Research Paper

INTRODUCTION

Animals face survival and reproductive consequences depending on the type and quality of habitat and resources that they use (Matthiopoulos et al. 2015). Resources include biotic and abiotic items, such as food or nesting material, and space needed for reproduction and survival (Johnson 1980, Aarts et al. 2008, Beyer et al. 2010, Gaillard et al. 2010). Theory predicts that individuals should use resources and thus habitat patches in such a way that their fitness is maximized (Fretwell 1969, Rosenzweig 1981, Northrup et al. 2022). Their requirements will influence how they move across the landscape, and movements will be further impacted by individual condition and the presence of other conspecifics or species (Avgar et al. 2020). Habitats or resources are selected when they are used disproportionately to their availability (Manly et al. 2002). An understanding of habitat use can inform management of populations, so habitat selection is of interest to managers and researchers.

An important consideration in habitat selection studies is that resource needs will change in seasonal environments, and animals will use different habitats depending on their stage in the annual cycle (Newton 2007, McLoughlin et al. 2010, Stanley et al. 2021). The seasonal changes in resource availability in time and space are ultimate reasons for the evolution of migratory behavior, particularly in birds (Cristol et al. 1999, Chapman et al. 2011). Migrants respond to environmental and habitat changes on their wintering and breeding areas to, in part, initiate migration, while stopover areas are used along the migration route and influence migratory decisions and duration (Newton 2007, Warnock 2010). Stopover areas are used for replenishing energy reserves through foraging and resting and sometimes as refugia from harsh conditions or predators (Schmaljohann et al. 2022). The energy acquired at stopover areas is critical for completing migration and, for some species, reproducing successfully during the following breeding season (Alisauskas 2002, Baker et al. 2004, Drent et al. 2007, Devries et al. 2008). Managing habitat or resources associated with energy acquisition and maintenance is of concern to many managers in key stopover areas for species of interest (Albanese and Davis 2015, Ballard et al. 2021).

In North America, there are six migratory populations of Sandhill Cranes (Antigone canadensis) managed by state and federal agencies (Collins et al. 2016, Gerber et al. 2020), and all of them use stopover areas during autumn and spring migration (Drewien and Bizeau 1974, Krapu et al. 2011, Fronczak et al. 2017). Several populations show high fidelity to certain stopover areas each year and season, and these areas have been the target of research and management (e.g., Krapu et al. 1984, Iverson et al. 1987, Bunting et al. 2022). For example, the San Luis Valley in Colorado is an important and well-known spring and autumn stopover area for Sandhill Cranes of the Rocky Mountain Population (RMP; Drewien and Bizeau 1974, Donnelly et al. 2021, Vanausdall et al. 2024). Sandhill Cranes rapidly increase lipid levels in spring by foraging on grain leftover from harvest, and these reserves are used to fuel the remainder of migration (Kauffeld 1981, Krapu et al. 1985) and, in part, assist in reproduction and egg laying (Krapu et al. 1985). Because of their status as a game species, a charismatic species, and a success story in terms of recovery from near-extirpation (Walkinshaw 1949), Sandhill Cranes and their use of stopover areas have been of interest in recent decades.

The habitat requirements of Sandhill Cranes during the non-breeding period have been well documented, with a particular emphasis on spring migration owing to its potential impact on reproduction (Lovvorn and Kirkpatrick 1982, Krapu et al. 1984, Boggie et al. 2018). Slightly less work has examined habitat selection for Sandhill Cranes in autumn (but see Littlefield 1992, Donnelly et al. 2021), but there is evidence to suggest that preferred habitat in autumn is more variable among individuals when compared to spring (Lovvorn and Kirkpatrick 1982). The distribution of Sandhill Cranes is largely determined by the proximity of roosting areas and foraging areas (Iverson et al. 1987, Sparling and Krapu 1994). Roosting habitats at stopover areas include wetlands, wet meadows, and riverine habitats with shallow water (approximately 30 cm deep or less) and low vegetation, channel widths wider than 200 m, and at least 25 m from disturbance (Kinzel et al. 2009, Kruse et al. 2017, Pearse et al. 2017). Sandhill Cranes will also use wetlands and pastures in between foraging bouts to loaf, which refers to behaviors such as resting, preening, and overall maintenance (Iverson et al. 1987, Tacha et al. 1987, Aborn 2010). Foraging habitat tends to include agricultural grains, but habitats such as wetlands, grasslands, and wet pastures are also important for providing other nutrients and minerals (Armbuster and Farmer 1982, Reinecke and Krapu 1986, Rowland et al. 1992). Sandhill Cranes show a preference for foods high in carbohydrates, such as corn and barley during migration, which allows for the synthesis and deposition of lipids (Krapu et al. 1985, Reinecke and Krapu 1986). Other crops, such as potatoes and alfalfa shoots, are also consumed in some areas (Kauffeld 1981, Armbuster and Farmer 1982, Walker and Schemnitz 1985). Sandhill Cranes will also forage on invertebrates and other plant parts during migration and these items provide protein, calcium, minerals, and other nutrients that agricultural crops do not (Lewis 1979, Walker and Schemnitz 1985, Reinecke and Krapu 1986).

The San Luis Valley (hereafter, SLV) in south-central Colorado is of key interest for the RMP because most of the population is in the SLV during peak migration in spring and autumn (Drewien and Bizeau 1974), and it is an area critical to the migratory connectivity of the RMP (Donnelly et al. 2021). The RMP is composed solely of the greater Sandhill Crane (A. c. tabida), but lesser Sandhill Cranes (A. c. canadensis) from the Midcontinent Population also use this area during autumn and spring migration (Krapu and Brandt 2008, Krapu et al. 2011). A challenge facing the management of the RMP is the increasingly limited water resources due to over-appropriation, prolonged drought, and warming temperatures (Barnett et al. 2005, 2008, Mix et al. 2011, Dettinger et al. 2015). Irrigated agriculture is an important economic driver in the SLV (San Luis Valley Development Resources Group and Council of Governments 2023), and most fields are irrigated using surface water runoff from the surrounding mountain ranges, along with water pumped from underground aquifers. Because of increased groundwater pumping in response to prolonged drought and lowered aquifer levels in the early 2000s, groundwater subdistricts were created and tasked with bringing groundwater levels back to a sustainable level (Rio Grande Basin Roundtable 2015, 2022, Cody et al. 2015). An irrigation season also limits the time that most irrigators can divert surface water, with diversions generally restricted to between 1 April and 31 October for most users (Rio Grande Basin Roundtable 2015). At the same time, public areas in the SLV distribute water on the landscape for migrating waterbirds, including Sandhill Cranes. The Monte Vista National Wildlife Refuge has adjudicated water rights to pump and divert water outside the irrigation season to provide roosting and foraging habitat for migrating Sandhill Cranes and waterbirds (U.S. Fish and Wildlife Service 2015). As a result, water availability differs spatially and temporally across the landscape in the SLV and can be dependent on whether property is privately or publicly owned. Temperatures are predicted to continue to increase in many parts of the Intermountain West, which will further reduce snowpack, the main provider of surface water for basins in the Southwest (Llewellyn and Vaddey 2013, Bureau of Reclamation 2021). Earlier spring runoff has already been documented (Barnett et al. 2008, Llewellyn and Vaddey 2013), which has the potential to create a mismatch between migration patterns and water availability in the SLV (Donnelly et al. 2019). Efficient management of water resources at a landscape scale on both public and private land will thus be important for migrating RMP Sandhill Cranes.

Examination of RMP habitat selection patterns at a key stopover area will provide a better understanding of where and when Sandhill Cranes most utilize habitat and inform how habitat could be managed to optimize limited resources (e.g., water) during the migratory period. Given continued water limitations, it would also be informative to understand the importance of publicly and privately managed lands to Sandhill Cranes, as the habitat management of each will require different strategies and include different stakeholders. We used data from global positioning system (GPS) transmitters fitted to RMP Sandhill Cranes between 2015 and 2022 to (1) identify habitat attributes and landcover types that influence selection of roosting, loafing, and foraging Sandhill Cranes in the SLV, (2) determine if there is a preference for private or public lands for roosting and loafing, and (3) identify areas with the strongest selection strength throughout the SLV. For the first objective, we predicted that surface water would have a positive influence on selection by both roosting and loafing RMP Sandhill Cranes. We expected distance to grain and surrounding vegetation height to be negatively correlated with roosting and loafing selection, with a stronger selection strength in autumn for distance to grain because of the depletion of grain resources closer to roosting and loafing sites over time. We also expected Sandhill Cranes to use a greater diversity of landcover types while loafing than while roosting, with a greater selection strength for pastures and grasslands for loafing and for open water and wetlands for roosting. In terms of foraging, we expected barley to be the crop type with the greatest selection strength in both seasons. For the second objective, we predicted Sandhill Cranes to use private areas (i.e., where more water is likely to be available) more than public areas for roosting and loafing because of limited water resources on public lands.

METHODS

Study area

The SLV is in the Upper Rio Grande Basin in southcentral Colorado, situated between the San Juan Mountains to the west and the Sangre De Cristo Mountains to the east (Fig. 1). The area of the valley floor is approximately 8200 km² and the average elevation is 2100 m (Emery et al. 1969). The average temperature is 6 °C but it can range between -20 °C and 20 °C throughout the year (Western Regional Climate Center 2013). Winds primarily come from the southwest and are highly variable. Considered an arid region, the SLV receives less than 20 cm of precipitation annually, while the surrounding mountains receive 115–127 cm per year (Cooper et al. 2006). Hydrology in the SLV is driven by runoff from snowpack in the mountains, along with water coming from and discharging to underground aquifers (Emery et al. 1969, Bexfield and Anderhold 2010). The headwaters of the Rio Grande are in the San Juan Mountains, and runoff from this mountain range provides 80% of the total runoff for the entire Upper Rio Grande Basin (Elias et al. 2015). Peak runoff occurs between April and June, but this has been shifting to earlier in the year in the Upper Rio Grande Basin and other parts of the Intermountain West (Barnett et al. 2008, Ficklin et al. 2013, Chavarria and Gutzler 2018). Monsoonal rains in the summer months can provide additional water, but surface water availability generally declines in autumn.

Agriculture is an important economic industry in the SLV. The main crops grown include alfalfa, potatoes, and barley. The main source of irrigation water in the Upper Rio Grande Basin comes from the Rio Grande but use in the last century has led to decreased water levels and depleted groundwater supplies utilized to compensate (Rio Grande Basin Roundtable 2022). Wetlands and wet meadows were historically more abundant throughout the SLV, but now large wetland complexes are mostly restricted to state and public lands (Beeton and Johnson 2020). Cottonwoods (Populus deltoides) and willows (Salix sp.) occur along riparian areas, and emergent vegetation, such as cattails (Typha sp.) and bulrush (Scirpus sp.), occur in wetlands. Wet meadows include moist, short grass areas near riverine habitat or as part of wetlands, and many occur in areas used for ranching, which is also common in the SLV.

GPS deployment and formatting

Sandhill Cranes were captured and fitted with global system for mobile communication (GSM) platform transmitter terminal (PTT) tags (n = 95; Evolution Series-400, 15-g; Cellular Tracking Technologies, Rio Grande, NJ, USA) or GPS PTT tags (n = 38; PTT-100, 22-g Solar Argos/GPS PTT, Microwave Telemetry, Columbia, MD, USA) between 2015 and 2022. Most Sandhill Cranes were captured in wintering areas in New Mexico at the Bosque del Apache National Wildlife Refuge and Bernardo Wildlife Area in Valencia and Socorro Counties, but a few were captured in Arizona, Colorado, Idaho, and Montana. Details on the capture methods can be found in Collins et al. (2016) and Boggie et al. (2018). Data were acquired at different rates between the two types of transmitters, as the original acquisition and use of the transmitters were for different projects. The GSM units acquired locations at 15- to 30-min intervals, and we resampled these to approximately two-hour intervals. The GPS units were programmed to acquire points at scheduled times (07:00, 08:00, 10:00, 14:00, 18:00, 24:00) and applied to data gathered between the spring of 2015 and the spring of 2017. The location error for the GSM units was ±22 m for the GSM units and ±18 m for the GPS units.

Sandhill Cranes in the SLV have predictable movement patterns during the autumn and the spring and this influenced how we formatted GPS data. Their behavior patterns at different times of the day, namely roosting, loafing, and foraging, are also distinct in terms of behavior and, generally, habitat preference. We separated locations into either roosting, loafing, or foraging, depending on the time of day (Table 1) and analyzed models separately for each behavior (Suraci et al. 2019). Sandhill Cranes roost along rivers and in wetlands at night and fly to foraging locations around sunrise to begin feeding. After about 2–3 hours, flocks will fly to nearby wetlands, ditches, pastures, or moist grassy areas to loaf, where they rest, preen, drink water, and socialize with some occasional foraging. Approximately 2–3 hours before sunset, they will go back to foraging areas before returning to a roost around sunset. These patterns were observed in the field, but we also examined trajectories of Sandhill Cranes with transmitters to confirm their movement patterns. Specifically, we examined the hourly average movement of Sandhill Cranes and identified times of the day when their movements were minimal (i.e., when they were roosting and loafing; see Appendix Fig. 1 for an example). We considered the roosting period to be between one hour after sunset and 1.5 hours before sunrise. However, Sandhill Cranes moved an average of 140.52 m (SE = 6.65 m) in autumn and 36.27 m (SE = 1.72 m) in spring per night, indicating that movement is minimal once they choose a roost. As a result, we used one location per crane per night for each season. We used the location acquired at or closest to midnight. We defined the foraging period to be between sunrise and two hours after sunrise and between two hours before sunset to sunset. The time in between was the loafing period, but we found considerable variation in the average distance moved during this period. The average distance between loafing locations within a day was 353.85 m (SE = 14.57 m) in autumn and 670.95 m (SE = 25.37 m) in spring. Sandhill Cranes were sometimes observed moving to different loafing locations during this time frame, which likely explains the greater movement between these locations. Additionally, the difference in distances between the seasons is likely due to within-season and individual variation in the timing of Sandhill Crane departure from and arrival to foraging locations. As a result, we decided to use one loafing location per day per individual at or closest to the time of day when movement was minimal, which was around noon. The remaining locations were considered foraging locations, and we resampled these locations to every two hours. The average distance between foraging locations within a day was 971.21 m (SE = 10.03 m) in autumn and 2505.58 m (SE = 39.77 m) in spring.

Habitat covariates

To estimate surface water and sandbar for roosting and loafing areas, we used constrained spectral mixture analysis (SMA), which is a technique that uses remotely sensed images to determine the proportional amount of habitat classes within image pixels (Adams et al. 1995, Adams and Gillespie 2006). We used Landsat 8 and 9 satellite imagery corrected for surface reflectance for the years 2015–2022. Landsat images have a spatial resolution of 30 x 30 m, so our estimates of water and sandbar are the proportion of a pixel classified as those habitat types. To determine the fractional amount of habitat classes within pixels of an image, SMA requires reference endmembers, which are representative areas composed of pure or well-characterized habitat classes. All pixels must also be classified to a habitat type, but it is recommended to not use more habitat classes than the number of wavelength bands in an image (Adams and Gillespie 2006). We chose the classes shrub, soil, sandbar, water, and vegetation, but our main interests were in water and sandbar. For surface water and vegetation, endmembers were generated using the median band values of images for each season and year. Although Landsat images are available approximately every 16 days, there can be considerable variation in the quality of the images because of cloud cover and other factors. To account for this variation, we chose to take the median of all images for each season and year during the periods between 1 October and 15 November for autumn and 15 February and 1 April for spring. We used the 98th percentile of Normalized Difference Water Index (NDWI) values to generate endmembers for each season-year. We used a similar approach for vegetation using the Normalized Difference Vegetation Index (NDVI). The endmembers for the remaining habitat classes were determined using static endmembers. Based on examination of various areas in the SLV, we chose polygons that did not significantly change in band values throughout the study period. The proportion of water and sandbar in each 30 x 30 m pixel was used in the resource selection analysis. We completed the SMA in the publicly available Google Earth Engine (Gorelick et al. 2017).

Vegetation height was also calculated using remote sensing techniques. We used Light Detection and Ranging (LiDAR) data collected in the SLV in 2011. To determine vegetation height, we generated a digital elevation model (DEM) and a digital surface model (DSM) at a spatial resolution of 10 m from the LiDAR data. The DEM was subtracted from the DSM to get the height of objects. To smooth out the resulting layer, we used a 3 x 3 cell window around each cell and calculated the majority height value within each window. We excluded urban areas from this analysis. We then estimated the mean vegetation height within a circular buffer with a 100 m radius around each used and available crane location (Boggie et al. 2018).

Distance to the nearest grain field was identified using data from the Rio Grande Water Conservation District (RGWCD) and the Cropland data layer from the National Agricultural Statistics Service (NASS). The RGWCD completes an annual census of crop types in the counties of Rio Grande, Alamosa, and Saguache. We used the Cropland data layer for Costilla County. We considered all the most dominant grain types (i.e., barley, rye, triticale, and wheat) and other grains (e.g., corn) that were not as prevalent. We also identified the crop types used by foraging Sandhill Cranes and included the classes: potatoes, alfalfa, grains, pasture, and other crops.

We classified used and available locations for the roosting and loafing period into other landcover types and into public or private property. Public lands in the SLV included areas owned by U.S. Fish and Wildlife National Wildlife Refuges, Bureau of Land Management, and Colorado Parks and Wildlife. We used both the RGWCD and the National Land Cover Database (NLCD) from 2016 (Dewitz 2019) and 2019 (Dewitz 2021) to determine landcover type: grassland, open water, pasture, wetland, and other. We used the 2016 NLCD data layer for crane locations from the years 2015 to 2017 and the 2019 NLCD layer for the remainder of the years.

Resource selection functions and habitat availability

We designed our study based on a use-availability framework to develop resource selection functions (RSF). A RSF estimates the selection strength by comparing a set of used locations, which were defined above, to a set of available locations (Boyce and McDonald 1999, Manly et al. 2002). Habitats or resources that are used disproportionately to availability are “selected” by that animal (Johnson 1980, Boyce et al. 2002, Manly et al. 2002). To define the available distribution, we examined the average and median distances that RMP Sandhill Cranes moved between roosting or loafing locations on days for each season (Appendix Table 1). Distances were heavily skewed right for both periods and seasons. Although the median values indicated that individuals typically moved within 1 km of roosting or diurnal locations between consecutive days, the average values are influenced by cranes that moved greater distances as they likely continued to migrate south or north in the SLV. We wanted to encapsulate their ability to move greater distances between days, so we used a 5-km buffer around each used location to define the available area for all periods. Although individuals moved less between roosting and loafing areas in autumn than in spring, we decided to keep the available area consistent between seasons. We further restricted the available area for roosting and loafing locations to only include areas defined by a wetland polygon layer that delineated wetland, riparian, and wet meadow features (Fig. 1). Restricting the sample space in this way excluded areas not typically used by roosting or loafing Sandhill Cranes (e.g., crop fields) and improves estimation of habitat features in wetland or similar habitat types, which may have a different spectral signature than crop types, urban areas, or other habitat types. For each used location, individual, season, and year, we randomly placed 20 available locations within the 5-km buffered region (Pearse et al. 2021).

We built RSFs using generalized linear mixed effects models (GLMM) with random intercepts for each location and random slopes for each roosting and loafing individual (Muff et al. 2019). Including random intercepts accounted for the unequal number of used and available locations per individual (Gillies et al. 2006, Muff et al. 2019), while random slopes allowed for individual variation on selection (Gillies et al. 2006, Hebblewhite and Merrill 2008). For roosting and loafing periods, we examined models that included landcover types as dummy covariates (i.e., 1 if the location was the specified landcover type, 0 otherwise), along with ownership type as a dummy covariate (i.e., 1 if location was specified as private, 0 otherwise). We included the habitat covariates proportion of water, proportion of sandbar, vegetation height, and distance to the nearest grain field. Proportions refer to the proportion of water and sandbar in the 30 x 30 m pixel in which the used and available points were located. We examined separate models with a linear effect of water and a quadratic effect of water, but we did not include any other interactions. We used a GLMM with a Poisson distribution and a log link following Muff et al. (2019),

(1)

where π_nj is the probability that point t from the jth set of used and available locations from Sandhill Crane n is used, α_nj is the stratum-specific intercept for set j and for Sandhill Crane n where α_nj~N(0, σ²_α), β_n is a vector of coefficients for covariates (i.e., sandbar, water, vegetation height, and distance to grain) for Sandhill Crane n where β_n~N(0, σ²_β), x_ntj is a vector of covariate values at point t from set j and Sandhill Crane n, and ε_ntj is a residual error term where ε_ntj~N(0,σ²). Using a Poisson distribution with a log link has a likelihood equivalent to a conditional logistic regression model that uses matched set data, but GLMMs can more easily be fit with random slope effects (Muff et al. 2019). Our global model could be written as

(2)

where we omitted the denominator here for readability. We used Akaike’s Information Criterion (AIC; Akaike 1973, Burnham and Anderson 2002) to compare a model with a linear trend for water to a model with a quadratic trend for water. We opted for examining models with all covariates because these habitat covariates are known to influence selection by Sandhill Cranes, and we were interested in the relative selection strength across landcover and ownership type. To examine covariate effects on selection strength, we removed the intercept from the top model, applied the coefficients to the observed values of the covariate of interest while holding the other covariates at their mean (i.e., zero), and exponentiated (Muff et al. 2019, Fieberg et al. 2021). We used the delta method to estimate the variance of the prediction. Next, we used a method described by Fieberg et al. (2021) to calculate the likelihood of a Sandhill Crane selecting one landcover type over another. Specifically, we calculated selection (above) across all observed covariate values and combinations, summed this value across all locations for the landcover type of interest, and divided it by the sum of selection for each of the other landcover types. Finally, using the population-level coefficient estimates to identify areas in the SLV where RMP Sandhill Cranes are most likely to roost and loaf, we projected our models to the most recent year (2022) of environmental data across the SLV. We applied each of the top models for roosting and loafing to the entirety of the SLV and identified areas with relatively high selection strength.

We built a separate, single model for foraging selection. Because we did not have continuous, measurable covariates for foraging areas, we only examined how crop type (alfalfa, potatoes, barley, rye, triticale, wheat, wetland, pasture, other crops) influenced selection in autumn and spring. Our model for foraging selection was

(3)

where the denominator is again omitted for readability. We predicted that grains would have the greatest selection strength for both seasons, with barley being the highest, followed by other grains, and then potatoes, alfalfa, and pastures.

Model validation

We used methods from Boyce et al. (2002) to ensure that our roosting and loafing models had good predictive capabilities using covariates from the top models. We used cross-validation by splitting the dataset into training and testing datasets, where 80% of the data was randomly assigned to the training dataset and the remainder was assigned to the testing dataset. The training data were used in the top model for each season, and the resulting model generated predictions using the testing data (Boyce et al. 2002, Fieberg et al. 2018). This was repeated for 1000 iterations. We then examined the correlation between the expected and the observed observations and considered a value > 0.90 to indicate a model with adequate predictive capabilities. We did not complete predictive checks for foraging models because we were only able to examine one covariate on selection and did not have continuous covariates.

Functional response

We used methods described by Holbrook et al. (2019) to examine if use of roosting and loafing areas changed with availability on public or private lands. We first converted proportion of water to area by multiplying NDWI values by 900 m² (i.e., area of a 30 x 30 m pixel), and we summed the total area each for used and available points on public and private lands for all individual Sandhill Cranes. We used approach 2 described in Holbrook et al. (2019) and fit a general linear model (GLM) with the log of water area at used points as the response variable and the log of water area at available points as the explanatory variable. We also included ownership type and fit both an additive model and a model with an interaction between water area and ownership type. Thus, our additive model was

(4)

where covariates were specific to point t. A slope of 1 indicates selection of water (or ownership type) proportional to its availability, but a significant deviation from 1 indicates increasing or decreasing selection as availability of water increases (i.e., a functional response).

RESULTS

Between 2015 and 2022, we used a total of 13,477, 17,953, and 10,357 locations for roosting, loafing, and foraging analyses, respectively, in autumn. We used a total of 29,218, 32,083, and 17,721 locations for roosting, loafing, and foraging analyses, respectively, in spring. Individual RMP Sandhill Cranes were monitored for an average of 2.10 autumn seasons (SD = 1.00 seasons) and 2.19 spring seasons (SD = 1.07 seasons; Table 1).

The roosting model with a quadratic effect of water performed better than the model with a linear effect for both seasons based on AIC (Appendix Table 2). The proportion of water had a positive effect on selection in both seasons (Fig. 2A, Fig. 3A), while the proportion of sandbar also had a positive effect (Fig. 2). Vegetation height had a negative effect on selection for both seasons, but there was a stronger effect in autumn than in spring (Fig. 2, Fig. 4A). Selection strength dropped to zero above approximately 2.5 m in height in autumn, while there appeared to be a greater tolerance for taller vegetation in spring. Distance to the nearest grain field had a negative effect on selection (Fig. 2, Fig. 5A), but the effect was stronger in autumn than in spring. Selection for private land was higher than for public land (Fig. 2). For both autumn and spring, RMP Sandhill Cranes were more likely to roost on open water than other landcover types, followed by wetland habitat (Table 2, Fig. 2).

The loafing model with a quadratic effect of water performed better than the model with a linear effect for both seasons based on AIC (Appendix Table 2). Covariate effects on selection for loafing areas followed similar patterns as for roosting except for proportion of water and landcover type (Fig. 2). In autumn, selection for the proportion of water peaked around 50% water, while it peaked around 30% water in spring (Fig. 3B). Vegetation height had a negative effect on selection for both seasons (Fig. 2, Fig. 4B). Distance to the nearest grain field also had a negative effect on selection similar to roosting, but selection dropped to zero more quickly (Fig. 2, Fig. 5B). Selection for private land was higher than for public land (Fig. 2). For both autumn and spring, RMP sandhill cranes were more likely to loaf on wetlands than other landcover types, followed by pastures (Table 2, Fig. 2).

Several areas showed a relatively high selection strength for both roosting and loafing in both seasons (Appendix Fig. 2). The MVNWR showed a higher area of locations with high selection strength than the reference private area along the Rio Grande in spring, but this pattern was less pronounced in autumn. Compared to roosting, there were fewer areas that showed a relatively high selection strength to loafing areas across the SLV.

From the foraging model, selection was highest for grain crops overall for both seasons, but there was a difference in the types of grain that had the highest selection strength between autumn and spring (Fig. 6). While barley had the highest selection strength in both seasons, other grains and triticale also had relatively high selection strengths in autumn. Selection for these grains was lower in spring. There was also higher variability among individuals in selection for grains other than barley. Compared to wheat and rye, potatoes, pasture, and alfalfa had higher selection strengths in both seasons.

Model validation

Overall, our models showed a relatively high predictive accuracy with the exception of the autumn roosting model. The correlation coefficient between the expected and observed RSF values for the roosting models were 0.66 and 0.92 for the autumn and spring models, respectively. The loafing models showed correlations of 0.94 for autumn and 0.95 for spring.

Functional response

We detected a functional response (i.e., slope coefficients deviated from 1) to water based on the ownership type for roosting selection in both autumn and spring but not for loafing selection (Fig. 7). In all cases except spring roosting, the model with an interaction between water area and ownership type did not improve model fit based on AIC values, so interpretations are from the additive models (Table 3). For autumn roosting, use was proportional to availability for both private and public land, but the rate of increase declined as availability increased. There was also higher use on private than public land (Fig. 7). We found that roost use also increased as water availability increased on public lands in spring, but use declined with increasing availability on private land, indicating avoidance of private lands and a stronger preference for public lands with increasing surface water. For the loafing period, we found little evidence of use as a function of availability. There was a slight apparent increase in the area used for private over public lands in autumn but no difference in spring.

DISCUSSION

Our results on the resource selection of RMP Sandhill Cranes in the SLV improve our understanding of the habitat features important to this population during autumn and spring migration. Understanding how Sandhill Cranes utilize water in the SLV helps identify where resources can be allocated to provide the most benefit throughout the year. In general, RMP Sandhill Cranes preferred to roost and loaf in wet areas with short vegetation and close to grain fields in both seasons. Roosting cranes showed a stronger selection for open water and wetlands, whereas wetlands and pasture were more preferred for loafing. There were also some differences in the strength of selection when comparing autumn to spring. For example, wetlands were preferred over other habitats for roosting in autumn, whereas there was greater preference for roosting in open water in spring. Wetlands were preferred for loafing in both seasons. Additionally, ownership type influenced selection and indicated that both public and private lands provide roosting and loafing habitat at different times of the year.

Consistent with other studies, vegetation, water, and sandbar influenced selection by roosting and loafing Sandhill Cranes in the SLV. As expected, surface water had an overall positive effect on selection by roosting Sandhill Cranes in both seasons, but selection strength peaked at around 25 to 50% for loafing areas. Surrounding vegetation heights below 2.5 m within 100 m of roosting and loafing locations had the highest selection strength in autumn, while there appeared to be a greater tolerance for taller (but still < 5 m) vegetation in spring. Other studies have identified areas with shallow water, shorter vegetation heights with wide channel or bank widths, and shallow water being selected by roosting Sandhill Cranes (Kauffeld 1981, Krapu et al. 1984, Folk and Tacha 1990, Pearse et al. 2017, Boggie et al. 2018). Loafing RMP Sandhill Cranes also preferred areas with shorter vegetation, but the combination of a positive effect of sandbar and a peak in the selection of proportion of water may indicate a preference for dryer conditions than for roosting areas. Our results indicate that managing the structural aspects of these habitats may not need to differ between seasons. Rather, focusing on providing habitats that meet both their roosting and loafing preferences (e.g., more surface water for roosting than for loafing) will optimize the allocation of resources.

Roosting and loafing Sandhill Cranes also showed preferences for certain landcover types and ownership types in the SLV. Most surface water in the SLV is on privately managed property and is especially limited in autumn compared to spring, when snowmelt provides an influx of available water from the surrounding mountains. Selection was higher for open water areas and for private areas overall for roosting Sandhill Cranes in both seasons, but selection for open water areas was slightly stronger in autumn than in spring. The water that is generally available in autumn tends to be permanent or semi-permanent waterbodies (Donnelly et al. 2021), which characterize open water landcover types. For loafing, pastures or wetland areas were more likely to be selected than open water, and private areas also had a higher selection strength than public areas. Sandhill Cranes have been documented roosting or loafing in areas identified as pasture, wet meadows, and grasslands in other studies (Krapu et al. 1984, Iverson et al. 1987, Davis 2003). Pastures in the SLV are generally characterized by shallow water or moist soil with relatively short vegetation. The reliance on habitat in irrigated pasture is not new for Sandhill Cranes, as their importance is evident during the nesting season in summer (Donnelly et al. 2024).

Although selection was higher for private areas, some public areas like the MVNWR manage water specifically for roosting and loafing Sandhill Cranes (U.S. Fish and Wildlife Service 2015, Wetland Dynamics 2019), which may be particularly beneficial for Sandhill Cranes during autumn. Our functional response analysis indicated that Sandhill Cranes in the SLV may be responding favorably to water management on public lands. Despite the high availability of water available on private areas in spring, there appears to be a preference for public areas for roosting RMP Sandhill Cranes, particularly in spring. Selection for loafing areas did not change with water availability, which may be because of less preference for large waterbodies for loafing than for roosting. These results may indicate that public areas play an important role in the roosting and loafing needs of RMP Sandhill Cranes, but private areas may be critical in preserving especially limited surface water used by Sandhill Cranes.

We found distance to grain fields to be influential for both roosting and loafing RMP Sandhill Cranes, with selection decreasing with increasing distance from grain fields. Having both roosting and loafing areas close to foraging areas can reduce energy expenditure during migration (Sparling and Krapu 1994, Jankowiak et al. 2015), which is a time when maintaining adequate energy reserves may be important for reproduction or survival (Ankney and MacInnes 1978, Alisauskas 2002, Devries et al. 2008). For autumn roosting and loafing and spring loafing, distances > 5 km showed a selection strength near zero, whereas the decline was more gradual for spring roosting. A distance of 5 km has been shown to be the maximum distance greater Sandhill Cranes travel between roosting or loafing areas and foraging areas (Sparling and Krapu 1994, Ivey et al. 2015, Boggie et al. 2018). Collins et al. (2023) also found that wintering greater Sandhill Cranes from the Lower Colorado River Valley Population flew farther (up to 10 km) to grain than other potential foraging sites. It may be that the higher tolerance for flying greater distances to foraging areas in spring than in autumn is due to both a higher energy demand and depleting resources near roosting areas. Sandhill Cranes rapidly increase their lipid reserves in spring in preparation for reproduction (Krapu et al. 1985, Reinecke and Krapu 1986). Additionally, departure from the SLV in spring requires Sandhill Cranes to fly over mountain ranges and fly greater distances, whereas there are fewer obstacles and a shorter total distance in autumn on their southbound migration from the SLV to their wintering areas, so energy requirements upon departure are likely lower. The effect of distance to grain fields was lower for choosing loafing areas than for roosting areas. Traveling shorter distances during the day, rather than going back to roosts, can help Sandhill Cranes reduce energy expenditure (Collins et al. 2023). Loafing areas allow Sandhill Cranes to forage on other food sources (e.g., invertebrates) that provide nutrients (e.g., protein, calcium) less available in agricultural foods (Reinecke and Krapu 1986, Ballard and Thompson 2000). Sandhill Cranes engage in other important behaviors during migration, such as courtship, preening, and drinking water, that are often observed more frequently during the loafing period (Tacha 1981, Tacha et al. 1987). Our results provide additional evidence that having adequate foraging areas nearby is important for both roosting and loafing RMP cranes in the SLV.

Grains were the dominant crop type used by foraging RMP Sandhill Cranes in both seasons, but barley was not the only important grain type, which was inconsistent with our original hypothesis. Autumn RMP Sandhill Cranes showed their highest selection strength for barley followed by triticale, while probability was highest for barley in spring. Grains provide a high amount of carbohydrates, which are used to synthesize the deposition of lipids and often dictate Sandhill Crane distribution throughout their range (Anteau et al. 2011, Boggie et al. 2023, Collins et al. 2023). The difference in the use of triticale in the SLV between spring and autumn is likely due to availability. Of the grains grown in the SLV, triticale makes up only approximately 1% of the total crop area (Rio Grande Water Conservation District 2022). Given the relatively low amount of triticale grown, these fields may be depleted soon after the arrival of Sandhill Cranes in autumn, leaving less triticale available the following spring. Preference for triticale may also be due to the chemical and physical properties of the grain type. Barley and triticale have similar energetic contents, but they can differ in carbohydrate content and other chemical characteristics (Rodehutscord et al. 2016, Siegert et al. 2022), which may make triticale more favorable than barley to Sandhill Cranes.

Although our study focused on the population-level decision patterns of RMP Sandhill Cranes, other methods exist that could elucidate individual-level behaviors related to habitat selection. Several covariates, including vegetation height and landcover type, showed high variation in effects across individuals, indicating that other factors may influence selection by RMP Sandhill Cranes. For example, if pairs of Sandhill Cranes are successful breeders, their offspring will generally stay with the parents until the following spring (Gerber et al. 2020), and adults with young in autumn may select habitat differently than those without young. Additionally, another important consideration is body condition, or the overall health of an individual bird. Body condition can influence habitat selection and overall stopover time, as birds may need to remain at a stopover until they reach a threshold in energetic content and may select habitats that more rapidly increase their lipid accumulation (Hedenström and Alerstam 1997, Alerstam and Hedenström 1998, Eikenaar and Schläfke 2013). Further examination of individual-level habitat use patterns in the future will add to the growing body of work on Sandhill Cranes land use patterns.

Finally, we recognize that there are limitations to our models and acknowledge several other methods exist to determine habitat selection. Our analysis and model building were specific to the SLV in Colorado, and caution should be taken if extrapolating the RSF to other areas. Additionally, we were not able to provide a measure of accuracy in our estimates of habitat covariates (e.g., surface water) derived using SMA. Several other studies have utilized SMA to determine habitat use by animals using similar methods and have reported a high degree of accuracy (Donnelly et al. 2020, Bunting et al. 2022). At the same time, our approach to validating our models showed a high degree of predictive ability in the SLV for spring roosting and both autumn and spring loafing Sandhill Cranes, which may indicate that our estimates for surface water and other covariates are suitable for predicting roosting and loafing areas for Sandhill Cranes in the SLV. However, the low predictive ability of the autumn roosting model could be due to missing covariates that may better explain roost selection during this time, such as disturbance from hunting for waterfowl or other covariates that may better explain roosting in autumn. There are several covariates that we did not consider because of a lack of data availability, but future work could assess the effects of attributes such as channel width and wetland type on habitat selection (Iverson et al. 1987, Pearse et al. 2017).

CONCLUSION

In summary, we identified habitat features important to the selection by RMP Sandhill Cranes for roosting, loafing, and foraging. We also found that ownership type was an important factor to consider, indicating that habitat characteristics and availability differ between public and private lands. Sandhill Cranes in the SLV are selecting for roosting and loafing areas with short vegetation (< 10 m), close to grain fields (< 5 km), and with > 25% open water in both autumn and spring and showed a preference for open water areas, wetlands, and pastures. Our functional response analysis also indicated that RMP Sandhill Cranes select more strongly for public areas for roosting as water becomes more available.

Understanding the factors that drive habitat selection is important because habitat quality and resources may influence fitness components of populations (Alisauskas 2002, Devries et al. 2008, Sedinger and Alisauskas 2014, Matthiopoulos et al. 2015). As a long-lived bird, Sandhill Cranes may be particularly susceptible to habitat changes that affect their ability to meet energetic demands during migration. It is also informative for land managers when resources are limited or costly to obtain and manage. Resources and jurisdictions are often limited for either private landowners or state and federal agencies managing property, which may reduce habitat availability for some species. Management on federal lands in the SLV, for example, is largely focused on Sandhill Cranes and waterfowl, but ranching and grazing greatly influence habitat on private lands even if water is available. Ultimately, we found that both public and private lands are important for the resource selection of RMP Sandhill Cranes, particularly for roosting and loafing cranes, and at different times of the year. Although we did not consider specific management actions, selection for public lands, particularly in spring, may indicate that management on public areas is providing the necessary resources migrating Sandhill Cranes need, particularly for roosting. At the same time, when roosting and loafing habitats are limited on public lands, private areas can potentially provide appropriate habitat conditions throughout the year.

LITERATURE CITED

Aarts, G., M. MacKenzie, B. McConnell, M. Fedak, and J. Matthiopoulos. 2008. Estimating space-use and habitat preference from wildlife telemetry data. Ecography 31:140-160. https://doi.org/10.1111/j.2007.0906-7590.05236.x

Aborn, D. A. 2010. Behavior and habitat use of greater sandhill cranes wintering in east Tennessee. Proceedings of the North American Crane Workshop 9-14.

Adams, J. B., and A. R. Gillespie. 2006. Remote sensing of landscapes with spectral images. Cambridge University Press, Cambridge, UK. https://doi.org/10.1017/CBO9780511617195

Adams, J. B., D. E. Sabol, V. Kapos, R. Almeida Filho, D. A. Roberts, M. O. Smith, and A. R. Gillespie. 1995. Classification of multispectral images based on fractions of endmembers: application to land-cover change in the Brazilian Amazon. Remote Sensing and Environment 52:137-154. https://doi.org/10.1016/0034-4257(94)00098-8

Akaike, H. 1973. Information theory and an extension of the maximum likelihood principle. Pages 276–281 in B. N. Petran and F. Csak, editors. Proceedings of the 2nd International Symposium on Information Theorys. Akadémiai Kiadó, Budapest, Hungary.

Albanese, G., and C. A. Davis. 2015. Characteristics within and around stopover wetlands used by migratory shorebirds: Is the neighborhood important? Condor 117:328-340. https://doi.org/10.1650/CONDOR-14-166.1

Alerstam, T., and A. Hedenström. 1998. The development of bird migration theory. Journal of Avian Biology 29:343-369. https://doi.org/10.2307/3677155

Alisauskas, R. T. 2002. Arctic climate, spring nutrition, and recruitment in Midcontinent Lesser Snow Geese. Journal of Wildlife Management 66:181-193. https://doi.org/10.2307/3802884

Ankney, C. D., and C. D. MacInnes. 1978. Nutrient reserves and reproductive performance of female Lesser Snow Geese. Auk 95:459-471.

Anteau, M. J., M. H. Sherfy, and A. A. Bishop. 2011. Location and agricultural practices influence spring use of harvested cornfields by cranes and geese in Nebraska. Journal of Wildlife Management 75:1004-1011. https://doi.org/10.1002/jwmg.135

Armbuster, M. J., and A. H. Farmer. 1982. Draft Sandhill Crane habitat suitability model. Proceedings 1981 Crane Workshop 136-143.

Avgar, T., G. S. Betini, and J. M. Fryxell. 2020. Habitat selection patterns are density dependent under the ideal free distribution. Journal of Animal Ecology 89:2777-2787. https://doi.org/10.1111/1365-2656.13352

Baker, A. J., P. M. González, T. Piersma, L. J. Niles, I. De Lima Serrano Do Nascimento, P. W. Atkinson, N. A. Clark, C. D. T. Minton, M. K. Peck, and G. Aarts. 2004. Rapid population decline in red knots: fitness consequences of decreased refuelling rates and late arrival in Delaware Bay. Proceedings of the Royal Society B: Biological Sciences 271:875-882. https://doi.org/10.1098/rspb.2003.2663

Ballard, B. M., and J. E. Thompson. 2000. Winter diets of sandhill cranes from central and coastal Texas. Wilson Bulletin 112:263-268. https://doi.org/10.1676/0043-5643(2000)112[0263:WDOSCF]2.0.CO;2

Ballard, D. C., O. E. Jones, and A. K. Janke. 2021. Factors affecting wetland use by spring migrating ducks in the Southern Prairie Pothole Region. Journal of Wildlife Management 85:1490-1506. https://doi.org/10.1002/jwmg.22096

Barnett, T. P., J. C. Adam, and D. P. Lettenmaier. 2005. Potential impacts of a warming climate on water availability in snow-dominated regions. Nature 438:303-309. https://doi.org/10.1038/nature04141

Barnett, T. P., D. W. Pierce, H. G. Hidalgo, C. Bonfils, B. D. Santer, T. Das, G. Bala, A. W. Wood, T. Nozawa, A. A. Mirin, D. R. Cayan, and M. D. Dettinger. 2008. Human-induced changes in the hydrology of the western United States. Science 319:1080-1083. https://doi.org/10.1126/science.1152538

Beeton, J. M. and B. G. Johnson. 2020. Geomorphology and climate change in the San Luis Valley. Pages 58-80 in J. M. Beeton, C. N. Saenz, and B. J. Waddell, editors. The geology, ecology, and human history of the San Luis Valley. University Press of Colorado, Louisville, Colorado, USA. https://doi.org/10.5876/9781646420407.c002

Bexfield, L. M., and S. K. Anderholm. 2010. Conceptual understanding and groundwater quality of the basin-fill aquifer in the San Luis Valley, Colorado and New Mexico. Pages 13-24 in S. A. Thiros, L. M. Bexfield, D. W. Anning, and J. M. Huntington, editors. Conceptual understanding and groundwater quality in selected basin-fill aquifers in the southwestern United States. Professional Paper 1781. U.S. Geological Survey, Reston, Virginia, USA.

Beyer, H. L., D. T. Haydon, J. M. Morales, J. L. Frair, M. Hebblewhite, M. Mitchell, and J. Matthiopoulos. 2010. The interpretation of habitat preference metrics under use-availability designs. Philosophical Transactions of the Royal Society B: Biological Sciences 365:2245-2254. https://doi.org/10.1098/rstb.2010.0083

Boggie, M. A., D. P. Collins, and S. A. Carleton. 2023. Availability of supplemental corn for Sandhill Cranes, light geese, and dabbling ducks wintering in New Mexico. Journal of Fish and Wildlife Management 14:51-61. https://doi.org/10.3996/JFWM-22-013

Boggie, M. A., D. P. Collins, J. P. Donnelly, and S. A. Carleton. 2018. Land use, anthropogenic disturbance, and riverine features drive patterns of habitat selection by a wintering waterbird in a semiarid environment. PLoS ONE 13: e0206222. https://doi.org/10.1371/journal.pone.0206222

Boyce, M. S., and L. L. McDonald. 1999. Relating populations to habitats using resource selection functions. Trends in Ecology and Evolution 14:268-272. https://doi.org/10.1016/S0169-5347(99)01593-1

Boyce, M. S., P. R. Vernier, S. E. Nielsen, and F. K. A. Schmiegelow. 2002. Evaluating resource selection functions. Ecological Modelling 157:281-300. https://doi.org/10.1016/S0304-3800(02)00200-4

Bunting, D. P., M. A. Boggie, D. P. Collins, P. P. Thorpe, and J. P. Donnelly. 2022. Linking ecological processes and animal movements to inform timing of long-term surveys of a migratory game bird. Ecosphere 13:e4298. https://doi.org/10.1002/ecs2.4298

Bureau of Reclamation. 2021. West-wide climate and hydrology assessment. U.S. Department of the Interior, Albuquerque, New Mexico, USA.

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

Chapman, B. B., C. Brönmark, J. Å. Nilsson, and L. A. Hansson. 2011. The ecology and evolution of partial migration. Oikos 120:1764-1775. https://doi.org/10.1111/j.1600-0706.2011.20131.x

Chavarria, S. B., and D. S. Gutzler. 2018. Observed changes in climate and streamflow in the Upper Rio Grande Basin. Journal of the American Water Resources Association 54:644-659. https://doi.org/10.1111/1752-1688.12640

Cody, K. C., S. M. Smith, M. Cox, and K. Andersson. 2015. Emergence of collective action in a groundwater commons: irrigators in the San Luis Valley of Colorado. Society and Natural Resources 28:405-422. https://doi.org/10.1080/08941920.2014.970736

Collins, D. P., M. A. Boggie, K. L. Kruse, C. M. Conring, J. P. Donnelly, W. C. Conway, and B. A. Grisham. 2023. Roost sites influence habitat selection of Sandhill Cranes (Antigone canadensis tabida) in Arizona and California. Waterbirds 45:428-439. https://doi.org/10.1675/063.045.0407

Collins, D. P., B. A. Grisham, C. M. Conring, J. M. Knetter, W. C. Conway, S. A. Carleton, and M. A. Boggie. 2016. New summer areas and mixing of two greater Sandhill Crane populations in the Intermountain West. Journal of Fish and Wildlife Management 7:141-152. https://doi.org/10.3996/042015-JFWM-036

Cooper, D. J., J. S. Sanderson, D. I. Stannard, and D. P. Groeneveld. 2006. Effects of long-term water table drawdown on evapotranspiration and vegetation in an arid region phreatophyte community. Journal of Hydrology 325:21-34. https://doi.org/10.1016/j.jhydrol.2005.09.035

Cristol, D. A., M. B. Baker, and C. Carbone. 1999. Differential migration revisited: latitudinal segregation by age and sex class. Pages 33-88 in V. Nolan, E. D. Ketterson, and C. F. Thompson, editors. Current ornithology. Springer, New York, New York, USA. https://doi.org/10.1007/978-1-4757-4901-4_2

Davis, C. A. 2003. Habitat use and migration patterns of Sandhill Cranes along the Platte River, 1998-2001. Great Plains Research 13:199-216.

Dettinger, M., B. Udall, and A. Georgakakos. 2015. Western water and climate change. Ecological Applications 25:2069-2093. https://doi.org/10.1890/15-0938.1

Devries, J. H., R. W. Brook, D. W. Howerter, and M. G. Anderson. 2008. Effects of spring body condition and age on reproduction in Mallards (Anas platyrhynchos). Auk 125:618-628. https://doi.org/10.1525/auk.2008.07055

Dewitz, J. 2019. National Landcover Database (NLCD). U.S. Geological Survey, Reston, Virginia, USA.

Dewitz, J. 2021. National Landcover Database (NLCD). U.S. Geological Survey, Reston, Virginia, USA.

Donnelly, J. P., D. P. Collins, J. M. Knetter, J. H. Gammonley, M. A. Boggie, B. A. Grisham, M. C. Nowak, and D. E. Naugle. 2024. Flood-irrigated agricultural mediates climate-induced wetland scarcity for summering Sandhill Cranes in western North America. Ecology and Evolution 14:e10998. https://doi.org/10.1002/ece3.10998

Donnelly, J. P., S. L. King, J. Knetter, J. H. Gammonley, V. J. Dreitz, B. A. Grisham, M. C. Nowak, and D. P. Collins. 2021. Migration efficiency sustains connectivity across agroecological networks supporting Sandhill Crane migration. Ecosphere 12:e03543. https://doi.org/10.1002/ecs2.3543

Donnelly, J. P., S. L. King, N. L. Silverman, D. P. Collins, E. M. Carrera-Gonzalez, A. Lafón-Terrazas, and J. N. Moore. 2020. Climate and human water use diminish wetland networks supporting continental waterbird migration. Global Change Biology 26:2042-2059. https://doi.org/10.1111/gcb.15010

Donnelly, J. P., D. E. Naugle, D. P. Collins, B. D. Dugger, B. W. Allred, J. D. Tack, and V. J. Dreitz. 2019. Synchronizing conservation to seasonal wetland hydrology and waterbird migration in semi-arid landscapes. Ecosphere 10:e02758. https://doi.org/10.1002/ecs2.2758

Drent, R. H., G. Eichhorn, A. Flagstad, A. J. Van der Graaf, K. E. Litvin, and J. Stahl. 2007, December. Migratory connectivity in Arctic geese: spring stopovers are the weak links in meeting targets for breeding. Journal of Ornithology 148:501-514. https://doi.org/10.1007/s10336-007-0223-4

Drewien, R. C., and E. G. Bizeau. 1974. Status and distribution of greater Sandhill Cranes in the Rocky Mountains. Journal of Wildlife Management 38:720-742. https://doi.org/10.2307/3800039

Eikenaar, C., and J. L. Schläfke. 2013. Size and accumulation of fuel reserves at stopover predict nocturnal restlessness in a migratory bird. Biology Letters 9(6):20130712. https://doi.org/10.1098/rsbl.2013.0712

Elias, E. H., A. Rango, C. M. Steele, J. F. Mejia, and R. Smith. 2015. Assessing climate change impacts on water availability of snowmelt-dominated basins of the Upper Rio Grande basin. Journal of Hydrology: Regional Studies 3:525-546. https://doi.org/10.1016/j.ejrh.2015.04.004

Emery, P. A., A. J. Boettcher, R. J. Snipes, and H. J. Mcintyre. 1969. Hydrology of the San Luis Valley, south-central Colorado. U.S. Department of the Interior, Geological Survey, Water Resources Division, Denver, Colorado, USA. https://doi.org/10.3133/70045464

Ficklin, D. L., I. T. Stewart, and E. P. Maurer. 2013. Climate change impacts on streamflow and subbasin-scale hydrology in the Upper Colorado River Basin. PLoS ONE 8:e71297. https://doi.org/10.1371/journal.pone.0071297

Fieberg, J. R., J. D. Forester, G. M. Street, D. H. Johnson, A. A. ArchMiller, and J. Matthiopoulos. 2018. Used-habitat calibration plots: a new procedure for validating species distribution, resource selection, and step-selection models. Ecography 41:737-752. https://doi.org/10.1111/ecog.03123

Fieberg, J., J. Signer, B. Smith, and T. Avgar. 2021. A ‘how to’ guide for interpreting parameters in habitat-selection analyses. Journal of Animal Ecology 90:1027-1043. https://doi.org/10.1111/1365-2656.13441

Folk, M. J., and T. C. Tacha. 1990. Sandhill Crane roost site characteristics in the North Platte River Valley. Journal of Wildlife Management 54:480-486. https://doi.org/10.2307/3809662

Fretwell, S. D. 1969. On territorial behavior and other factors influencing habitat distribution in birds. Institute of Statistics, Mimeograph Series No. 642.

Fronczak, D. L., D. E. Andersen, E. E. Hanna, and T. R. Cooper. 2017. Distribution and migration chronology of Eastern population Sandhill Cranes. Journal of Wildlife Management 81:1021-1032. https://doi.org/10.1002/jwmg.21272

Gaillard, J. M., M. Hebblewhite, A. Loison, M. Fuller, R. Powell, M. Basille, and B. Van Moorter. 2010. Habitat-performance relationships: finding the right metric at a given spatial scale. Philosophical Transactions of the Royal Society B 365:2255-2265. https://doi.org/10.1098/rstb.2010.0085

Gerber, B. D., J. F. Dwyer, S. A. Nesbitt, R. C. Drewien, C. D. Littlefield, T. C. Tacha, and P. A. Vohs. 2020. Sandhill Crane (Antigone canadensis), version 1.0. In A. F. Poole, editor. Birds of the world. Cornell Lab of Ornithology, Ithaca, New York, USA. https://doi.org/10.2173/bow.sancra.01

Gillies, C. S., M. Hebblewhite, S. E. Nielsen, M. A. Krawchuk, C. L. Aldridge, J. L. Frair, D. J. Saher, C. E. Stevens, and C. L. Jerde. 2006. Application of random effects to the study of resource selection by animals. Journal of Animal Ecology 75:887-898. https://doi.org/10.1111/j.1365-2656.2006.01106.x

Gorelick, N., M. Hancher, M. Dixon, S. Ilyushchenko, D. Thau, and R. Moore. 2017. Google Earth Engine: planetary-scale geospatial analysis for everyone. Remote Sensing of Environment 202:18-27. https://doi.org/10.1016/j.rse.2017.06.031

Hebblewhite, M., and E. Merrill. 2008. Modelling wildlife-human relationships for social species with mixed-effects resource selection models. Journal of Applied Ecology 45:834-844. https://doi.org/10.1111/j.1365-2664.2008.01466.x

Hedenström, A., and T. Alerstam. 1997. Optimum fuel loads in migratory birds: distinguishing between time and energy minimization. Journal of Theoretical Biology 189:227-234. https://doi.org/10.1006/jtbi.1997.0505

Holbrook, J. D., L. E. Olson, N. J. DeCesare, M. Hebblewhite, J. R. Squires, and R. Steenweg. 2019. Functional responses in habitat selection: clarifying hypotheses and interpretations. Ecological Applications 29:e01852. https://doi.org/10.1002/eap.1852

Iverson, G. C., P. A. Vohs, and T. C. Tacha. 1987. Habitat use by mid-continent Sandhill Cranes during spring migration. Journal of Wildlife Management 51:448-458. https://doi.org/10.2307/3801033

Ivey, G. L., B. D. Dugger, C. P. Herziger, M. L. Casazza, and J. P. Fleskes. 2015. Wintering ecology of sympatric subspecies of Sandhill Crane: correlations between body size, site fidelity, and movement patterns. Condor 117:518-529. https://doi.org/10.1650/CONDOR-14-159.1

Jankowiak, Ł., P. Skórka, Ł. Ławicki, P. Wylegała, M. Polakowski, A. Wuczyński, and P. Tryjanowski. 2015. Patterns of occurrence and abundance of roosting geese: the role of spatial scale for site selection and consequences for conservation. Ecological Research 30:833-842. https://doi.org/10.1007/s11284-015-1282-2

Johnson, D. H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61:65-71. https://doi.org/10.2307/1937156

Kauffeld, J. D. 1981. Management of migratory crane habitat on Alamosa and Monte Vista National Wildlife Refuges. Proceedings of the 1981 Crane Workshop 117-121.

Kinzel, P. J., J. M. Nelson, and A. K. Heckman. 2009. Response of Sandhill Crane (Grus canadensis) riverine roosting habitat to changes in stage and sandbar morphology. River Research and Applications 25:135-152. https://doi.org/10.1002/rra.1103

Krapu, G. L., and D. A. Brandt. 2008. Spring and migratory habits and breeding distribution of lesser Sandhill Cranes in west-central New Mexico and Arizona. North American Crane Workshop Proceedings 43-49.

Krapu, G. L., D. A. Brandt, K. L. Jones, and D. H. Johnson. 2011. Geographic distribution of the mid-continent population of Sandhill Cranes and related management applications. Wildlife Monographs 175:1-38. https://doi.org/10.1002/wmon.1

Krapu, G. L., D. E. Facey, E. K. Fritzell, and D. H. Johnson. 1984. Habitat use by migrant Sandhill Cranes in Nebraska. Journal of Wildlife Management 48:407-417. https://doi.org/10.2307/3801172

Krapu, G. L., G. C. Iverson, K. J. Reinecke, and C. M. Boise. 1985. Fat deposition and usage by arctic-nesting Sandhill Cranes during spring. Auk 102:362-368. https://doi.org/10.2307/4086780

Kruse, K. L., D. P. Collins, C. M. Conring, B. A. Grisham, W. C. Conway, and J. M. Knetter. 2017. Summer habitat selection of the Lower Colorado River Valley population of greater Sandhill Cranes. Journal of Fish and Wildlife Management 8:436-448. https://doi.org/10.3996/042017-JFWM-037

Lewis, J. C. 1979. Taxonomy, food, and feeding habitat of Sandhill Cranes, Platte Valley, Nebraska. Proceedings of the 1978 Crane Workshop 21-27.

Littlefield, C. D. 1992. Autumn sandhill crane migration in southeastern Oregon. Proceedings of the 1988 Crane Workshop 47-54.

Llewellyn, D., and S. Vaddey. 2013. West-wide climate risk assessment: Upper Rio Grande impact assessment. Bureau of Reclamation, U.S. Department of the Interior, Washington, D.C., USA.

Lovvorn, J. R., and C. M. Kirkpatrick. 1982. Field use by staging eastern Greater Sandhill Cranes. Journal of Wildlife Management 46:99-108. https://doi.org/10.2307/3808412

Manly, B., 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. Kluwer Academics, Dordrecht, The Netherlands.

Matthiopoulos, J., J. Fieberg, G. Aarts, H. L. Beyer, J. M. Morales, and D. T. Haydon. 2015. Establishing the link between habitat selection and animal population dynamics. Ecological Monographs 85:413-436. https://doi.org/10.1890/14-2244.1

McLoughlin, P. D., D. W. Morris, D. Fortin, E. Vander Wal, and A. L. Contasti. 2010. Considering ecological dynamics in resource selection functions. Journal of Animal Ecology 79:4-12. https://doi.org/10.1111/j.1365-2656.2009.01613.x

Mix, K., V. L. Lopes, and W. Rast. 2011. Annual and growing season temperature changes in the San Luis Valley, Colorado. Water, Air, and Soil Pollution 220:189-203. https://doi.org/10.1007/s11270-011-0746-4

Muff, S., J. Signer, and J. Fieberg. 2019. Accounting for individual-specific variation in habitat-selection studies: efficient estimation of mixed-effects models using Bayesian or frequentist computation. Journal of Animal Ecology 89:80-92. https://doi.org/10.1111/1365-2656.13087

Newton, I. 2007. The migration ecology of birds. Academic, Cambridge, Massachusetts, USA.

Northrup, J. M., E. Vander Wal, M. Bonar, J. Fieberg, M. P. Laforge, M. Leclerc, C. M. Prokopenko, and B. D. Gerber. 2022. Conceptual and methodological advances in habitat-selection modeling: guidelines for ecology and evolution. Ecological Applications 32:e02470. https://doi.org/10.1002/eap.2470

Pearse, A. T., G. L. Krapu, and D. A. Brandt. 2017. Sandhill crane roost selection, human disturbance, and forage resources. Journal of Wildlife Management 81:477-486. https://doi.org/10.1002/jwmg.21215

Pearse, A. T., K. L. Metzger, D. A. Brandt, J. A. Shaffer, M. T. Bidwell, and W. Harrell. 2021. Migrating Whooping Cranes avoid wind-energy infrastructure when selecting stopover habitat. Ecological Applications 31:e02324. https://doi.org/10.1002/eap.2324

Reinecke, K. J., and G. L. Krapu. 1986. Feeding ecology of Sandhill Cranes during spring migration in Nebraska. Journal of Wildlife Management 50:71-79. https://doi.org/10.2307/3801490

Rio Grande Basin Roundtable. 2015. Rio Grande Basin Implementation Plan 1-164. Rio Grande Basin Roundtable, Alamosa, Colorado, USA.

Rio Grande Basin Roundtable. 2022. Rio Grande Basin Implementation Plan Volume 2. Rio Grande Basin Roundtable, Alamosa, Colorado, USA.

Rodehutscord, M., C. Rückert, H. P. Maurer, H. Schenkel, W. Schipprack, K. E. Bach Knudsen, M. Schollenberger, M. Laux, M. Eklund, W. Siegert, and R. Mosenthin. 2016. Variation in chemical composition and physical characteristics of cereal grains from different genotypes. Archives of Animal Nutrition 70:87-107. https://doi.org/10.1080/1745039X.2015.1133111

Rosenzweig, M. L. 1981. A theory of habitat selection. Ecology 62:327-335. https://doi.org/10.2307/1936707

Rowland, M. M., L. Kinter, T. Banks, and D. C. Lockman. 1992. Habitat use by greater Sandhill Cranes in Wyoming. Proceedings of the 1988 North American Crane Workshop 82-86.

San Luis Valley Development Resources Group and Council of Governments. 2023. 2023 San Luis Valley Statistical Profile. San Luis Valley Development Resources Group, Alamosa, Colorado, USA.

Schmaljohann, H., C. Eikenaar, and N. Sapir. 2022. Understanding the ecological and evolutionary function of stopover in migrating birds. Biological Reviews 97:1231-1252. https://doi.org/10.1111/brv.12839

Sedinger, J. S., and R. T. Alisauskas. 2014. Cross-seasonal effects and the dynamics of waterfowl populations. Wildfowl 4:277-304.

Siegert, W., T. Zuber, and M. Rodehutscord. 2022. Variability and prediction of metabolisable energy of wheat, triticale, and rye in caecectomised laying hens. European Poultry Science 86:1-16. https://doi.org/10.1399/eps.2022.354

Sparling, D. W., and G. L. Krapu. 1994. Communal roosting and foraging behavior of staging sandhill cranes. Wilson Bulletin 106:62-77.

Stanley, C. Q., M. R. Dudash, T. B. Ryder, W. G. Shriver, K. Serno, S. Adalsteinsson, and P. P. Marra. 2021. Seasonal variation in habitat selection for a Neotropical migratory songbird using high-resolution GPS tracking. Ecosphere 12:e03421. https://doi.org/10.1002/ecs2.3421

Suraci, J. P., L. G. Frank, A. Oriol-Cotterill, S. Ekwanga, T. M. Williams, and C. C. Wilmers. 2019. Behavior-specific habitat selection by African lions may promote their persistence in a human-dominated landscape. Ecology 100:e02644. https://doi.org/10.1002/ecy.2644

Tacha, T. C. 1981. Behavior and taxonomy of Sandhill Cranes from Mid-continental North America. Dissertation. Oklahoma State University, Stillwater, Oklahoma, USA.

Tacha, T. C., P. A. Vohs, and G. C. Iverson. 1987. Time and energy budgets of Sandhill Cranes from Mid-Continental North America. Journal of Wildlife Management 51:440-448. https://doi.org/10.2307/3801032

U.S. Fish and Wildlife Service. 2015. Comprehensive conservation plan: San Luis Valley National Wildlife Refuge Complex. Lakewood, Colorado, USA.

Vanausdall, R. A., W. L. Kendall, D. P. Collins, and Q. R. Hays. 2024. Time of year and weather influence departure decisions of Sandhill Cranes at a primary stopover. Frontiers in Ecology and Evolution 12:1279279. https://doi.org/10.3389/fevo.2024.1279279

Walker, D. L., and S. D. Schemnitz. 1985. Food habits of Sandhill Cranes in relation to agriculture in central and southwestern New Mexico. Proceedings of the 1985 Crane Workshop 201-211.

Walkinshaw, L. H. 1949. The Sandhill Cranes. Cranbrook Institute of Science, Bloomfield Hills, Michigan, USA.

Warnock, N. 2010. Stopping vs. staging: the difference between a hop and a jump. Journal of Avian Biology 41:621-626. https://doi.org/10.1111/j.1600-048X.2010.05155.x

Western Regional Climate Center. 2013. Cooperative climatological data summaries. Western Regional Climate Center, Reno, Nevada, USA. https://wrcc.dri.edu/Climate/summaries.php

Wetland Dynamics. 2019. San Luis Valley wetland and wildlife conservation assessment. Monte Vista, Colorado, USA.