INTRODUCTION

The tribe Numeniini (Aves) includes waders and shorebirds, and over half of the species in this group are currently of conservation concern (Pearce-Higgins et al. 2017). The Long-billed Curlew (Numenius americanus) is North America’s largest shorebird and is dependent on arid short and mixed grassland habitats in the western portion of the continent (Dugger and Dugger 2020). The species is considered “Highly Imperiled” in the United States Shorebird Conservation Plan (Brown et al. 2001), “Critically Important” in the Intermountain West Regional Shorebird Plan (Oring et al. 2013), and a “Bird of Conservation Concern” by the U.S. Fish and Wildlife Service (USFWS 2021). Loss of grassland breeding habitat is considered to be the greatest threat to the stability of Long-billed Curlew populations (Dugger and Dugger 2020).

The Intermountain West in North America ranges from the Rocky Mountains in the east to the Sierra and Cascade Mountains in the west, and from Canada in the north to Mexico in the south. This is the most important region in North America for breeding Long-billed Curlews (Oring et al. 2013). In several U.S. states, including Oregon, Idaho, Utah, and Nevada, uplands associated with wetlands and riparian areas provide crucial nesting habitat for curlews (Oring et al. 2013).

Because of the importance of the Intermountain West to nesting curlews, understanding these birds’ density and abundance in this region is important to their conservation. Although several studies have been conducted in this region related to breeding ecology and habitat use of curlews (e.g., Pampush and Anthony 1993, Stocking et al. 2010), few studies have focused on density or abundance. In the Ruby Valley, Nevada, numbers of breeding pairs slightly fluctuated (59–66 pairs) between 2004 and 2007 (Hartman 2008). At one site in western Idaho, a breeding population of Long-billed Curlews was relatively stable over a seven-year period in the late 1970s and early 1980s (Redmond and Jenni 1986). However, at this same site, curlew abundance decreased by > 90% over the subsequent ~30 years, and declines appear to be ongoing (Coates et al. 2021). In the Columbia Basin in Oregon, Long-billed Curlew distribution and abundance was last studied in 1980 (Pampush 1980, Fellows and Jones 2009), so updated information in this study area would be useful in understanding the current conservation needs of the species.

Lands managed by the U.S. Government, including in the Intermountain West, provide habitat for many wildlife species of conservation concern, and these lands are becoming increasingly important to sustaining the nation’s biodiversity (Stein et al. 2008). The U.S. Department of Defense (DoD) manages < 2% of all federal lands (Esri 2022), yet the densities of imperiled species are at least three times higher on military lands than they are on lands managed by any other federal agency in the country (Tazik and Martin 2002, Stein et al. 2008). Military lands throughout the world, if well-managed, can benefit biodiversity (Warren et al. 2007, Zentelis and Lindenmayer 2015, Caudal and Gallet 2023). Thus, the DoD plays an important role in the conservation of wildlife (Tazik and Martin 2002). Although the primary purpose of DoD-managed natural resources is to support mission-related activities, conserving those resources and maintaining the integrity of ecosystems enhances the military’s ability to train on those lands in a long-term, sustainable manner (Lankow et al. 2022). Such conservation work sometimes can be implemented through the establishment of collaborative partnerships between the DoD and other agencies or institutions (Lankow et al. 2022).

We conducted a two-year study of Long-billed Curlews on DoD-managed sagebrush steppe habitat in the Columbia Basin in Oregon. Our objective was to estimate density and abundance of curlews during the breeding season. We predicted curlew density would be positively associated with the amount of grassland habitat and with less rugged, lower-elevation landscapes. Because a wildfire burned a substantial portion of the study area during the latter part of the breeding season in the first year, we further predicted curlews would respond negatively to this large-scale disturbance and their density would decrease in the second year.

METHODS

Study area

We studied Long-billed Curlews in 2015 and 2016 on the U.S. Navy’s 192.39-km² Naval Weapons Systems Training Facility (NWSTF) near Boardman, Oregon (Fig. 1). The installation is under the command of the Naval Air Station Whidbey Island and is used as a U.S. military bombing and gunnery range. The NWSTF has a history of livestock grazing, frequent fires from military activities and lightning strikes, and invasive plants, especially cheatgrass (Bromus tectorum).

The NWSTF is located within the sagebrush steppe region of the Columbia Basin. Shrub species present include big sagebrush (Artemisia tridentata), antelope bitterbrush (Purshia tridentata), gray rabbitbrush (Ericameria nauseosa), and green rabbitbrush (Chrysothamnus viscidiflorus). Bitterbrush is sparse and restricted to the sandy soils in the northern end of the study area, and sagebrush communities are restricted to the southern half. Grasses on NWSTF include cheatgrass, Sandberg’s bluegrass (Poa secunda), needle-and-thread (Hesperostipa comata), and bluebunch wheatgrass (Pseudoroegneria spicata). The central area of the installation is a mosaic of needle-and-thread, rabbitbrush shrublands, and extensive open areas dominated by cheatgrass.

Weather for the region tends to be cool at night with warmer days and high winds. Annual precipitation during the two study years was 18.9 cm and 21.1 cm in 2015 and 2016, respectively (Western Regional Climate Center 2022). Average monthly temperatures ranged from -1.8 °C in December 2016 to 25.7 °C in July 2015 (Western Regional Climate Center 2022).

Field surveys

To survey Long-billed Curlews, we randomly established point count locations in each year throughout the property (Fig. 1). Points were established an average of ~1000 m apart to reduce the probability of double counting (Stanley and Skagen 2007, Jones et al. 2008). We targeted a narrow survey time period to coincide with the pre-incubation period when breeding birds are most conspicuous, i.e., they display and vocalize more during this period (Redmond et al. 1981, Stanley and Skagen 2007, Jones et al. 2008). In the Columbia Basin, this period approximately occurs from mid-March to mid-April (Stanley and Skagen 2007, Jones et al. 2008). Our survey methods assumed the local breeding population was closed (i.e., individuals were not entering or exiting the NWSTF population) because, during our survey period, breeding birds were already on their territories.

We surveyed 78 locations 1 time over 6 days between 1 April and 9 April in 2015, and 78 different locations 1 time over 9 days between 4 April and 13 April in 2016. We chose different survey points between the two years so that we could conduct more surveys in 2016 within an area that burned in 2015 (almost 30% more). We surveyed all points with teams of two observers, using the double-observer method (Nichols et al. 2000), to adhere to an NWSTF regulation that does not allow surveyors to be alone on the property. We conducted surveys from half an hour after local sunrise until approximately 13:00 PDT, and we did not survey during rainy conditions or when winds were above 25 km/hr. We surveyed at these times of day and under these weather conditions to optimize detections when territorial birds are most active (Hartman 2008).

Surveys lasted for seven minutes, and the primary observer detected curlews by sight and sound. This observer estimated distance to the detection by using either a laser rangefinder or unaided visual estimation and noted curlew behavior, number of birds, and azimuth to the detection. In a few cases, the observer recorded the distance band rather than the actual distance. The other team member recorded observations made by the primary observer, also noting other curlews observed that went undetected by the primary observer, and recorded location-specific data, including coordinates of the survey location and weather data. The weather data recorded included wind speed (m/s), temperature (°C), and cloud cover categorized as clear (no clouds), scattered (light clouds spread out over the sky), broken (clear in part of the sky and heavy clouds in other parts of the sky), or overcast (heavy clouds throughout the sky).

Analytical approach

We used multinomial-Poisson mixture models to estimate Long-billed Curlew density and abundance on NWSTF (Royle et al. 2004, Chandler 2020). This type of analysis requires four underlying assumptions (Buckland et al. 2015, Schmidt et al. 2022): (1) birds are detected with certainty at the point; (2) birds are detected at their initial location; (3) distances to birds are measured accurately; and (4) placement of survey points is random. Violation of the first assumption is uncommon for this species and habitat because curlews are easy to detect. To reduce the biases associated with the second assumption, we did not count birds that were flying over the survey area or those that flew into the survey area and landed after surveys began. To reduce the chances of erroneous distance estimates (i.e., to address the third assumption), we grouped estimated distances to birds into distance bands of 0–400 m or 401–800 m (Stanley and Skagen 2007, Jones et al. 2008), and we removed from analyses data from birds observed at distances > 800 m. Lastly, for the fourth assumption, we randomly placed survey point locations across the entire property, regardless of access or habitat.

Environmental covariates

We associated 2.01-km² survey plots (based on an 800-m radius around each survey point) with land cover, topographic, and meteorological characteristics. We considered covariates with ecological relevance to Long-billed Curlews and to the grassland habitat within our study area. For example, land cover is known to influence distribution and abundance of this species (Duggar and Duggar 2020).

First, we obtained 30-m resolution data on 2016 land cover from the National Land Cover Database (Dewitz 2019). We used ArcGIS Pro 2.9.1 (Esri, Redlands, California) to calculate the proportion of each survey plot that consisted of grassland. We used the proportion of a plot in grassland as the only land cover covariate because > 80% of NWSTF consisted of grassland, and Long-billed Curlews are dependent on grassland habitat for nesting.

We also classified land cover at NWSTF according to recent fire history. A wildland fire burned 68.3 km² of NWSTF (~36%; Fig. 1) on 1 June 2015, after the 2015 survey period. In 2016, 31 of the 78 survey plots overlapped with the area burned. We assigned a “1” to these plots and a “0” to the remaining 47 plots in 2016.

Second, we linked each 30-m cell within the survey plots to elevation data and to a measurement of a terrain ruggedness index (TRI; Riley et al. 1999). We considered these two topographic variables because curlews tend to prefer nesting in cheatgrass, which is located in lower elevations and smoother terrain (Pampush and Anthony 1993, Peeler and Smithwick 2018). We obtained elevation data from a digital elevation model (USGS 2015). Terrain ruggedness index, estimated with geomorphometry and gradient metrics tools (Evans et al. 2014), reflects landscape roughness and is calculated as the square root of the sum of the squared differences between the elevation in a cell and the elevation of its neighboring cells (Riley et al. 1999). In analyses, we used the mean elevation and TRI value within each survey plot. Finally, we associated the weather data recorded during the surveys (i.e., wind speed, temperature, and cloud cover) with each survey plot.

Data analysis

We estimated density of Long-billed Curlews from the survey data with the distance sampling technique, using multinomial-Poisson mixture models that simultaneously model both detection probability and density, and that allow the incorporation of covariate effects on both parameters (Royle et al. 2004, Chandler 2020). We ran separate models for each of the two years (2015 and 2016) using the “unmarked” package in R v.4.1 (Fiske and Chandler 2011, R Core Team 2021). Although this method does not allow us to statistically compare detection probability and density estimates between years in the same model, separating the two years allowed us to generate separate abundance estimates before and after the 2015 fire.

We first tested correlations between pairs of continuous environmental variables to ensure that no model contained two variables that had a correlation ≥ 0.50 (Table 1). Only elevation and TRI were correlated (r = 0.50), and thus we did not include both variables in the same model. Next, we ran null models (i.e., no predictor variables for detection or density) for each year using each of the detection functions available for distance sampling: half-normal, hazard rate, and negative exponential (Chandler 2020). The half-normal models were the only ones that converged; thus, we moved forward with this detection function in our models.

Because of the large number of potential covariates in our models, we began analyses by identifying priority predictor variables. To do this, we ran univariate models (i.e., only one predictor variable at a time) for detection and density for each year, and we used Akaike information criterion (AIC) to rank the models for each year (Burnham and Anderson 2002, Anderson 2008). In both years, our models evaluated covariates describing proportion of grassland, mean elevation, and mean TRI as predictors of detection and density. Additionally, we considered wind speed, temperature, and cloud cover as predictors of detection only. For 2016, we also considered the binary term describing the effect of fire as a covariate for both detection and density.

Next, we constructed a global, multivariate model for each year that only included the three highest-ranked variables for detection and for density, unless elevation and TRI were both included in the three highest-ranked variables, in which case the lower-ranked variable was removed. This resulted in the global model for 2015 including TRI, wind speed, and cloud cover as predictors of detection, and proportion of grassland and TRI as predictors of density. Elevation was one of the three top-ranked variables for density but was ranked lower than TRI and thus removed. The global model for 2016 included wind speed, temperature, and cloud cover as predictors of detection, and proportion of grassland, fire, and elevation as predictors of density. We then used the “dredge” function in the “MuMIn” R package (Bartoń 2020) to evaluate all possible submodels for each year (n = 32 for 2015, n = 64 for 2016; Doherty et al. 2012). We used AIC corrected for small sample size to rank the models for each year (Burnham and Anderson 2002, Anderson 2008), and we averaged the models with weights ≥ 0.01.

Finally, we used the averaged models to predict detection functions and density estimates for each survey plot in each year. Because many of the plots within a year overlapped, we removed the overlap between adjacent plots by creating a border that evenly distributed the overlapping area between plots and then calculated the revised area of each plot. We used the predicted detection functions to calculate the detection probability for each plot at each distance band (400 and 800 m) for each year. We then calculated curlew abundance (with a 95% confidence interval; CI) separately for each of the 78 plots in each year by multiplying the predicted density estimate (and the lower and upper CI estimates) for each plot by the revised plot area. We summed these products to estimate abundance across the surveyed area. Additionally, we estimated abundance within NWSTF by multiplying the predicted density estimate for each plot by the NWSTF area and calculating the average value across all plots. We repeated this procedure for the density estimate at the lower and upper 95% CI. Lastly, we calculated the average density for each day to evaluate the degree to which observers conducted the point counts consistently across days. We interpreted these daily counts with caution because plots were not surveyed in a random order and, thus, clusters of high-density plots may have been surveyed on one day and clusters of low-density plots on another day.

RESULTS

Curlew surveys

After removing overlapping areas, the area of a survey plot ranged from 0.66 km² to 2.01 km² in 2015, and from 0.91 km² to 2.01 km² in 2016. Total area surveyed was 115.02 km² in 2015 and 119.63 km² in 2016. This represented approximately 60% of NWSTF.

We detected 151 Long-billed Curlews across 50 (64% of the 78 surveyed) survey plots in 2015, and 125 curlews across 46 (59%) plots in 2016. The number of birds detected per plot ranged from 0 to 8 in 2015 and from 0 to 6 in 2016 (Fig. 2a). We observed curlews at locations throughout NWSTF, but curlew density appeared to be highest in the northern portion of the range, an area characterized by open cheatgrass-dominated landscapes. In 2015, we detected more birds in the 0–400 m distance band, but in 2016, we observed more birds in the 401–800 m band (Fig. 2b).

Detection probabilities and abundance estimates

As expected, detection probability decreased with distance in both years (Fig. 3). The mean detection probability at distances of 400 m and 800 m was 0.45 and 0.07, respectively, in 2015, and 0.61 and 0.18, respectively, in 2016. Our multivariate model suggested cloud cover was the only variable associated with detection in 2015, with clear skies increasing detection probability (Table 2a). No variables were associated with detection in 2016 (Table 2b).

Density estimates were different in the two years of the study. In 2015, mean predicted density across all surveyed plots was 3.3 curlews/km² (CI: 2.4–4.7), and daily predicted density ranged between 2.8 and 4.3 curlews/km². Only TRI was associated with density of curlews. This relationship was negative, indicating higher curlew density in areas of lower landscape roughness (Table 2a, Fig. 4a). The proportion of grassland within a survey plot was not associated with curlew density (Table 2a, Fig. 4b). In 2016, mean predicted density was 1.8 curlews/km² (1.2–2.7), and daily predicted density ranged between 1.1 and 2.7 curlews/km². Only elevation was associated with density of curlews. This relationship also was negative, indicating higher curlew density in lower-elevation areas (Table 2b, Fig. 5a). Neither the proportion of grassland nor the fire covariate were associated with curlew density (Table 2b, Fig. 5b, c).

Based on predicted density estimates for each plot, the estimated abundance of curlews in 2015 for the sampled area was 376 birds (CI: 268–538), and for the NWSTF was 639 birds (CI: 456–912). Estimated abundance of curlews in 2016 for the sampled area was 219 birds (CI: 148–324), and for the NWSTF was 350 birds (CI: 237–520).

The structure and ranking of models describing detection and density also varied between years. In 2015, the top-ranked model had 22% of model weights and included cloud cover as a predictor of detection and TRI as a predictor of density (Table 3a). We averaged the top 10 models (weights ≥ 0.01). The top 8 models included cloud cover, and all 10 models included TRI, indicating that these 2 variables were particularly important to detection and density, respectively (Table 3a). In 2016, the top-ranked model had 14% of model weights and included wind speed, temperature, and cloud cover as predictors of detection, and elevation as a predictor of density (Table 3b). We averaged the top 22 models (weights ≥ 0.01). Cloud cover was included in 20 of these models (although averaged effect estimates were not different from 0), and elevation was included in 18 models (Table 3b).

DISCUSSION

Mean estimated abundance of the breeding population of Long-billed Curlews at NWSTF was 45% higher in 2015 than in 2016, although the CIs of the abundance estimates in each year overlapped each other. Although environmental covariates associated with density were different between the two years, elevation and landscape roughness were positively correlated at NWSTF, indicating that curlews generally were found in the same areas in each year, i.e., in areas with flatter, less rugged terrain. These results are consistent with previous findings that curlews prefer to nest in grasslands, especially cheatgrass habitat (Pampush and Anthony 1993, Stocking et al. 2010), which is usually located in lower-elevation, smoother areas (Leger et al. 2009, Peeler and Smithwick 2018).

Although fire was not an important predictor in our 2016 models, curlew abundance appeared to be substantially lower at NWSTF after the June 2015 wildfire. Because most curlews occupied the northern portion of the range, and the fire occurred in the southern portion (Fig. 1), the lack of a fire effect, as we measured it, on abundance was not surprising. However, although curlews may prefer nesting habitat that has been burned (Cannings 1999), this large fire, which occurred during the latter part of the breeding season, still may have affected post-breeding dispersal, and thus also abundance during the following breeding season. In southwestern Idaho, a similar effect was documented in which female curlews were more likely to disperse from their breeding territories after excessive disturbance during the breeding season (Redmond and Jenni 1982, 1986). This behavior also has been documented in other bird species, such as Whinchats (Saxicola rubetra) in Switzerland (Grüebler et al. 2015) and a community of forest birds in eastern France (Bötsch et al. 2017).

Our density estimates (3.3 and 1.8 curlews/km² in 2015 and 2016, respectively) were lower than those reported in western Idaho in the late 1970s and early 1980s (approximately 6–7 male curlews/km²; Redmond and Jenni 1986) but higher than reported densities in three sites in Idaho in 2021 (0.04–2.3 curlews/km²; Coates et al. 2021). Curlew densities in the mid-2000s were also low in central Utah (0.20 curlews/km²) and in several breeding areas in mainland Washington (0.02–1.5 curlews/km²; Fellows and Jones 2009). Although comparisons between our study and these earlier studies are limited due to differences in methodology, our relatively higher densities than those in nearby states could indicate the importance of the NWSTF, and the Columbia Basin in general, to breeding Long-billed Curlews.

Our curlew abundance estimates may have been influenced by the point count survey design. Survey points generally were chosen so that each point was located an average distance of ~1000 m from another point. Curlew observations were then recorded up to distances of 800 m, a design that may result in substantial overlap among survey plots. We corrected our abundance estimates by adjusting areas of plots so they did not overlap, but our density, and thus abundance, values still may have been over-estimated. That said, the fact that two different mechanisms to estimate curlew density (daily and annually) were roughly similar gives us confidence in our annual abundance estimates.

Generally, detection probability was influenced more by weather variables than by topographic or land cover variables. In 2015, we are unsure why clear skies increased detection probability, but this may have been because greater amounts of light increased visibility or caused the curlews’ behavior to change, such as an increase in vocalizations (Bruni et al. 2014). In 2016, we did not find a similar influence of clear skies, but our detection probability estimates were somewhat higher than in 2015, so these estimates were likely influenced by variables we did not measure, e.g., change in height of vegetation.

Our two-year study provided valuable information on the abundance of Long-billed Curlews at NWSTF and the Columbia Basin in general, all of which can be used to understand the current conservation needs of the species. Our findings confirm that curlew populations can fluctuate from year to year, as was found in Ruby Valley, Nevada (Hartman 2008). During our study, a large-scale disturbance (wildfire) may have contributed to changes in local density of curlews. That said, other factors could also be influential, for example, those occurring in the non-breeding season in migration and winter habitats. Whether density increased in the following breeding seasons is currently unknown; however, cheatgrass, a habitat type preferred by curlews (Pampush and Anthony 1993, Stocking et al. 2010), likely increased post-fire. Additionally, the higher densities of curlews found on the NWSTF, relative to those reported in other recent studies in surrounding areas, suggest that military lands might be important for breeding curlews. One possible explanation for this is that disturbance regimes on military installations are heterogeneous rather than homogeneous, which differs from the majority of anthropogenic disturbances (Warren et al. 2007). Although curlews are not always recognized as benefiting from efforts by military land managers, if they respond favorably to military-style disturbance regimes, management strategies for the species could be implemented accordingly. Lands managed by the DoD are often of high conservation value to many species (Stein et al. 2008). Thus, continued collaboration in conservation planning between the DoD and other institutions may benefit curlews and other wildlife species.

LITERATURE CITED

Anderson, D. R. 2008. Model based inference in the life sciences: a primer on evidence. Springer, New York, New York, USA. https://doi.org/10.1007/978-0-387-74075-1

Bartoń, K. 2020. MuMIn: multi-model inference. R package, version 1.43.17. R Foundation for Statistical Computing, Vienna, Austria. https://cran.r-project.org/package=MuMIn

Bötsch, Y., Z. Tablado, and L. Jenni. 2017. Experimental evidence of human recreational disturbance effects on bird-territory establishment. Proceedings of the Royal Society B: Biological Sciences 284:20170846. https://doi.org/10.1098/rspb.2017.0846

Brown, S., C. Hickey, B. Harrington, and R. Gill. 2001. United States shorebird conservation plan. Second edition. Manomet Center for Conservation Sciences, Manomet, Massachusetts, USA. https://www.shorebirdplan.org/plan-and-council/

Bruni, A., D. J. Mennill, and J. R. Foote. 2014. Dawn chorus start time variation in a temperate bird community: relationships with seasonality, weather, and ambient light. Journal of Ornithology 155:877–890. https://doi.org/10.1007/s10336-014-1071-7

Buckland, S. T., E. A. Rexstad, T. A. Marques, and C. S. Oedekoven. 2015. Distance sampling: methods and applications. Springer, Cham, Switzerland. https://doi.org/10.1007/978-3-319-19219-2

Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: a practical information-theoretic approach. Second edition. Springer, New York, New York, USA.

Cannings, R. J. 1999. Status of the Long-billed Curlew in British Columbia. Wildlife Working Report No. WR-96. Ministry of Environment, Lands and Parks, Wildlife Branch, Victoria, British Columbia, Canada.

Caudal, P., and S. Gallet. 2023. Khaki conservation: a review of the effects on biodiversity of worldwide military training areas. Environmental Reviews 31:527–541. https://doi.org/10.1139/er-2023-0014

Chandler, R. 2020. Distance sampling analysis in unmarked. R Foundation for Statistical Computing, Vienna, Austria. https://cran.r-project.org/web/packages/unmarked/vignettes/distsamp.html

Coates, S., H. Hayes, and J. Carlisle. 2021. IBO Long-billed Curlew research and community education: 2021 status report. Intermountain Bird Observatory, Boise State University, Boise, Idaho, USA. https://doi.org/10.13140/RG.2.2.17649.61282

Dewitz, J. 2019. National land cover database (NLCD) 2016 products. U.S. Geological Survey data release. USGS, Reston, Virginia, USA. https://doi.org/10.5066/P96HHBIE

Doherty, P. F., G. C. White, and K. P. Burnham. 2012. Comparison of model building and selection strategies. Journal of Ornithology 152:S317–S323. https://doi.org/10.1007/s10336-010-0598-5

Dugger, B. D., and K. M. Dugger. 2020. Long-billed Curlew (Numenius americanus). Version 1.0. In A. F. Poole and F. B. Gill, editors. Birds of the world. Cornell Lab of Ornithology, Ithaca, New York, USA. https://doi.org/10.2173/bow.lobcur.01

Esri. 2022. USA federal lands spatial dataset. Esri, Redlands, California, USA. https://www.arcgis.com/home/item.html?id=5e92f2e0930848faa40480bcb4fdc44e

Evans, J. S., J. Oakleaf, S. A. Cushman, and D. Theobald. 2014. An ArcGIS toolbox for surface gradient and geomorphometric modeling. Version 2.0-0. https://evansmurphy.wixsite.com/evansspatial/arcgis-gradient-metrics-toolbox#:~:text=The%20Geomorphometry%20and%20Gradient%20Metrics,vegetation%2C%20and%20remote%20sensing%20studies

Fellows, S. D., and S. L. Jones. 2009. Status assessment and conservation action plan for the Long-billed Curlew (Numenius americanus). Biological Technical Publication, BTP-R6012-2009. U.S. Department of the Interior, Fish and Wildlife Service, Washington, D.C., USA. https://whsrn.org/wp-content/uploads/2019/02/statusassessment_conservationactionplan_lbcu.pdf

Fiske, I. J., and R. B. Chandler. 2011. unmarked: an R package for fitting hierarchical models of wildlife occurrence and abundance. Journal of Statistical Software 43(10):1–23. https://doi.org/10.18637/jss.v043.i10

Grüebler, M. U., H. Schuler, R. Spaar, and B. Naef-Daenzer. 2015. Behavioural response to anthropogenic habitat disturbance: indirect impact of harvesting on whinchat populations in Switzerland. Biological Conservation 186:52–59. https://doi.org/10.1016/j.biocon.2015.02.031

Hartman, C. A. 2008. Behavioral ecology and population biology of Long-billed Curlews in northeastern Nevada. Dissertation. University of Nevada, Reno, Nevada, USA.

Jones, S. L., C. S. Nations, S. D. Fellows, and L. L. McDonald. 2008. Breeding abundance and distribution of Long-billed Curlews (Numenius americanus) in North America. Waterbirds 31:1–14. https://doi.org/10.1675/1524-4695(2008)31[1:BAADOL]2.0.CO;2

Lankow, A. J., B. T. Hazen, and M. A. Douglas. 2022. Converging on symbiosis: developing a collaborative framework for guiding successful United States Department of Defense conservation partnerships. Human Dimensions of Wildlife 27:489–505. https://doi.org/10.1080/10871209.2021.1963017

Leger, E. A., E. K. Espeland, K. R. Merrill, and S. E. Meyer. 2009. Genetic variation and local adaptation at a cheatgrass (Bromus tectorum) invasion edge in western Nevada. Molecular Ecology 18:4366–4379. https://doi.org/10.1111/j.1365-294X.2009.04357.x

Nichols, J. D., J. E. Hines, J. R. Sauer, F. W. Fallon, J. E. Fallon, and P. J. Heglund. 2000. A double-observer approach for estimating detection probability and abundance from point counts. Auk 117:393–408. https://doi.org/10.1093/auk/117.2.393

Oring, L. W., L. Neel, and K. E. Oring. 2013. Intermountain West regional shorebird plan. Version 1.0. The U.S. Shorebird Conservation Partnership, U.S. Fish and Wildlife Service, Lakewood, Colorado, USA. https://www.shorebirdplan.org/wp-content/uploads/2013/01/IMWEST4.pdf

Pampush, G. J. 1980. Status report on the Long-billed Curlew in the Columbia and northern Great Basins. U.S. Department of the Interior, Fish and Wildlife Service, Portland, Oregon, USA.

Pampush, G. J., and R. G. Anthony. 1993. Nest success, habitat utilization and nest-site selection of Long-billed Curlews in the Columbia Basin, Oregon. Condor 95:957–967. https://doi.org/10.2307/1369431

Pearce-Higgins, J. W., D. J. Brown, D. J. T. Douglas, J. A. Alves, M. Bellio, P. Bocher, G. M. Buchanan, R. P. Clay, J. Conklin, N. Crockford, P. Dann, J. Elts, C. Friis, R. A. Fuller, J. A. Gill, K. Gosbell, J. A. Johnson, R. Marquez-Ferrando, J. A. Masero, D. S. Melville, S. Millington, C. Minton, T. Mundkur, E. Nol, H. Pehlak, T. Piersma, F. Robin, D. I. Rogers, D. R. Ruthrauff, N. R. Senner, J. N. Shah, R. D. Sheldon, S. A. Soloviev, P. S. Tomkovich, and Y. I. Verkuil. 2017. A global threats overview for Numeniini populations: synthesizing expert knowledge for a group of declining migratory birds. Bird Conservation International 27:6–34. https://doi.org/10.1017/S0959270916000678

Peeler, J. L., and E. A. H. Smithwick. 2018. Exploring invasibility with species distribution modeling: how does fire promote cheatgrass (Bromus tectorum) invasion within lower montane forests? Diversity and Distributions 24:1308–1320. https://doi.org/10.1111/ddi.12765

R Core Team. 2021. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/

Redmond, R. L., T. K. Bicak, and D. A. Jenni. 1981. An evaluation of breeding season census techniques for Long-billed Curlews (Numenius americanus). Studies in Avian Biology 6:197–201.

Redmond, R. L., and D. A. Jenni. 1982. Natal philopatry and breeding area fidelity of Long-billed Curlews (Numenius americanus): patterns and evolutionary consequences. Behavioral Ecology and Sociobiology 10:277–279. https://doi.org/10.1007/BF00302817

Redmond, R. L., and D. A. Jenni. 1986. Population ecology of the Long-billed Curlew (Numenius americanus) in western Idaho. Auk 103:755–767. https://doi.org/10.1093/auk/103.4.755

Riley, S. J., S. D. DeGloria, and R. Elliot. 1999. A terrain ruggedness index that quantifies topographic heterogeneity. Intermountain Journal of Sciences 5:23–27. https://download.osgeo.org/qgis/doc/reference-docs/Terrain_Ruggedness_Index.pdf

Royle, J. A., D. K. Dawson, and S. Bates. 2004. Modeling abundance effects in distance sampling. Ecology 85:1591–1597. https://doi.org/10.1890/03-3127

Schmidt, J. H., W. L. Thompson, T. L. Wilson, and J. H. Reynolds. 2022. Distance sampling surveys: using components of detection and total error to select among approaches. Wildlife Monographs 210:e1070. https://doi.org/10.1002/wmon.1070

Stanley, T. R., and S. K. Skagen. 2007. Estimating the breeding population of Long-billed Curlew in the United States. Journal of Wildlife Management 71:2556–2564. https://doi.org/10.2193/2007-023

Stein, B. A., C. Scott, and N. Benton. 2008. Federal lands and endangered species: the role of military and other federal lands in sustaining biodiversity. BioScience 58:339–347. https://doi.org/10.1641/B580409

Stocking, J., E. Elliott-Smith, N. Holcomb, and S. M. Haig. 2010. Long-billed curlew breeding success on Mid-Columbia River National Wildlife Refuges, south-central Washington and north-central Oregon, 2007–08. Open-File Report 2010-1089. U.S. Geological Survey, Reston, Virginia, USA. https://pubs.usgs.gov/of/2010/1089/pdf/ofr20101089.pdf

Tazik, D. J., and C. O. Martin. 2002. Threatened and endangered species on U.S. Department of Defense lands in the arid west, USA. Arid Land Research and Management 16:259–276. https://doi.org/10.1080/153249802760284801

U.S. Fish and Wildlife Service (USFWS). 2021. Birds of conservation concern 2021. Migratory Bird Program, U.S. Fish and Wildlife Service, Washington, D.C., USA. https://www.fws.gov/sites/default/files/documents/birds-of-conservation-concern-2021.pdf

U.S. Geological Survey (USGS). 2015. The national map. 3D elevation program. U.S. Geological Survey, Reston, Virginia, USA. https://www.usgs.gov/programs/national-geospatial-program/national-map

Warren, S. D., S. W. Holbrook, D. A. Dale, N. L. Whelan, M. Elyn, W. Grimm, and A. Jentsch. 2007. Biodiversity and the heterogeneous disturbance regime on military training lands. Restoration Ecology 15:606–612. https://doi.org/10.1111/j.1526-100X.2007.00272.x

Western Regional Climate Center. 2022. Boardman, Oregon (350858): total of precipitation and average of average temperature. Western Regional Climate Center, Reno, Nevada, USA. https://wrcc.dri.edu/cgi-bin/cliMAIN.pl?or0858

Zentelis, R., and D. Lindenmayer. 2015. Bombing for biodiversity – enhancing conservation values of military training areas. Conservation Letters 8:299–305. https://doi.org/10.1111/conl.12155