Citation
Mountain plover breeding ecology : home-range size, habitat use, and nest survival in an agricultural landscape

Material Information

Title:
Mountain plover breeding ecology : home-range size, habitat use, and nest survival in an agricultural landscape
Creator:
Woolley, Colin Allan
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( Master of Science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Integrative Biology, CU Denver
Degree Disciplines:
Biology

Record Information

Source Institution:
University of Colorado Denver Collections
Holding Location:
Auraria Library
Rights Management:
Copyright COLIN ALLAN WOOLLEY. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
MOUNTAIN PLOVER BREEDING ECOLOGY: HOME-RANGE SIZE, HABITAT USE,
AND NEST SURVIVAL IN AN AGRICULTURAL LANDSCAPE
by
COLIN ALLAN WOOLLEY B.A., Prescott College, 2006
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Biology Program
2016


This thesis for the Master of Science degree by Colin Allan Woolley has been approved for the Department of Integrative Biology by
Michael Wunder, Chair Rafael Moreno Jennifer Blakesley
Date: 17 December 2016


Woolley, Colin Allan (M.S., Biology Program)
Mountain Plover Breeding Ecology: Home-range Size, Habitat Use, and Nest Survival in an Agricultural Landscape
Thesis directed by Associate Professor Michael B. Wunder
ABSTRACT
Mountain Plovers (Charadrius montanus), a species of conservation concern throughout their breeding range, frequently nest on agricultural crop fields that have replaced native short-grass prairie habitat. To understand how plovers nesting on crop fields use habitats during nesting, I estimated home-range size and habitat use during the nest incubation period using GPS loggers. I also sampled available invertebrate prey species from across a range of habitat types to quantify potentially available biomass and caloric density. Average home-range size during nest incubation was 155.4 54.9 (SE) ha. Plovers foraged almost exclusively on fallow crop fields and avoided fields with growing crops. I found no significant difference in invertebrate prey biomass or caloric density between grassland and agricultural fields. I also modelled daily nest survival of plovers to evaluate the influence of various weather variables and individual traits. Nest survival averaged 56.1% over the three year study period, and was best explained by a model including a quadratic time trend, a year effect, and an interaction effect of precipitation and maximum daily temperature.
The form and content of this abstract are approved. I recommend its publication.
Approved: Michael B. Wunder
m


ACKNOWLEDGEMENTS
Funding for this study was provided by University of Colorado-Denver, Bird Conservancy of the Rockies, Nebraska Prairie Partners, Lois Webster Fund of the Audubon Society of Greater Denver, and Colorado Field Ornithologists. I would like to thank the landowners of Kimball County, NE and Weld County, CO for granting access to their land. I especially thank Larry Snyder, local landowner and employee of Bird Conservancy of the Rockies, for his assistance in coordinating with other landowners and providing logistical support. Craig Strieker assisted in use of the bomb calorimeter. Sara Oyler-McCance processed feather samples for molecular sexing. I also thank Clay Edmonson, Tasha Blecha, Molly Elderbrook, Kelly Surgalski, Lacey Clarke and Eric Matechak for their work in the field. All fieldwork followed handling protocols approved by University of Colorado Institutional Animal Care and Use Committee protocol #102244.
IV


TABLE OF CONTENTS
CHAPTER
I. HOME-RANGE SIZE AND HABITAT USE OF MOUNTAIN
PLOVERS DURING NEST INCUBATION...................................1
Introduction.....................................................1
Materials and Methods............................................3
Results..........................................................7
Discussion.......................................................8
Figures.........................................................11
II. NEST SURVIVAL.....................................................14
Introduction....................................................14
Materials and Methods...........................................16
Results.........................................................19
Discussion......................................................21
Figures and Tables..............................................24
REFERENCES......................................................28
APPENDIX........................................................32
v


CHAPTER I
HOME-RANGE SIZE AND HABITAT USE OF MOUNTAIN PLOVERS DURING
NEST INCUBATION Introduction
Over the last 150 years, grasslands of the Great Plains have undergone many changes, including conversion to croplands and loss of native grazers (e.g. bison, prairie dogs). During that period, over 80% of North American grasslands have been altered by agriculture, energy production and urbanization (White et al. 2000). This has, in part, led to the decline of numerous grassland-endemic bird populations (Knopf 1994; Vickery & Herkert 2001). The Mountain Plover (Charadrius montanus) is a ground-nesting migratory shorebird that breeds primarily in the short-grass prairie of the western Great Plains. It is a species of conservation concern in our study area (Schneider et al. 2011, Colorado Parks and Wildlife 2015) and throughout its range. Mountain Plovers are a bare-ground nesting specialist (Graul 1975, Knopf and Miller 1994, Knopf and Wunder 2006, Augustine and Derner 2012), drawn to areas of localized disturbance from grazing, fire, or agricultural tillage. Conversion of short-grass prairie for crop production has greatly reduced availability of native breeding habitat (Samson et al. 2004). However, Mountain Plover readily nest in privately-owned agricultural crop fields, particularly in fields that are fallow or with low-growing crops (Shackford et al. 1999; Knopf & Rupert 1999; Bly et al. 2008). Conservation and management of at-risk populations require an understanding of the spatial context in which they forage, reproduce, and meet their needs for survival (Levin 1992). Mountain Plovers nesting on agricultural
1


crop fields are presented with a spatial context that differs from the native short-grass prairie in which they evolved.
Mountain Plover home-range size and habitat use patterns have been studied during the post-hatch breeding season when adult plovers are moving around the landscape with precocial broods (see Knopf and Rupert 1996, Dreitz et al. 2005) but nothing is known about these variables during the incubation period. This is a potentially substantial omission because of the unusual breeding system of plovers, which can be described as split clutch. Female plovers lay (typically) three eggs in one nest scrape for the male to independently incubate, before laying another three eggs in a different nest scrape for her to incubate. There is no food provisioning during incubation; males and females independently forage and tend the nests through fledging of the broods. Therefore, the nesting adult does not benefit from mate-feeding and must balance incubation duties with its own foraging needs (Martin 1987). Habitat use and home-range size during incubation will likely differ from that during broodrearing, as the adults are not tethered to a mobile brood but rather to a stationary nest. For species with precocial young, the incubation period represents a time of high energetic investment when adults are at increased risk of starvation and mortality (Martin 1987). Daytime high temperatures in our study site typically reach upwards of 30C and adults must shade the eggs to prevent overheating. However, when temperatures are moderate, Mountain Plovers may leave the nest unattended to seek foraging opportunities. Long-distance commutes between distinct nesting and foraging grounds have been found in other species of Charadriidae (Whittingham et al. 2000) though it is unclear if this is also the case for Mountain Plovers.
2


Mountain Plover diets consist primarily of terrestrial invertebrates (Knopf and Wunder 2006), including beetles (Coleoptera), grasshoppers (Orthoptera), and ants (.Hymenoptera) as the most common prey items. Foraging preferences for one habitat type over another may be driven by differing food resource availability (Peek 1986, Manly et al. 2002). Our objectives were (1) to estimate plover home-range size during the incubation period; (2) to quantify foraging habitat preference in a variable landscape; and (3) to evaluate any habitat preferences in relation to biomass and caloric content of invertebrate prey sources in each habitat type. Understanding how adult plovers use agricultural landscapes during this vulnerable incubation period can inform management strategies for this species of conservation concern.
Materials and Methods
Study Area. Our study occurred mostly on privately-owned dryland agriculture in Kimball County, Nebraska and Weld County, Colorado (Figure 1). The dominant crops were winter wheat (Triticum aestivum) and millet (Panicum miliaceum). Interspersed among the agricultural fields were small patches of rangeland characterized by two native short-grass prairie species; blue grama (Bouteloua gracilis) and buffalograss (Bouteloua dactyloides). Additionally, some land was enrolled in the Conservation Reserve Program (CRP); a federal program administered by the Farm Service Agency in which land is taken out of agricultural production and planted to both native and non-native grasses to enhance wildlife habitat and various ecosystem services.
Field Methods. I gathered data on plover home ranges and habitat use during the 2014 and 2015 breeding seasons. Plover movements during the incubation period have traditionally been difficult to document using radio-telemetry; foraging bouts off the nest
3


may be short and observer bias is introduced when plovers arent located quickly; plovers close to their nest are more easily located while a plover foraging far from the nest may outrange the telemetry reception and be difficult to locate until it returns to the nest area. This results in a bias towards over-representation of observations close to the nest when using radio-telemetry. I deployed GPS-tags (Pinpoint 50, Lotek, Newmarket, Ontario, Canada) on a programmed schedule to avoid this observer bias.
This study was conducted alongside a nest-marking program administered by Bird Conservancy of the Rockies, in which nests are located and marked with wooden stakes to prevent accidental tillage. After nests were located, the incubating adult was caught using a walk-in box trap placed over the nest. When nests were discovered I estimated nest age using egg floatation (Dinsmore et al. 2002) and assumed a 29-day incubation period (Knopf and Wunder 2006). To reduce risk of nest abandonment, I tagged plovers that were at least seven days into incubation. The tags collect 50 locational points on a programmable schedule, after which the bird must be re-caught to retrieve the data. When possible, I redeployed tags a second time on each incubating plover for a total of 100 points per bird. I was unable to retrieve tags from plovers if their nest failed during the deployment period, as they left the area after nest loss. Because Skrade and Dinsmore (2012) found that incubating Mountain Plovers are off the nest more frequently during crepuscular and nocturnal times, I programmed the tags to take proportionately more locational fixes during these times.
Data Analysis. Spatial and temporal autocorrelation are problematic when estimating home ranges because most models assume that each successive locational fix is independent of previous fixes (Swihart and Slade 1986). To reduce the problems of serial autocorrelation,
I randomized the fix schedule according to the following ruleset (all times in MST): 4 fixes
4


taken at random from 0500 to 0759; 1 from 0800 to 1759; 4 from 1800 to 2059; 1 from 2100 to 0459; all with at least 20 min between each fix. Any autocorrelation still present may lead to underestimated home-range sizes due to biased bandwidth selection (Swihart and Slade 1986). Temporal autocorrelation tends to decrease distances between fixes, leading to a smaller kernel bandwidth estimate and smaller associated home-range estimates.
The GPS-tags record metadata for each locational fix, including the number of satellites used in acquiring the fix. Prior in situ field-testing suggested that four or more satellites were necessary for accuracy under 10 meters. Thus, I censored fixes with fewer than four satellites (10.0% of fixes). I also censored fixes within 10 meters of the nest (64.2% of fixes), as these fixes potentially represented a bird attending its nest and not on a foraging bout away from the nest.
I estimated home-range size during the incubation and brood-rearing periods using a fixed-kernel method with bandwidth selection through least squares cross validation (Seaman and Powell 1996). This method is non-parametric and has been shown to have less bias than other bandwidth estimators when used when used with non-normally distributed data and small sample sizes (Seaman and Powell 1996). I defined home-range size using the 95% isopleth within the kernel density estimate. Home-ranges were estimated using ArcGIS 10.1 (ESRI2012) with Geospatial Modelling Environment (Beyer 2012) and program R (R Core Team 2015).
I characterized habitat surrounding each nest on the following nominal scale: fallow (including stubble fields), growing crop (including cover crops), grassland (including grazed rangeland or land enrolled in the CRP) or other (consisting of residential houses, yards, or unknown habitat we could not visually confirm). I estimated habitat preference by comparing
5


differences in proportion of habitat types used to that of habitat types available. This comprises design III habitat selection (Manly et al. 2002, Johnson 1980) as the proportions of used and available habitat types are estimated separately for each individual plover. I used the log-ratio differences of used and available habitat proportions to determine preference (Aebischer et al. 1993). A positive value indicates habitat selection while a negative value indicates habitat avoidance. Use was defined by the observed occurrence within each habitat type. Typically, availability is defined by the proportional occurrence of habitat types within some defined home range (Aebischer et al. 1993). However, movement patterns for an incubating bird differ from typical use within a home-range in that all foraging bouts begin and end at the same location (the nest). In this case, areas closer to the nest will show greater use simply due to their proximity to the nest and not necessarily due to a habitat preference. To account for this, I used a bootstrapping approach (Efron 1979) to define availability for each bird by generating 10,000 random sets of points (where n = no. of observed points for that bird) and summing habitat distributions across them. A locational fix can be defined as a distance and direction from the nest. I generated the randomly chosen availability points by separately pooling all observed distances and directions and sampling with replacement from these pooled data. This results in a model of availability that does not assume a uniform distribution across a home range. To evaluate any directional preference in off-nest movements I used a Rayleigh Test of Uniformity (Batschelet 1981). This test compares observed directional data with a null hypothesis of directional uniformity. These analyses were performed using program R (R Core Team 2015).
To estimate available food resources across habitat types, I sampled the terrestrial invertebrate prey species using pitfall traps. I randomly selected three sites on rangeland and
6


three sites on fallow fields. I placed a 4 by 4 array of pitfall traps spaced 1 meter apart at each site. Pitfall traps were 16-oz Solo cups (95mm diameter, 115mm depth) placed into the ground such that the rim of the cup was flush with the ground surface. Sampling efforts consisted of a 24-hour period, with three sampling efforts per site. Each sampling effort also included a 200m walking transect using a bug net to sweep for orthopterans that may not be captured in the pitfall traps. I dried the invertebrate samples for 24hr at 60 C and used a bomb calorimeter (IKA C2000 Basic) to measure caloric density for each sampling effort.
Results
Home Ranges. I attached GPS-tags to 15 plovers across 2014 and 2015.1 recovered GPS-tags from nine plovers, five of which were tagged a 2nd time for additional points.
Home range analyses were based on a mean of 21.2 4.4 (SE) locational fixes per bird. Mean home range size was 155.4 54.9 (SE) ha (see Fig. 1 for examples). Mean distance from the nest for locational fixes of all birds combined was 385.4 26.4 (SE) meters. The Rayleigh Test of Einiformity found no significant directional preference to plover movements (p=0.238).
Habitat Use. I estimated the log-ratio differences of used and available habitat proportions for each of the nine tagged plovers (See Appendix). This method allows for habitats to be ranked from most to least preferred. For all individuals, fallow fields ranked as the most preferred habitat type. Overall, fallow fields accounted for 92.7% of all observed locational fixes. I used the observed and expected proportions for each plover to create bootstrapped 95% confidence intervals of plover habitat use (Fig. 2). Non-overlapping 95% confidence intervals of Observed and Available for fallow and growing crop habitats further suggest use and avoidance of cover types, respectively. Plovers also used crop fields nearly
7


exclusively for nest-site location throughout the duration of our study, with 168 of 170 found nests on agricultural crop fields.
Invertebrate Prey Densities and Caloric Content. The estimate for mean biomass of invertebrates captured per 24hr sampling period was slightly higher on rangeland (1.411 g/day .353 SE, n=6) than on agricultural fields (0.873 g/day .162 SE, n=8) though the difference was not statistically significant (Welch two-sample t-test; ti,\= -1.39, p= .207).
The mean caloric density of invertebrates captured on rangeland (5467 cal/g 99.2 SE) was similar to agricultural fields (5502 cal/g 81.8 SE) (Welch two-sample t-test; tio.6= -28, p=788).
Coleoptera were by far the most commonly captured invertebrates on both habitat types and overall proportions of taxa captured were also similar, with slightly more Hymenoptera and fewer Orthoptera on rangeland (Fig. 3). Orthoptera are a common prey item throughout the plovers range (Knopf and Wunder 2006) though they are scarce at the study site. No Orthopterans were captured in sweep-netting during the walking transects associated with each invertebrate sampling effort.
Discussion
Home Range Size. The point estimate for mean home range size during incubation was very similar to a previously published home range size (131.6 ha; Dreitz et al. 2005) based on the brood-rearing period for Mountain Plovers that nested on fallow crop fields.
This suggests that Mountain Plover home range size remains similar throughout the nesting period, regardless of whether the adult is tending to eggs or a mobile brood. It is possible that home ranges during incubation and brood-rearing may not overlap, which would result in a larger home range size overall. I had insufficient sample size to detect any differences in
8


home range size by sex. It is possible that home-range size may differ by sex, because females may have to overcome a greater energy deficit associated with egg-laying. Skrade and Dinsmore (2012), however, found no difference in incubation patterns by sex and given that there is no sexual size dimorphism it is just as plausible that home-range sizes between the sexes are similar.
Habitat Use and Invertebrate Prey Densities. It is clear that fallow fields play an important role for Mountain Plovers nesting in agricultural settings. All nine tagged plovers used fallow fields nearly exclusively and avoided areas with growing crops. Patches of native grassland, while not common in the study area, showed no preferential use. The biomass, caloric content, and proportional distribution of taxa for invertebrate prey species on rangeland and crop fields were similar. Farmers in the study area typically plant crops in alternating strips that are 100 to 120 meters wide, with fallow strips in between. These results suggest that these alternating fallow strips provide sufficient foraging opportunities for nesting plovers. I cannot rule out the possibility of an energetic cost associated with foraging on fallow fields as compared to native short-grass prairie. Weight loss during incubation is fairly common in species with precocial young (Martin 1987). There was some evidence of weight loss over the incubation period (CAW unpublished data) but this hasnt been well quantified for Mountain Plovers nesting in any habitat.
Management Implications. The population-level implications of plovers nesting on crop fields are still not fully understood. Dreitz and Knopf (2007) found that nest survival on agricultural fields was similar to nest survival on native rangeland, though the cause of nest mortality differed. Nests on crops were destroyed primarily by mechanical farming operations (e.g., tilling, planting) while rangeland nests were destroyed by predators.
9


Agricultural fields and grassland appear to be interchangeable in terms of nest survival, home-range size, and available food resources. These results suggest that the fallow fields are supporting both the nest-site and foraging requirements of incubating plovers and that plovers are not using the interspersed patches of grassland for foraging opportunities, nor for nest site location. This suggests that an agrarian landscape with interspersed fallow areas can support the full breeding cycle of Mountain Plovers and a lack of native grassland patches will not preclude plover use.
10


JHI
FIG. 1. Ninety-five percent fixed kernel untilization distributions for two different incubating Mountain Plovers (blue shading). The plover locations are shown as blue circles and the nest locations are shown as black crosses.
11


Proportion of total
o
Fallow Growing Crop Grassland
FIG. 2. Bootstrapped ninety-five percent confidence intervals of plover habitat observed use and availability by habitat type.
12


% of total
cq
o
to
o
o
OJ
o

o
Other
Orthoptera
Araneae
Hymenoptera
Coleoptera
FIG. 3. Proportional distribution of prey taxa biomass by habitat type.


CHAPTER II
NEST SURVIVAL Introduction
Short-grass prairies of the western Great Plains are among the most heavily altered ecosystems in North America (Samson et al. 2004). These changes are characterized by conversion of grassland to agricultural production and the loss of native grazers such as bison {Bison bison) and prairie dogs (Cynomis spp.). Over the last century, avifauna of the Great Plains grasslands has experienced population declines greater than those found in any other North American biome (Knopf 1994). Mountain Plovers {Charadrius montanus) in particular have experienced population declines averaging an estimated 3.15% per year from 1966 to 2012, according to Breeding Bird Survey data (Sauer et al. 2014).
Mountain Plover is a mid-sized, migratory shorebird species that breeds in arid grasslands of the western Great Plains and winters from southern California to Texas and into northern Mexico (Knopf and Wunder 2006). While conversion of native short-grass prairie to agriculture has likely contributed to the species decline, Mountain Plovers are one of just a few short-grass prairie specialists that will nest on agricultural fields (Shackford et al. 1999; Knopf & Rupert 1999). Mountain Plovers use a split-clutch mating system, in which the female lays 3 eggs in a nest that will be tended by the male before laying 3 more eggs in a separate nest that she alone will tend (Knopf and Wunder 2006). Mountain Plovers are arid, bare-ground nesting specialists (Knopf and Wunder 2006) that were historically associated with bare ground created by grazing of native herbivores. Modem farming, particularly the practice of leaving some fields fallow during the spring and summer, has replaced the native grazers in providing this bare-ground habitat in many parts of the plovers range.
14


Climate models for across the Mountain Plovers breeding range predict higher daily temperatures and lower precipitation levels overall though with more frequent and severe precipitation events such as hail (Matthews 2008). Nest survival of Mountain Plover has been previously studied across their range (see Graul 1975, Knopf and Rupert 1996, Dinsmore et al. 2002, Mettenbrink et al. 2006, Dreitz and Knopf 2007, Dreitz et al. 2012) with population-level seasonal survival estimates ranging from 26% (Knopf and Rupert 1996) up to 65% (Graul 1975). Several of these studies have considered the influence of daily weather measures on nest survival. Dinsmore et al. (2002) found that the daily survival rate (DSR) of nests decreased following precipitation events and that there was no evidence for an effect of maximum daily temperatures on DSR. Additionally, male-tended nests had higher survival probabilities than female-tended nests. Dreitz et al (2012) found that plover DSR increased during long (>10 day) dry periods and decreased at higher maximum daily temperatures. Whereas previous studies have included cropland-nesting plovers in their samples, this study is the first to exclusively consider cropland-nesting plovers.
My objectives were to estimate daily nest survival of cropland-nesting plovers to compare with previous studies of mostly grassland-nesting plovers. I compared individual-level traits such as sex of the incubating adult with daily climate variables and seasonal time trends to determine which covariates best explain whether a nest fails or is successful. Understanding the impacts of the modern agricultural landscape and a warming and increasingly volatile climate on plover nest survival and long-term population dynamics will aid future management decisions regarding this increasingly rare species.
15


Materials and Methods
Study Area. My study area consisted of mostly private agricultural land in Kimball County, Nebraska and northern Weld County, Colorado. Dryland agriculture dominates the landscape in this area (-80%), where winter wheat (Triticum aestivum) and millet (Panicum miliaceum) are the primary crops. Small patches of grazed rangeland and grassland enrolled in the Conservation Reserve Program (CRP) are interspersed throughout the agricultural matrix. The rangeland most closely resembles native shortgrass prairie habitat, consisting of blue grama (Bouteloua gracilis) and buffalograss (Bouteloua dactyloides). CRP grasslands in the study area have been planted to taller mixed-grass species such as western wheat-grass (Pascopyrum smithii) and crested wheatgrass (.Agropyron cristatum). In other parts of their range, Mountain Plover are highly associated with prairie dog colonies (Dinsmore et al.
2005, Tipton et al. 2008, Augustine 2011) but prairie dogs were nearly absent from the study area. The study area is along the eastern periphery of the Mountain Plovers breeding range.
Nest Monitoring. From 2013 to 2015 I began surveying for plover nests in late April and continued until the last nest hatched in late July. I located active nests by dragging a rope between two ATVs and systematically searching fallow crop fields (including standing stubble fields). In areas of low-growing crop I walked through the area and scanned for plovers. When plovers were located, I observed them until they returned to their nest.
This study was conducted alongside a nest-marking program administered by Bird Conservancy of the Rockies in partnership with local landowners. When we located an active nest, we marked the nest by placing 4 wooden stakes around the nest about 10 meters out. This was to prevent accidental tillage of nests when farmers later plowed these fields.
16


Mountain Plovers are typically very tolerant of heavy farming equipment operating in close proximity to their nest (Knopf and Wunder 2006) so long as the nest itself is not destroyed.
We checked nests every 2-7 days until the eggs hatched or the nest failed. Upon finding a nest, we floated the eggs to estimate nest age using guidelines provided by Dinsmore et al. (2002). We continued to float eggs about every 7 days until pipping cracks developed. A nest was considered successful if at least one of the eggs hatched. In cases of asynchronous hatch, the nest attempt was considered complete on the day of first hatch. For hatched nests, I estimated nest age by using the hatch date and assuming a 29 day incubation period (Knopf and Wunder 2006). If a nest failed, I estimated nest age by egg floatation. This is accurate to within several days except for one floatation stage covering 13 days. If a failed nest only had egg floatation data for this stage then I used the midpoint to estimate nest age. Sex of the incubating adult was determined by molecular sexing from a feather sample as there is no reliable field method for sexing this species.
Weather Variables. I obtained daily measures of precipitation and temperature values from the weather station nearest to the field site, located in Kimball, Nebraska and administered by the National Oceanic and Atmospheric Administration (NOAA 2016). All nests included in analyses were within 25km of this station.
Data Analysis. I estimated daily survival rate (DSR) of nests using program R (R Core Team 2015) and the package rmark which also utilizes program MARK (White and Burnham 1999). I used Akaikes information criterion for small samples (AICc) to evaluate which models best explained DSR (Akaike 1973, Burnham and Anderson 2002). I developed a set of models based on a priori biological hypotheses including variables that have
17


previously been found to influence DSR of Mountain Plovers and other grassland birds (see Dinsmore et al. 2002, Mettenbrink et al. 2006, Dreitz et al. 2012).
The sex of the incubating adult was determined for 114 out of 154 total nests. To begin constructing the set of models, I first compared a model with sex of the incubating bird to a constant survival model. The model with sex of the incubating bird performed poorly compared to the constant survival model. Thus, all further models did not include sex and were analyzed including the full sample of nests with adults of unknown sex. The a priori hypotheses used to construct models included in the analysis are described below:
1. Year. Environmental stochasticity can vary greatly from year to year. I included year as a variable to capture any annual variation that was not described by other listed hypotheses. For example, predator activity and frequency of farming operations can vary from year to year and they were not measured in this study.
2. Seasonal time trends. I compared a model of constant survival across the breeding season with both linear and quadratic time trends to allow for DSR to vary within a season. Higher DSR in the early season has been found in Mountain Plovers and other species (Ainley and Schlatter 1972, Dinsmore et al. 2002).
3. Nest age. For precocial species, DSR may increase as the nest approaches hatching as nests more vulnerable to predation are depredated early and increased time investment of the incubating adult may reduce nest abandonment.
4. Daily maximum temperatures. On sunny days plovers shade their eggs to prevent overheating. High temperatures could lead to lower DSR by direct overheating of eggs or increased energetic stress of adults foraging less in order to shade the eggs.
18


5. Precipitation. Dreitz et al. (2012) found that Mountain Plover DSR increased during drought periods (defined as 10+ days of < 1mm rain) and Dinsmore et al. (2002) found lower DSR following rain events. I modelled daily precipitation as a continuous variable measured in mm. I also included a one-day lag effect. Precipitation events may destroy nests directly (e. g. hail) or increase activity of olfactory predators following precipitation events (hence the one-day lag effect).
6. Tagged birds. In this study, 36 adult birds were fitted with GPS tags or radiotransmitters for tracking purposes. These birds experienced greater (although brief) disturbance at their nest site as they were trapped multiple times over the course of the nesting cycle. I included tagged birds as a group to examine whether this increased disturbance influenced their DSR.
In addition to including tagged birds into the modelling analysis, I also used a Fishers exact test comparing nest fates for tagged vs. untagged birds. Fishers exact test is less biased than chi-squared tests when sample sizes (and expected values) are small. To reduce risk of nest abandonment, I only tagged birds that were at least seven days into nest incubation. Thus, nests that failed during the first seven days of incubation were censored from this test.
All models except for the constant survival model included a year effect and either a linear or quadratic time trend (see table 1 for candidate models).
Results
Nest Survival. The analysis included 154 plover nests. Of these, 106 (68.8%) hatched, 16 (10.4%) were abandoned, 10 (6.5%) were destroyed by farming practices (e.g. tilling and planting of fields), 15 (9.7%) were depredated, and 7 failed due to other reasons (hail
19


damage, infertile eggs, or unknown cause). Females incubated 72 nests, 41 were incubated by males, and 41 were incubated by an unknown sex plover. I could not determine the sex of plovers I was unable to catch; it is possible this disparity in the distribution of sex is partly due to differing capture probabilities by sex. Additionally, nests with an unknown sex adult had lower apparent survival (51.2%) likely because in many cases the nest failed before I was able to catch the bird and collect a tissue sample for molecular sexing. Mean nest initiation date (Fig. 4) was similar for both sexes (males, mean = 24 May 1.9 days SE; females, mean = 25 May 1.4 days SE). The earliest and latest days of active nesting across the three study year were 24 April and 27 July.
The estimated survival to hatch from the simple constant survival model Sq was 56.1% (95% Cl: 46.3% to 64.8%), obtained by raising the DSR to the power of 29 for the number of incubation days in this species. See table 2 for full summary of the model selection results.
The variance in Mountain Plover nest DSR was best predicted by a year effect with a quadratic time trend and an interaction effect of rain and maximum temperature. Models with quadratic time trends performed better than models with linear time trends. Models with the one-day precipitation lag effect were less parsimonious than precipitation models with no lag effect. Also, there was not strong evidence to suggest that tagging birds had an influence on their nest survival, as these models were among the least supported by the data. However, Fishers exact test for nest survival of tagged vs. untagged birds produced an odds ratio estimate of 2.48 (95% CT 1.05 to 5.85, p = 0.0241), indicating that nests tended by untagged birds were about twice as likely to have hatched than those tended by tagged birds (Table 3).
20


In all 3 years, DSR gradually peaked in the mid-season and dropped considerably for late-season nests (Fig. 5). I used the quadratic time trend to estimate survival to hatch probabilities for five nesting periods throughout a breeding season based on earliest, latest,
1st, 2nd and 3rd quartiles of nest initiation (Table 4).
Discussion
This study represents an effort to model Mountain Plover nest survival in agrarian setting. All 154 nests included in the study occurred on privately owned farmland and were marked to prevent loss through tilling or other farming practices. The apparent nest survival of 68.8% is high compared to nest survival in other studies of plovers which ranges from 27.2% to 70% (Graul 1975, Dinsmore et al. 2002, Mettenbrink et al. 2006, Dreitz and Knopf 2007, Dreitz et al. 2012). In a previous study comparing plovers nesting in agrarian vs. grassland settings, Dreitz and Knopf (2007) found that overall nest loss was similar although the mechanism differed. On crop fields, mechanical farming practices were the leading cause of loss whereas in grasslands nest loss occurred primarily due to predation. The results of my study similarly suggest that predation on crop fields is low.
There was some evidence for effects of rain and temperature on nest survival. Individually or in additive models these variables were not strongly supported but the top model included an interaction term with rain and precipitation. The most parsimonious nest survival models of both Dinsmore et al. (2002) and Dreitz et al. (2012) included a precipitation variable although they did not consider interactions of precipitation and temperature. Precipitation events across the western Great Plains can be very localized, severe, and brief. Those previous studies included six to seven seasons of data; it may be that the temporal and geographic span of my study did not include weather events severe enough
21


to influence nest survival and overcome the modelling penalty for additional parameters. For example, Dreitz et al. (2012) reported having drought periods of 10 days or more in each season and droughts lasting up to 45 days. In my study the longest drought period was 19 days and there were no drought periods exceeding 10 days in 2015. Furthermore, if rainy weather events increase predator activity as had been suggested (Dinsmore et al. 2002) then this effect may have been depressed on crop fields where nest loss to predation is lower overall.
Although the modelling results showed no evidence of an effect on daily nest survival probabilities from attaching tracking tags to incubating adults, a direct comparison of nest fates between tagged and untagged birds suggested otherwise. Although the sample size of tagged birds was comparatively low and the 95% confidence interval of the odds ratio nearly contained 1 (no association), these results suggest further inquiry is needed in studies where tags are applied to nesting birds. Although many of these birds were caught at the nest twice or even three times, all but one of nine nest abandonment incidents occurred after just one trapping attempt. These observations suggest that if there is an effect of tagging, it may stem from the addition of the tag itself rather than the stress associated with repeated capturing and handling of the birds.
There was variation in nest survival from year to year; the highest rates were observed in 2013 and they decreased each subsequent year. This variation was not well-captured by the weather covariates included here, as only the interaction of precipitation and daily maximum temperatures improved models beyond a simple quadratic time trend and year effect. These results suggest that although there was evidence of a sharp drop in nest survival in the late season, it cannot be attributed entirely to weather variables that have
22


explained variance in plover nest survival in other locations and times. This late-season drop in nest survival may be attributed to differing prey availability over the season or to renesting efforts of inexperienced breeders.
For plovers nesting in crop fields, weather can shape their environment not only through direct impacts such as hail or flooding of lowlands, but also by influencing how and when farmers work their fields. Rain events delay tilling and planting operations as farmers wait for their fields to dry. In 2014 and 2015 tilling was often delayed until after plover nesting periods were complete. This study did not attempt to quantify or model farming practices or any associations it may have with climatic stochasticity, but expected increases in both drought and severe precipitation events will no doubt impact farming practices. Mountain plovers have shown adaptability in nesting on the crop fields that have supplanted much of their native breeding range; continued study will be needed to assess plover adaptability to predicted climate change.
23


TABLE 1. Candidate models and notations
Candidate models Notation
Constant survival S(.)
Year + linear trend S(year+T)
Year + quadratic trend S (year+TT)
Year + linear trend + prep S(year+T+prcp)
Year + quadratic trend + prep S(year+TT+prcp)
Year + linear trend + prep lag effect S(year+T+lag)
Year + quadratic trend + prep lag effect S(year+TT+lag)
Year + linear trend + nest age S(year+T+age)
Year + quadratic trend + nest age S(year+TT+age)
Year + linear trend + max temp S (year+T+maxtemp)
Year + quadratic trend + max temp S (year+TT+maxtemp)
Year + linear trend + tagged S(year+T+tag)
Year + quadratic trend + tagged S(year+TT+tag)
Year + quadratic trend + max temp + prep S (year+TT+maxtemp+prcp)
Year + quadratic trend + max temp*prcp S(year+TT+maxtemp*prcp)
Global model S(year+TT+rain+maxtemp+age+tag+maxtemp*rain)
24


CO
Q.
<
o
CO
Q.
<
FIG. 4. Nest Initiation by date
25
Jun26


TABLE 2. Summary of model selection results
Model AAICc Wi K Deviance
S(year+TT+ maxtemp*prcp) 0 0.158 6 338.42
S(year+TT) 0.415 0.129 5 340.85
S(year+TT+prcp) 1.060 0.0931 6 339.48
S(year+TT+maxtemp+prcp) 1.107 0.0910 7 337.52
S(year+T) 1.272 0.0837 4 343.71
S (year+T+maxtemp) 1.859 0.0624 5 342.29
S(year+TT+maxtemp) 1.972 0.0590 6 340.39
S(year+TT+tag) 2.016 0.0577 6 340.44
S(year+TT+prcp lag effect) 2.271 0.0508 6 340.69
S(year+TT+age) 2.378 0.0482 6 340.80
S(year+T+prcp) 2.761 0.0398 5 343.19
S(year+T+tag) 2.857 0.0379 5 343.29
S(year+T+age) 3.097 0.0336 5 343.53
S(year+T+ prep lag effect) 3.239 0.0313 5 343.67
S(year+TT+prcp+maxtemp+age+tag+maxtemp*prcp) 4.903 0.0316 10 335.27
S(.) 5.280 0.0112 1 353.74
TABLE 3. Contingency table for nest outcomes of tagged and untagged birds
Failed Hatched
Tagged 17 19
Untagged 28 78
26


TABLE 4. Probability of surviving to hatch for five nesting periods throughout a season
Nesting period Probability of hatch 95% Cl
Earliest .618 .264 to .816
1st quartile (early) .638 .442 to .772
2nd quartile (mid) .630 .491 to .739
3rd quartile (late) .573 .447 to .680
Latest .196 .026 to .469
FIG. 5. Estimate of nest Daily Survival Rate throughout the season with 95% Cl
27


REFERENCES
AebischerN., P. A. Robertson, andR. E. Kenward. 1993. Compositional analysis of habitat use from animal radio-tracking data. Ecology 74:1313-1325.
Ainley, D. G., and R. P. Schlatter. 1972. Chick raising ability in Adelie Penguins. The Auk 89:559-566.
Akaike, H. 1973. Information theory as an extension of the maximum likelihood principle. Pages 267-281 in B. N. Petrov and F. Csaki, editors. Second International Symposium on Information Theory. Akademiai Kiado, Budapest, Hungary.
Augustine, D.J., 2011. Habitat selection by mountain plovers in shortgrass steppe. Journal of Wildlife Management 75:297-304.
Augustine, D.J., and J. D. Demer. 2012. Disturbance regimes and mountain plover habitat in shortgrass steppeLarge herbivore grazing does not substitute for prairie dog grazing or fire: Journal of Wildlife Management 76:721-728.
Batschelet E. 1981. Circular statistics in biology. Academic Press, London, England.
Beyer, H. L. 2012. Geospatial Modelling Environment. Version 0.7.3.0. http://www.spatialecology.com/gme
Bly, B. L., L. Snyder, and T. VerCauteren. 2008. Migration chronology, nesting ecology, and breeding distribution of mountain plover (Charadrius montanus) in Nebraska. Nebraska Bird Review 76:120-128.
Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: a practical information-theoretic approach. Springer, New York, New York, USA.
Colorado Parks and Wildlife. 2015. State Wildlife Action Plan: A Strategy for Conserving Wildlife in Colorado. Denver, Colorado, USA.
Dinsmore, S. J., G. C. White and F. L. Knopf. 2002. Advanced techniques for modeling avian nest survival. Ecology 83:3476-3488.
Dinsmore, S.J., G. C. White, and F. L. Knopf, 2005. Mountain Plover population responses to black-tailed prairie dogs in Montana. The Journal of Wildlife Management 69:1546-1553.
Dreitz, V. J., and F. L. Knopf. 2007. Mountain plovers and the politics of research on private lands: BioScience 57:681-687.
28


Dreitz, V. J., R. Y. Conrey, and S. K. Skagen. 2012. Drought and cooler temperatures are associated with higher nest survival in mountain plovers. Avian Conservation and Ecology 7(1):6.
Dreitz, V.J., M. B. Wunder, and F. L. Knopf. 2005. Movements and home ranges of mountain plovers raising broods in three Colorado landscapes. The Wilson Bulletin 117:128-132.
Efron, B. 1979. Bootstrap methods: another look at the jackknife. Annals of Statistics 7: 1-26.
Environmental Systems Research Institute (ESRI). 2012. ArcGIS. Version 10.1. ESRI, Redlands, California, USA.
Graul, W. D., 1975. Breeding biology of the Mountain Plover. The Wilson Bulletin 87:6-31.
Johnson, D. H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61:65-71.
Knopf, F. L. 1994. Avian assemblages on altered grasslands. Studies in Avian Biology 15:247-257.
Knopf, F. L., andB. J. Miller. 1994. Charadrius montanusMontane, grassland, or bare-ground plover? Auk 111: 5 04-5 06.
Knopf, F. L. and J. R. Rupert. 1996. Reproduction and movements of Mountain Plovers breeding in Colorado. The Wilson Bulletin 108:28-35.
Knopf, F. L. and J. R. Rupert. 1999. Use of cultivated fields by breeding mountain plovers in Colorado. Studies in Avian Biology 19:81-86.
Knopf, F. L., and M. B. Wunder. 2006. Mountain plover (Charadrius montanus). The birds of North America. Number 211.
Levin, S. A. 1992. The Problem of pattern and scale in ecology. Ecology 73: 1943-1967.
Manly, B. F. J., L. L. McDonald, D. L. Thomas, T. L. McDonald, and W. P. Erickson. 2002. Resource selection by animals: statistical design and analysis for field studies. Second Edition. Springer Science+Business Media, Dordrecht, the Netherlands.
Martin, T. E. 1987. Food as a Limit on Breeding Birds: A Life-History Perspective. Annual Review of Ecology and Systematics 18:453-487.
Matthews, J. H. 2008. Anthropogenic climate change in the Playa Lakes Joint Venture Region: understanding impacts, discerning trends, and developing responses. World Wildlife Fund, Corvallis, Oregon, USA.
29


Mettenbrink, C. W., V. J. Dreitz, and F. L. Knopf. 2006. Nest success of mountain plovers relative to anthropogenic edges in eastern Colorado. The Southwestern Naturalist 51:191-196.
National Oceanic and Atmospheric Administration. 2016. National Centers for
Environmental Information. http://www.ncdc.noaa.gOv/cdo-web/datasets#GHCND Accessed 04 June 2016.
Peek, J. M. 1986. A review of wildlife management. Prentice-Hall, Englewood Cliffs, USA.
R Core Team. 2015. R: a language and environment for statistical computing. Version 3.2.2. R Foundation for Statistical Computing, Vienna, Austria, http://www.r-proiect.org/
Samson, F.B., F. L. Knopf, and W. R. Ostlie. 2004. Great Plains ecosystemsPast, present, and future: Wildlife Society Bulletin 32:6-15.
Sauer, J. R., J. E. Hines, J. E. Fallon, K. L. Pardieck, D. J. Ziolkowski Jr., and W. A. Link, 2014. The North American Breeding Bird Survey, Results and Analysis 1966 2012. Version 01.30.2015 USGS Patuxent Wildlife Research Center. Laurel, Maryland. Accessed 02 Nov 2016.
Schneider, R., K. Stoner, G. Steinaure, M. Panella, and M. Humpert. 2011. The Nebraska Natural Legacy Project: State Wildlife Action Plan. Second edition. The Nebraska Game and Parks Commission, Lincoln, USA.
Seaman, D. E. and R. A. Powell. 1996. An evaluation of the accuracy of kernel density estimators for home range analysis. Ecology 77:2075-2085.
Shackford, J.S., D. M. Leslie, and W. D. Harden. 1999. Range-wide use of cultivated fields by mountain plovers during the breeding season. Journal of Field Ornithology 70:114-120.
Skrade, P.D.B. and S. J. Dinsmore. 2012. Incubation patterns of a shorebird with rapid multiple clutches, the Mountain Plover (Charadrius montanus). Canadian Journal of Zoology 90:257-266.
Swihart, R. K. and N. A. Slade. 1986. The importance of statistical power when testing for independence in animal movements. Ecology 67:255-258.
Tipton, H. C., V. J. Dreitz, and P. F. Doherty, Jr., 2008. Occupancy of mountain plover and burrowing owl in Colorado. Journal of Wildlife Management 72:1001-1006.
Vickery, P.D. and J. R. Herkert. 2001. Recent advances in grassland bird research: where do we go from here? Auk 118:11-15.
30


White, G. C, and K. P. Burnham. 1999. Program MARK: survival estimation from populations of marked animals. Bird Study 46:120-138.
White, R. P., S. Murray, and M. Rohweder. 2000. Grassland Ecosystems. World Resources Institute, Washington DC, USA.
Whittingham, M. J., S. M. Percival, and A. F. Brown. 2000. Time budgets and foraging of breeding golden plover (Pluvialis apricaria). Journal of Applied Ecology 37:632-646.
31


APPENDIX
Log-ratio differences of used and available habitats for each plover; a positive value indicates habitat selection of the row over the column while a negative value indicates habitat avoidance. Ranks are from 1 (most preferred) to 4 (least preferred).
Bird ID:XWRO Fallow Growing Crop Grassland Other Total Rank
Fallow - 2.584 11.820 10.622 25.026 1
Growing Crop -2.584 - 9.235 8.038 14.689 2
Grassland -11.820 -9.235 - -1.197 -22.252 4
Other -10.622 -8.038 1.197 -17.463 3
Bird ID: ORIX Fallow Growing Crop Grassland Other Total Rank
Fallow - 12.985 10.302 7.889 31.176 1
Growing Crop -12.985 - -2.684 -5.096 -20.765 4
Grassland -10.302 -2.684 - -2.412 -10.030 3
Other -7.889 5.096 2.412 -0.381 2
Bird ID: WWIX Fallow Growing Crop Grassland Other Total Rank
Fallow - 1.819 11.506 11.440 24.764 1
Growing Crop -1.819 - 9.687 9.621 17.490 2
Grassland -11.506 -9.687 - -0.066 -21.259 4
Other -11.440 -9.621 0.066 - -20.995 3
32


Bird ID: IXIO Fallow Growing Crop Grassland Other Total Rank
Fallow - 1.644 10.012 7.200 18.855 1
Growing Crop -1.644 - 8.368 5.556 12.281 2
Grassland -10.012 -8.368 - -2.812 -21.191 4
Other -7.200 -5.556 2.812 -9.944 3
Bird ID: RIRX Fallow Growing Crop Grassland Other Total Rank
Fallow - 2.478 10.100 8.454 21.028 1
Growing Crop -2.478 - 7.618 5.976 11.115 2
Grassland -10.100 -7.618 - -1.642 -19.356 4
Other -8.454 -5.976 1.642 -12.787 3
Bird ID: XORW Fallow Growing Crop Grassland Other Total Rank
Fallow - 13.283 12.302 8.026 33.611 1
Growing Crop -13.283 - -0.981 -5.257 -19.520 4
Grassland -12.302 -0.981 - -4.276 -15.597 3
Other -8.026 5.257 4.276 - 1.507 2
Bird ID: XOIR Fallow Growing Crop Grassland Other Total Rank
Fallow - 13.150 12.374 8.504 34.028 1
Growing Crop -13.150 - -0.776 -4.646 -18.572 4
Grassland -12.374 0.776 - -3.871 -15.469 3
Other -8.504 4.646 3.871 - 0.013 2
33


Bird ID: XRRW Fallow Growing Crop Grassland Other Total Rank
Fallow - 13.375 11.808 9.972 35.155 1
Growing Crop -13.375 - -1.568 -3.403 -18.346 4
Grassland -11.808 1.568 - -1.836 -12.076 3
Other -9.972 3.403 1.836 -4.733 2
Bird ID: XWIO Fallow Growing Crop Grassland Other Total Rank
Fallow - 12.697 0.413 8.688 21.797 1
Growing Crop -12.697 - -12.284 -4.009 -28.989 4
Grassland -0.413 12.284 - 8.275 20.146 2
Other -8.688 4.009 -8.275 - -12.955 3
34


Full Text

PAGE 1

MOUNTAIN PLOVER BREEDING ECOLOGY: H OME RANGE SIZE, HABITAT USE, AND NEST SURVIVAL IN AN AGRICULTURAL LANDSCAPE by COLIN ALLAN WOOLLEY B.A., Prescott College, 2006 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment o f the requirements for the degree of Master of Science Biology Program 2016

PAGE 2

ii This thesis for the Master of Science degree by Colin Allan Woolley h as been approved for the Department of Integrative Biology b y Michael Wunder, Chair Rafael Moreno Jennifer Blakesley Date: 17 December 2016

PAGE 3

iii Woolley, Colin Allan (M.S. Biology Program ) Mountain Plover Breeding Ecology: Home range Size, Habitat Use, and Nest Survival in an Agricultural Landscape Thesis directed by Associate Professor Michael B. Wunder ABSTRACT Mountain Plovers ( Charadrius montanus ) a species of conservation co ncern throughout their breeding range, frequently nest on agricultural crop fields that have replaced native short grass prairie habitat. To understand how plovers nesting on crop field s use habitats during nesting, I estimated home range size and habitat use during the nest incubation period using GPS loggers I also sampled available invertebrate prey species from across a range of habitat types to quantify potentially available biomass and calori c density Average h ome range size during nest incubation was 155.4 54.9 (SE) ha Plovers foraged almost exclusively on fallow crop fields and avoided fields with growing crop s I found no significant difference in invertebrate prey biomass or calor ic density between grassl and and agricultural fields. I also modelled daily nest survival of plovers to evaluate the influence of various weather variables and individual traits. Nest survival averaged 56.1% over the three year study period, and was best explained by a model including a quadratic time trend, a year effect, and an interaction effect of precipitation and maximum daily temperature. The form and content of this abstract are approved. I recommend its publication. Approved: Michael B. Wunder

PAGE 4

iv ACKNOWLEDGEMENTS Funding for this study was provide d by University of Colorado Denver, Bird Conservancy of the Rockies, Nebraska Prairie Partners, Lois Webster Fund of the Audubon Society of Greater Denver, and Colorado Field Ornithologists. I would like to thank the landowners of Kimball County, NE and We ld County, CO for granting access to their la nd. I especially thank Larry Snyder local landowner and employee of Bird Conservancy of the Rockies, for his assistance in coordinating with other landowners and providing logistical support. Craig Stricker assisted in use of the bomb calorimeter. Sara Oyler McCance processed feather samples for molecular sexing. I also thank Clay Edmonson, Tasha Blecha, Molly Elderbrook, Kelly Surgalski, Lacey Clarke and Eric Matechak for their work in the field. All fieldw ork followed handling protocols approved by University of Colorado Institutional Animal Care and Use Committee protocol #102244.

PAGE 5

v TABLE OF CONTENTS CHAPTER I. HOME RANGE SIZE AND HABITAT USE OF MOUNTAIN PLOVERS DURING NEST INCUBATION 1 3 7 8 11 II. NEST SURVIVAL Introduction Materials and Methods Results 19 Discussion 21 Figures and Tables 24 R E FERENCES A PPENDIX 32

PAGE 6

1 CHAPTER I HOME RANGE SIZE AND HABITAT USE OF MOUNTAIN PLOVERS DURING NEST INCUBATION Introduction Over the last 150 years, grasslands of the Great Plains have undergone many changes, including conversion to croplands and loss of native grazers (e.g. bison, prairie dogs). During that period, o ver 80% of North American grasslands have been altered by agriculture, energy production and urbanization (White et al. 2000). This has, in part, led to the decline of numerous grassland endemic bird population s (Knopf 1994; Vickery & Herkert 2001) The Mountain Plover ( Charadrius montanus ) is a ground nesting migratory s horebird that breeds primarily in the short grass prairie of the western Great Plains. It is a species of conservation concern in our study area (Schneider et al. 2011 Colorado Parks and Wildlife 2015 ) and throughout its range Mountain Plovers are a bare ground nesting specialist (Graul 1975, Knopf and Miller 1994, Knopf and Wunder 2006, Augustine and Derner 2012), drawn to areas of localized disturbance from grazing, fire, or agricultural tillage. Conversion of short grass prairie for crop production has greatly reduced availability of native breeding habitat (Samson et al. 2004) However, Mountain Plover readily nest in privately owned agricultural crop fields, particularly in fields that are fallow or with low growing crop s (Shackford et al. 1999; Knopf & Rupert 1999 ; Bly et al. 2008 ) Conservation and management of at risk populations require an understanding of the spatial context in which they forage, reproduce, and meet their needs for survival (Levin 1992). Mountain Plovers nesting on agricultural

PAGE 7

2 crop fields are presented with a spatial context that differs from the native short grass prairie in which they evolved. Mountain Plover home range size and habitat use patterns have been studied during the post hatch breeding season when adult plovers are moving around the landscape with precocial broods (see Knopf and Rupert 1996, Dreitz et al. 2005) but nothing is known about these variables during the incubation period. This is a potentially substantial omission because of the unusual breeding system of plovers, which can be described as split clutch Female plovers lay (typically) three eggs in one nest scrape for the male to independently incubate, before laying another three eggs in a different nest scrape for her to incubate. There is no food provisioning during incubation; males and females independently forage and tend the nests through fledging of the broods. Therefore, the nesting adult does not benefit from mate feeding and must balance incubation duties with its own foraging needs (Martin 1987). Habitat use and home range size during incubation will likely differ from that during brood rearing as the adults are not tethered to a mobile brood but rather to a stationary nest. For species with precocial young, the incubation period represents a time of high energetic investment when adults are at increased risk of starvation and mortality (Martin 1987). Daytime high temperatures in our study site typically reach upwards of 30 C and adults must shade the eggs to prevent overheating. However, when temperatures are moderate, Mountain Plovers may leave the nest unattended to seek foraging opportunitie s. Long distance commutes between distinct nesting and foraging grounds have been found in other species of Charadriidae (Whittingham et al. 2000) though it is unclear if this is also the case for Mountain Plovers.

PAGE 8

3 Mountain Plover diet s consist primarily of terrestrial invertebrates (Knopf and Wunder 2006), including beetles ( Coleoptera ), grasshoppers ( Orthoptera ), and ants ( Hymenoptera ) as the most common prey items. Foraging p references for one habitat type over another may be driven by differing food re source availability (Peek 1986, Manly et al. 2002) Our objectives were (1) to estima te plover home range size during the incubation period; (2) to quantify foraging habitat preference in a variable landscape; and (3) to evaluate any habitat preferences i n relation to biomass and calor i c content of invertebrate prey sources in each habitat type Understanding how adult plovers use agricultural landscapes during this vulnerable incubation period can inform management strategies for this species of conservat ion concern. Materials and Methods Study Area Our study occurred mostly on privately owned dryland agriculture in Kimball County, Nebraska and Weld County, Colorado (Figure 1) The dominant crops were winter wheat ( Triticum aestivum ) and millet ( Panicum miliaceum ). Interspersed a mong the agricultural fields were small patches of rangeland characterized by two native short grass prairie species; blue grama ( Bouteloua gracilis ) and buffalograss ( Bouteloua dactyloides ). Additionally, some l and was enrolled i n the Conservation Reserve Program (CRP); a federal program administered by the Farm Service Agency in which land is taken out of agricultural production and planted to both native and non native grasses to enhance wildlife habitat and various ecosystem se rvices. Field Methods I gathered data on plover home range s and habitat use during the 2014 and 2015 breeding seasons. Plover movements during the incubation period have traditionally been difficult to document using radio telemetry; for aging bouts off t he nest

PAGE 9

4 may be short and observer bias is introduced when lovers close to their nest are more easily located while a plover foraging far from the nest may outrange the telemetry reception and be difficult to locate until it returns to the nest area This results in a bias towards over representation of observations close to the nest when using radio telemetry I deployed GPS tags (Pinpoint 50, Lotek, Newmarket, Ontario, Canada) on a programmed schedule to avoid this observer bias. This study was conducted alongside a nest marking program administered by Bird C onservancy of the Rockies in which nests are located and marked with wooden stakes to prevent accidental tillage. After nests were located, the incubating adult was caught using a walk in box trap placed over the nest. When nes ts were discovered I estimated nest age using egg floatation (Dinsmore et al. 2002) and assumed a 29 day incubation period (Knopf and Wunder 2006 ). To re duce risk of nest abandonment, I tagged p lovers that were at lea st seven days into incubation The tags collect 50 locational points on a programmable schedule, after which the bird must be re caught to retr ieve the data. When possible, I redeployed tags a second time on each incubating plover fo r a total of 100 points per bird. I was unable to retrieve tags from plovers if their nest fail ed during the deployment period as they left the area after nest loss Because Skrade and Dinsmore (2012) found that incubating Mountain Plovers are off the nes t more frequently during crepuscular and nocturnal times I programmed the tags to take proportionately more locational fixes during these times. Data Analysis Spatial and temporal autocorrelation are problematic when estimating home ranges because most models assume that each successive locational fix is independent of previous fixes (Swihart and Slade 1986). To reduce the problems of serial autocorrelation, I randomized the fix schedule according to the fo llowing ruleset (all times in MST) : 4 fixes

PAGE 10

5 take n at random from 0500 to 07 59 ; 1 from 0800 to 17 59; 4 from 1800 to 2059; 1 from 21 00 to 0459; all with at least 20 min between each fix. Any autocorrelation still present may lead to underestimated home range sizes due to biased bandwidth selection (Swihar t and Slade 1986). Temporal autocorrelation tends to decrease distances between fixes, leading to a smaller kernel bandwidth estimate and smaller associated home range estimates. The GPS tags record metadata for each locational fix, including the number of satellites used in acquiring the fix. Prior in situ field testing suggested that four or more satellites were necessary for ac curacy under 10 meters. Thus, I censored fixes with fewer than four satellites (10.0% of fixes). I also censored fixes within 10 meters of the nest (64.2% of fixes) as these fixes potentially represented a bird attending its nest and not on a foraging bout away from the nest I estimated home range size during the incubation and brood rearing periods using a fixed kernel method with bandwidth selection through least squares cross validation (Seaman and Powell 1996). This method is non parametric and has been shown to have less bias than other bandwidth estimators when used when used with non normally distributed data and small sample si zes (Seaman and Powell 1996). I defined home range size using the 95% isopleth within the kernel density estimate. Home ranges were estimated using ArcGIS 10.1 (ESRI 2012) with Geospatial Modelling Environment (Beyer 201 2) and program R (R Core Team 2015). I characterized habitat surrounding each nest on the following nominal scale : fallow (including stubble fields), growing crop (including cover crops), grassland (including grazed rangeland or land enrolled in the CRP) or other (consisting of residential houses, yards, or unknown habitat we could not visually confirm). I estimated habitat preference by comparing

PAGE 11

6 differences in proportion of habitat types used to that of habitat types available This comprises design III habitat selection (Manly et al. 2002, Johnson 1980) as the proportions of used and available habitat types are estimated separately for each individual plover. I used the log ratio differences of used an d available habitat proportions to determine preferen ce (Aebischer et al. 1993) A positive value indicates habitat selection while a negative value indicates habitat avoidance. Use was defined by the observed occurrence within each habitat type. Typically, availability is defined by the proportional occurre nce of habitat types within some defined home range (Aebischer et al. 1993). However, movement patterns for an incubating bird differ from typical use within a home range in that all foraging bouts begin and end at the same location (the nest). In this cas e, areas closer to the nest will show greater use simply due to their proximity to the nest and not necessarily due to a habitat pre ference. To account for this, I used a bootstrapping approach (Efron 1979) to define availability for each bird by generatin g 10,000 random sets of points (where n = no. of observed points for that bird) and summing habitat distributions across them. A locational fix can be defined as a distance and direction from the nest. I generated the random ly chosen availability points by separately pooling all observed distances and directions and sampling with replacement from these pooled data. This results in a model of availability that does not assume a uniform distribution across a home range. To evaluate any directional p reference in off nest move ments I used a Rayleigh Test of Uniformity ( Batschelet 1981). This test compares observed directional data with a null hypothesis of directional uniformity. These analyses were performed using program R (R Core Team 2015). To esti mate available food resou rces across habitat types, I sampled the terrestrial invertebrate prey species using pitfall traps. I randomly selected three sites on rangeland and

PAGE 12

7 three sites on fallow fields. I placed a 4 by 4 array of pitfall traps spaced 1 me ter apart at each site. Pitfall traps were 16 oz Solo cups (95mm diameter, 115mm depth) placed into the ground such that the rim of the cup was flush with the ground surface. Sampling efforts consisted of a 24 hour period, with t hree sampling efforts per site. Each sampling effort also included a 200m walking transect using a bug net to sweep for orthopterans that may not be captured in the pitfall traps. I dried the invertebrate samples for 24h r at 6 0 C and used a bomb calorimeter (IKA C2000 Basic) to measure calor ic density for each sampling effort. Results Home Ranges I attached GPS tags to 15 plovers across 2014 and 2015 I recovered GPS tags from nine plovers, five of which were tagged a 2 nd time for additional points. Home range analyses were based on a mean of 21.2 4.4 (SE) locational fixes per bird. Mean home range size was 155.4 54.9 (SE) ha (see Fig. 1 for examples) Mean distance from the nest for locational fixes of all birds combined was 385.4 26.4 (SE) mete rs. The Rayleigh Test of Uniformity found no significant directional preference to plover movements (p=0.238). Habitat Use I estimated the log ratio differences of used and available habitat proportions for each of the nine tagged plovers (See Appendix) This method allows for habitats to be ranked from most to least preferred. For all individuals, fallow fields ranked as the most preferred habitat type Overall, fallow fields accounted for 92.7% of a ll observed locational fixes. I used the observed and expected proportions for each plover to create bootstrapped 95% confidence intervals of plover habitat use (Fig. 2 ). Non overlapping 95% confidence intervals of Observed and Available for fallow and growing crop habitats further suggest us e and avoidance o f cover types respectively. Plovers also used crop fields nearly

PAGE 13

8 exclusively for nest site location throughout the duration of our study, with 168 of 170 found nests on agricultural crop fields. I nvertebrate Prey Densities and Caloric Content The estimate for mean biomass of invertebrates captured per 24h r sampling period was slightly higher on rangeland ( 1.411 g/day .353 SE n=6 ) than on agricultural fields (0.873 g/day .162 SE n=8 ) though the difference was not statistically significant (We lch two sample t test; t 7.1 = 1.39, p= .207). The mean calor ic density of invertebrates captured on rangeland ( 5467 cal/g 99.2 SE ) was similar to a gric ul tural fields (5502 cal/g 81.8 SE ) ( Welch two sample t test; t 10.6 = .28, p=.788). Coleoptera were by far the most commonly captured invertebrates on both habitat types and overall proportions of taxa captured were also similar with slightly more Hymenoptera and fewer Orthoptera on rangeland ( Fig. 3 ) Orthoptera are a common prey item throughout the pl scarce at the study site. No Orthopterans were captured in sweep netting during the walking transects associated with each invertebrate sampling effort. Discussion Home Range Size The point estimate for mean home range size during incubation was very similar to a previously published home range size (131.6 ha ; Dreitz et al. 2005 ) based on the brood rearing period for Mountain Plovers that nested on fallow crop fields This suggests that Mountain Plover home range size remains similar throughout the nesting period, regardless of whether the adult is tending to eggs or a mobile brood. It is possible that home ranges during incubation and brood rearing may not overlap, which would result in a larger home ra nge size overall. I had insufficient sample size to detect any differences in

PAGE 14

9 home range size by sex. It is possible that home range size may differ by sex, because females may have to overcome a greater energy deficit associated with egg laying. Skrade an d Dinsmore (2012), however, found no difference in incubation patterns by sex and given that there is no sexual size dimorphism it is just as plausible that home range sizes between the sexes are similar. H abitat Use and Invertebrate Prey Densities It is clear that fallow fields play an important role for Mountain Plovers nesting in ag ricultural settings All nine tagged plovers used fallow fields nearly exclusively and avoided areas with growing crop s Patches of native grassland, while not common i n the study area, showed no preferential use. The biomass, calor ic content, and proportional distribution of taxa for invertebrate prey species on rangeland and crop fields were similar. Farmers in the study area typically plant crops in alternating strips that are 100 to 120 meters wide, wi th fallow strips in between. These results suggest that these alternating fallow strips provide sufficient foraging oppor tunities for nesting plovers. I cannot rule out the possibility of an energetic cost associated wit h foraging on fallow fields as compared to native short grass prairie. Weight loss during incubation is fairly common in species with precocial young (Martin 1987). There was some evidence of weight loss over the incubation period ( CAW unpublished data) bu quantified for M ountain P lovers nesting in any habitat M anagement Implications The population level implications of plovers nesting on crop fields are still not fully understood. Dreitz and Knopf (2007) found that nest survival on agricultural fields was similar to nest survival on native rangeland, though the cause of nest mortality differed. Nests on crops were destroyed primarily by mechanical farming operations (e.g. tilling, planti ng) while rangeland nests were destroyed by predators.

PAGE 15

10 Agricultural fields and grassland appear to be interchangeable in terms of nest survival, home range size, and available food resources. These results suggest that the fallow fields are supporting both the nest site and foraging requirements of incubating plovers and that plovers are not using the interspersed patches of grassland for foraging opportuni ties, nor for nest site location. This suggests that an agrarian landscape with interspersed fallow areas can support the full breeding cycle of Mountain Plovers and a lack of native grassland patche s will not preclude plover use

PAGE 16

11 FIG. 1. Ninety five percent fixed kernel untilization distributions for two different incubating Mountain Plovers (blue shading) The plover locations are shown as blue circles and the nest locations are shown as black crosses.

PAGE 17

12 FIG. 2. Bootstrapped ninety five percent confidence intervals of plover habitat observed use and availability by habitat type.

PAGE 18

13 FIG. 3. Proportional distribution of prey taxa biomass by habitat type.

PAGE 19

14 CHAPTER II NEST SURVIVAL Introduction Short grass prairies of the western Great Plains are among the most heavily altered ecosystems in North America (Samson et al. 2004). These changes are characterized by conversion of grassland to agricultural production and the loss of native grazers such as bison ( Bison bison ) and prairie dogs ( Cynomis spp. ). Over the last century, avifauna of the Great Plains grasslands has experienced population declines greater than those found in any other North American biome (Knopf 1994). Mountain Plovers ( Charadriu s montanus ) in particular have experienced population declines averaging an estimated 3.15% per year from 1966 to 2012, according to Breeding Bir d Survey data (Sauer et al. 2014 ). Mountai n Plover is a mid sized, migratory shorebird species that breeds in arid grasslands of the western Great Plains and winters from southern California to Texas and into northern Mexico (Knopf and Wunder 2006). While conversion of native short grass prairie to ain Plovers are one of just a few short grass prairie specialists that will nest on agricultural fields (Shackford et al. 1999; Knopf & Rupert 1999). Mountain Plovers use a split clutch mating system, in which the female lays 3 eggs in a nest that will be tended by the male before laying 3 more eggs in a separate nest that she alone will tend (Knopf and Wunder 2006). Mountain Plovers are arid, bare ground nesting specialists (Knopf and Wunder 2006) that were historically associated with bare ground created by grazing of native herbivores. Modern farming, particularly the practice of leaving some fields fallow during the spring and summer, has replaced the native grazers in providing this bare

PAGE 20

15 Climate mode temperatures and lower precipitation levels overall though with more frequent and severe precipitation events such as hail (Matthews 2008). Nest survival of Mountain Plover has been pr eviously studied across their range (see Graul 1975, Knopf and Rupert 1996, Dinsmore et al. 2002, Mettenbrink et al. 2006, Dreitz and Knopf 2007, Dreitz et al. 2012) with population level seasonal survival estimates ranging from 26% (Knopf and Rupert 1996) up to 65% (Graul 1975). Several of these studies have considered the influence of daily weath er measures on nest survival. Dinsmore et al. (2002) found that the daily survival rate (DSR ) of nests decreased following precipitation events and that there was no evidence for an effect of m aximum daily temperatures on DSR Additionally, male tended nests had higher survival probabilities than female tended nests. Dreitz et al (2012) found that plover DSR increased aximum daily temperatures. Whereas previous st udies have included cropland nesting plovers in their samples, this study is the first to e xclusively consider cropland nesting plovers. My objectives were to estimate daily nest survival of cropland nesting plovers to compare with previous studies of most ly grassland nesting plovers. I compared individual level traits such as sex of the incubating adult with daily climate variables and seaso nal time trends to determine which covariates best explain whether a nest fails or is successful. U nderstanding the impacts of the modern agricultural landscape and a warming and increasingly volatile climate on plover nest survival and long term populatio n dynamics will aid future management decisions regarding this increasingly rare species.

PAGE 21

16 Materials and Methods Study Area. My study area consisted of mostly private agricultural land in Kim ball County, Nebraska and northern Weld County, Colorado. Dryla nd agriculture dominates the lan dscape in this area (~80%), where winter wheat ( Triticum aestivum ) and millet ( Panicum miliaceum ) are the primary crops. Small patches of grazed rangeland and grassland enrolled in the Conservation Reserve Program (CRP) are interspersed throughout the agricultural matrix. The rangeland most closely resembles native shortgrass prairie habitat, consisting o f blue grama ( Bouteloua gracili s ) and buffalograss ( Bouteloua dactyloides ). CRP grasslands in the study area have been planted to taller mixed grass species such as western wheat grass ( Pascopyrum smithii ) and crested wheatgrass ( Agropyron cristatum ). In o ther parts of their range, Mountain Plover are high ly associated with prairie dog colonies (D insmore et al. 2005, Tipton et al. 2008 Augu stine 2011) but prairie dogs were nearly absent from the study area. The study area is along the eastern periphery of Nest Monitoring. From 2013 to 2015 I began surveying for plover nests in late April and continued until the la st nest hatched in late July. I located active nests by dragging a rope between two ATVs and systematicall y searching fallow crop fields (including standing stubble fields). In areas of low growing crop I walked through the area and scanned for plovers. When plovers were located, I observed them until they returned to their nest. This study was conducted alon gside a nest marking program administered by Bird Conservancy of the Rockies in partnershi p with local landowners. When we located an active nest, we marked the nest by placing 4 wooden stakes around the nest about 10 meters out. This was to prevent accidental tillage of nests when farmers later plowed these fields.

PAGE 22

17 Mountain Plovers are typically very tolerant of heavy farming equipment operating in cl ose proximity to their nest (Knopf and Wunder 2006) so long as the nest itself is not destroyed. We checked nests every 2 7 days until the eggs hatched or the nest failed. Upon fin ding a nest, we floated the eggs to estimate nest age using guidelines provi d ed by Dinsmore et al. (2002). We continued to float eggs about every 7 days until pipping cracks developed. A nest was considered successful if at least one of the eggs hatched. In cases of asynchronous hatch, the nest attempt was considered complete on t he day of first hatch. For hatched nests, I estimated nest age by using the hatch date and assuming a 29 day incubation period (Knopf and Wunder 2006). If a nest failed, I estimated nest age by egg floatation. This is accurate to within several days except for one floatation stage covering 13 days. If a failed nest only had egg floata tion data for this stage then I used the midpoint to estimate nest age. Sex of the incubating adult was determined by molecular sexing from a feather sample as there is no reli able field method for sexing this species. Weather Variables. I obtained daily measures of precipitation and temperature values from th e weather station nearest to the field site, located in Kimball, Nebraska and administered by the National Oceanic and At mospheric Administration (NOAA 2016). All nests included in analyses were within 25km of this station. Data Analysis. I estimated daily survival rate ( DSR ) of nests using program R (R AR K (White and Burnham 1999). I c ) to evaluate which models best explained DSR (Akaike 1973, Burnham and Anderson 2002). I developed a set of models based on a priori biological hypotheses including v ariables that have

PAGE 23

18 previo usly been found to influence DSR of Mountain Plovers and other grassland birds (see Dinsmore et al. 2002, Mettenbrink e t al. 2006, Dreitz et al. 2012) The sex of the incubating adult was determined for 114 out of 154 total nests. To begin constructing the set of models, I first compared a model with sex of the incubating bird to a constant survival model. The model with sex of the incubating bird performed poorly compared to the constant survival model. Thus, all further models di d not include sex and were analyzed including the full sample of nests with adults of unknown sex. The a priori hypotheses used to construct models included in the analysis are described below: 1. Year. Environmental stochasticity can vary greatly from year to year. I included year as a variable to capture any annual variation that was not described by other listed hypotheses. For example, predator activity and frequency of farming operations can vary from year to year and they were not measured in this study 2. Seasonal time trends. I compared a model of constant survival across the breeding season with both linear and quadratic time trends to allow for DSR to vary within a season. Higher DSR in the early season has been found in Mountain Plovers and other spec ies (Ainley and Schlatter 1972, Dinsmore et al. 2002). 3. Nest age. For precocial species, DSR may increase as the nest approaches hatching as nests more vulnerable to predation are depredated early and increased time investment of the incubating adult may reduce nest abandonment. 4. Daily maximum temperatures. On sunny days plovers shad e their eggs to prevent overheating. High temperatur es could lead to lower DSR by direct overheating of eggs or increased energetic stress of adults foraging less in order to shade the eggs.

PAGE 24

19 5. Precipitation. Dreitz et al. (2012) found that Mountain Plover D SR increased during drought periods (defined as 10+ days rain) and Dinsmore et al. (2002) found lower DSR following rain events. I modelled daily precipitation as a continuous variable measured in mm I also included a one day lag effect. Precipit ation events may destroy nests directly (e. g. hail) or increase activity of olfactory predators following precipitation events (hence the one day lag effect) 6. Tagged birds. In this study, 36 adult birds were fitted with GPS tags or radio transmitters for tracking purposes. These birds experienced greater (although brief) disturbance at their nest site as they were trapped multiple times over the course of the nesting cycle. I included tagged birds as a group to examine whether this increased disturbance influenced their DSR. In addition to including tagged birds into the modelling analysis, I also used a t is less biased than chi squared tests when sample sizes (and expected values) are small. To reduce risk of nest abandonment, I only tagged birds that were at least seven days into nest incubation. Thus, nests that failed during the first seven days of in cubation were censored from this test. All models except for the constant survival model included a year effect and either a linear or quadratic time trend (see table 1 for candidate models). Results Nest Survival. The analysis included 154 plover nests Of these, 106 (68.8%) hatched, 16 (10.4%) were abandoned, 10 (6.5%) were destroyed by farming practices (e.g. tilling and planting of fields), 15 (9.7%) were depredated, and 7 failed due to other reasons (hail

PAGE 25

20 damage, infertile eggs, or unknown cause). Fem ales incubated 72 nests 41 were incubated by males, and 41 were incubated by an unknown sex plover. I could not determine the sex of plovers I was unable to catch; it is possible this disparity in the distribution of sex is partly due to differing capture probabilities by sex. Additionally, nests with an unknown sex adult had lower apparent survival (51.2%) likely because in many cases the nest failed before I was able to catch the bird and collect a tissue sample for molecular sexing. Mean nest initiation date ( F ig. 4 ) was simi lar for both sexes (males, mean = 24 May 1.9 days SE; females, mean = 25 May 1.4 days SE). The earliest and latest days of active nesting across the t hree study year were 24 April and 27 July. The estimated survival to hatch from the simple constant su rvival model S (.) was 56.1% (95% CI: 46.3% to 64.8%), obtained by raising the DSR to the power of 29 for the number of incubation days in this species. See table 2 for full summary of the model selection results. The variance in Mountain Plover nest DSR wa s best predicted by a year effect with a quadratic time trend and an interaction effect of rain and maximum temperature Models with quadratic time trends performed better than models with linear time trends. Models with the one day precipitation lag effect were less parsimonious than precipitation models with no lag effect Also, there was not strong evidence to suggest that tagging birds had an influence on their nest sur vival, as these model s were among the least supported by the data. Howeve r, for nest survival of tagged vs. untagged birds produced an odds ratio estimate of 2.48 (95 % CI: 1.05 to 5.85, p = 0.0241 ), indicating that nests tended by untag ged birds were about twice as likely to have hatched than those tende d by tagged birds (Table 3).

PAGE 26

21 In all 3 years, DSR gradually peaked in the mid season and dropped considerably for late season nests (Fig. 5 ) I used the quadratic time trend to estimate s urvival to hatch proba bilities for five nesting periods throughout a breeding season based on earliest, latest, 1 st 2 nd and 3 rd quartiles of nest initiation (Table 4). Discussion This study represents an effort to model Mountain Plover ne st survival in agrarian setting. All 154 nests included in the study occurred on privately owned farmland and were marked to prevent loss through tilling or other farming practices. The apparent nest survival of 68.8% is high compared to nest survival in other studies of plovers which ranges from 27.2% to 70% (Graul 1975, Dinsmore et al. 2002, Mettenbrink et al. 2006, Dreitz and Knopf 2007, Dreitz et al. 2012). In a previous study comparing plovers nesting in agrarian vs. grassland settings, Dreitz and Knopf (2007) found that overall nest loss was similar although the mechanism differed. On crop fields, mechanical farming practices were the leading cause of loss whereas in grasslands nest loss occurred primarily due to predation. The results of my study similarly suggest t hat predation on crop fields is low. There was some evidence for eff ects of rain and temperature on nest survival. Individually or in additive models these variables were not strongly supported but the top model included an interaction term with rain and precipitation. The most parsimonious nest survival models of both Dinsmore et al. (2002) and Dreitz et al. (2012) inc luded a precipitation variable although they did not consider interactions of precipitation and temperature. Precipitation events across th e western Great Plains can be very localized, severe, and brief. Those previous studies included six to seven seasons of data; it may be that the temporal and geographic span of my study did not include weather events severe enough

PAGE 27

22 to influence nest surviv al and overcome the modelling penalty for additional parameters. For example, Dreitz et al. (2012) reported having drought periods of 10 days or more in each season and droughts lasting up to 45 days. In my study the longest drought period was 19 days and there were no drought periods exceeding 10 days in 2015. Furthermore, if rainy weather events increase predator activity as had been suggested (Dinsmore et al. 2002) then this effect may have be en depressed on crop fields where nest loss to predation is lo wer overall Although the modelling results showed no evidence of an effect on daily nest survival probabilities from attaching tracking tags to incubating adults, a direct comparison of nest fates between tagged and untagged birds suggested otherwise. Al though the sample size of tagged birds was comparatively low and the 95% confidence interval of the odds ratio nearly contained 1 (no association), these results suggest further inquiry is needed in studies where tags are applied to nesting birds. Although many of these birds were caught at the nest twice or even three times, all but one of nine nest abandonment incidents occurred after just one trapping attempt. These observations suggest that if there is an effect of tagging, it may stem from the addition of the tag itself rather than the stress associated with repeated capturing and handling of the birds. There was v ariation in nest survival from year to year; the highest rates were observed in 2013 and they decreased each subsequent year. This v ariation was not well captured by the weather covariates included here, as only the interaction of precipitation and daily maximum temperatures improved models beyond a simple quadratic time trend and year effect. These results suggest that although there was evidence of a sharp drop in nest survival in the late season, it cannot be attri buted entirely to weather variables that have

PAGE 28

23 explained variance in plover nest survival in other locations and times. This late se ason drop in nest survival may be attributed to differing pr ey availability over the season or to re nesting efforts of inexperienced br eeders. For plovers nesting in crop fields, weather can shape their environment not only through direct impacts such as hail or flooding of lowlands, but also by influencing how and when farmers work their fields. Rain events delay tilling and planting operations as farmers wait for their fields to dry. In 2014 and 2015 tilling was often delayed until after plover nesting periods were complete. This study did not attempt to quantify or model farming practices or any associations it may have with climatic stochasticity, but expected increases in both drought and severe precipitation events will no doubt impact farming practices. Mountain plovers have shown a daptability in nesting on the crop fields that have supplanted much of their native breeding range; continued study will be needed to assess plover adaptability to predicted climate change.

PAGE 29

24 TABLE 1 Candidate models and notations Candidate models Notation Constant survival S (.) Year + linear trend S (year+T) Year + quadratic trend S (year+TT) Year + linear trend + prcp S (year+ T+prcp ) Year + quadratic trend + prcp S (year+TT+prcp ) Year + linear trend + prcp lag effect S (year+T+lag) Year + quadratic trend + prcp lag effect S (year+TT+lag) Year + linear trend + nest age S (year+T+age) Year + quadratic trend + nest age S (year+TT+age) Year + linear trend + max temp S (year+T+maxtemp) Year + quadratic trend + max temp S (year+TT+maxtemp) Year + linear trend + tagged S (year+T+tag) Year + quadratic trend + tagged S (year+TT+tag) Year + quadratic trend + max temp + prcp S (year+TT+ maxtemp+prcp ) Year + quadratic trend + max temp*prcp S ( year+TT +maxtemp*prcp ) Global model S (year+TT+rain+maxtemp+age+tag+maxtemp*rain)

PAGE 30

25 FIG. 4 Nest Initiation by date

PAGE 31

26 TABLE 2. Summary of model selection results Model AIC c W i K Deviance S (year+TT+ maxtemp*prcp) 0 0.158 6 338.42 S (year+TT ) 0.415 0.129 5 340.85 S (year+TT+prcp) 1.060 0.0931 6 339.48 S (year+TT+maxtemp+prcp) 1.107 0.0910 7 337.52 S (year+T) 1.272 0.0837 4 343.71 S (year+T+maxtemp) 1.859 0.0624 5 342.29 S (year+TT+maxtemp) 1.972 0.0590 6 340.39 S (year+TT+tag) 2.016 0.0577 6 340.44 S (year+TT+prcp lag effect) 2.271 0.0508 6 340.69 S (year+TT+age) 2.378 0.0482 6 340.80 S (year+T+prcp) 2.761 0.0398 5 343.19 S (year+T+tag) 2.857 0.0379 5 343.29 S (year+T+age) 3.097 0.0336 5 343.53 S (year+T+ prcp lag effect) 3.239 0.0313 5 343.67 S (year+TT+ prcp + maxtemp+age+tag+maxtemp*prcp ) 4.903 0.0316 10 335.27 S (.) 5.280 0.0112 1 353.74 TABLE 3. Contingency table for nest outcomes of tagged and untagged birds Failed Hatched Tagged 17 19 Untagged 28 78

PAGE 32

27 TABLE 4. Probability of surviving to hatch for five nesting periods throughout a season Nesting period Probability of hatch 95% CI Earliest .618 .264 to .816 1 st quartile (early) .638 .442 to .772 2 nd quartile (mid) .630 .491 to .739 3 rd quartile (late) .573 .447 to .680 Latest .196 .026 to .469 FIG. 5 Estimate of n est Daily Survival Rate throughout the season with 95% CI

PAGE 33

28 REFERENCES Aebischer N ., P. A. Robertson and R. E. Kenward. 1993. Composition al analysis of habitat use from anima l radio tracking data. Ecology 74:1313 1325. Ainley, D. G., and R. P. Schlatter. 1972. Chick raising ability in Adelie Penguins. The Auk 89:559 566. Akaike, H. 1973. Information theory as an extension of the maximum likelihood principle. Pages 267 281 in B. N. Petrov and F. Csaki, editors. Second International Symposium on Information Theory. Akademiai Kiado, Budapest, Hungary. Augustine, D.J., 2011. Habitat selection by mountain plovers in shortgrass steppe. Journal of Wildlife Management 75: 297 304. Augustine, D.J., and J. D. Derner. 2012 Disturbance regimes and mountain plover habitat in shortgrass steppe Large herbivore grazing does not substitute for prairie dog grazing or fire: Journal of Wildlife Management 76 : 721 728. Batschelet E. 1981. Circular statistics in biology. Academic Press, London, England. Beyer, H. L. 2012. Geospatial Modelling Environment. Version 0.7.3.0. http://www.spatialecology.com/gme Bly, B. L., L. Snyder, and T. VerCauteren. 2008. Mig ration chronology, nesting ecology, and breeding distribution of mountain plover ( Charadrius montanus ) in Nebraska. Nebraska Bird Review 76 : 120 128. Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: a practical informatio n theoretic approach. Springer, New York, New York, USA. Colorado Parks and Wildlife. 2015. State Wildlife Action Plan: A Strategy for Conserving Wildlife in Colorado. Denver, Colorado, USA. Dinsmore, S. J. G. C. White and F. L. Knopf. 2002. Advanced techniques for modeling avian nest survival. Ecology 83: 3476 3488. Dinsmore, S.J., G. C. White, and F. L. Knop f, 2005. Mountain Plover population responses to black tailed prairie d ogs in Montana The Journal of Wildlife Management 6 9: 1546 1553. Dreitz, V. J., and F. L. Knopf. 2007 Mountain plovers and the politics of research on private lands: BioScience 57 : 681 687.

PAGE 34

29 Dreitz, V. J., R. Y. Conrey, and S. K. Skagen. 2012. Drought and cooler temperatures are associated with higher nest survival in mountain plovers. Avian Conservation and Ecology 7(1):6. Dreitz, V.J., M. B. Wunder and F. L. Knop f. 2005. Movements and home ranges of mountain plovers raising broods in three Colorado landscapes The Wilson Bulletin 117 : 128 132. Efron, B. 1979. Bootstrap methods: another look at the jackknife Annals of Statistics 7: 1 26. Environmental Systems Research Institute (ESRI). 2012. ArcGIS. Version 10.1. ESRI, Redlands, California, USA. Graul, W. D. 1975. Breeding biology of the Mountain Plover. The Wilson Bulletin 87: 6 31. Johnson, D. H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61:65 71. Knopf, F. L. 1994. Avian assemblages on altered g rasslands Studies in Avian Biology 15: 247 257. Knopf, F. L., and B. J. Miller. 1994 Charadrius montanus Montane, grassland, or bare ground plover? Auk 111 : 504 506. Knopf, F. L. and J. R. Rupert. 1996 Reproduction and movements of Mountain Plovers bre eding in Colorado The Wilson Bulletin 108:28 35 Knopf, F. L. and J. R. Rupert. 1999. Use of cultivated fields by breeding mountain plovers in Colorado. Studies in Avian Biology 19: 81 86. Knopf F. L., and M. B. Wunder. 2006. Mountain plover ( Charadrius montanus ) The birds of North America. Number 211. Levin, S. A. 1992. The Problem of pattern and scale in ecology. Ecology 73: 1943 1967. Manly, B. F. J., L. L. McDonald, D. L. Thomas, T. L. McDonald, and W. P. Erickson. 2002. Resource selection by ani mals: statistical design and analysis for field studies. Second Edition. Springer Science+Business Media, Dordrecht, the Netherlands. Martin, T. E. 1987 Food as a Limit on Breeding Birds: A Life History Perspective Annual Re view of Ecology and Systemati cs 18 : 453 487 Matthews, J. H. 2008. Anthropogenic climate change in the Playa Lakes Joint Venture Region: understanding impacts, discerning trends, and developing responses. World Wildlife Fund, Corvallis, Oregon, USA.

PAGE 35

30 Mettenbrink, C. W., V. J. Dreitz, and F. L. Knopf. 2006. Nest success of mountain plovers relative to anthropogenic edges in eastern Colorado. The Southwestern Naturalist 51:191 196. National Oceanic and Atmospheric Administration. 2016. National Centers for Environmental Information. http://www.ncdc.noaa.gov/cdo web/datasets#GHCND Accessed 04 June 2016. Peek J. M. 1986. A review of wildlife management. Prentice Ha ll, Englewood Cliffs, USA. R Core Team. 2015. R: a language and environment for statistical computing. Version 3.2.2. R Foundation for Statistical Computing, Vienna, Austria. http://www.r project.org/ Samson, F.B., F. L. Knopf and W. R. Os tlie. 2004 Great Plains ecosystems Past, present, and future: Wildlife Society Bulletin 32 : 6 15. Sauer, J. R., J. E. Hines, J. E. Fallon, K. L. Pardieck, D. J. Ziolkowski Jr., and W. A. Link, 2014. The North American Breeding Bird Survey, Results and Analysis 1966 2012. Version 01.30.2015 USGS Patuxent Wildlife Research Center Laurel, Maryland. Accessed 02 Nov 2016 Schneider, R., K. Stoner, G. Steinaure M. Panella, and M. Humpert. 2011. The Nebraska Natural Legacy Project: State Wildlife Action Plan Second edition. The Nebraska Game and Parks Commission, Lincoln, USA. Seaman, D. E. and R. A. Powell. 1996. An evalu ation of the accuracy of kernel density estimators for home range analysis. Ecology 77:2075 2085. Shackford, J.S., D. M. Leslie, and W. D. Harden. 1999. Range wide use of cultivated fields by mountain plovers during the breeding season. Journal of Field O rnithology 70: 114 120. Skrade, P.D.B. and S. J. Dinsmore. 2012. Incubation patterns of a shorebird with rapid multiple clutches, the Mountain Plover ( Charadrius montanus ). Canadian Journal of Zoology 90: 257 266. Swihart, R. K. and N. A. Slade. 1986. The importance of statistical power when testing for independence in animal movements. Ecology 67:255 258. Tipton, H. C. V. J. Dreitz, and P. F. Doherty, Jr. 2008. Occupancy of mountain plover and burrowing owl in Colorado. Journal of Wildlife Management 72: 1001 1006. Vickery, P.D. and J. R. Herkert. 2001. Recent advances in grassland bird research: where do we go from here? Auk 118: 11 15.

PAGE 36

31 White, G. C, and K. P. Burnham. 1999. Program MARK: survival estimation from populations of marked an imals. Bird St udy 46 :120 138 White, R. P., S. Murray, and M. Rohweder. 2000 Grassland Ecosystems World Resources Institute, Washington DC, USA Whittingham, M. J., S. M. Percival, and A. F. Brown. 2000. Time budgets and foraging of breeding golden plover ( Pluvialis apricaria ) Journal of Applied Ecology 37: 632 646.

PAGE 37

32 APPENDIX Log ratio differences of used and available habitats for each plover; a positive value indicates habitat selection of the row over the column while a negative value indicates habitat avoidance. Ranks are from 1 (most preferred) to 4 (least preferred). Bird ID:XWRO Fallow Growing Crop Grassland Other Total Rank Fallow 2.584 11.820 10.622 25.026 1 Growing Crop 2.584 9.235 8.038 14.689 2 Grassland 11.820 9.235 1.197 22.252 4 Other 10.622 8.038 1.197 17.463 3 Bird ID: ORIX Fallow Growing Crop Grassland Other Total Rank Fallow 12.985 10.302 7.889 31.176 1 Growing Crop 12.985 2.684 5.096 20.765 4 Grassland 10.302 2.684 2.412 10.030 3 Other 7.889 5.096 2.412 0.381 2 Bird ID: WWIX Fallow Growing Crop Grassland Other Total Rank Fallow 1.819 11.506 11.440 24.764 1 Growing Crop 1.819 9.687 9.621 17.490 2 Grassland 11.506 9.687 0.066 21.259 4 Other 11.440 9.621 0.066 20.995 3

PAGE 38

33 Bird ID: IXIO Fallow Growing Crop Grassland Other Total Rank Fallow 1.644 10.012 7.200 18.855 1 Growing Crop 1.644 8.368 5.556 12.281 2 Grassland 10.012 8.368 2.812 21.191 4 Other 7.200 5.556 2.812 9.944 3 Bird ID: RIRX Fallow Growing Crop Grassland Other Total Rank Fallow 2.478 10.100 8.454 21.028 1 Growing Crop 2.478 7.618 5.976 11.115 2 Grassland 10.100 7.618 1.642 19.356 4 Other 8.454 5.976 1.642 12.787 3 Bird ID: XORW Fallow Growing Crop Grassland Other Total Rank Fallow 13.283 12.302 8.026 33.611 1 Growing Crop 13.283 0.981 5.257 19.520 4 Grassland 12.302 0.981 4.276 15.597 3 Other 8.026 5.257 4.276 1.507 2 Bird ID: XOIR Fallow Growing Crop Grassland Other Total Rank Fallow 13.150 12.374 8.504 34.028 1 Growing Crop 13.150 0.776 4.646 18.572 4 Grassland 12.374 0.776 3.871 15.469 3 Other 8.504 4.646 3.871 0.013 2

PAGE 39

34 Bird ID: XRRW Fallow Growing Crop Grassland Other Total Rank Fallow 13.375 11.808 9.972 35.155 1 Growing Crop 13.375 1.568 3.403 18.346 4 Grassland 11.808 1.568 1.836 12.076 3 Other 9.972 3.403 1.836 4.733 2 Bird ID: XWIO Fallow Growing Crop Grassland Other Total Rank Fallow 12.697 0.413 8.688 21.797 1 Growing Crop 12.697 12.284 4.009 28.989 4 Grassland 0.413 12.284 8.275 20.146 2 Other 8.688 4.009 8.275 12.955 3