INTRODUCTION

Precise mapping of ecologically relevant landcover characteristics represents a critical component for assessing wildlife habitat associations accurately (Roloff and Kernohan 1999, Gallant 2009). However, widely available geospatial landcover databases regularly differ in their demarcation of fine-scale but ecologically relevant features such as riparian corridors and associated wetlands (Cunningham 2006, Gergel et al. 2007, Fremgen-Tarantino et al. 2021). Differences in landscape definitions and classifications, as well as mapping scale, mapping techniques, and when the data were collected can result in misleading conclusions and, ultimately, poor management decisions (Langford et al. 2006, Kleindl et al. 2015). Integrating several geospatial data sources can increase mapping accuracy of finer-scale habitat features and therefore improve wildlife habitat models (Gallant 2009). This approach can also help identify distinct and biologically relevant ecotopes, which are the smallest homogenous mappable units of analysis in landscape ecology considering spatially explicit abiotic (e.g., flooding) and biotic (e.g., vegetation) factors (Klijn and Udo de Haes 1994, Naveh 1994, Hong et al. 2004, Ingegnoli 2013). Ecotope-level landscape classifications have been helpful in demarcating potential habitat for wildlife species that require a relatively specific set of conditions (Hong et al. 2004, Caven et al. 2017).

Whooping Cranes (Grus americana) were once at the precipice of extinction when fewer than 20 individuals were in existence in the mid-20th century, but numbers in the Aransas-Wood Buffalo population (AWBP) have slowly increased because of concerted long-term conservation efforts (Urbanek and Lewis 2020). The AWBP of Whooping Cranes, the only wild, self-sustaining population in the world, migrates through Central Flyway twice annually. Migration is an important life history event for many birds, and it is critically important that individuals are able to locate safe resting areas with sufficient food resources that will fuel migration and ensure birds arrive at breeding sites in good physical condition (Piersma et al. 2005, O’Neal et al. 2012). In the Great Plains, the Central Platte River Valley (CPRV) serves as an important stopover area for Whooping Cranes, where, between 1995 and 2015, more than 27% of all Whooping Crane records in Nebraska have been recorded (U.S. Fish and Wildlife Service, unpublished data). The CPRV is one of only five areas designated as Critical Habitat in the Central Flyway for the endangered Whooping Crane (CWS and USFWS 2005). The CPRV is important to Whooping Cranes and other migratory birds, both historically and currently, because of the diverse array of landcover classes and rich food resources it provides (NRC 2004, Caven et al. 2021). The CPRV is now dominated by row-crop agriculture (Dappen et al. 2008). Natural landscapes, including palustrine wetlands such as wet meadows, are limited but serve as important foraging sites for Whooping Cranes and Sandhill Cranes (Antigone canadensis) based on studies throughout the Central Flyway (Reinecke and Krapu 1986, Lingle et al. 1991, Austin and Richert 2005, Geluso et al. 2013, Baasch et al. 2019). Recent research by Howlin and Nasman (2017) in the CPRV did not detect a significant association between wet meadow landcover and Whooping Crane use. This result raises questions as to whether Whooping Crane habitat use in the CPRV is different than elsewhere in the Central Flyway.

Herbaceous plant communities in the CPRV are influenced by ground water levels and vary meaningfully across relatively small elevational gradients (e.g., < 2.0 m; Currier 1989, Henszey et al. 2004). Sub-irrigated herbaceous landcovers in the CPRV have been classified multiple ways. Brie and Bishop (2008) categorized the drier portions of these sub-irrigated herbaceous systems dominated by big bluestem (Andropogon gerardi) and switchgrass (Panicum virgatum) as xeric wet meadows. By contrast, Rolfsmeier and Steinauer (2010) defined these systems as Sandhills mesic tallgrass prairies. Brie and Bishop (2008) categorized the wetter portions of sub-irrigated herbaceous landcovers that are dominated by sedges, such as Emory’s sedge (Carex emoryi) and woolly sedge (C. pellita), as mesic wet meadows. Additional terms used to classify these wetter ecotopes include eastern sedge meadows (Rolfsmeier and Steinauer 2010), wet meadows (Currier 1982, Chávez-Ramírez and Weir 2010), and sedge meadows (Henszey et al. 2004).

Considering nationally recognized wetland definitions, wet meadows represent a subtype of palustrine wetland dominated by emergent herbaceous vegetation that are sustained by fluctuating moisture regimes (Wilcox et al. 2007, Tiner 2016). Annually, these ecosystems experience periodic surface inundation or saturation during the growing season for durations long enough to allow the development of wetland soil profiles (Tiner 2016). Regionally, most wet meadows are sustained by fluctuating groundwater levels linked to riparian corridors (Hurr 1981, Wesche et al. 1994, Brinley Buckley et al. 2021). Variable hydrology is a key feature of this landcover type because sustained inundation or a lack of periodic inundation results in the system transitioning to an alternative community (i.e., shallow marsh or lowland tallgrass prairie; Boswell and Olyphant 2007, Wilcox et al. 2007, Tiner 2016). In Nebraska, Platte River wet meadows are classified into the hydro-geomorphic subclass of riverine floodplain, rapid permeability with minimal out of bank flooding (LaGrange 2015). The plant and soil profiles used by Brie and Bishop (2008) to map mesic wet meadow correspond closely to the conception of a wet meadow put forth by Tiner (2016) as a subtype of palustrine wetland with a seasonally saturated water regime. Brie and Bishop’s (2008) definition of xeric wet meadow similarly corresponds to the general description of sub-irrigated lowland tallgrass prairie put forth by Kaul et al. (2006) as well as Rolfsmeier and Steinauer (2010).

Recent research suggests that Whooping Cranes select for aquatic and wetland habitats for diurnal use over all other land cover types including cornfields (Baasch et al. 2019). Howlin and Nasman’s (2017) contradictory conclusion that Whooping Cranes do not select wet meadows over agricultural lands in the CPRV may be the result of imprecise regional definitions, mapping, or spatial consideration of wetland components (i.e., wet meadows and shallow marshes) embedded within a lowland tallgrass prairie matrix. Howlin and Nasman (2017) employed a highly inclusive and arguably arbitrary definition of wet meadow following the Platte River Recovery Implementation Program’s (PRRIP 2012) definition that did not correspond to standard wetland classification schemes (Tiner 2016). The Platte River Recovery Implementation Program (2012) considered nearly all grasslands that possessed some proportion of sub-irrigated herbaceous vegetation entirely as wet meadow, regardless of its overarching hydrologic regime (Howlin and Nasman 2017). However, many of the landscapes classified by PRRIP (2012) as wet meadow were defined as non-wetland habitats (e.g., lowland tallgrass prairie and sand ridge prairie) via alternative habitat classification and mapping efforts (Currier 1989, Henszey 2001, Henszey et al. 2004, Brei and Bishop 2008, Rolfsmeier and Steinauer 2010, Kaul et al. 2006). Brei and Bishop’s (2008) map of the CPRV, as well as 2011 National Wetland Inventory mapping, may help us better assess the degree to which Whooping Cranes select for herbaceous wetland components (e.g., wet meadows, shallow marsh, etc.) compared to analogous upland land cover types. As Chávez-Ramírez and Weir (2010) noted, wet meadows have been variously conceptualized within the CPRV but have always been considered important as diurnal Whooping Crane habitat.

We used multiple fine-scale geospatial data sources to reassess Whooping Crane occurrence in the CPRV at the ecotope level. We also examined landscape features associated with diurnal terrestrial occurrence at multiple spatial scales to determine which was most predictive of Whooping Crane occurrence (Johnson 1980, Mayor et al. 2009, Niemuth et al. 2018). We hypothesized that wetland landcover types would be predictive of Whooping Crane terrestrial habitat use at multiple spatial scales considering recent studies in other areas of the Great Plains (e.g., Niemuth et al. 2018, Baasch et al. 2019). Results of our study will help conservation practitioners and decision makers prioritize areas that provide important resources to Whooping Cranes during migratory stopovers and could guide future conservation actions regionally.

METHODS

Study area

Our study area included a 7 km buffer surrounding a 56 km reach of the main channel of the Platte River between Lexington (40°73’ N, 99°74’ W) and Chapman (40°98’ N, 98°15’ W) Nebraska, USA (184,776 km²; Fig. 1). A large portion of this reach has been designated as Critical Habitat for the endangered Whooping Crane (USFWS 1978). Our study area is entirely encompassed within the current 95% migration corridor (Pearse et al. 2018) and is located 1500 km north of wintering grounds and 2500 km south of the breeding grounds. The CPRV is dominated by row-crop and other agriculture including corn (Zea mays), soybeans (Glycine max), alfalfa (Medicago sativa), sorghum (Sorghum bicolor), and wheat (Triticum aestivum; Dappen et al. 2008). However, native herbaceous landcover types, including wet meadows and lowland tallgrass prairies, are distributed throughout the study area and have been protected, enhanced, and restored for 40 years as part of cooperative conservation efforts by non-governmental organizations, government agencies, and private landowners (Currier and Henszey 1996, Krapu et al. 2014). Research indicates that meadow-marsh landcover is more abundant in the eastern portion of the CPRV (Krapu et al. 2014, Caven et al. 2019). The CPRV extends the range of tallgrass prairie westward into the central mixed-grass eco-region of the Great Plains by providing sub-irrigated moisture to a deep-rooted herbaceous perennial plant community (Currier 1989, Henszey et al. 2004, Kaul et al. 2006). Drier portions of these sub-irrigated herbaceous systems (i.e., lowland tallgrass prairies) are dominated by big bluestem and switchgrass, whereas the wetter portions of sub-irrigated herbaceous ecosystems (i.e., wet meadows) are dominated by prairie cordgrass (Spartina pectinata) and sedges, such as Emory’s sedge and woolly sedge. The CPVR is also an area of sustained restoration and management action by conservation organizations (e.g., Crane Trust, Audubon, and The Nature Conservancy) and state and federal natural resource agencies (e.g., U.S. Fish and Wildlife Service and Nebraska Game and Parks Commission) to improve Whooping Crane habitat conditions since the mid-1970s as well as the ongoing Platte River Recovery Implementation Program (Currier and Pfeiffer 2005, PRRIP 2006, Renfrew et al. 2006, Meyer et al. 2010).

Locational data

We used the U.S. Fish and Wildlife Service (USFWS) public sightings database, which included verified sightings of Whooping Cranes reported by the public, state and federal agencies, conservation organizations, and others from 1995–2015 to determine Whooping Crane use locations (Austin and Richert 2005, Tacha et al. 2010, Caven et al. 2020). All sightings in the public sightings database were confirmed via visual verification of photos, videos, or in-field assessment by a qualified observer including biologists, natural resource professionals, or experienced bird watchers (Austin and Richert 2005, Tacha et al. 2010, Caven et al. 2020). We updated the associated geodatabase by digitizing secondary, tertiary, and additional use locations that previously existed only in hard copy reports because most analyses to date have simply used the first observed use location (Tacha et al. 2010, Niemuth et al. 2018). We undertook this effort because most of the first documented use locations in the CPRV correspond to night roosts; therefore, diurnal use locations important to foraging were not well represented within the primary geodatabase.

The USFWS public sightings database contains inherent biases (Tacha et al. 2010, Hefley et al. 2013). For instance, Niemuth et al. (2018) found that Whooping Cranes had a higher probability of being sighted near roads and population centers. Within our statistical models, we included distance to nearest paved road and developed area as fixed effects and population density within a 2.6 km² area of use and random locations as random effects to account for the variation explained by these variables in our modeling effort (Seirup and Yetman 2006, Niemuth et al. 2018). We also included unbiased use locations from Platform Transmitter Terminal-marked (PTT) cranes, collected during 2010 to 2015, to make up a comprehensive set of known use locations and increase the size of our database (Baasch et al. 2019, Pearse et al. 2017, 2018, 2020). To facilitate our analysis, we generated 20 random locations distributed throughout the study area per use location for a total of 6940 random locations per 306 public sightings and 41 PTT-marked crane locations, which equates to 3.8 random locations per km² and generally reflects or exceeds the number of random points relative to study area size or known use location number used in similar space-use studies (e.g., Aarts et al. 2007, Hebblewhite and Merrill 2008). Use locations were included in the analysis based on three criteria: location > 10 m outside of the high banks of any river channel, data quality (location accuracy estimated at ≤ 400 m), and data within 10 years of 2005 (i.e., 1995-2015), which was when the CPRV landcover was mapped by Brei and Bishop (2008).

Landcover mapping

To assess Whooping Crane occurrence in relation to landscape characteristics, we used secondary data sources including the 2005 vegetation and landcover classifications created by the Rainwater Basin Joint Venture (Brei and Bishop 2008), the National Wetlands Inventory Project (NWI; Wilen and Bates 1995, USFWS 2021; Tande and Michaelson 2011, unpublished data), and flooding frequency data from the USDA-NRCS (2019, 2020; Table 1). All input data were converted to 10-m raster to match the USDA-NCRS (2020) gridded soil survey geographic database. All palustrine wetland types, with the exception of excavated pits, stock ponds, and impoundments, were dissolved together to get total palustrine wetland area and overlaid with the 2005 landcover (Wilen and Bates 1995, USFWS 2021). Soil survey data included the following flooding frequency categories: never, very rarely, rarely, occasionally, frequently, and very frequently, which we assessed in relation to terrestrial Whooping Crane occurrence patterns (USDA-NRCS 2019, 2020). Frequently and very frequently flooded categories were merged into one category because very frequently flooded landcover types were rarely observed in the study area. We hypothesized that Whooping Cranes would be more likely to use wetter components of the landscape per hydrological mapping efforts. However, we further hypothesized that landscape designations that integrated landcover types (e.g., herbaceous, woodland, etc.) with hydrological categorizations (e.g., flooding frequency, wetland designation, etc.) into ecotope classifications would better predict Whooping Crane use locations. Meadow-marsh was intended to represent wetland herbaceous landcover thought to be preferred by Whooping Cranes (Baasch et al. 2019). The meadow-marsh landcover class combined multiple categories from Brei and Bishop (2008) including mesic wet meadow, floodplain marsh, and warm-water slough (Table 1). We defined categorically non-wetland herbaceous landcovers in the CPRV as prairie, which included meadow sand ridge, xeric wet meadow, upland grassland, and undisturbed grassland per Brei and Bishop (2008; Table 1). However, some landcover types classified as prairie were defined as wetlands through the NWI (USFWS 2021; Tande and Michaelson 2011, unpublished data). As Rolfsmeier and Steinauer (2010) noted, the Platte Valley contains a significant amount of eastern cordgrass wet prairie, which can serve as a transitional landcover type between sub-irrigated lowland tallgrass prairies and wet meadows. Therefore, we defined the prairie landcover type per Brie and Bishop (2008) with a wetland footprint per the NWI (Tande and Michaelson 2011, unpublished data) as wet prairies (Table 1). Given rotational practices of farming within the CPRV (i.e., soybean and cornfields are frequently rotated on an annual basis), all agricultural landcover types were lumped together and classified as agricultural wetlands or agricultural uplands based on flooding frequency. Binary rasters were generated for each landscape variable and the mean of each binary was calculated at 400 m (minimum locational accuracy of our data) and 1000 m (relevant to Whooping Cranes per Belaire et al. 2014, Niemuth et al. 2018, Baasch et al. 2019, Caven et al. 2022) radius scales with the focal statistics tool in ArcGIS version 10.6.1 (Environmental Systems Research Institute, Redlands, CA.) using a circular moving window. The mean values were multiplied by 100 to give the percent of landcover at each scale. Scaled ecotope, flood frequency, and palustrine wetland values were extracted to Whooping Crane use and available locations for statistical analysis.

Statistical analyses

We used binomial family Generalized Linear Mixed-Effects Models (GLMM) within the lme4 package of Program R with a logit link function and a BOBYQA nonlinear optimizer to assess the effects of ecotope composition, flooding frequency, and wetland status on the probability of Whooping Crane occurrence across multiple spatial scales (Dean and Nielsen 2007, Powell 2009, Bates et al. 2015, Bates et al. 2021). Using the scale function in Program R version 4.2.0 (R Core Team 2020), all integer and continuous predictor variables were scaled to z-scores (standard deviations above or below the mean) and centered at zero to improve model convergence (Bring 1994). This approach had the added benefit of producing standardized coefficients with model results, which provide a coarse indication of each covariate’s relative importance within a statistical model (Bring 1994, Afifi et al. 2020). To avoid issues of multicollinearity, we examined bivariate relationships between all ordinal, interval, and continuous predictor variables using Pearson product-moment correlation coefficients with the cor function in the stats package (R Core Team 2020). No two variables with > |0.6| correlation or association (binary) were included in the same model (Dormann et al. 2013) but, rather, were tested in separate models. To ensure that standard errors in our models were not inflated as a result of collinearity between more than two variables, we conducted variance inflation tests on each candidate model using the vif function in the car package and dropped all models scoring > 5.0 from our analyses (Fox and Weisberg 2019).

We developed 54 a priori candidate models based on landscape mapping efforts that separated the study area across gradients of wetness to determine if Whooping Cranes differentially utilized the wetland components of herbaceous or agricultural habitats in the CPRV. Candidate models were based on three themes including ecotope, flooding frequency, and palustrine wetland models. Models included all possible uncorrelated variable combinations within each model theme. Like models were run at two scales including landcover within 400-m and 1000-m buffers around each use or random location to determine which landcovers were most predictive of terrestrial Whooping Crane occurrence. All models shared a common structure including a set of uncorrelated and thematically related (e.g., flooding frequency) predictor variables at the same spatial scale (i.e., 400 m or 1000 m) along with developed landcover and distance to nearest road as fixed effects control variables as well as landcover classification and categorical human population density as random effects control variables. Our model structure allowed for intercepts to vary across categories of landcover and population, which helped ensure that these variables potentially associated with detection probability did not represent a large amount of unexplained variance that could have biased independent variable parameter estimates. Population was converted to a categorical variable and treated as a random effect because it was highly correlated to developed landcover at both spatial scales. We used natural breaks from a histogram analysis to derive population categories from local census data (Seirup and Yetman 2006). Abundance categories included population as 0 people (15.8% of observations), 1 person (35.0%), 2 people (21.6%), 3–5 people (16.0%), 6–10 people (3.6%), 11–30 people (3.4%), 31–100 people (2.2%), and > 100 people (2.4%) per 2.6 km². In total, 26 models were run at each scale in addition to two null models, including one with random effects (y ~ 1 + (1|Landcover Class) + (1|Categorical Population)) and a simple null model (y ~ 1). Models were compared using AIC corrected for small sample sizes (AICc) using the model.sel function in the MuMIn package (Burnham and Anderson 2002, Wagenmakers and Farrell 2004, Barton 2020). We assessed the model fit of top models using marginal and conditional R² values per Nakagawa and Schielzeth (2013). We also used predictor effects plots from the effects package (Fox and Weisberg 2018, 2019) and conditional density plots from the graphics package (R Core Team 2020) to describe the relationship between key predictor variables and Whooping Crane occurrence.

RESULTS

Ecotope-based models at the 1000-m scale performed the best and represented the top eight of 54 a priori models tested (Appendix 1). They were followed by palustrine wetland models at the 1000-m scale and ecotope models at the 400-m scale, which in total accounted for the top 19 of 54 models (Appendix 1). However, all flooding frequency models also outperformed both null models, which ranked 53rd and 54th out of the 54 models, respectively. Scales were intermixed for flooding frequency models, but the top couple flooding frequency models were at the 400-m scale and the four lowest ranked flooding frequency models corresponded to the 1000-m scale. The top six models overall, which all corresponded to ecotope models at the 1000-m scale, represented 100% of the cumulative model weight (Appendix 1).

The top model included agricultural wetland (B ± SE = 0.090 ± 0.034, p = 0.008), meadow-marsh (B ± SE = 0.193 ± 0.053, p < 0.001), prairie (B ± SE = -0.155 ± 0.091, p = 0.086), river channel (B ± SE = 0.541 ± 0.042, p = < 0.001), woodland (B ± SE = -0.195 ± 0.067, p = 0.004), development (B ± SE = -0.770 ± 0.190, p < 0.001), and distance to nearest road (B ± SE = 0.094 ± 0.058, p = 0.104) as fixed effects and landcover class and categorical population density as random effects (Appendix 1, Fig. 2). The only other model with a Δ AICc ≤ 2 represented the same set of predictors with prairie removed.

Predictor effect plots (Fig. 3) and our probability of occurrence heat map (Fig. 4) clearly demonstrate an increasing likelihood of Whooping Crane occurrence with increasing values of meadow marsh, agricultural wetland, and riverine landcovers and a decreasing likelihood of occurrence with increased woodland and developed landcovers within a 1000-m buffer. Top model results indicate that a 1% increase in meadow-marsh landcover would increase the probability of Whooping Crane use by 5.1% (Fig. 3). Our top model indicates that a 1% increase in agricultural wetland landcover would result in a 17.4% increase in the probability of Whooping Crane occurrence (Fig. 3). Our top model predicted that a 1% increase in riverine landcover would result in a 27.9% increase in Whooping Crane occurrence (Fig. 3). A 1% increase in woodland was predicted to result in a 3.8% decrease in the probability of Whooping Crane use (Fig. 3). Similarly, a 1% increase in developed landcover was predicted to result in a 3.9% decrease in Whooping Crane occurrence (Fig. 3). Theoretical pseudo R² values were 0.39 for the model as a whole including random effects (conditional) and 0.19 regarding fixed effects only (marginal).

DISCUSSION

Similar to other studies, we found areas with increased wetland components (i.e., meadow-marsh and wetland agriculture) located near the Platte River and decreased densities of roads and development had a higher likelihood of occupancy for diurnal use by Whooping Cranes than drier components of the landscape (Krapu 1981, Belaire et al. 2014, Hefley et al. 2015, Baasch et al. 2019). Our top model explained 39% of the variance in Whooping Crane occurrence within terrestrial landscapes in the CPRV. This indicates that finer scale habitat features, defined here as ecotopes, are likely an important factor in defining migratory stopover habitat for Whooping Cranes. Contrastingly, Howlin and Nasman (2017) found Whooping Cranes avoided wet meadows as compared to agricultural fields within the CPRV. However, their demarcations of wet meadows were based on PRRIP’s (2012) delineations, which included all riparian grassland landcover types regardless of relative wetland composition or wetland size. Contrastingly, we found Whooping Crane occurrence was positively associated with meadow-marsh and negatively linked to upland prairie habitats. Therefore, the coarser mapping scale used by Howlin and Nasman (2017) likely masked ecologically important differences in landcover types that are predictive of Whooping Crane occurrence. We found nuanced differences in ecotope classification that captured variations in wetness (e.g., prairie, wet prairie, and meadow-marsh) were important to consider when determining the likelihood of Whooping Crane occurrence. Whooping Cranes appear to be more likely to occupy wetland components of both herbaceous and agricultural landscapes, so occurrence and selection models that treat all grasslands or agricultural lands equally likely fail to detect important and biologically relevant habitat associations for this species.

Our study demonstrates the importance of using biologically relevant mapping approaches when evaluating species’ habitat use patterns. This is especially important when individual species depend on finer-scale habitat features, such as wetlands, that are challenging to precisely map (Cunningham 2006, Roloff and Kernohan 1999, Gallant 2009, Fremgen-Tarantino et al. 2021). Our ecotope classifications considered not only broad landcover categories, but also included hydrologically significant gradients of wetness (Brei and Bishop 2008, USDA-NRCS 2019, 2020, Tiner 2016; Tande and Michaelson 2011, unpublished data). Accordingly, ecotope classifications were more predictive of Whooping Crane occurrence than their component parts including flooding frequencies, wetland status, and landcover designations alone. Most studies do not adequately consider the effects of classification precision and accuracy on model results (Cunningham 2006, Gergel et al. 2007, Gallant 2009, Lechner et al. 2012, Howlin and Nasman 2017, Fremgen-Tarantino et al. 2021). Recent research suggests that the landcover classifications and resolutions selected for use in modeling can impact estimates of species distributions (Lawler et al. 2004, Roloff and Kernohan 1999, McKerrow et al. 2018, Neimuth et al. 2018, Fremgen-Tarantino et al. 2021). A potential limitation of our study is that all agricultural landcovers were lumped, which could have masked the effects of different crop types on Whooping Crane occurrence. However, most sites were farmed rotationally, and several different cycles of annual crops were likely planted on each field across the study period. Additionally, Whooping Cranes generally respond positively to most herbaceous agricultural landcover classes, especially pre- and post-harvest (Caven et al. 2022). Ecologically relevant and accurate mapping is expressly important when studies are focused on a critically endangered species and when findings can influence management decisions with significant long-term implications (e.g., land protection and restoration). Our finding that the 1000-m buffer scale was most predictive of Whooping Crane occurrence was similar to Niemuth et al. (2018), which suggested that landcover classifications at the 1200-m scale were most predictive. Future research should attempt to assess variations in Whooping Crane occurrence at an ecotope scale including real-time information regarding habitat conditions (e.g., grazing, flooding, etc.) and behavior (e.g., foraging, preening, etc.; Rasool et al. 2021). In addition, assessing food resource availability within different ecotopes in conjunction with behavioral studies may provide valuable insights into the provisions provided to Whooping Cranes by different ecotopes (e.g., diet, rest, etc.; Lingle et al. 1991, Jorgensen and Dinan 2016, Hou et al. 2021).

Given that we used opportunistically collected data obtained through public sightings, consistent information does not exist regarding the condition of habitats during the time birds were using them (i.e., whether the landscape was inundated or not). Furthermore, data used in our study do not include Whooping Crane behaviors associated with each use location. Whereas use of one landcover class over another is an indication of its importance for Whooping Cranes, behavioral activities performed in various landcover classes, such as foraging or loafing, represent other important considerations when assessing the relative importance of each landcover class with regards to fitness and survival during migration (Lingle et al. 1991, Jorgensen and Dinan 2016, Hale et al. 2020). Research from the CPRV suggests that Sandhill Cranes spend a higher proportion of their time foraging while using lowland grasslands (i.e., lowland tallgrass prairie and wet meadows combined) as compared to other landscapes in the CPRV (i.e., agricultural fields; Sparling and Krapu 1994, VerCauteren 1998, Davis 2003). Davis and Vohs (1993) found Sandhill Cranes exhibited differential use patterns within the CPRV in that they extensively used areas composed of wet meadow and lowland tallgrass prairie but minimally used areas composed entirely of lowland tallgrass prairie, highlighting the importance of wetland components in supporting Sandhill Crane diets, which may serve as a useful indicator of their dietary value for Whooping Cranes as well. Research suggests that corn cannot meet a crane’s daily physiological needs in terms of nutrients, such as protein, calcium, and phosphorus, necessary for successful migration and reproduction (Reinecke and Krapu 1986, Sparling and Krapu 1994, Caven et al. 2021). Neri (2011) demonstrated that plant matter, such as seeds and berries, are high in calories but low in protein, whereas animal matter is low in calories but high in protein value for Whooping Cranes. Wetland habitats provide access to an array of animal protein sources, such as invertebrates, fish, amphibians, and reptiles (Neri 2011, Geluso et al. 2013, Urbanek and Lewis 2020, Caven et al. 2021). Recent behavioral research conducted in Nebraska indicates Whooping Cranes spend a disproportional amount of time foraging in wetland landcover classes, including wet meadows (Jorgensen and Dinan 2016, Baasch et al. 2021). Although waste grains are known to provide essential caloric requirements to complete migration, recent studies have also documented the importance of aquatic vertebrates and invertebrates in the diets of migrating Whooping Cranes (Neri 2011, Geluso et al. 2013, Urbanek and Lewis 2020, Caven et al. 2021).

Stopover habitats with quality foraging opportunities are thought to be essential for migrating Whooping Cranes to build energy reserves that enable them to complete their 4000 km biannual migration and to arrive at the breeding ground in adequate fitness for successful reproduction (Baasch et al. 2019, Caven et al. 2022 ). Interestingly, Caven et al. (2022) found that stopover duration was predicted by landcover classes associated with what appear to be important Whooping Crane foraging habitats (e.g., palustrine wetlands). Lingle (1987) described diurnal habitat use from 51 Whooping Crane sightings within the CPRV and reported that corn stubble received 37% of the documented diurnal use whereas tilled and natural wetlands received 35%. Given there are wetlands that are farmed, many of these wetland areas may have a more diverse set of food resources because they can sometimes be to too wet to successfully produce a crop and become more grass– and forb–dominated, especially during wet years. Though we did not directly observe or study foraging behaviors, wetland portions of low-elevation grasslands, such as wet meadows and wetland areas in agricultural fields, clearly provide important food resources for Whooping Cranes during migration (Neri 2011, Geluso et al. 2013, Urbanek and Lewis 2020). However, lowland grasslands and embedded wetlands have experienced substantial losses and degradation because of agricultural expansion and development during the past half century (Sidle et al. 1989, Samson and Knopf 1994, Noss et al. 1995, Ricketts et al. 1999, Samson et al. 2010; WWFC 1988, personal communication). Protection of these landscapes, as well as the hydrological regime that sustains them, remain critical objectives for Whooping Cranes and, broadly, waterbird conservation in the CPRV and similar riverine ecosystems throughout the Great Plains (Currier 1989, Henszey et al. 2004, Baasch et al. 2019, Caven et al. 2020, Brinley Buckley et al. 2021).

Protecting expanses of agricultural and herbaceous landcovers free of human development with limited woodland landcover near to the Platte River that contains significant and functional wetland footprints may provide the most desirable set of conditions for supporting Whooping Cranes during migration stopovers in the CPRV. Similar to other research, we found having riverine landcover within 1000-m buffer had a strong influence on the likelihood of Whooping Crane occurrence (Howe 1989, Austin and Richert 2005, Niemuth et al. 2018, Baasch et al. 2019). Conversely, we found woodland and developed landcover, including both structures and roads, within a 1000-m buffer had a negative association with the likelihood of Whooping Crane occurrence. These findings are similar to other studies that reported Whooping Cranes generally select foraging and roosting locations that are located away from woodland edges, roads, and human development (Pearse et al. 2017, Neimuth et al. 2018). This suggests that maintaining relatively treeless and undeveloped expanses of land, especially adjacent to wetland habitats and the main channel of the Platte River, could increase the likelihood of Whooping Crane occurrence. Protecting or restoring large expanses of habitat near key Whooping Crane use sites through easement or purchase will be essential to ensure that the CPRV remains a valuable stopover site for the Aransas-Wood Buffalo population of Whooping Cranes. Whooping Cranes do not exhibit a significant association with upland agricultural landcover and have a marginal negative association with upland prairie landcover. Nonetheless, protecting these habitats with the goal of preserving embedded wetlands and limiting human development is probably necessary to effectively maintain undisturbed Whooping Crane habitat within the CPRV.

The historic drainage and conversion of wetland to agricultural landcover and human development in the CPRV have reduced the availability of terrestrial wetland habitats that are most likely to be occupied by Whooping Cranes (Krapu 1981, Currier et al. 1985, Sidle et al. 1989, NRC 2004). Additional reductions in stream flows resulting from water development will continue to reduce the frequency and duration of wetland inundation, which will further reduce the availability and productivity of wetland landcovers (Brinley Buckley et al. 2021). Conservation efforts have retained remnant native wet meadows within the CPRV and restored other areas back to a prairie-wetland mosaic (Pfeiffer 1999, Meyer et al. 2010). Management efforts during the last 50 years have focused on the restoration of native plants and wetland topography to former agricultural lands, as well as the conversion of forest-dominated accretion to herbaceous habitats adjacent to the Platte River (Pfeiffer 1999, Krapu et al. 2014). To date, little research has been implemented to compare and evaluate wet meadow restoration efforts to determine whether they contain functional components found in native wet meadows (Pfeiffer 1999, Meyer and Whiles 2008, Meyer et al. 2010). Future efforts should evaluate if remnant wet meadow and shallow marsh landcover is more predictive of Whooping Crane occurrence than restored analogs as well as the length of time required for successful restoration of functional wetlands, both of which may influence long-term regional conservation strategies.

MANAGEMENT IMPLICATIONS

Given the continued developmental pressures, the protection of undeveloped landcover near the main channel of the Platte River may be essential for maintaining the long-term importance of the CPRV to Whooping Cranes. In addition, wetland components of agricultural fields provide an important resource to Whooping Cranes. As such, draining or filling these wetlands to improve agricultural production generally reduces the suitability of these landcover types for Whooping Cranes. Similarly, degradation of base flows or future water development, which decreases groundwater levels in adjacent wetlands, may also reduce the suitability of Whooping Crane habitats in the CPRV. Conservation efforts in the CPRV should not abandon investments in the conservation of herbaceous wetland habitats but, rather, focus such efforts on protecting and enhancing landscapes with a higher proportion landcovers with a diverse array of seasonal, shallow wetland habitats (e.g., wet meadows and shallow marshes). Our results also indicate that Whooping Crane terrestrial occurrences take place primarily in areas of lower human development near the Platte River, where most Whooping Cranes roost within the CPRV. As such, protecting these lands from development through the use of land acquisitions and conservation easements may be the best way to ensure functional Whooping Crane habitat within the CPRV into the future. Our findings also emphasize the importance of the scale of landscape considered when assessing habitat use and species occurrence. These findings could be used to prioritize landcover types for intensified conservation efforts and further study to ensure functional critical habitat is maintained within the CPRV for the Aransas-Wood Buffalo Whooping Crane population.

LITERATURE CITED

Aarts, G., M. MacKenzie, B. McConnell, M. Fedak, and J. Matthiopoulos. 2007. 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

Afifi, A., S. May, R.A. Donatello, and V.A. Clark. 2020. Practical multivariate analysis. Sixth Edition. Chapman and Hall/CRC, Boca Raton, Florida, USA.

Austin, J., A. Richert. 2005. Patterns of habitat use by Whooping Cranes during migration: summary from 1977-1999 site evaluation data. Proceedings of the North American Crane Workshop 9:79-104. https://digitalcommons.unl.edu/usgsnpwrc/6/

Baasch, D. M., A. J. Caven, and B. Krohn. 2021. Whooping Crane diurnal behavior and natural history during migration in the Central Great Plains: interim report–Fall 2020. Platte River Whooping Crane Maintenance Trust, Wood River, Nebraska, USA.

Baasch, D. M., P. D. Farrell, A. T. Pearse, D. A. Brandt, A. J. Caven, M. J. Harner, G. D. Wright, and K. L. Metzger. 2019. Diurnal habitat selection of migrating Whooping Crane in the Great Plains. Avian Conservation and Ecology 14(1):6. https://doi.org/10.5751/ACE-01317-140106

Barton, K. 2020. MuMIn: multi-Model inference. CRAN repository, R version 1.43.17. https://CRAN.R-project.org/package=MuMIn

Bates, D., M. Maechler, B. Bolker, and S. Walker. 2015. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67:1-48. https://doi.org/10.18637/jss.v067.i01

Bates, D., M. Maechler, B. Bolker, S. Walker, R. H. B. Christensen, H. Singmann, B. Dai, F. Scheipl, G. Grothendieck, P. Green, J. Fox, A. Bauer, and P. N. Krivitsky. 2021. Package ‘lme4’: Linear Mixed-Effects Models using 'Eigen' and S4. CRAN, R Package Version 1.1-27.1. https://cran.r-project.org/package=lme4

Belaire, J. A., B. J. Kreakie, T. Keitt, and E. Minor. 2014. Predicting and mapping potential Whooping Crane stopover habitat to guide site selection for wind energy projects. Conservation Biology 28:541-550. https://doi.org/10.1111/cobi.12199

Bishop, A., R. Grosse, E. Nugent, and D. Varner. 2015. Rainwater Basin 2004 wetland vegetation map. Rainwater Basin Joint Venture, Grand Island, Nebraska, USA. https://www.sciencebase.gov/catalog/item/570e693be4b0ef3b7ca24d00

Boswell, J. S., and G. A. Olyphant. 2007. Modeling the hydrologic response of groundwater dominated wetlands to transient boundary conditions: implications for wetland restoration. Journal of Hydrology 332:467-476. https://doi.org/10.1016/j.jhydrol.2006.08.004

Brei, J., and A. Bishop. 2008. Platte River vegetation mapping project 2005 land cover methods summary. Platte River recovery implementation program report, Kearney, Nebraska, USA. https://www.rwbjv.org/wp-content/uploads/Platte-River-2005-Land-Cover-Methods.pdf

Bring, J. 1994. How to standardize regression coefficients. American Statistician 48:209-213. https://doi.org/10.2307/2684719

Brinley Buckzley, E. M., A. J. Caven, J. D. Wiese, M. J. Harner. 2021. Assessing the hydroregime of an archetypal riverine wet meadow in the central Great Plains using time-lapse imagery. Ecosphere 12:e03829 https://doi.org/https://doi.org/10.1002/ecs2.3829

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

Canadian Wildlife Service (CWS) and U.S. Fish and Wildlife Service (USFWS). 2005. International recovery plan for the Whooping Crane. Third revision. Recovery of Nationally Endangered Wildlife (RENEW), Ottawa, Ontario, Canada, and U.S. Fish and Wildlife Service, Albuquerque, New Mexico, USA. https://savingcranes.org/wp-content/uploads/2022/05/whooping_crane_recovery_plan.pdf

Caven, A. J., E. M. Brinley Buckley, K. C. King, J. D. Wiese, D. M. Baasch, G. D. Wright, M. J. Harner, A. T. Pearse, M. Rabbe, D. M. Varner, B. Krohn, N. Arcilla, K. D. Schroeder, and K. F. Dinan. 2019. Temporospatial shifts in Sandhill Crane staging in the Central Platte River Valley in response to climatic variation and habitat change. Monographs of the Western North American Naturalist 11:33-76. https://doi.org/10.3398/042.011.0104

Caven, A. J., K. C. King, J. D. Wiese, and E. M. Brinley Buckley. 2017. A descriptive analysis of Regal Fritillary (Speyeria idalia) habitat utilizing biological monitoring data along the big bend of the Platte River, NE. Journal of Insect Conservation 21(2):183-205. https://doi.org/10.1007/s10841-017-9968-0

Caven, A. J., K. D. Koupal, D. M. Baasch, E. M. Brinley Buckley, J. Malzahn, M. L. Forsberg, and M. Lundgren. 2021. Whooping Crane (Grus americana) family consumes a diversity of aquatic vertebrates during fall migration stopover at the Platte River, Nebraska. Western North American Naturalist 81(4):592-607. https://doi.org/10.3398/064.081.0411

Caven, A. J., A. T. Pearse, D. A. Brandt, M. J. Harner, G. D. Wright, D. M. Baasch, E. M. Brinley Buckley, K. L. Metzger, M. R. Rabbe, and A. E. Lacy. 2022. Whooping crane stay length in relation to stopover site characteristics. Proceedings of the North American Crane Workshop 15:6-33

Caven, A. J., M. Rabbe, J. Malzahn, J., and A. E. Lacy. 2020. Trends in the occurrence of large Whooping Crane groups during migration in the Great Plains, USA. Heliyon 6(4):e03549. https://doi.org/10.1016/j.heliyon.2020.e03549

Chávez-Ramírez, F., and E. Weir. 2010. Wet meadow literature and information review. Draft report Platte River Recovery and Implementation Program. Kearney, Nebraska, USA. https://platteriverprogram.org/sites/default/files/PubsAndData/ProgramLibrary/PRRIP%202010_Wet%20Meadow%20Literature%20and%20Info%20Review.pdf

Cunningham, M. A. 2006. Accuracy assessment of digitized and classified land cover data for wildlife habitat. Landscape and Urban Planning 78:217-228. https://doi.org/10.1016/j.landurbplan.2005.08.002

Currier, P. J. 1982. The floodplain vegetation of the Platte River: phytosociology, forest development, and seedling establishment. Dissertation, Iowa State University, Ames, Iowa, USA.

Currier, P. J. 1989. Plant species composition and groundwater levels in a Platte River wet meadow. Pages 19–24 in T. B. Bragg and J. Stubbendieck, editors. Proceedings of the 11th North American Prairie Conference. University of Nebraska Printing. Lincoln, Nebraska, USA. https://digitalcommons.unl.edu/napcproceedings/31/

Currier, P. J., and R. J. Henszey. 1996. Platte River wet meadows: a primer on their flora, fauna, hydrology, management, and restoration. Draft submitted to the U.S. Fish and Wildlife Service (Project No. 14-16-0006-90-917). Grand Island, Nebraska, USA.

Currier, P. J., G. R. Lingle, and J. G. VanDerwalker. 1985. Migratory bird habitat on the Platte and North Platte Rivers in Nebraska. Platte River Whooping Crane Critical Habitat Maintenance Trust, Grand Island, Nebraska, USA.

Currier, P. J., and K. Pfeiffer. 2005. An adaptive approach to river channel management on the Platte River in central Nebraska. Pages 151-154 in F. Chávez-Ramírez, editor. Proceedings of the North American Crane Workshop, January 17-20, 2003. Sacramento, California, USA. https://digitalcommons.unl.edu/nacwgproc/39/

Dappen, P. R., I. C. Ratcliffe, C. R. Robbins, and J. W. Merchant. 2008. Mapping agricultural land cover for hydrologic modeling in the Platte River watershed of Nebraska. Great Plains Research 18:39-52. https://www.jstor.org/stable/23779769

Davis, C. A. 2003. Habitat use and migration patterns of Sandhill Cranes along the Platte River, 1998-2001. Great Plains Research 13:199-216. https://www.jstor.org/stable/23779597

Davis, C. A., and P. A. Vohs. 1993. Role of macroinvertebrates in spring diet and habitat use of Sandhill Cranes. Transactions of the Nebraska Academy of Sciences 20:81-86. https://digitalcommons.unl.edu/tnas/118/

Dean, C. B., J. D. Nielsen. 2007. Generalized linear mixed models: a review and some extensions. Lifetime data analysis 13:497-512. https://doi.org/10.1007/s10985-007-9065-x

Dormann, C. F., J. Elith, S. Bacher, C. Buchmann, G. Carl, G. Carré, J. R.G. Marquéz, B. Gruber, B. Lafourcade, P. J. Leitão, T. Münkemüller, C. McClean, P. E. Osborne, B. Reineking, B. Schröder, A. K. Skidmore, D. Zurell, and S. Lautenbach. 2013. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36:27-46. https://doi.org/10.1111/j.1600-0587.2012.07348.x

Fox, J., and S. Weisberg. 2018. Visualizing fit and lack of fit in complex regression models with predictor effect plots and partial residuals. Journal of Statistical Software 87:1-27. https://doi.org/10.18637/jss.v087.i09

Fox, J., and S. Weisberg. 2019. An R companion to applied regression. Third edition. Sage, Thousand Oaks, California, USA.

Fremgen-Tarantino, M. R., P. J. Olsoy, G. G. Frye, J. W. Connelly, A. H. Krakauer, G. L. Patricelli, and J. S. Forbey. 2021. Assessing accuracy of GAP and LANDFIRE land cover datasets in winter habitats used by greater sage-grouse in Idaho and Wyoming, USA. Journal of Environmental Management 280:111720. https://doi.org/10.1016/j.jenvman.2020.111720

Gallant, A. L. 2009. What you should know about land-cover data. Journal of Wildlife Management 73:796-805. https://doi.org/10.2193/2007-509

Geluso, K., B. T. Krohn, M. J. Harner, and M. J. Assenmacher. 2013. Whooping Cranes consume Plains Leopard Frogs at migratory stopover sites in Nebraska. Prairie Naturalist 45:91-93. https://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1050&context=tpn

Gergel, S. E., Y. Stange, N. C. Coops, K. Johansen, and K. R. Kirby. 2007. What is the value of a good map? An example using high spatial resolution imagery to aid riparian restoration. Ecosystems 10:688-702. https://doi.org/10.1007/s10021-007-9040-0

Hale, R., D. T. Blumstein, R. Mac Nally, and S. E. Swearer. 2020. Harnessing knowledge of animal behavior to improve habitat restoration outcomes. Ecosphere 11(4):e03104. https://doi.org/10.1002/ecs2.3104

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

Hefley, T. J., A. J. Tyre, D. M. Baasch, and E. E. Blankenship. 2013. Nondetection sampling bias in marked presence-only data. Ecology and Evolution 3:5225-5236. https://doi.org/10.1002/ece3.887

Henszey, R. J. 2001. Platte River Whooping Crane Maintenance Trust. Pages 112-113 in C.L. Convis, editor. Conservation geography: case studies in GIS, computer mapping, and activism. ESRI, Inc., Redlands, California, USA.

Henszey, R. J., K. Pfeiffer, and J. R. Keough. 2004. Linking surface-and ground-water levels to riparian grassland species along the Platte River in Central Nebraska, USA. Wetlands 24:665-687. https://doi.org/10.1672/0277-5212(2004)024[0665:LSAGLT]2.0.CO;2

Hong, S. K., S. W. Kim, K. H. Cho, J. E. Kim, S. K. Kang, and D. W. Lee. 2004. Ecotope mapping for landscape ecological assessment of habitat and ecosystem. Ecological Research 19:131-139. https://doi.org/10.1111/j.1440-1703.2003.00603.x

Hou, J., L. Li, Y. Wang, W. Wang, H. Zhan, N. Dai, and P. Lu. 2021. Influences of submerged plant collapse on diet composition, breadth, and overlap among four crane species at Poyang Lake, China. Frontiers in Zoology 18:1-17. https://doi.org/10.1186/s12983-021-00411-2

Howe, M. A. 1989. Migration of radio-marked Whooping Cranes from the Aransas-Wood Buffalo population: patterns of habitat use, behavior, and survival. U.S. Fish and Wildlife Technical Report 21. Patuxent Wildlife Research Center, Laurel, Maryland, USA.

Howlin, S., and K. Nasman. 2017. Correlates of Whooping Crane habitat selection and trends in use in the Central Platte River, Nebraska. Prepared for the Platte River Recovery Implementation Program. Western EcoSystems Technology, Inc., Cheyenne, Wyoming, USA, 140 pp. https://platteriverprogram.org/sites/default/files/PubsAndData/ProgramLibrary/Correlates%20of%20Whooping%20Crane%20Habitat%20Selection%20and%20Trends%20in%20Use%20in%20the%20Central%20Platte%20River.pdf

Hurr, T. R. 1981. Ground-water hydrology of the Mormon Island Crane Meadows wildlife area near Grand Island, Hall County, Nebraska. U.S. Geological Survey Open-File Report 81-1109. https://doi.org/10.3133/ofr811109

Ingegnoli, V. 2013. Landscape ecology: a widening foundation. Springer-Verlag, Berlin, Germany. https://doi.org/10.1007/978-3-662-04691-3

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

Jorgensen, J. G., and L. R. Dinan. 2016. Whooping Crane (Grus americana) behavior, habitat use and wildlife watching visitation during migratory stopover at two wildlife management areas in Nebraska 2015-2016. Nongame Bird Program, Nebraska Game and Parks Commission, Lincoln, Nebraska, USA. https://digitalcommons.unl.edu/nebgamestaff/77/

Kaul, R. B., D. Sutherland, and S. Rolfsmeier. 2006. The flora of Nebraska: keys, descriptions, and distributional maps of all native and introduced species that grow outside cultivation. Conservation and Survey Division, School of Natural Resources, University of Nebraska-Lincoln, Lincoln, Nebraska, USA.

Kleindl, W. J., S. L. Powell, F. R. Hauer. 2015. Effect of thematic map misclassification on landscape multi-metric assessment. Ecological Monitoring and Assessment 187, 321. https://doi.org/10.1007/s10661-015-4546-y

Klijn, F., and H. A. Udo de Haes. 1994. A hierarchical approach to ecosystems and its implications for ecological land classification. Landscape Ecology 9:89-104. https://doi.org/10.1007/BF00124376

Krapu, G.L. 1981. The Platte River Ecology Study: special research report. Northern Prairie Wildlife Research Center, Jamestown, North Dakota, USA. https://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1248&context=usgsnpwrc

Krapu, G. L., D. A. Brandt, P. J. Kinzel, and A. T. Pearse. 2014. Spring migration ecology of the mid-continent sandhill crane population with an emphasis on use of the Central Platte River Valley, Nebraska. Wildlife Monographs 189:1-41. https://doi.org/10.1002/wmon.1013

LaGrange, T. 2015. Wetland program plan for Nebraska. Nebraska Game and Parks Commission, Lincoln, Nebraska, USA. https://digitalcommons.unl.edu/nebgamewhitepap/38/

Langford, W. T., S. E. Gergel, T. G. Dietterich, and W. Cohen. 2006. Map misclassification can cause large errors in landscape pattern indices: examples from habitat fragmentation. Ecosystems 9:474-488. https://doi.org/10.1007/s10021-005-0119-1

Lawler, J. J., R. J. O’Connor, C. T. Hunsaker, K. B. Jones, T. R. Loveland, and D. White. 2004. The effects of habitat resolution on models of avian diversity and distributions: a comparison of two land-cover classifications. Landscape Ecology 19:515-530. https://doi.org/10.1023/B:LAND.0000036151.28327.01

Lechner, A. M., W. T. Langford, S. A. Bekessy, and S. D. Jones. 2012. Are landscape ecologists addressing uncertainty in their remote sensing data?. Landscape Ecology 27:1249-1261. https://doi.org/10.1007/s10980-012-9791-7

Lingle, G. R. 1987. Status of Whooping Crane migration habitat within the Great Plains of North America. Pages 331-340 in J. C. Lewis, editor. Proceedings of the 1985 Crane Workshop. Platte River Whooping Crane Trust and U.S. Fish and Wildlife Service, Grand Island, Nebraska, USA.

Lingle, G. R., G. A. Wingfield, and J. W. Ziewitz. 1991. The migration ecology of Whooping Cranes in Nebraska, USA. Pages 395-401 in J. Harris, editor. Proceedings of the 1987 International Crane Workshop. International Crane Foundation, Baraboo, Wisconsin, USA.

Mayor, S. J., D. S. Schneider, J. A. Schaefer, and S. P. Mahoney. 2009. Habitat selection at multiple scales. Ecoscience 16:238-247. https://doi.org/10.2980/16-2-3238

McKerrow, A. J., N. M. Tarr, M. J. Rubino, and S. G. Williams. 2018. Patterns of species richness hotspots and estimates of their protection are sensitive to spatial resolution. Diversity and Distributions, 24:1464-1477. https://doi.org/10.1111/ddi.12779

Meyer, C. K., and M. R. Whiles. 2008. Macroinvertebrate communities in restored and natural Platte River slough wetlands. Journal of the North American Benthological Society 27:626-639. https://doi.org/10.1899/07-086.1

Meyer, C. K., M. R. Whiles, and S. G. Baer. 2010. Plant community recovery following restoration in temporally variable riparian wetlands. Restoration Ecology 18:52-64. https://doi.org/10.1111/j.1526-100X.2008.00451.x

Nakagawa, S., and H. Schielzeth. 2013. A general and simple method for obtaining R² from generalized linear mixed-effects models. Methods in Ecology and Evolution 4(2): 133-142. https://doi.org/10.1111/j.2041-210x.2012.00261.x

National Research Council (NRC). 2004. Endangered and threatened species of the Platte River. National Academy of Science, Washington, D.C., USA.

Naveh, Z. 1994. From biodiversity to ecodiversity: a landscape-ecology approach to conservation and restoration. Restoration Ecology 2:180-189. https://doi.org/10.1111/j.1526-100X.1994.tb00065.x

Neri, H. 2011. A characterization of the diets of wild and reintroduced Whooping Cranes (Grus americana). Thesis. Hood College, Frederick, Maryland, USA.

Niemuth, N. D., A. J. Ryba, A. T. Pearse, S. M. Kvas, D. A. Brandt, B. Wangler, J. E. Austin, and M. J. Carlisle. 2018. Opportunistically collected data reveal habitat selection by migrating Whooping Cranes in the U.S. Northern Plains. Condor: Ornithological Applications 120:343-356. https://doi.org/10.1650/CONDOR-17-80.1

Noss, R. F., E. T. LaRoe III, and J. M. Scott. 1995. Endangered ecosystems of the United States: a preliminary assessment of loss and degradation. Biological Report 28. National Biological Service, Washington, D.C., USA. https://ecos.fws.gov/ServCat/DownloadFile/101448?Reference=61075

O’Neal, B. J., J. D. Stafford, and R. P. Larkin. 2012. Stopover duration of fall-migrating dabbling ducks. Journal of Wildlife Management 76:285-293. https://doi.org/10.1002/jwmg.283

Pearse, A. T., D. A. Brandt, D. M. Baasch, M. T. Bidwell, J. A. Conkin, M. J. Harner, W. Harrell, and K. L. Metzger. 2020. Location data for Whooping Cranes of the Aransas-Wood Buffalo population, 2009-2018. Data Release, U.S. Geological Survey, Reston, Virginia, USA. https://doi.org/10.5066/P9Y8KZJ9

Pearse, A. T., M. J. Harner, D. M. Baasch, G. D. Wright, A. J. Caven, and K. L. Metzger. 2017. Evaluation of nocturnal roost and diurnal sites used by Whooping Cranes in the Great Plains, United States. Open-File Report 2016-1209, U.S. Geological Survey, Jamestown, North Dakota, USA. https://doi.org/10.3133/ofr20161209

Pearse, A. T., M. Rabbe, L. M. Juliusson, M. T. Bidwell, L. Craig-Moore, D. A. Brandt, and W. Harrell. 2018. Delineating and identifying long-term changes in the Whooping Crane (Grus americana) migration corridor. PLoS ONE 13:e0192737. https://doi.org/10.1371/journal.pone.0192737

Pfeiffer, K. 1999. Evaluation of wet meadow restorations in the Platte River Valley. Pages 202-206 in J. T. Springer, editor. Proceedings of the 16th North American Prairie Conference, Kearney, Nebraska, USA.

Piersma, T., D. I. Rogers, P. M. González, L. Zwarts, L. J. Niles, I. De Lima Serrano, S. Donascimento, C. D. T. Minton, and A. J. Baker. 2005. Fuel storage rates before northward flights in Red Knots worldwide. Pages 262-273 in R. Greenberg and P. P. Marra, editors. Birds of two worlds: the ecology and evolution of migration. Johns Hopkins University Press, Baltimore, Maryland, USA. https://www.roebuckbay.org.au/wp-content/uploads/2012/01/Book-Birds2Worlds2005-comparative-knot-fueling.pdf

Platte River Recovery Implementation Program (PRRIP). 2006. Final Platte River Recovery Implementation Program Adaptive Management Plan. White paper, U.S. Department of the Interior, State of Wyoming, State of Nebraska, and State of Colorado. https://platteriverprogram.org/sites/default/files/PubsAndData/ProgramLibrary/PRRIP%202006_Attachment%203_AMP.pdf

Platte River Recovery Implementation Program (PRRIP). 2012. Platte River Wet Meadow geohydrology and management through flow releases. Office of the Executive Director of the PRRIP, Kearney, Nebraska, USA. https://platteriverprogram.org/sites/default/files/PubsAndData/ProgramLibrary/PRRIP%202012_Wet%20Meadow%20Geohydrology%20and%20Management%20Through%20Flow%20Releases.pdf

Powell, M. J. D. 2009. The BOBYQA algorithm for bound constrained optimization without derivatives. Report No. DAMTP 2009/NA06. Centre for Mathematical Sciences, University of Cambridge, Cambridge, UK. https://www.damtp.cam.ac.uk/user/na/NA_papers/NA2009_06.pdf

R Core Team. 2020. R: a language and environment for statistical computing, version 4.0.2. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/

Rasool, M. A., M. A. Hassan, X. Zhang, Q. Zeng, Y. Jia, L. Wen, and G. Lei. 2021. Habitat quality and social behavioral association network in a wintering waterbirds community. Sustainability 13(11):6044. https://doi.org/10.3390/su13116044

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

Renfrew, R. B., D. H. Johnson, G. R. Lingle, and W. D. Robinson. 2006. Avian response to meadow restoration in the central Great Plains. Pages 313-324 in J. T. Springer and E. C. Springer, editors. Proceedings of the 20th North American Prairie Conference. Kearney, Nebraska, USA. https://digitalcommons.unl.edu/usgsnpwrc/236/

Ricketts, T. H., E. Dinerstein, D. M. Olson, C. J. Loucks, W. Eichbaum, D. A. DellaSala, K. Kavanagh, P. Hedao, P. Hurley, K. Carney, R. Abell, and S. Walters. 1999. Terrestrial ecoregions of North America: a conservation assessment. Island Press, Washington, D.C., USA.

Rolfsmeier, S. B., and G. Steinauer 2010. Terrestrial ecological systems and natural communities of Nebraska (version IV). Nebraska Natural Heritage Program, Nebraska Game and Parks Commission, Lincoln, Nebraska, USA. https://outdoornebraska.gov/wp-content/uploads/2015/09/TerrestrialEcologicalSystemsAndNaturalCommunitiesOfNebraska_03092010.pdf

Roloff, G. J., and B. J. Kernohan. 1999. Evaluating reliability of habitat suitability index models. Wildlife Society Bulletin 27:973-985. https://www.jstor.org/stable/3783655

Samson, F. B., and F. L. Knopf. 1994. Prairie conservation in North America. Bioscience 44:428-421. http://dx.doi.org/10.2307/1312365

Samson, F. B., F.L. Knopf, and W.R. Ostlie. 2010. Great Plains ecosystems: past, present, and future. Wildlife Society Bulletin 32:6-16. https://doi.org/10.2193/0091-7648(2004)32[6:GPEPPA]2.0.CO;2

Seirup, L., and G. Yetman. 2006. U.S. Census Grids (summary file 3), 2000. NASA Socioeconomic Data and Applications Center (SEDAC), Palisades, New York, USA. https://doi.org/10.7927/H42R3PMN

Sidle, J. G., E. D. Miller, and P. J. Currier. 1989. Changing habitats in the Platte River Valley of Nebraska. Prairie Naturalist 21:91-104.

Sparling, D. W., and G. L. Krapu. 1994. Communal roosting and foraging behavior of staging Sandhill Cranes. Wilson Bulletin 106:62-77. https://www.jstor.org/stable/4163386

Tacha, M., A. Bishop, and J. Brei. 2010. Development of the Whooping Crane tracking project geographic information system. Pages 98-104 in B. K. Hartup, editor. Proceedings of the Eleventh North American Crane Workshop. North American Crane Working Group, Baraboo, Wisconsin, USA. https://digitalcommons.unl.edu/nacwgproc/144/

Tiner, R. W. 2016. Wetland indicators: a guide to wetland formation, identification, delineation, classification, and mapping. Second edition. CRC press, Boca Raton, Florida, USA. https://doi.org/10.1201/9781315374710

U.S. Department of Agriculture-Natural Resources Conservation Service (USDA-NRCS). (2019). Web Soil Survey. Soil Survey Division. http://websoilsurvey.nrcs.usda.gov/

U.S. Department of Agriculture-Natural Resources Conservation Service (USDA-NCRS). 2020. Soil Survey Geographic (SSURGO) Database. https://sdmdataaccess.sc.egov.usda.gov/Query.aspx

U.S. Fish and Wildlife Service (USFWS). 1978. Determination of critical habitat for the Whooping Crane. Federal Register 43:20938-20942. https://digitalcommons.unl.edu/endangeredspeciesbull/44/

U.S. Fish and Wildlife Service (USFWS). 2021. National Wetlands Inventory website. Website. U.S. Fish and Wildlife Service, Washington, D.C., USA. http://www.fws.gov/wetlands

Urbanek, R. P., and J. C. Lewis. 2020. Whooping Crane (Grus americana). Version 1.0. In A. F. Poole, editor. Birds of the world (A. F. Poole, Editor). Cornell Lab of Ornithology, Ithaca, NY, USA. https://doi.org/10.2173/bow.whocra.01

VerCauteren, T. L. 1998. Local scale analysis of Sandhill Crane use of lowland grasslands along the Platte River, Nebraska. Thesis. University of Nebraska-Lincoln, Lincoln, Nebraska, USA.

Wagenmakers, E. J., and S. Farrell. 2004. AIC model selection using Akaike weights. Psychonomic Bulletin and Review 11:192-196. https://doi.org/10.3758/BF03206482

Wesche, T. A., Q. D. Skinner, and R. J. Henszey. 1994. Platte River wetland hydrology study: final report. Technical report QQRC-94-07, U.S. Bureau of Reclamation.

Wilcox, J. C., M. T. Healy, and J. B. Zedler. 2007. Restoring native vegetation to an urban wet meadow dominated by reed canarygrass (Phalaris arundinacea) in Wisconsin. Natural Areas Journal 27:354-366. https://doi.org/10.3375/0885-8608(2007)27[354:RNVTAU]2.0.CO;2

Wilen, B. O., and M. K. Bates. 1995. The U.S. Fish and Wildlife Service’s National Wetlands Inventory Project. Pages 153-169 in C. M. Finlayson and A. G. Valk, editors. Classification and inventory of the world’s wetlands. Springer, Dordrecht, The Netherlands. https://doi.org/10.1007/978-94-011-0427-2