nest as indications of nest success. Nest failure was determined by empty nests early in the incubation period, nest disturbance, and/or destroyed eggs. Nests were categorized as depredated due to either the loss of eggs or a known depredation of an incubating adult, and either would result in total nest failure.

Vegetation Parameters:

After nests were no longer active, habitat was measured at the nest site and associated random points. We defined the microhabitat (nest site) scale as the area within a 1 m diameter circle centered on the nest. In each godwit territory, we placed 25 circular plots at randomly selected points as well as at the nest site. Points were selected from within a 200 m radius of the nest using a random number generator. All points were within the wet sedge dominated bog and study boundaries.

For each circular plot, we measured the distance to the nearest water body (≥ 2 cm deep) from the center of the circle. Within the 1 m diameter circular plots, we recorded the percent species composition for all species present. From this, we summarized the percentage of the circle that was covered by shrubs, sedges and grasses, and forbs as well as the percent of bare ground (water, mud, or rocks). We also summed the number of species present in the circular plot as a metric of species richness.

Ponds and black spruce tree islands, both of which are used by godwits, were outlined as polygons from satellite imagery (Google Earth) and then converted into centroid points for use in point-pattern analyses (see below), using ArcGIS, version 10.3.1 (ESRI, Redlands, California).

Centroid points provide a crude estimation of the distribution of these features within the study plots while not accounting for the area they cover.

Statistical Methods:

To test the null hypothesis that Hudsonian Godwit nests were distributed randomly within the study plots, we used a combination of first- and second-order point-pattern tests. Point- pattern analysis is the study of the spatial arrangements of points in (usually 2-dimensional) space, where the datum of interest is the location of the point itself (Diggle 1979, 2003). Point- pattern analysis assumes a complete census of the study area, and most statistical tests also assume that the data are both stationary and isotropic (Fortin and Dale 2005). First-order tests are related to the mean number of nests per unit area (density), and they allow a crude assessment of clustering at a local scale. Second-order methods reflect the covariance between nests per unit area and allow investigation of interactions among pairs of nests (Perry et al. 2006). In other words, structured spatial patterns may arise from spatially structured habitat features (first-order effects) or from biological processes such as competition or conspecific attraction (second-order effects) promoting active spacing or clustering (Wiegand et al. 2007). All tests were computed in program R (R Development Core Team 2015) using the package ‘SPATSTAT’ (Baddeley and

Turner 2005), and each plot was analyzed separately to comply with the assumption of a complete census. We used a combination of these first- and second-order tests to determine the robustness of our results.

Our first-order test was the Clark and Evans aggregation index, which we used as a preliminary tool to assess whether the spatial distribution of godwit nests differed significantly from the null hypothesis of complete spatial randomness. The Clark and Evans aggregation index

(R) is calculated as the ratio of the mean nearest-neighbor distance (NND) for all nests to the

mean NND expected for a Poisson point process of the same intensity (Clark and Evans 1954).

We then ran 999 Monte-Carlo simulations of the Poisson null model to estimate p-values for R.

A value of R = 1 is the expected value for complete spatial randomness, a value significantly less than 1 suggests clustering, and a value significantly greater than 1 suggests dispersed or repulsed placement.

We used Ripley’s K (Ripley 1976, 1979, 1981, 1988), which is used to detect spatial randomness at successively larger scales based upon the cumulative distribution function (i.e. the number of additional nests within a distance r of a random nest; Baddeley and Turner 2005). A second-order test, Ripley’s K was derived from the nest data set and then compared with the theoretical curve of the Poisson point pattern, which represents complete spatial randomness. We used the linearized form of K, L(r) = (K[r]) – πr2, to aid in interpretation and to stabilize the variance (Besag 1977, Haase 1995). Here, the expected number of nests in an area with radius r is subtracted from (K[r]), the observed number of nests in a circle with radius r. Under complete spatial randomness, the number of nests in a circle follows a Poisson distribution and L(r) = 0 for all distances.

Because variability in user-defined distances for this test can affect the outcome of

Ripley’s K, so we ran each test using the default range as prescribed by SPATSTAT. The recommended range for the distance lags (r) was 0 – 852 m for the North plot and 0 – 367 m for the South plot. To determine whether there was evidence of clustering or repulsion, we generated simulation confidence envelopes of K from 999 Monte-Carlo simulations of the null model

(Poisson process) with an α level of 0.05. Simulation envelopes specify the critical points for a

Monte Carlo test and allow us to reject the null hypothesis if the observed function lies outside of these values in a similar manner to confidence intervals (Baddeley and Turner 2005). Values of

the empirical K function above the simulation confidence envelope indicate clustering, and those below the simulation confidence envelope indicate repulsion. Evidence of significant attraction suggests either conspecific attraction or habitat clustering. Significant repulsion implies that territoriality determines spacing. These tests were performed for all four years of data combined and for each year separately.

Though Ripley’s K-function is widely recognized as a useful tool for detecting spatial aggregations, the cumulative character of this statistic often hampers the detection of scale- dependent patterns (Condit et al. 2000, Schurr et al. 2004). If clumping occurs on a relatively small scale, the point density at larger scales will be above average as well because increasing circular scales are cumulative. Consequently, we also performed the pair-correlation function

(PCF) (Ripley 1981, Stoyan and Stoyan 1994), which tests for interactions between points (i.e. nests) separated by a distance r. The PCF can be thought of as a circle centered at a given nest, where the only nests counted are those that lie on the circle boundary (i.e. a ring); whereas,

Ripley’s K function counts all nests that were contained within the circle. The PCF is the probability of observing a pair of nests separated by a distance r, divided by the corresponding probability for a Poisson process (Baddeley 2008). Interpretation of the PCF is similar to that of

Ripley’s K in that values above the upper bounds of the confidence envelope indicate clustering.

An important consideration in point-pattern analyses is the possibility of edge effects, which can result in an underestimate of clustering. Edge effects arise because the points located near the edge of the study area have fewer neighbors available in all directions than points located in the interior, and consequently the value of R will be positively biased as the observed nearest neighbor distances tend to be larger than the true nearest neighbor distances (Perry et al.

2006). Edge corrections mathematically account for the unsurveyed area that lies outside the

focal area. Many edge corrections exist, and the shape and extent of the study area can be used as a guide for choosing among them (Haase 1995, Diggle 2003, Lancaster and Downes 2004).

We had mixed justification for using edge correction methods in our system. On the one hand, many boundaries did not need corrections as they could be considered hard boundaries because godwits did not occur in adjacent habitats (e.g., boreal forest); however, in other cases there were soft or artificial boundaries, such as roads bisecting the bog on the other side of which we have not searched for godwit nests. Because our system contained this ambiguity, we conducted the Clark and Evans aggregation index with and without the cumulative distribution function edge correction. We report differences between the two sets of results as a test of the sensitivity to edge-related bias. The cumulative distribution function edge correction calculates the nearest neighbor distribution and uses the mean of the distribution to correct for edges.

However, we only performed the Ripley’s K test with the Ripley’s isotropic correction as uncorrected tests are biased with sample sizes less than 1,000 points (Baddeley and Turner

2005). In the Ripley’s K isotropic correction, the naive K estimate is scaled by the ratio of the circumference of the circle that lies outside the study area to the circumference of the circle that lies inside the study area.

We used Moran’s I test (Moran 1948) to examine if spatial patterning of nest locations was due to an underlying spatial pattern of habitat features. If certain vegetation characteristics drove settlement decisions, then clusters of nests should correspond to patches of especially favorable habitat. The focal vegetation parameters were selected based on previous work

(Chapter Two) showing that godwits selected areas with greater numbers of plant species, more sedge/grass, forb, and tall shrubby cover between 30cm and 1m tall, less bare ground, and closer to shallow water than random sites. Therefore, to test if habitat was driving an observed pattern

of aggregation, we examined the distribution of sedges/grasses, forbs, shrubs, bare ground, the number of species present, and the distance to water within the bog. Moran’s I is used to test the null hypothesis of no systematic pattern, or spatial autocorrelation, in the distribution of a quantitative variable (Cliff and Ord 1981); clustering is indicated by positive values of the coefficient, dispersion by negative values, and randomness by values close to 0.

We tested for spatial autocorrelation of habitat characteristics at three different scales of analysis. To explore spatial autocorrelation, each vegetation parameter was tested using a different number of distance classes (20, 50, 100) in the freely available software SAM (Rangel et al. 2010), with greater numbers of distance classes representing a finer-scale analysis. Each distance class was defined such that an approximately equal number of pairs of points were considered in each distance class. Significance of Moran’s I was determined for each distance class using a randomization procedure with 199 simulations (Fortin and Dale 2005). Vegetation data for nest locations and randomly selected points were analyzed in both a combined dataset and a nests-only dataset. To account for non-independence among distance classes, significance for each class was assessed using a Bonferroni correction. Moran’s I values were then plotted as a correlogram against k distance classes to aid in interpretation (Fortin and Dale 2005). A positive, significant, Moran’s I value indicated a patch of similarly structured vegetation; a negative, significant, value indicated dissimilar vegetation characteristics and was interpreted as a space between patches (Amico et al. 2008).

We then applied spatial point processes and spatial regression modeling to analyze the influence of predation on nest location. Not all nests were monitored through completion in

2012, so all nests from that year were excluded from this analysis. Nests from the south plot were omitted due to small sample size (n = 5 for each year). Generalized additive models were

used to test how location was associated with nest fate using the package ‘mgcv’ in program R

(Wood 2000). Regression models were created using the logistic family within the generalized linear model formula; Akaike Information Criteria corrected for small sample size (AICc) was used to assess different models (Burnham and Anderson 2002). Models tested the distance to the nearest conspecific neighbor and the distance to the road as a predictor of nest fate (Bayard and

Elphick 2010; Klaver et al. 2012). Nearest neighbor distances were calculated in the SPATSTAT package using the ‘nndist’ function; whereas, distance to the road was calculated in ArcGIS using a multi-line road layer and the near tool.

Results:

Our analyses of nest aggregation used 112 nests between the two study plots (North plot:

92, Fig. 1a; South plot: 20, Fig. 1b). Godwit nests were spatially clustered within each plot, based on first- and second-order tests (Table 1, Fig. 2-5). The spatial pattern of clustering was evident in the North plot based on the Clark and Evans aggregation index, irrespective of edge correction, for all years except in 2009 when a cumulative distribution function edge correction was used (R = 0.83, p = 0.27; Table 1). The clustering effect was not significant in most years on the south plot (Table 1). Consistent with a clustering pattern seen in the Clark and Evans aggregation index, comparison of the Ripley’s K function with 99% confidence intervals of the

Poisson point-process null model showed clustering across the combined years/nests for each plot (Fig. 2), for 2010, 2011, and 2012 on the North plot (Fig. 4a), and each year for the South plot (Fig. 4b). For the North plot, clustering occurred specifically at scales between 150 and 400 m within a year; there was also evidence for significant clustering just above the 99% simulation confidence envelope at the 500 m and 700-800 m scale in some years (Fig. 5a). For the South

plot, significant clustering occurred between 100-200 m each year (Fig. 5b). Collectively these results provide strong evidence that Hudsonian Godwit nests are clustered within bogs.

Distribution of ponds and tree islands within the bogs was random. In the North plot, ponds were randomly distributed according to the Clark and Evans aggregation index with or without the edge correction (n = 1,013, Table 2) and the PCF test (Fig. 7a). The Ripley’s K test

(Fig. 6a) found that ponds were clustered at intermediate scales of 100 - 400 m, but became randomly distributed as scales increased. Ponds were randomly distributed on the South plot (n =

113, Table 2, Fig. 6b and 7b). Tree islands were also randomly distributed according to the Clark and Evans aggregation index and PCF test, regardless of whether edge effects were included

(North plot: n = 305, South Plot: n = 123, Table 2, Fig. 9). However, the Ripley’s K test identified a clustering effect that accumulated across large distances on the North plot, likely due to the boreal forest bordering the edges of the plot (Fig. 8a, Table 2).

Vegetation attributes varied in the degree to which they were spatially autocorrelated

(i.e., patchily distributed) (Table 3). Of the 72 tests of spatial autocorrelation conducted (6 vegetation variables × 3 distance classes × 2 point subsets x 2 sites), 51% (n = 72) yielded no significant autocorrelation. The best support for patchiness in vegetation features was found for sedge/grass cover (9 of 12 tests), and the least support was found for shrubby cover between

30cm and 1m tall (2 of 12 tests). Significant spatial autocorrelation was detected at distances ranging from 34 - 3,476 m for North plot and 18 - 1,225 m for South plot, depending on the number of classes and which vegetation features were considered.

Greater levels of patchiness tended to be detected when all points were included than when restricted to nest locations, suggesting that godwit nests occurred in areas with similar vegetation structure, regardless of where in the study area they occur. Nonetheless, patchiness

was detected under all combinations of distance classes and vegetation parameters considered when all points were analyzed. Additionally, when we examined the densities of each of these vegetation parameters, there was no correlating peak (aggregation) or trough (dispersion) associated with areas with high densities of godwit nests (Fig. 10). For example, because godwits nest in areas with more forb cover than random points, we would expect a cluster of godwit nests to have higher densities of forb cover than other areas of the bog. However, on both plots there was a peak or cluster of high forb cover locations away from any godwit nests (Fig. 6).

Collectively, these results suggest that vegetation patchiness did not drive aggregation of godwit nests.

Between 2009-2011, 64 godwit nests were monitored to completion on the North plot

(Table 4). Location did not predict nest fate (North plot: p-value for x = 0.319, p-value for y =

0.065). Successfully hatched nests showed a clustered pattern; depredated nests a random pattern

(Table 4, Fig. 11). The clustered pattern observed in hatched nests is most likely attributed to the underlying clustered pattern of all nests. However, the random pattern found for depredated nests supports the GAM model that found no influence of location on nest fate, since nests were equally likely to be depredated in any part of the plot. Among the five possible models predicting nest fate, the null model – which only includes year as a random effect – had the lowest AICc score (Table 5). Therefore, depredated nests were randomly distributed throughout the bog, and there was no apparent influence of conspecific neighbors or proximity to the road system on nest fate.

Discussion:

We found strong evidence of spatial clustering of Hudsonian Godwit nests within bogs, but contrary to our original expectations, this clustering was not explained by habitat heterogeneity or predation risk. Interestingly, the clustering pattern of nests was strongest at intermediate distances. The pronounced pattern of clustering is somewhat paradoxical to the strong territorial behavior of godwits, which defend distinct territories that would seem to promote more uniform distributions (Walker et al. 2011). However, other territorial species have clustered distributions of breeding individuals across the landscape, including Snowy Plovers

(Charadrius nivosus, Patrick 2013), as well as many forest songbirds (Bourque and Desrochers

2006, Melles et al. 2009). This pattern is often believed to convey some alternate benefit to individuals typically through antipredator behaviors, or habitat, food, or mate availability (Tarof and Ratcliffe 2004, Melles et al. 2009). For godwits, access to mates may be particularly important as migratory species breeding in the Arctic and sub-Arctic have short periods of time to find a mate and initiate clutches within the short breeding season (Piersma et al. 2003,

Morrison 2006, Forsman et al. 2007, Morrison et al. 2007).

The lack of clustering of the habitat attributes known to be important to godwits suggests that social cues were driving aggregations of nests. Individuals aggregate nests in response to many types of social cues. For example, conspecific attraction, where individuals settle preferentially near other individuals, could explain why godwits are present in high densities in some bogs, but not in neighboring ones. Conspecific attraction has been shown to influence breeding distributions of species with diverse mating systems, including swallows (Brown and

Brown 2000), raptors (Alonso et al. 2004, Serrano et al. 2004, Serrano et al. 2005, Sergio and

Penteriani 2005), flycatchers (Tarof and Ratcliffe 2004), wrens (Muller et al. 1997), vireos

(Ward and Schlossberg 2004), chickadees (Ramsay et al. 1999), sparrows (Ahlering et al. 2006,

Nocera et al. 2006), warblers (Hahn and Silverman 2006), shrikes (Etterson 2003), and owls

(Seamans and Gutiérrez 2006). The presence of conspecifics as an indicator of habitat quality is especially useful to migratory birds that have limited time to select breeding sites and for which early breeding initiates egg laying earlier (Forsman et al. 2007), which may have important phenological consequences (Visser et al. 1998, Both and Visser 2001, Senner 2013).

Additionally, understanding the role that conspecific attraction plays in the breeding distributions of threatened or endangered taxa is necessary within the context of reestablishing breeding populations (Ahlering and Faaborg 2006).

Social cues related to the presence of other species, known as heterospecific attraction, also can influence nest locations and result in spatial aggregations (Quinn and Ueta 2008,

Kleindorfer et al. 2009). The majority of described heterospecific aggregations are protective nesting associations in which one or more species place nests near a more aggressive species that provides indirect protection through early warning of predators and/or an aggressive response

(Quinn and Ueta 2008). For example, Brent geese, which are unable to protect nests from arctic foxes, nest closer to Snowy Owls when lemming populations are low or arctic fox numbers are high (Kharitonov et al. 2008). Snowy Owls readily defend a large area around their nests from arctic foxes, thereby protecting goose nests. Godwits could be nesting near larid or other shorebird species that more readily engage in mobbing behaviors of predators. Indeed, field observations in our study area suggest that godwits nest near the loud and defensive Mew Gulls

(Larus canus), but then, after hatching, move chicks away from gulls, which frequently depredate chicks (R. J. Swift, personal observation).

Contrary to our original expectations, nest location was not related to fate of godwit nests. Although predation is one of the main causes of nest failure for many species (Ricklefs

1969), predation risk to godwit nests was low (15-25% per year, Senner 2013), and, therefore, may not be an important consideration for individuals as they select nest sites. The random pattern of depredated nest locations and the lack of influence of the distance to the nearest conspecific neighbor in predicting nest fate may mean that predators are not using area- concentrated search techniques. Because predators may respond to nest density (Ratti and Reese

1988), we might expect clustered nests to have higher predation rates. For instance, if predators develop a search image for nests (Nams 1997), shift their focus towards eggs in a functional response (Holling 1959, 1961, Holt 1977), or increase search effort in areas where they have found nests previously (Tinbergen et al. 1967), then predation rates could be higher in clusters.

The random pattern of nest predation observed here may be due to an overall low density of nests or because predators are not functionally responding to the higher densities of nests in the observed clusters (Neimuth and Boyce 1995). However, antipredator benefits proposed as an explanation of nest aggregations (Mönkkönen and Forsman 2002, Tarof and Ratcliffe 2004,

Kleindorfer et al. 2009) could be offsetting the risk caused by nest density. The lack of effect of the distance to the nearest conspecific neighbor in predicting nest fate in combination with the lower proportional predation rates observed in the cluster of nests imply that heterospecific interactions may have antipredator benefits for godwits. For example, individuals may benefit from the differential sensory capabilities of heterospecific group members to detect predators

(Sieving et al. 2004, Griffin et al. 2005, Semeniuk and Dill 2006) or from aggressive mobbing behaviors of heterospecific group members (Nguyen et al. 2006, Kharitonov et al. 2008, Quinn and Ueta 2008).

Proximity to roads failed to explain variation in nest fate, though we may have failed to detect a relationship given that few nests were closer than 100 m to a road. In fact, the closest we

detected a godwit nest to the main road and parallel transmission line is 62.5 m, and only five nests were within 200 m of the main road during the course of the study (N. R. Senner unpubl. data). There are several reasons why godwits might avoid roads, including noise, dust/pollution, physical vibrations/motion, disturbance from human activity, or other edge effects (Trombulak and Frissell 2000, Fahrig and Rytwinski 2009, Ware et al. 2015). For example, noise from compressor stations in natural gas fields alters the breeding distribution and richness of songbird communities (Habib et al. 2007, Bayne et al. 2008, Francis et al. 2009), and Lesser Prairie-

Chickens nest farther away from oil/gas infrastructure (Pitman et al. 2005). Indeed, our study area includes areas with active natural gas wells within one kilometer of the North plot, which could make portions of the study plot unsuitable for nest sites. Further, the amount of disturbance associated with any given road would depend upon use, which varies between years (R. J. Swift personal observation) with employment numbers and activity of drilling operations. This preliminary analysis of the proximity of nests to the road does not account for road use or disturbances associated with the road.

One of the limitations of our analysis was that we did not account for shape and area of ponds and tree islands, which varied considerably in size (Ponds: n = 1,126, range = 5 - 20,752 m2, mean = 1,152 m2; Tree islands: n = 428, range = 24 - 47,857 m2, mean = 1,645 m2). This limitation resulted from our conversion of habitat polygons at the landscape scale to centroid point locations when assessing the distribution of breeding habitat. Although area covered by those habitats is important, we were most interested in the spatial distribution of habitat features throughout the study area, which was assessable using our centroid points. If ponds and/or trees centroid points showed defined patchiness within the bog, one would expect the physical attributes of the bog to differ between those nesting clusters and unoccupied areas as well as in

terms of habitat heterogeneity. For instance, if there were differences in topography or soil that influenced the formation of ponds or tree growth, one would expect a highly significant clustering of ponds or tree islands.

Understanding why species of conservation concern fail to occupy presumably suitable habitat has important implications for management and population forecasting, both of which often rely upon models of species occurrence. While our study identified strong clustering patterns of godwit nests, we failed to find evidence that vegetation characteristics, roads, or predation risk were important drivers of spatial patterns of nest placement. Rather, we suspect that conspecific or heterospecific social cues may explain nest distribution. Future work should identify how conspecific and heterospecific interactions such as with a mobbing species or other shorebird species affect nest location and settlement decisions (Nguyen et al. 2006, Kharitonov et al. 2008, Reiter and Andersen 2013). Either could have important consequences for breeding distributions and conservation. Determining when the absence of Hudsonian Godwits at the landscape level should be ascribed to a deficiency in the physical habitat, rather than to other factors, is critical for management recommendations. Ultimately, understanding the drivers of species distributions and habitat selection remains a crucial step in predicting occurrence of breeding individuals.

LITERATURE CITED

Ahlering, M. A., and J. Faaborg (2006). Avian habitat management meets conspecific attraction: If you build it, will they come? The Auk 123:301–312.

Ahlering, M. A., D. H. Johnson, and J. Faaborg (2006). Conspecific attraction in a grassland bird, the Baird's Sparrow. Journal of Field Ornithology 77:365–371.

Alonso, J. C., C. A. Martin, J. A. Alonso, C. Palacin, M. Magana, and S. J. Lane (2004). Distribution dynamics of a great bustard metapopulation throughout a decade: Influence of conspecific attraction and recruitment. Biodiversity and Conservation 13:1659–1674.

Amico, G., D. Garcia, and M. A. Rodriguez-Cabal (2008). Spatial structure and scale-dependent microhabitat use of endemic “tapaculos” (Rhinocryptidae) in a temperate forest of southern South America. Ecología Austral 18:169–180.

Armstrong, A. R., and E. Nol (1993). Spacing behavior and reproductive ecology of the Semipalmated Plover at Churchill, Manitoba. The Wilson Bulletin 105:455–464.

Baddeley, A. (2008). Analysing spatial point patterns in R. In Workshop Notes, version 3. CSIRO, Australia. [Online.] Available at www.csiro.au/resources/pf16h.html.

Baddeley, A., and R. Turner (2005). SPATSTAT: An R package for analyzing spatial point patterns. Journal of Statistical Software 12:1–42.

Bayard, T. S., and C. S. Elphick (2010). Spatial point-pattern assessment to understand the social and environmental mechanisms that drive avian habitat selection. The Auk 127:485–494.

Bayne, E. M., L. Habib, and S. Boutin (2008). Impacts of chronic anthropogenic noise from energy-sector activity on abundance of songbirds in the boreal forest. Conservation Biology 22:1186–1193.

Besag, J. (1977). Contribution to the discussion of Dr. Ripley’s paper. Journal of the Royal Statistical Society, Series B 39:193–195.

Both, C., and M. E. Visser (2001). Adjustment to climate change is constrained by arrival date in a long-distance migrant bird. Nature 411:296–298.

Bourque, J., and A. Desrochers (2006). Spatial Aggregation of Forest Songbird Territories and Possible Implications for Area Sensitivity. Avian Conservation and Ecology 1:3.

Brown, C. R., and M. B. Brown (1996). Coloniality in the cliff swallow: the effect of group size on social behavior. University of Chicago Press. Chicago.

Brown, C. R., and M. B. Brown (2000). Nest spacing in relation to settlement time in colonial Cliff Swallows. Animal Behaviour 59:47–55.

Bulmer, M. G. (1984). Risk avoidance and nesting strategies. Journal of Theoretical Biology 106:529–535.

Burger, J. (1984). Grebes nesting in gull colonies: Protective associations and early warning. The American Naturalist 123:327–337.

Burnham, K. P., and D. R. Anderson (2002). Model selection and multimodel inference: a practical information-theoretic approach. Springer Science & Business Media.

Clark, P. J., and F. C. Evans (1954). Distance to nearest neighbor as a measure of spatial relationships in populations. Ecology 35:445–453.

Cliff, A., and J. Ord (1981). Spatial Processes: Models and Applications. Pion, London.

Condit, R., P. S. Ashton, P. Baker, S. Bunyavejchewin, S. Gunatilleke, N. Gunatilleke, S. P. Hubbell, R. B. Foster, A. Itoh, J. V. LaFrankie, H. S. Lee, E. Losos, N. Manokaran, R. Sukumar, and T. Yamakura (2000). Spatial patterns in the distribution of tropical tree species. Science 288:1414–1418.

Cornell, K. L., and T. M. Donovan (2010). Scale-dependent mechanisms of habitat selection for a migratory passerine: an experimental approach. The Auk 127:899–908.

Danchin, E., L. A. Giraldeau, T. J. Valone, and R. H. Wagner (2004). Public information: from nosy neighbors to cultural evolution. Science 305:487–491.

Diggle, P. J. (1979). Statistical methods for spatial point patterns in ecology. International Cooperative Publishing House, Fairland, Maryland.

Diggle, P. J. (2003). Statistical Analysis of Spatial Point Patterns, 2nd ed. Arnold, London.

ESRI (2015). ArcView® 10.3.1 GIS. Environmental Systems Research Institute Inc., Redlands, California, USA.

Etterson, M. A. (2003). Conspecific attraction in loggerhead shrikes: implications for habitat conservation and reintroduction. Biological Conservation 114:199–205.

Fahrig, L., and T. Rytwinski (2009). Effects of roads on animal abundance: an empirical review and synthesis. Ecology and Society 14:21.

Forsman, J. T., R. L. Thomson, and J. T. Seppänen (2007). Mechanisms and fitness effects of interspecific information use between migrant and resident birds. Behavioral Ecology 18:888–894.

Forstmeier, W., and I. Weiss (2004). Adaptive plasticity in nest-site selection in response to changing predation risk. Oikos 104:487–499.

Fortin, M. J., and M. R. T. Dale (2005). Spatial Analysis: A Guide for Ecologists. Cambridge University Press, Cambridge, United Kingdom.

Francis, C. D., C. P. Ortega, and A. Cruz (2009). Noise pollution changes avian communities and species interactions. Current Biology 19:1415–1419.

Frid, A., and L. Dill (2002). Human-caused disturbance stimuli as a form of predation risk. Conservation Ecology 6:11.

Gascoigne, J. C., and R. N. Lipcius (2004). Allee effects driven by predation. Journal of Applied Ecology 41:801–810.

Goodale, E., G. Beauchamp, R. D. Magrath, J. C. Nieh, and G. D. Ruxton (2010). Interspecific information transfer influences animal community structure. Trends in Ecology & Evolution 25:354–361.

Götmark, F., and M. Andersson (1984). Colonial breeding reduces nest predation in the common gull (Larus canus). Animal Behaviour 32:485–492.

Griffin, A. S., R. S. Savani, K. Hausmanis, and L. Lefebvre (2005). Mixed-species aggregations in birds: Zenaida doves, Zenaida aurita, respond to the alarm calls of carib grackles, Quiscalus lugubris. Animal Behaviour 70:507–515.

Haase, P. (1995). Spatial pattern analysis in ecology based on Ripley’s K-function: Introduction and methods of edge correction. Journal of Vegetation Science 6:575–582.

Habib, L., E. M. Bayne, and S. Boutin (2007). Chronic industrial noise affects pairing success and age structure of ovenbirds Seiurus aurocapilla. Journal of Applied Ecology 44:176–184.

Hackney, A. D., R. F. Baldwin, and P. G. R. Jodice (2013). Mapping risk for nest predation on a barrier island. Journal of Coastal Conservation 17:615–621.

Hahn, B. A., and E. D. Silverman (2006). Social cues facilitate habitat selection: American redstarts establish breeding territories in response to song. Biology Letters 2:337–340.

Hamilton, W. D. (1971). Geometry for the Selfish Herd. Journal of Theoretical Biology 31:295– 311.

Holling, C. S. (1959). The components of predation as revealed by a study of small-mammal predation on the European pine sawfly. Canadian Entomologist 91:293–320.

Holling, C. S. (1961). Principles of insect predation. Annual Review Entomology 6:163–182.

Holt, R. D. (1977). Predation, apparent competition, and the structure of prey communities. Theoretical Population Biology 12:197–229.

Howe, M. A. (1982). Social organization in a nesting population of eastern Willets (Catoptrophorus semipalmatus). The Auk 99:88–102.

Kharitonov, S. P., A. E. Volkov, F. Willems, H. Van Kleef, R. H. G. Klaassen, D. J. Nowak, A. I. Nowak, and A. G. Bublichenko (2008). Brent goose colonies near snowy owls: Internest distances in relation to abundance of lemmings and arctic foxes. Biology Bulletin 35:270– 278.

Kiester, A. R., and M. Slatkin (1974). A strategy of movement and resource utilization. Theoretical Population Biology 6:1–20.

Klaver, R. W., D. Backlund, P. E. Bartelt, M. G. Erickson, C. J. Knowles, P. R. Knowles, and M. C. Wimberly (2012). Spatial analysis of Northern Goshawk territories in the Black Hills, South Dakota. The Condor 114:532–543.

Klein, M. L. (1993). Waterbird behavioral responses to human disturbances. The Wildlife Society Bulletin 21:31–39.

Kleindorfer, S., F. J. Sulloway, and J. O'Connor (2009). Mixed species nesting associations in Darwin's tree finches: nesting pattern predicts predation outcome. Biological Journal of the Linnean Society 98:313–324.

Knight, R. L., and R. E. Fitzner (1985). Human disturbance and nest site placement in Black- billed Magpies. Journal of Field Ornithology. 56:153–157.

Kristan, W. B., and W. I. Boarman (2007). Effects of anthropogenic developments on common raven nesting biology in the west Mojave Desert. Ecological Applications 17:1703–1713.

Lancaster, J., and B. J. Downes (2004). Spatial point pattern analysis of available and exploited resources. Ecography 27:94–102.

Lengyel, S. (2006). Spatial differences in breeding success in the Pied Avocet Recurvirostra avosetta: effects of habitat on hatching success and chick survival. Journal of Avian Biology 37:381–395.

Mabee, T. J. (1997). Using eggshell evidence to determine nest fate of shorebirds. The Wilson Bulletin 109:307–313.

Martin, T. E. (1988). Processes organizing open-nesting bird assemblages: competition or nest predation? Evolutionary Ecology 2:37–50.

Martin, T. E. (1993). Nest predation and nest sites. BioScience 43:523–532.

Melles, S. J., D. Badzinski, M. J. Fortin, F. Csillag, and K. Lindsay (2009). Disentangling habitat and social drivers of nesting patterns in songbirds. Landscape Ecology 24:519–531.

Møller, A. P. (1987). Advantages and disadvantages of coloniality in the swallow, Hirundo rustica. Animal Behaviour 35:819–832.

Mönkkönen, M., and J. T. Forsman (2002). Heterospecific attraction among forest birds: a review. Ornithological Science 1:41–51.

Moran, P. A. P. (1948). The interpretation of statistical maps. Journal of the Royal Statistical Society, Series B 10:243–251.

Morrison, R. I. G. (2006). Body transformations, condition, and survival in Red Knots Calidris canutus travelling to breed at Alert, Ellesmere Island, Canada. Ardea 94:607–618.

Morrison, R. I. G., N. C. Davidson, and J. R. Wilson (2007). Survival of the fattest: Body stores on migration and survival in Red Knots Calidris canutus islandica. Journal of Avian Biology 38:479–487.

Muller, K. L., J. A. Stamps, V. V. Krishnan, and N. H. Willits (1997). The effects of conspecific attraction and habitat quality on habitat selection in territorial birds (Troglodytes aedon). The American Naturalist 150:650–661.

Nams, V. O. (1997). Density-dependent predation by skunks using olfactory search images. Oecologia 110:440–448.

Niemuth, N. D., and M. S. Boyce (1995). Spatial and temporal patterns of predation of simulated sage grouse nests at high and low nest densities: an experimental study. Canadian Journal of Zoology 73:819–825.

Nelson, Z. J. (2007). Conspecific attraction in the breeding distribution of the Western Snowy Plover (Charadrius alexandrines nivosus). M.S. Thesis. Humboldt State University.

Nguyen, L. P., K. F. Abraham, and E. Nol (2006). Influence of Arctic Terns on survival of artificial and natural Semipalmated Plover nests. Waterbirds 29:100–104.

Nocera, J. J., G. J. Forbes, and L. A. Giraldeau (2006). Inadvertent social information in breeding site selection of natal dispersing birds. Proceedings of the Royal Society of London B: Biological Sciences 273:349–355.

Patrick, A. M. K. (2013). Semi-colonial nesting in the Snowy Plover. M.S. Thesis. Humboldt State University.

Perry, E. F., and D. E. Andersen (2003). Advantages of clustered nesting for least flycatchers in north-central Minnesota. The Condor 105:756–770.

Perry, E. F., J. C. Manolis, and D. E. Andersen (2008). Reduced predation at interior nests in clustered all-purpose territories of least flycatchers (Empidonax minimus). The Auk 125:643– 650.

Perry, G. L. W., B. P. Miller, and N. J. Enright (2006). A comparison of methods for the statistical analysis of spatial point patterns in plant ecology. Plant Ecology 187:59–82.

Piersma, T., Å. Lindström, R. H. Drent, I. Tulp, J. Jukema, R. I. G. Morrison, J. Reneerkens, H. Schekkerman, and G. H. Visser (2003). High daily energy expenditure of incubating shorebirds on High Arctic tundra: a circumpolar study. Functional Ecology 17:356–362.

Pitman, J. C., C. A. Hagen, R. J. Robel, T. M. Loughin, and R. D. Applegate (2005). Location and success of lesser prairie-chicken nests in relation to vegetation and human disturbance. Journal of Wildlife Management 69:1259–1269.

Quinn, J. L., and M. Ueta (2008). Protective nesting associations in birds. Ibis 150:146–167.

R Development Core Team (2015). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. [Online.] Available at www.R-project.org.

Ramsay, S. M., K. Otter, and L. M. Ratcliffe (1999). Nest-site selection by female black-capped chickadees: settlement based on conspecific attraction? The Auk 116:604–617.

Rangel, T. F., J. A. F. Diniz-Filho, and L. M. Bini (2010). SAM: a comprehensive application for spatial analysis in macroecology. Ecography 33:46–50.

Ratti, J. T., and K. P. Reese (1988). Preliminary test of the ecological trap hypothesis. The Journal of Wildlife Management 52:484–491.

Reed, J. M., and A. P. Dobson (1993). Behavioural constraints and conservation biology: conspecific attraction and recruitment. Trends in Ecology and Evolution 8:253–256.

Reiter, M. E., and D. E. Andersen (2013). Evidence of territoriality and species interactions from spatial point-pattern analyses of subarctic-nesting geese. PloS ONE 8:e81029.

Restani, M., J. M. Marzluff, and R. E. Yates (2001). Effects of anthropogenic food sources on movements, survivorship, and sociality of Common Ravens in the Arctic. The Condor 103:399–404.

Reynolds, J. D. (1985). Sandhill Crane use of nest markers as cues for predation. The Wilson Bulletin 97:106–108.

Ricklefs, R. E. (1969). An analysis of nesting mortality in birds. Smithsonian Institution Press.

Ripley, B. D. (1976). 2nd-order analysis of stationary point processes. Journal of Applied Probability 13:255–266.

Ripley, B. D. (1979). Simulating spatial patterns: Dependent samples from a multivariate density. Applied Statistics 28:109–112.

Ripley, B. D. (1981). Spatial Statistics. Wiley, New York.

Ripley, B. D. (1988). Statistical Inference for Spatial Processes. Cambridge University Press, Cambridge, United Kingdom.

Saalfeld, S. T., W. C. Conway, D. A. Haukos, and W. P. Johnson (2012). Snowy Plover nest site selection, spatial patterning, and temperatures in the Southern High Plains of Texas. The Journal of Wildlife Management 76:1703–1711.

Schurr, F. M., O. Bossdorf, S. J. Milton, and J. Schumacher (2004). Spatial pattern formation in semi-arid shrubland: a priori predicted versus observed pattern characteristics. Plant Ecology 173:271–282.

Seamans, M. E., and R. J. Gutiérrez (2006). Spatial dispersion of spotted owl sites and the role of conspecific attraction on settlement patterns. Ethology Ecology & Evolution 18:99–111.

Semeniuk, C. A., and L. M. Dill (2006). Anti!predator benefits of mixed!species groups of cowtail stingrays (Pastinachus sephen) and whiprays (Himantura uarnak) at rest. Ethology 112:33–43.

Senner, N. R. (2010). Conservation Plan for the Hudsonian Godwit. Version 1.1. Manomet Center for Conservation Science, Manomet, MA.

Senner, N. R. (2012). One species but two patterns: Populations of the Hudsonian Godwit (Limosa haemastica) differ in spring migration timing. The Auk 129:670–682.

Senner, N. R. (2013). The Effects of Global Climate Change on Long-Distance Migratory Birds. PhD Dissertation. Cornell University.

Sergio, F., and V. Penteriani (2005). Public information and territory establishment in a loosely colonial raptor. Ecology 86:340–346.

Serrano, D., M. G. Forero, J. A. Donazar, and J. L. Tella (2004). Dispersal and social attraction affect colony selection and dynamics in lesser kestrels. Ecology 85:3438–3447.

Serrano, D., D. Oro, E. Ursua, and J. L. Tella (2005). Colony size selection determines adult survival and dispersal preferences: Allee effects in a colonial bird. The American Naturalist 166:22–31.

Sieving, K. E., T. A. Contreras, and K. L. Maute (2004). Heterospecific facilitation of forest boundary crossing by mobbing understory birds in north-central Florida. The Auk 121:738– 751.

Sih, A., G. Englund, and D. Wooster (1998). Emergent impacts of multiple predators on prey. Trends in Ecology & Evolution 13:350–355.

Smith, A. T., and M. M. Peacock (1990). Conspecific attraction and the determination of metapopulation colonization rates. Conservation Biology 4:320–323.

Stamps, J. A. (1988). Conspecific attraction and territorial aggregation: a field experiment. The American Naturalist 131:329–347.

Stoyan, D., and H. Stoyan (1994). Fractals, Random Shapes and Point Fields: Methods of Geometrical Statistics. Wiley, New York.

Tarof, S. A., and L. M. Ratcliffe (2004). Habitat characteristics and nest predation do not explain clustered breeding in least flycatchers (Empidonax minimus). The Auk 121:877–893.

Tinbergen, N., M. Impekoven, and D. Franck (1967). An experiment on spacing-out as a defence against predation. Behaviour 28:307–321.

Trombulak, S. C., and C. A. Frissell (2000). Review of ecological effects of roads on terrestrial and aquatic communities. Conservation Biology 14:18–30.

Visser, M. E., A. J. van Noordwijk, J. M. Tinbergen, and C. M. Lessells (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.

Ward, M. P., and S. Schlossberg (2004). Conspecific attraction and the conservation of territorial songbirds. Conservation Biology 18:519–525.

Ware, H. E., C. J. McClure, J. D. Carlisle, and J. R. Barber (2015). A phantom road experiment reveals traffic noise is an invisible source of habitat degradation. Proceedings of the National Academy of Sciences 112:12105–12109.

Wiegand, T., S. Gunatilleke, and N. Gunatilleke (2007). Species associations in a heterogeneous Sri Lankan dipterocarp forest. The American Naturalist 170:77–95.

Wood, S. N. (2000). Modelling and smoothing parameter estimation with multiple quadratic penalties. Journal of the Royal Statistical Society (B) 62:413–428.

TABLES AND FIGURES

Table 1. Clark and Evans aggregation index for North and South plots with and without a cumulative distribution function edge correction applied. Significant index values <1 indicate clustering, while significant index values >1 indicate repulsion. Nearest neighbor is the mean distance to the nearest conspecific neighbor and its associated standard error. Sample sizes for some individual years were too small to generate p-values.

Nearest neighbor No Correction Edge Correction Year n (m ± SE) R p-value R p-value North 2009-2012 92 71.4 (7.2) 0.5832 0.001 0.5366 0.001 2009 18 232.0 (35.3) 0.8382 0.036 0.8342 0.25 2010 23 195.9 (29.8) 0.8002 0.007 0.6766 0.02 2011 23 190.1 (25.3) 0.7766 0.003 0.6997 0.04 2012 28 169.0 (22.6) 0.7614 0.002 0.7254 0.03 South 2009-2012 20 66.3 (13.1) 0.5579 0.001 0.4607 0.001 2009 5 118.2 (16.0) 0.4825 0.004 0.4786 - 2010 5 182.8 (2.6) 0.7461 0.05 0.7433 - 2011 5 198.1 (44.8) 0.8088 0.09 0.7627 - 2012 5 231.4 (61.8) 0.9446 0.16 0.5534 -

Table 2. Clark and Evans aggregation index for landscape level habitat variables with and without a cumulative distribution function edge correction. Significant index values <1 indicate clustering, while significant index values >1 indicate repulsion. Nearest neighbor is the mean distance to the nearest conspecific neighbor and its associated standard error.

Nearest neighbor No Correction Edge Correction North n (m ± SE) R p-value R p-value Ponds 1,013 42.4 (0.5) 1.148 1.0 1.135 1.0 Trees 305 83.6 (1.8) 1.243 1.0 1.212 1.0 South Ponds 113 55.7 (2.3) 1.081 0.72 1.042 0.78 Trees 123 64.2 (2.2) 1.300 1.0 1.229 1.0

Table 3. Results of Moran’s I tests of spatial autocorrelation for six vegetation features associated with Hudsonian Godwit nests for the North and South plots. Each variable was evaluated using 20, 50, and 100 distance classes and two subsets of point vegetation data: all points and nests only. The significance of Moran’s I coefficients for each distance class was evaluated using a Bonferroni correction. When significant spatial autocorrelation was detected at a given distance class, the median distance (m) of that class is reported; “NS” indicates that the result was not significant. When significant spatial autocorrelation was detected for multiple distance classes, the range of the median distance of the closest and farthest distance classes is reported, along with the p-value associated with those classes. The percentage of distance classes with a significant Moran’s I value is given as the % significant. The percent of significant distance classes that had a positive Moran’s I value indicating a cluster is given as the % positive.

North Plot All points Nests only

% % % % Distance p-value Distance p-value signif positive signif positive Water depth (cm) 20 56-1,102 15 66 0.005 NS NS NS NS 50 28-1,168 8 75 0.005 NS NS NS NS 100 19-1,153 4 75 0.005 NS NS NS NS

Sedge and grass cover 20 91-3,031 35 29 0.005 NS NS NS NS 50 51-2,443 45 33 0.005 NS NS NS NS 100 34-2,821 14 36 0.005 NS NS NS NS Shrub cover 20 NS NS NS NS NS NS NS NS 50 NS NS NS NS NS NS NS NS 100 NS NS NS NS NS NS NS NS

Forb cover 20 91-985 35 29 0.005 NS NS NS NS 50 51-843 18 44 0.005 NS NS NS NS 100 34-822 8 75 0.005 NS NS NS NS Bare ground 20 91-3,030 15 66 0.005 1,110 5 0 0.005

50 51-1,058 6 33 0.005 NS NS NS NS

100 34-2,821 5 40 0.005 NS NS NS NS Richness 20 91 5 100 0.005 NS NS NS NS 50 51-3,311 6 66 0.005 NS NS NS NS 100 34-2,821 3 66 0.005 NS NS NS NS

South Plot All points Nests only

% % % % Distance p-value Distance p-value signif positive signif positive Water depth (cm) 20 35-1,068 55 45 0.005 NS NS NS NS 50 21-867 36 44 0.005 NS NS NS NS 100 13-972 21 52 0.005 NS NS NS NS Sedge and grass cover 20 45-843 20 50 0.005 397-429 10 50 0.005 50 27-849 14 57 0.005 395-419 4 50 0.005 100 18-859 9 56 0.005 430 1 0 0.005 Shrub cover 20 NS NS NS NS NS NS NS NS 50 385 2 100 0.005 NS NS NS NS 100 381 1 100 0.005 NS NS NS NS Forb cover 20 45-590 40 38 0.005 NS NS NS NS 50 27-612 30 33 0.005 NS NS NS NS 100 18-945 22 50 0.005 286 1 0 0.005 Bare ground 20 45-590 15 66 0.005 NS NS NS NS 50 27-849 10 40 0.005 NS NS NS NS 100 18-859 3 66 0.005 NS NS NS NS Richness 20 45-507 20 25 0.005 NS NS NS NS 50 27-419 6 66 0.005 NS NS NS NS 100 18-491 5 80 0.005 NS NS NS NS

Table 4. Clark and Evans aggregation index for hatched nests and depredated nests on the North plot. Significant index values <1 indicate clustering, while significant index values >1 indicate repulsion. Nearest neighbor is the mean distance to the nearest conspecific neighbor and its associated standard error. Sample sizes for some individual years were too small to generate p- values.

North Plot Nearest neighbor No Correction Edge Correction Year n (m ± SE) R p-value R p-value Hatched 2009-2011 53 109.4 (13.1) 0.6132 0.001 0.6298 0.001 2009 17 238.3 (37.0) 0.8366 0.03 0.8327 0.29 2010 18 261.4 (48.7) 0.9443 0.11 0.6222 0.03 2011 18 233.0 (33.9) 0.8419 0.02 0.7295 0.07 Depredated 2009-2011 11 238.1 (41.2) 0.8601 0.05 0.7345 0.13 2009 1 - - - - - 2010 5 278.1 (108.1) 0.5296 0.01 0.2019 - 2011 5 412.4 (221.2) 0.7853 0.08 0.3664 -

Table 5. Summary of model selection results for an analysis of Hudsonian Godwit nest fate in Beluga River, Alaska between 2009-2011using Akaike's Information Criteria corrected for small sample sizes (AICc). Two variables (distance to the nearest conspecific neighbor nest and distance to the road) were modeled for an analysis of nest fate. Number of parameters (k), the AIC values corrected for small sample size (AICc), the difference in AICc values, and Akaike's weights (wi) are presented for all models.

Models k AICc ΔAICc wi Nest Fate~1 + Year 3 19.92 0.00 0.742 Nest Fate~NND + Year 4 23.40 3.48 0.130 Nest Fate~Distance to Road + Year 4 23.66 3.74 0.115 Nest Fate~Distance to Road + NND + Year 5 28.04 8.12 0.013 Nest Fate~Distance to Road * NND + Year 6 32.81 13.19 0.001

Figure 1. Hudsonian Godwit nest locations and study plot boundaries in 2009, 2010, 2011, and 2012 for the North plot (a) and South plot (b).

Figure 2. Ripley’s K (transformed to L(r)) for all Hudsonian Godwit nests at each plot found between 2009 and 2012. Solid black lines represent values for the point patterns (observed), dashed red lines represent the expectation under complete spatial randomness (theoretical), and gray lines represent the 99% confidence interval based on 999 randomizations of a Poisson point process. Values above the upper bounds of the confidence envelope indicate clustering at distance r, and values below the lower bounds indicate regularity.

All Nests a North Plot b SouthAll Nests Plot 4e+05 3e+05 1500000 2e+05 1e+05 1000000 0e+00 L(r)=K(r)-pi*r^2 L(r)=K(r)-pi*r^2 500000 -1e+05 0 -2e+05 -3e+05

0 200 400 600 800 0 100 200 300 r r

Figure 3. PCF test results for all nests found on the North plot (a) and South plot (b). Solid black lines represent values for the point patterns (observed), dashed red lines represent the expectation under complete spatial randomness (theoretical), and gray lines represent the 99% confidence interval based on 999 randomizations of a Poisson point process. Values above the upper bounds of the confidence envelope indicate clustering at distance r, and values below the lower bounds indicate regularity.

a envelope(mypatternall, North fun = pcf, nsim = 999) b envelope(mypatternall, South fun = pcf, nsim = 999) 5 10 g^ r ^ obs( ) gobs(r) g r theo( ) gtheo(r) g^ r ^ hi( ) ghi(r) 4 g^ r 8 ^ lo( ) glo(r) 3 6 ) ) r r ( ( g g 2 4 1 2 0 0

0 200 400 600 800 0 100 200 300

r r

Figure 4. Ripley’s K (transformed to L(r)) for Hudsonian Godwit nests for North plot (a) and South plot (b) for each year between 2009 and 2012. Solid black lines represent values for the point patterns (observed), dashed red lines represent the expectation under complete spatial randomness (theoretical), and gray lines represent the 99% confidence interval based on 999 randomizations of a Poisson point process. Values above the upper bounds of the confidence envelope indicate clustering at distance r, and values below the lower bounds indicate regularity.

a 2009 2010 2500000 1000000 0 1000000 L(r)=K(r)-pi*r^2 L(r)=K(r)-pi*r^2 -500000 -1000000 0 200 400 600 800 0 200 400 600 800

r r

2011 2012 1000000 1000000 0 0 L(r)=K(r)-pi*r^2 L(r)=K(r)-pi*r^2 -1000000 0 200 400 600 800 -1000000 0 200 400 600 800

r r

b 2009 2010 8e+05 4e+05 4e+05 0e+00 L(r)=K(r)-pi r^2 L(r)=K(r)-pi r^2 L(r)=K(r)-pi 0e+00 -4e+05 -4e+05 0 100 200 300 0 100 200 300

r r

2011 2012 4e+05 0e+00 0e+00 L(r)=K(r)-pi r^2 L(r)=K(r)-pi r^2 L(r)=K(r)-pi -4e+05 -4e+05 0 100 200 300 0 100 200 300

r r

Figure 5. PCF test results for all nests on the North plot (a) and South plot (b) for each year between 2009 and 2012. Solid black lines represent values for the point patterns (observed), dashed red lines represent the expectation under complete spatial randomness (theoretical), and gray lines represent the 99% confidence interval based on 99 randomizations of a Poisson point process. Values above the upper bounds of the confidence envelope indicate clustering at distance r, and values below the lower bounds indicate regularity.

a envelope(mypattern09, 2009 fun = pcf, nsim = 9) envelope(mypattern10, 2010 fun = pcf, nsim = 99) 5 5 ^ ^ gobs(r) gobs(r) gtheo(r) gtheo(r) g^ r g^ r hi( ) hi( ) 4 4 ^ ^ glo(r) glo(r) 3 3 ) ) r r ( ( g g 2 2 1 1 0 0

0 200 400 600 800 0 200 400 600 800

r r

envelope(mypattern11, 2011 fun = pcf, nsim = 99) envelope(mypattern12, 2012 fun = pcf, nsim = 99) 5 5 g^ r ^ obs( ) gobs(r)

gtheo(r) gtheo(r) g^ r ^ hi( ) ghi(r) 4 g^ r 4 ^ lo( ) glo(r) 3 3 ) ) r r ( ( g g 2 2 1 1 0 0

0 200 400 600 800 0 200 400 600 800

r r

b envelope(mypattern09, 2009 fun = pcf, nsim = 9) envelope(mypattern10, 2010 fun = pcf, nsim = 9) 10 10 g^ r ^ obs( ) gobs(r) g r theo( ) gtheo(r) g^ r ^ hi( ) ghi(r) 8 g^ r 8 ^ lo( ) glo(r) 6 6 ) ) r r ( ( g g 4 4 2 2 0 0

0 100 200 300 0 100 200 300

r r

envelope(mypattern11, 2011 fun = pcf, nsim = 9) envelope(mypattern12, 2012 fun = pcf, nsim = 9) 10 10 ^ ^ gobs(r) gobs(r)

gtheo(r) gtheo(r) ^ ^ ghi(r) ghi(r) 8 8 ^ ^ glo(r) glo(r) 6 6 ) ) r r ( ( g g 4 4 2 2 0 0

0 100 200 300 0 100 200 300

r r

Figure 6. Ripley’s K (transformed to L(r)) for pond centroid locations for each study plot. Solid black lines represent values for the point patterns (observed), dashed red lines represent the expectation under complete spatial randomness (theoretical), and gray lines represent the 99% confidence interval based on 999 randomizations of a Poisson point process. Values above the upper bounds of the confidence envelope indicate clustering at distance r, and values below the lower bounds indicate regularity.

a NorthPonds - North Plot Plot b SouthPonds - South Plot Plot 60000 80000 60000 40000 40000 20000 0 20000 L(r)=K(r)-pi*r^2 L(r)=K(r)-pi*r^2 0 -20000 -20000 -40000 -40000 0 200 400 600 800 0 100 200 300

r r

Figure 7. PCF test results for ponds on the (a) North plot and (b) South plot. Solid black lines represent values for the point patterns (observed), dashed red lines represent the expectation under complete spatial randomness (theoretical), and gray lines represent the 99% confidence interval based on 999 randomizations of a Poisson point process. Values above the upper bounds of the confidence envelope indicate clustering at distance r, and values below the lower bounds indicate regularity.

a envelope(mypattern, North fun = pcf, nsim = 999) b envelope(mypattern, South fun = pcf, nsim = 999) 2.0 2.0 ^ g^ r gobs(r) obs( ) g r gtheo(r) theo( ) ^ g^ r ghi(r) hi( ) ^ g^ r glo(r) lo( ) 1.5 1.5 ) ) r r ( ( 1.0 1.0 g g 0.5 0.5 0.0 0.0

0 200 400 600 800 0 100 200 300

r r

Figure 8. Ripley’s K (transformed to L(r)) for black spruce tree islands centroid locations for two study plots. Solid black lines represent values for the point patterns (observed), dashed red lines represent the expectation under complete spatial randomness (theoretical), and gray lines represent the 99% confidence interval based on 999 randomizations of a Poisson point process. Values above the upper bounds of the confidence envelope indicate clustering at distance r, and values below the lower bounds indicate regularity.

a NorthTrees - NorthPlot Plot b SouthTrees - South Plot Plot 60000 1e+05 40000 5e+04 20000 0 L(r)=K(r)-pi*r^2 L(r)=K(r)-pi*r^2 0e+00 -5e+04 -20000 -40000 -1e+05 0 200 400 600 800 0 100 200 300

r r

Figure 9. PCF test results for black spruce tree islands on the (a) North plot and (b) South plot. Solid black lines represent values for the point patterns (observed), dashed red lines represent the expectation under complete spatial randomness (theoretical), and gray lines represent the 99% confidence interval based on 999 randomizations of a Poisson point process. Values above the upper bounds of the confidence envelope indicate clustering at distance r, and values below the lower bounds indicate regularity.

a envelope(mypattern, North fun = pcf, nsim = 999) b envelope(mypattern, South fun = pcf, nsim = 999) 2.0 2.0 ^ ^ gobs(r) gobs(r)

gtheo(r) gtheo(r) ^ ^ ghi(r) ghi(r) ^ ^ glo(r) glo(r) 1.5 1.5 ) ) r r ( ( 1.0 1.0 g g 0.5 0.5 0.0 0.0

0 200 400 600 800 0 100 200 300

r r

Figure 10. Density plots of six habitat variables known to be important at predicting Hudsonian Godwit nest locations for North plot (a) and South plot (b). Open circles represent nest locations.

a Water depth (cm) Number of species (Richness)

Sedge/Grass cover 30cm

Forb cover Bare ground

b Water depth (cm) Number of species (Richness)

Sedge/Grass cover 30cm

Forb cover Bare ground

Figure 11. Ripley’s K test transformed into L(r) for all nests that hatched successfully or were depredated during incubation for the North plot between 2009 and 2011. Solid black lines represent values for the point patterns (observed), dashed red lines represent the expectation under complete spatial randomness (theoretical), and gray lines represent the 99% confidence interval based on 999 randomizations of a Poisson point process. Values above the upper bounds of the confidence envelope indicate clustering at distance r, and values below the lower bounds indicate regularity.

Hatched 1000000 L(r)=K(r)-pi*r^2

-500000 0 200 400 600 800

Depredated 2e+06 0e+00 L(r)=K(r)-pi*r^2

-2e+06 0 200 400 600 800

APPENDIX

Figure S1. Vegetation sampling locations used in Moran’s I analysis within the bounded study area for North plot (a) and South plot (b).

a North Plot b South Plot

Figure S2. Pond centroid locations with study plot boundaries for the North plot (a) and South plot (b).

a North Plot b South Plot

Figure S3. Black spruce tree island centroid locations with study plot boundaries for the North plot (a) and South plot (b).

a North Plot b South Plot

Figure S4. Satellite imagery (Google Earth) with polygons for ponds (blue) and black spruce tree island (green) within study plot boundaries for the North plot (a) and South plot (b).

a North Plot b South Plot

Figure S5. Satellite imagery (Google Earth) with polygons for ponds (blue) within study plot boundaries for the North plot (a) and South plot (b).

a North Plot b South Plot

Figure S6. Satellite imagery (Google Earth) with polygons for black spruce tree island (green) within study plot boundaries for the North plot (a) and South plot (b).

a North Plot b South Plot

Figure S7. Hudsonian Godwit nest locations categorized by nest fate within study plot boundaries for 2009-2011 for the North plot (a) and South plot (b).

a North Plot b South Plot

Figure S8. Hudsonian Godwit nest locations categorized by nest fate within study plot boundaries in 2009 (a), 2010 (b), and 2011 (c) for the North plot.

a 2009 b 2010

c 2011

100