Skip to main content
NCBI home page
Springer logoLink to Springer
. 2025 Dec 5;198(1):11. doi: 10.1007/s10661-025-14784-y

Climate change refugia in Canadian prairies: assessing range shifts and identifying breeding habitats for grassland songbirds

Rinjan Shrestha 1,✉, Joyce Arabian 2, Caroline Martin 3, Will Merrit 4, Emily Giles 1, James Snider 1
PMCID: PMC12680839  PMID: 41348246

Abstract

The identification of climate change refugia is fundamental for climate-smart conservation planning, especially in highly altered landscapes, such as temperate grasslands. Our study aimed to identify breeding refugia for three focal grassland birds: Baird’s sparrow (Centronyx bairdii), Sprague’s pipit (Anthus spragueii), and thick-billed longspur (Rhynchophanes mccownii) across the Canadian prairies. We used species distribution models to identify breeding refugia within the climatically suitable range for two time periods (2050 and 2080) under two of the most likely climatic scenarios (“intermediate scenario” RCP 4.5 and “worst-case scenario” RCP 8.5). In doing so, we demonstrate the importance of incorporating species-specific dispersal ability and projected shifts in grassland habitats in the analyses. Our study predicts a northward shift in the breeding ranges of all three bird species under both climate scenarios, with almost 100% loss of their current breeding habitat. However, all species are expected to gain bioclimatic space outside of their current range under RCP 4.5 in 2050 and 2080. Further increases in emissions under the RCP 8.5 scenario will likely cause Baird’s sparrow to lose bioclimatic space both in 2050 and 2080, and the same is true for the other two species only in 2080. Approximately 80% of currently suitable habitats for the focal species are located outside protected areas. As the climate warms, almost 100% of future breeding refugia for all birds are likely to reside outside protected areas in all climate change scenarios. Our study provides a framework for climate-integrated conservation planning for the wide-ranging migratory species.

Supplementary information

The online version contains supplementary material available at 10.1007/s10661-025-14784-y.

Keywords: Climate-smart conservation planning, Climate change refugia, Grassland birds, Range shifts, Species distribution

Introduction

Human-induced climate change, occurring at rates faster than any climate change over the past 65 million years (Diffenbaugh & Field, 2013), is affecting the distribution, abundance, and persistence of species and ecosystems around the world (Turner et al., 2020). Thus, there is an urgent need for pragmatic strategies to secure vulnerable species and their ecosystems now and into the future. Such strategies call for identifying resilient bioclimatic niches by accounting for species-specific ecological and behavioral needs under a changing climate (Groves et al., 2012). In this regard, the detection, protection, and restoration of climate change refugia (Keppel et al., 2012) has been acknowledged as one of the fundamental approaches for climate-smart conservation planning (Carroll & Noss, 2020; Langham et al., 2015).

Climate change refugia are the potentially suitable habitats where species take refuge and may grow under changing climatic conditions (Brambilla et al., 2022; Keppel et al., 2012). As such, spatially explicit climate change refugia can represent both “in situ refugia”—specific habitats within species’ current distributional range that are expected to be more capable of resisting influence of climate change than other areas—and “ex situ refugia”—sites outside current species’ distributional range where they may more readily disperse as climate conditions change over time (Ashcroft, 2010; Keppel et al., 2012). The fundamental prerequisites for identifying both in situ and ex situ climate change refugia are understanding the bioclimatic niche of a species, i.e., the range of bioclimatic conditions that it can withstand (Ashcroft, 2010), and its ability to track climate velocity, i.e., a measure of the speed of travel required to match up with climate changing conditions (Williams & Blois, 2018).

The temperate grasslands are one of the most intensively developed landscapes in the world (Hoekstra et al., 2005), with losses due to conversion into agricultural land, urban development, oil and gas extraction, and fire suppression (Bernath-Plaisted et al., 2023a). Historically, these grasslands were regulated by wildfire and grazing by bison in Canada (Anderson, 2006). By the late 1800 s, bison were hunted to extirpation, and the subsequent European settlement resulted in the massive conversion of native grasslands to crop-dominated landscapes.

Critically endangered Canadian prairies (Olson & Dinerstein, 2002) are habitat for many birds, including 26 avian grassland obligates (Partners in Flight, 2021) known to be highly sensitive to climate and weather conditions in the region (Morelli et al., 2020). These grasslands also constitute heavily altered ecosystems due to conversion into agricultural land (World Wildlife Fund, 2021). Consequently, over 70% of native mixed and short grass prairie have already been converted to row-crop agriculture in Canada, the greatest conversion reported for any major ecological community on the continent (Gauthier & Wiken, 2003). Moreover, ongoing global climate change is expected to adversely affect the existing prairie region (Sauchyn et al., 2020), as the greatest climate velocities are negatively correlated with slope and positively correlated with latitude (Loarie et al., 2009). Accordingly, recent studies suggest that more than 70% of grassland birds are threatened by climate-induced range shifts (Wilsey et al., 2019). These birds are predicted to suffer more during the breeding season compared to the non-breeding season (Langham et al., 2015) due to the substantial loss of their breeding ranges that are mostly located in Canada and the northern USA (Wilsey et al., 2019).

Species may be able to keep pace with climate change in large, protected areas with greater connectivity (Paterson et al., 2024; Elsen et al., 2020; Roberts et al., 2020). Accordingly, a recently adopted global biodiversity framework embraces area-based conservation measures to guide conservation actions, highlighting the importance of connectivity conservation (Convention on Biological Diversity, 2021). However, there is currently a mismatch in projected climate vulnerabilities and existing protected area coverage in North America (Carroll et al., 2017; Gahbauer et al., 2022). This is particularly true for three grassland specialist birds of continental conservation concern: Baird’s sparrow (Centronyx bairdii; hereafter referred to as BAIS), Sprague’s pipit (Anthus spragueii; hereafter referred to as SPPI), and thick-billed longspur (Rhynchophanes mccownii, previously known as the McCown’s longspur; hereafter referred to as TBLO), which are undergoing persistent population declines (Sauer et al., 2014) and are also highly threatened by climate change impacts (Wilsey et al., 2019).

In this study, we aim to identify breeding bioclimatic refugia for the three aforementioned grassland songbirds by (i) using species distribution models (SDMs) to predict the current distribution of climatically suitable habitats across their breeding ranges in Canadian prairies, (ii) assessing the effect of species dispersal ability and shifts in grassland in identifying bioclimatic refugia for grassland birds, and (iii) identifying bioclimatic refugia within the climatically suitable range for two time periods (2050 and 2080) under two of the most likely climatic scenarios: RCP 4.5 “intermediate scenario” and RCP 8.5 “worst-case scenario.” We also (iv) examine the current protected area coverage with respect to the spatial extent of present and future bioclimatic refugia.

Previous studies aimed at assessing climate-induced range shifts have largely ignored species dispersal ability, thereby leading to potential overestimates of range expansions and underestimates of range contractions (Schloss et al., 2012; Wilsey et al., 2019). Likewise, there is also a dearth of studies that incorporated an independent assessment of potential shifts in grassland, even though wildfire and drought are predicted to facilitate the transition from boreal forest to grassland as climate change increases the frequency and severity of forest fires (Scheffer et al., 2012; Stralberg et al., 2018). Changes from boreal forest into deciduous forests, shrublands, and grasslands have already been detected (Wang et al., 2020), and the vegetation community shifts along this grassland-boreal ecotone will particularly impact the grassland specialist birds included in this study (Nixon et al., 2016). Furthermore, to our knowledge, studies undertaken so far have used the same set of environmental variables for all species, rather than the selection of variables for each species to model climatic niches, which is one of the most critical aspects for any conservation planning (Michalak et al., 2020). This is especially true considering that individual species vary greatly in their capacity to adapt to and cope with rates of change, suggesting that the range shift of each species depends on the cumulative effect of several behavioral or internal traits and environmental drivers of change (Chen et al., 2011).

This study adopts robust techniques to address these pertinent issues and incorporates both intrinsic (e.g., species ability to disperse) and extrinsic (e.g., shifts in grassland habitats) factors. In addition, we also perform species-specific selection of environmental variables to model climatic niches and build maps of bioclimatic refugia for the three focal grassland bird species across their breeding ranges in Canada.

Materials and methods

Study area

Our study area comprises the temperate grassland ecoregion of Canada (Fig. 1), characterized by strong seasonality, with cold, dry winters followed by periods of drought during the warmer months. The region generally receives relatively low annual precipitation between 250 and 600 mm, mostly in the form of rain in the spring and summer. The average annual temperatures range between 4 and 16 °C with extreme variability between seasons; the summer maximum and minimum temperatures are 22 and 28 °C higher, respectively, than those in winter across the prairie region (McGinn, 2010).

Fig. 1.

Fig. 1

Open in a new tab

Study area encompassing the temperate grasslands of Canada across the provinces of Alberta (AB), Saskatchewan (SK), and Manitoba (MB). Insets a), b), and c) show the current breeding ranges of Baird’s sparrow (BAIS), Sprague’s pipit (SPPI), and thick-billed longspur (TBLO), respectively.

In this study, we focused on the northern breeding ranges of grassland birds across Alberta, Manitoba, and Saskatchewan (Fig. 1) for two primary reasons. First, the northern boundary of the grassland ecozone provides insight into the earliest impact of climate change (Schwarz & Wein, 1990). Range edge dynamics will influence changes in populations and/or communities of birds as climate change causes geographic shifts in natural ecoregions (Davidson et al., 2020). Second, ranges of grassland birds are predicted to shift northward, and we aim to identify areas of potential in situ and ex situ refugia to guide conservation management now and into the future.

Study species

BAIS, SPPI, and TBLO are grassland specialist birds endemic to the northern grasslands of North America with differing natal dispersal abilities (Bateman et al., 2020). All these birds migrate to Canada during summer (Fig. 1). While BAIS and SPPI breed in southern Alberta, Saskatchewan, and Manitoba (Fig. 1a and b), the historic breeding range of TBLO also encompassed these provinces but has since contracted to only southern Alberta and southwestern Saskatchewan over the past century (Fig. 1c; Stewart, 1975). Canada currently contains approximately 76%, 56%, and 14% of the total breeding populations of BAIS, SPPI, and TBLO, respectively (ECCC & BC, 2024). Due to rapid loss of habitats, 74%, 90%, and 98% decline in populations of BAIS, SPPI, and TBLO, respectively, have been reported since 1970 (ECCC & BC, 2024). Consequently, BAIS is designated as “Special Concern,” and both TBLO and SPPI are assessed as “Threatened” in Canada (COSEWIC, 2010, 2012, 2016).

BAIS breed in native mixed grass and fescue prairie with few shrubs, and may tolerate a wider variety of anthropogenic habitats than other grassland-obligate species (Davis, 2004). SPPI breed in the moist mixed and mixed grassland ecoregions of Canada in open native grasslands, and are found rarely in non-native grassland cover. They are considered area-sensitive and have the most breeding success in larger patches of grassland (Davis, 2004). TBLO breeds in grassland habitats with short grasses maintained by grazing, fire, and drought (With, 2010). Features of native grasslands that TBLO depends on can sometimes be met by non-native habitats, such as tame hay and croplands, but non-native habitats can reduce breeding success and negatively affect long-term survival of the species (COSEWIC, 2016).

Data collection

Presence data

We compiled presence data from the North American Breeding Bird Survey (BBS), one of the most comprehensive and systematic breeding bird surveys in the Northern Hemisphere (Sauer et al., 2014). These surveys are conducted annually during the breeding season (late May to early July) by skilled observers once per season along the approximately 40 km survey routes, each with 50 stops. At each stop, observers record every bird seen within a 0.4-km radius or heard during a 3-min point count. We used start points of each survey route to reflect the presence records (Phillips et al., 2004; Veech et al., 2017), for a total of 1137 points for BAIS, 1543 for SPPI, and 369 for TBLO, and summarized across years for each survey route between 1981–2010 (Nixon et al., 2016; Supplementary Information 1).

Climate data

Current and future climate data were downloaded from AdaptWest—a climate projection database for North America (AdaptWest Project, 2015). Ensemble projections, an average of 15 Coupled Model Intercomparison Project Phase 5 (CMIP5) models, were used for this assessment, and data were generated by downscaling projections from the CMIP5 database. These data are based on Representative Concentration Pathways (RCP) scenarios used in the IPCC Fifth Assessment Report (Pachauri et al., 2014). For our study, the two most likely climate change scenarios were considered, RCP 4.5 and RCP 8.5, for the years 2050 and 2080 (Hébert et al., 2020). These future projections were compared to the baseline scenario of 1981 to 2010.

Twenty-seven climatic variables mainly consisting of seasonal and annual means for temperature and precipitation, evapotranspiration, extremes, snowfall, and drought indices were included in the analyses (Supplementary Information 2; Stralberg et al., 2015). These climate data were used both to predict the spatial extent of grasslands as well as the climatically suitable habitat for three focal bird species under the baseline and future climate scenarios at a resolution of 1 km × 1 km.

To assess the potential impact of spatial scale mismatch between occurrences and climate data, we conducted a sensitivity analysis by extracting climate summaries at the route start point, within a 5-km buffer of the route start point, and within a 40-km buffer of the route start point (i.e., approximate length of each BBS route). Results indicated no statistically significant differences among these scales (Supplementary Information 6), supporting the use of start points as proxies for route-level climate conditions.

Modelling climatic niche

We first performed species-specific selection of environmental variables and identified the optimal beta-multiplier by using the “MaxentVariableSelection” R package (Jueterbock et al., 2016). Excessively complex, as well as simple models, are known to have reduced performance in modelling species range shifts (Warren et al., 2014). By performing species-specific selection, we used a stepwise process to identify a set of environmental variables with a relative contribution of > 5% and a correlation coefficient (Pearson) of < 0.9 or > − 0.9 among the 27 environmental variables (Jueterbock et al., 2016). The beta-multiplier was optimized by running the analyses for values ranging from 0 to 15 in increments of 0.5. Subsequently, the model with the lowest sample corrected Akaike Information Criterion (AICc; Akaike, 1974) was selected as the best model (Warren et al., 2014; Warren & Seifert, 2011; Supplementary information 3).

We then employed the maximum entropy method of climatic niche modeling (hereafter referred to as “Maxent”; Phillips et al., 2004) to predict climate suitability for the baseline period and projections for both 2050 and 2080 time periods under RCP 4.5 and 8.5 emission scenarios for each species of birds. Maxent associates occurrences of a species to environmental variables following the principles of maximum entropy (Phillips et al., 2009). In doing so, it uses presence-only data, thereby not requiring data of confirmed absences, and is known to provide robust results even with sparse and irregularly sampled data with minor location errors (Elith et al., 2010; Phillips et al., 2009).

The program “MaxEnt” (Version 3.4.1; Phillips & Dudík, 2008) was used to construct species distribution models by allocating 30% of the presence data for model testing and 70% as a training dataset in each of 15 bootstrapped replicates. In order to account for potential nonlinearities in interactions among environmental (predictor) variables, we applied linear, quadratic, product, and threshold features. We then used average models to project into current and future climate space to predict climate suitability for each bird species across their breeding range in Canada.

To address spatial bias in sample representation (Phillips et al., 2009) and spatio-temporal autocorrelation (Veloz, 2009) that primarily affects delineating sites with high conservation significance (Reddy & Dávalos, 2003), we ensured that background samples were distributed in a way that reflected the sampling effort (Elith et al., 2010). Accordingly, we used the overlapped target group background data approach by including the entire set of presence data of the three bird species into the background data set (Supplementary information 1; Phillips et al., 2009). This technique improves the predictive performance across a wide range of species distribution modelling techniques by ensuring a more reliable prediction in less sampled areas and reducing the dependency of sampling effort (Phillips et al., 2009). We used the area under the receiver operating characteristic curve (AUC) to evaluate the goodness of fit of the model, where AUC values range from 0 to 1, and AUC > 0.7 indicates good model performance.

High-resolution predictive maps (1 km × 1 km) were generated for each time period, emissions scenario, and species. Projected climate suitability was converted into binary suitable/unsuitable climate maps by applying a true skill statistic (TSS) based maximum sensitivity and specificity threshold (Liu et al., 2016).

Modelling grassland shift

To evaluate how grassland habitats may be affected by a changing climate, we modelled potential shifts in the spatial extent of the grassland bioclimatic envelope. While numerous model types have previously been applied to model climate-related ecosystem shifts, the Random Forest (RF) has shown itself to be flexible and robust in this application, as well as capable of slightly outperforming other options (Roberts & Hamann, 2012). For this reason, we employed a RF model that was implemented through sklearn 0.20 for Python 3.7 to project shifts in the bioclimatic envelope suitable for Canadian grasslands (refer to Supplementary information 4 for details).

The model was run using the same suite of 27 annual environmental variables which was used in the climate projections for bird species for the same time periods and resolution as described above. Sample training points were randomly selected from the current extent of grassland ecosystems. As a means of balancing the tradeoff between sufficient sample size and processing time, the number of training points drawn for each ecosystem type was calculated based on a logarithmic function with a minimum sample size of 2000 points (Hamann & Wang, 2006).

ni=2000+n·ln106.110100 1

where n is the maximum number of sample points which could be taken from an ecoregion based on the raster resolution and ecoregion size (1 sample/km2). Once the base RF model was trained using the baseline historic climate data, it was then applied to future climate scenarios RCP 4.5 and RCP 8.5 for the study region to produce spatial predictions on the extent of the grassland bioclimatic envelope. The RF model was subsequently assessed through the Out-of-Bag (OOB) score and classification accuracy when applied to partitioned testing data.

Since studies indicate that northern wetland/peatlands are likely to be resilient to climate change (Schneider et al., 2015), regions classified as wetlands within the agri-food Canada land cover data were held constant in future projections. Additionally, anthropogenic land cover types (e.g., agriculture, urban areas) were similarly held constant through future projections.

Determining dispersal limit

Following Wilsey et al. (2019) and Bateman et al. (2020), we derived the annual dispersal rate by dividing the mean natal dispersal distance (Bateman et al., 2020) by the generation time (Bird, 2020). The dispersal limit of each species of birds was then obtained by multiplying the annual dispersal rate by the number of years away between the current and future projection, depending on the climate scenario (Table 1).

Table 1.

Calculation of annual dispersal rate (km/y) using mean natal dispersal distance (Bateman et al., 2020) and generation length (Bird, 2020) for Baird’s Sparrow (BAIS), Sprague’s Pipit (SPPI), and Thick-billed Longspur (TBLO)

Species Mean natal dispersal distance (km) Generation length (yrs) Annual dispersal rate (km/yr)
BAIS 3 2.14 1.41
SPPI 37 1.94 19.07
TBLO 14.5 2.50 5.8
Open in a new tab

Mapping bioclimatic refugia

We first evaluated the significance of dispersal ability and grassland shift in mapping bioclimatic refugia for all three species of birds. In doing so, we ran four separate analyses: (i) raw projections that did not include grassland shift and species dispersal ability, (ii) projections that included grassland shift only, (iii) projections that included dispersal ability only, and (iv) projections that included both grassland shift and dispersal ability.

Species dispersal ability and grassland shift were included in the analyses by masking out areas that were not projected to be grassland and were beyond dispersal limits for each species, time period, and emissions scenario. The projections that included both grassland shift and dispersal ability were identified as bioclimatic refugia, and the sites located within and outside the current distributional range were termed as “in situ refugia” and “ex situ refugia,” respectively. The resulting maps thus reflect not only the future bioclimatic niche of the species but also species’ biological limits to dispersal.

To evaluate projected bioclimatic refugia in terms of protected areas coverage, we used the Canadian Protected and Conserved Areas Database (Environment and Climate Change Canada, 2021). This database includes provincial and national parks, as well as other protected areas such as Ecological Reserves, Wilderness Areas, Heritage Rangelands, and Provincial Recreational Areas. The extent of bioclimatic refugia that occurs within the protected area for each species of birds and emission scenario was obtained by masking out the corresponding geospatial layer of bioclimatic refugia that occurs outside of the protected areas’ network.

Results

Projection of climatically suitable space

The model-averaged AUC (Test) for each of all three species was > 0.8, suggesting a good fit of the selected models (Table 2). The climate variables such as summer precipitation, degree-days below 18 °C, the number of frost-free days, mean coldest month temperature, and winter mean temperature were common in the selected models of BAIS, SPPI, and TBLO. Except for summer precipitation and degree-days below 18 °C, climate variables were positively associated with the fundamental niche of these birds. However, the most influential climate variables governing their distribution differed, as the climate moisture deficit (60%), evaporation (40%), and summer precipitation (90%) contributed most to BAIS, SPPI, and TBLO, respectively (Supplementary information 2).

Table 2.

Predictive performance of models as shown by AUC (test) values in the average Maxent models for Baird’s sparrow (BAIS), Sprague’s pipit (SPPI), and thick-billed longspur (TBLO) in the Canadian prairies. N and SD denote the number of presence points and standard deviations, respectively

Species N Test AUC SD No of variables Beta-multiplier
BAIS 1137 0.817 0.008 27 1.5
SPPI 1543 0.806 0.009 17 1.5
TBLO 369 0.935 0.007 6 1.5
Open in a new tab

Likewise, there was an effect of grassland shift on the projected suitable bioclimatic space of all three species under all climate emission scenarios, with a decrease of projected suitable bioclimatic space of approximately 30 to 70% when incorporating grassland shift in the models (Supplementary information 5). The same was not true, however, with regard to the dispersal ability, as only the projected bioclimatic space of BAIS was found to be affected, and there was almost no effect on SPPI, while TBLO was only moderately affected (Supplementary information 5). The combined effect of grassland shift and dispersal ability on projected bioclimatic space followed a similar pattern, with a difference between the models with and without combined grassland shift and dispersal ability for BAIS (Fig. 2a) and TBLO (Fig. 2c), while the suitable range of SPPI (Fig. 2b) remained unchanged when grassland shift was accounted for.

Fig. 2.

Fig. 2

Open in a new tab

The effect of grassland shift, and dispersal ability/grassland shift combined in estimating suitable bioclimatic niche for Baird’s Sparrow (BAIS; a), Sprague’s pipit (SPPI; b), and thick-billed longspur (TBLO; c) in 2050 and 2080 under RCP 4.5 (intermediate) and RCP 8.5 (worst-case) emission scenarios

Mapping of current and future extent of bioclimatic refugia

All three birds, BAIS, SPPI, and TBLO, have less than 80,000 km2 of currently suitable bioclimatic space in Canada, with BAIS occupying a slightly larger area than the other two birds (Table 3). As the climate warms, our study suggests that there will be a pronounced northward range shift outside of the current breeding ranges of these three species under all climate emissions scenarios considered for this study (Fig. 3). Projections after accounting for combined effect of dispersal ability and grassland shift indicate that, under RCP 4.5, all species will gain bioclimatic space, with approximately 100 and 50% more than the current extent for BAIS in 2050 and 2080, respectively, and about 300% each for SPPI and TBLO in the same period (Fig. 4). Furthermore, the increase in emissions predicted under RCP 8.5 will only result in gains of bioclimatic space for SPPI and TBLO in 2050, with progressive losses of bioclimatic space for BAIS in both 2050 and 2080, and SPPI and TBLO in 2080 (Figs. 3 and 4). When comparing the overlap between current and future species distributions (in situ refugia), there are changes in the location and amount of area predicted to be bioclimatically suitable between emission scenarios. In 2050, BAIS has 12% less ex situ refugia identified in RCP 8.5 compared to RCP 4.5, while SPPI and TBLO have approximately 300 and 150% more ex situ refugia identified in the same period, due to northward shifts of their suitable bioclimatic space (Fig. 3). Under the most extreme scenario, by 2080 under RCP 8.5, all species are predicted to lose bioclimatic space, resulting in no in situ refugia available for BAIS or SPPI, and < 5% of its current distribution available for TBLO (Table 3 and Fig. 3).

Table 3.

Projected area (km2) of baseline bioclimatic space and in situ refugia for Baird’s sparrow (BAIS), Sprague’s pipit (SPPI), and thick-billed longspur (TBLO) in 2050 and 2080 under RCP 4.5 (intermediate) and RCP 8.5 (worst-case) emissions scenarios, compared to baseline (2010)

Species Baseline 2050 2080
RCP 4.5 RCP 8.5 RCP 4.5 RCP 8.5
BAIS 77 147.10 281.34 40.28 80.37 0
SPPI 66 493.23 338.12 151.52 187.47 0
TBLO 48 951.50 52.42 61.36 58.20 41.89
Open in a new tab

Fig. 3.

Fig. 3

Open in a new tab

Projected baseline (2010) and bioclimatic refugia (both in situ and ex situ refugia) for Baird’s Sparrow (BAIS), Sprague’s pipit (SPPI), and thick-billed longspur (TBLO) in 2050 and 2080 under RCP 4.5 (intermediate) and RCP 8.5 (worst-case) emission scenarios. Light yellow areas represent the predicted present distributions of each species, blue areas represent predicted future distributions of each species (ex situ refugia), and dark yellow areas represent the overlap between predicted present and future distributions of each species (in situ refugia)

Fig. 4.

Fig. 4

Open in a new tab

Proportional change in suitable bioclimatic space for Baird’s sparrow (BAIS), Sprague’s pipit (SPPI), and thick-billed longspur (TBLO) in 2050 and 2080 under RCP 4.5 (Intermediate) and RCP 8.5 (Worst-case) emission scenarios with respect to the baseline of 2010. Error bars denote standard deviation

Approximately 80% of the currently suitable habitat for the three bird species is situated outside protected areas in their breeding range in Canada (Table 4). Under the projected climate change scenarios, almost 100% of the ranges for BAIS and TBLO will reside outside current protected areas in Canada. In comparison, SPPI will have 70% of its suitable habitat outside protected areas in 2080 under RCP 8.5 (Table 4).

Table 4.

Suitable bioclimatic space (%) projected to be outside protected areas for Baird’s sparrow (BAIS), Sprague’s pipit (SPPI), and thick-billed longspur (TBLO) in 2050 and 2080 under RCP 4.5 (intermediate) and RCP 8.5 (worst-case) emission scenarios, compared to baseline (2010)

Species Baseline 2050 2080
RCP 4.5 RCP 8.5 RCP 4.5 RCP 8.5
BAIS 79.07 96.73 95.44 95.15 95.75
SPPI 78.07 96.25 91.62 93.16 69.91
TBLO 78.89 97.03 95.86 95.09 93.97
Open in a new tab

Discussion

Modelling approach

We followed a comprehensive approach to modeling current and future distributions of three important grassland birds by accounting for both extrinsic (i.e., bioclimatic) and intrinsic (i.e., behavioral traits) factors. Typically, mapping refugia at large scales has mainly incorporated landscape features and general measures of climatic conditions (Michalak et al., 2020; Carroll et al., 2017; Lawler et al., 2015). In addition, our approach of incorporating species-specific climate niches (Michalak et al., 2020), which is not accounted for by species-neutral or blanket approaches, can provide a more biologically meaningful characterization of refugia (Willis et al., 2015). As individual species differ in their tolerance of climate change, incorporating species-specific information to identify refugia may enhance our understanding of vulnerable wildlife populations in the face of ongoing anthropogenic climate change and inform management decisions (Michalak et al., 2020).

We included dispersal abilities of individual species as a key behavioral trait affecting their access to climatically suitable areas as their ranges shift over time. Grassland songbirds are highly sensitive to climate and weather fluctuations (Wiens, 1974), and their ability to respond to short-term changes in habitat conditions (caused by inter-annual weather variations) may facilitate range shifts in response to climate change as new sites comprising suitable land cover and climate emerge (Skagen & Adams, 2012). However, the inherent variability in dispersal ability across species may ultimately govern the range shift by these birds. Therefore, it is imperative to consider dispersal abilities in species distribution models to avoid overestimating range expansion and underestimating range contraction (Schloss et al., 2012). For example, in our study, dispersal ability ranged from 1.4 km y−1 for BAIS to 5.8 km y−1 for TBLO to 19.07 km y−1 for SPPI, and our analyses clearly demonstrate the dependence of the extent of accessible bioclimatic niche on species dispersal ability, thereby underscoring the likely importance of including dispersal ability in the predictive modelling of migratory species’ ranges.

External factors, such as the rate of transition of boreal and parkland vegetation to grassland along the grassland-boreal ecotone, will have a significant effect on species’ abilities to colonize future habitats (Nixon et al., 2016). Species that depend heavily on native grasslands such as TBLO and SPPI will be especially impacted by these changes in land cover. Therefore, we included projected shifts in grassland habitats into our modeling framework as one of the key filters to assess the spatial extent of bioclimatic refugia now and into the future.

These grassland bird species will rely on either natural dispersal or assisted relocation to keep pace with the geographic shifts in suitable climatic conditions. However, even if species can synchronize with shifts in their climatic niche, they may still be negatively affected by shifts in other anthropogenic and ecological perturbances, such as trophic forcing mechanisms (Alonso-Crespo & Hernández-Agüero, 2023), differences in species life history, and other local-scale effects (Rapacciuolo et al., 2014). Conversely, these species may also benefit from new interactions, such as accessing novel food sources or gaining competitive advantages for nesting. As well, future work should investigate the temporal dynamics of climate change, especially relating to the life cycle and demography of species and the potential colonization by different populations and environments of the current distributional range of the species.

We acknowledge the fact that we have used presence data from the Canadian portion of the breeding range of the three species of birds to model the climate projections, whereas accounting for the climate envelope of the entire range would provide more robust results compared to the parts of the range. Nevertheless, our reason for including Canadian portion of the distributional range is mainly due to the fact that all most all of the confirmed evidence of breeding of the three species of the birds were from their northern distributional range in Canada (Birds Canada, 2018), and we believed that restricting data to only that part of the range would provide a more realistic predictive maps of species distributions. We also found similar results to other previous studies, suggesting that our methods have not overestimated potential loss of bioclimatic space for these species (Langham et al., 2015; Nixon et al., 2016; Wilsey et al., 2019). Moreover, our approach of correcting sampling bias by incorporating “overlapped target group background data” into maxent models is known to efficiently handle the sparse, irregularly sampled data that are often spatially biased towards better-surveyed areas (Kramer-Schadt et al., 2013; Barber et al., 2022; Elith et al., 2010).

We also acknowledge that the Breeding Bird Survey (BBS) data are collected along fixed 40-km routes, which do not represent precise point locations, but rather route-level detections. Climate variables were retained at 1 km2 resolution because they represent continuous environmental gradients that influence habitat suitability. While this creates a nominal scale mismatch, the effective resolution of our predictions is constrained by the coarser occurrence data, and all interpretations were made at the route scale. In doing so, we believe it also helps align with the space requirements and dispersal ecology of birds. In addition, Maxent is known to effectively handle fine-resolution predictors even if occurrence data are coarser, particularly if sampling bias is accounted for (e.g., Keil et al., 2013; Kramer-Schadt et al., 2013; Syfert et al., 2013). Using finer-resolution climate data is a standard practice in SDMs to capture local temperature and precipitation variability (Karger et al., 2017; Fick & Hijmans, 2017; Fronzek et al., 2011). Moreover, aggregating detections across years was necessary to reduce interannual variability and improve detection reliability particularly for species with low detection probabilities and is commonly used in climate-based habitat modeling (e.g., Nixon et al., 2016; Phillips et al., 2004).

Bioclimatic refugia

All three focal bird species, BAIS, SPPI, and TBLO, will likely lose nearly all their currently suitable bioclimatic space as their breeding ranges in Canada shift northward by mid-century and into the end of the century. Likewise, Langham et al. (2015) reported similar losses, specifically that by 2080, BAIS is predicted to lose more than 95% of its potential current range, while TBLO is predicted to lose approximately 85% of its potential current range. Wilsey et al. (2019) also predicted similar range changes, with over 95% loss of current breeding ranges for these three species. Furthermore, while more than half of the grassland bird species in Alberta are projected to have > 50% stable suitable climatic space, BAIS, SPPI, and TBLO have some of the smallest projected areas of stable climate due to limitations in the availability of suitable land cover (Nixon et al., 2016).

We assumed that the current distribution of birds is primarily governed by climate (Nixon et al., 2016), and this system is operating at equilibrium (Araújo & Pearson, 2005) which will continue to hold in the future (Wiens et al., 2009). However, there are many factors that may limit the ability of bird populations to fully realize projections of climate suitability in the future, such as site fidelity, interspecific species interactions, limitations in dispersal ability, grassland successional dynamics and lags in vegetation transition (Stralberg et al., 2015), micro-climatic complexity (Bernath-Plaisted et al., 2023b), and human-induced changes in land use and management (e.g., grazing intensity, timing, frequency, and duration; Monroe et al., 2017). Projections of suitable bioclimatic space might address these issues, as it includes areas that are predicted to be both climatically and biologically suitable for species.

We also assumed that the baseline areas would remain unchanged and remain available to the different populations and environments as they shift northward in the future. Changes in climate may also lead to agricultural expansion, which may further constrain species that are reliant on native grasslands (Nixon et al., 2016).

Moreover, the use of agricultural land cover differs among grassland bird species, which may affect the occupancy of suitable bioclimatic space by certain species if these areas are converted to agriculture in the future. For example, BAIS may be more tolerant than other grassland-obligate species to a wider range of anthropogenic habitats, such as seeded pasture, hay, or cropland (McMaster & Davis, 2001). In contrast, TBLO is known to strongly prefer native and non-native grasslands with short grasses but is negatively associated with man-made habitats such as crop and hayfields (With, 2010), and SPPI is rarely associated with cultivated lands or non-native grassland (McMaster & Davis, 2001). Future increases in the conversion of grassland to agriculture could have further negative effects on these species. Although BAIS can tolerate a wider range of anthropogenic habitats, they are projected to gain less bioclimatic space in all scenarios compared to SPPI and TBLO. Our analyses suggest that one of the primary reasons for this discrepancy could be the inability of BAIS to track climate velocity compared to the other two species (BAIS has 4 × less dispersal ability than TBLO and 14 × less than SPPI). However, the dependence of SPPI and TBLO on less disturbed habitats suggests that human encroachment into suitable habitats following the warming climate will further reduce their preferred native grassland habitat (Paterson et al., 2024). Further examination into the potential for human-mediated habitat conversion is warranted, as well as continuous monitoring of these species to detect changes in habitat use and range expansion.

Changes in species’ habitat use and potential range expansion may affect future bird communities. By 2080, under a high-emissions scenario (SRES A2), northern boreal forests in Canadian provinces are predicted to gain as many as 80 species and lose as many as 69 species (Langham et al., 2015). Most breeding season species losses will also occur near the USA-Canada border, in the Eastern deciduous forest, prairie pothole, Rockies, and Sierra ranges (Langham et al., 2015). Passerines in particular have the lowest heat tolerances (compared to galliforms and columbids; Smith et al., 2017), and smaller passerine species exhibit much lower heat tolerance than larger species (McKechnie et al., 2017), which suggests that grassland passerines may be less adaptive to climate changes than other taxa. Physiological studies to determine heat tolerances of different bird taxa could determine how bird community composition changes as temperatures continue to rise, especially due to evidence that grassland bird populations are affected by extreme weather and changes in rainfall patterns (Maresh Nelson et al., 2023).

These changes could also lead to unforeseen shifts in the structure and functioning of the grassland ecosystem. As such, in the northern prairie region, hydrologic features (e.g., springs, riparian areas, and lakeshores) that influence soil moisture and surface water availability, instead of factors like topography, may emerge as important factors for fine-scale refugia (McLaughlin et al., 2017; Wang et al., 2020). Furthermore, because relatively flat, fertile landscapes are often dominated by intensive agricultural land uses, habitat quality and availability may be the main limiting factors for where refugia can exist (Paterson et al., 2024). For example, over 1 million hectares of grassland were plowed in the Great Plains in 2018–2019, primarily for row crop agriculture (World Wildlife Fund, 2021), with approximately 60% of the ecoregion at risk of conversion (Olimb & Robinson, 2019) and likely further increases in the future. Conversion analyses to predict the amount and location of grassland conversion are needed to further inform conservation planning.

In the future, under climate change, forests may transform into grasslands. Although precipitation is predicted to increase in boreal regions, it is unlikely to offset the increased evaporative demand, leading to reduced moisture availability in the future (Price et al., 2013). In fact, the prolonged drought and increase in frequency and intensity of wildfire are expected to trigger the transitions from conifer-dominated forests to deciduous forests, shrublands, or grasslands in North America (Rupp et al., 2016; Scheffer et al., 2012; Wang et al., 2020). Specifically, wildfires may facilitate the conversion of treed ecosystems to deciduous forest and grassland by 2100 in Alberta, if the incidence and severity of fires follow future predictions (Stralberg et al., 2018), but more studies are needed to determine the specific effects of wildfire on the creation and maintenance of grasslands. The uncertainty of climate pattern changes, extreme weather events, and other cascading effects of global warming when attempting to protect and manage climate refugia necessitates long-term monitoring to design and implement climate-integrated conservation plans over time (Morelli et al., 2016).

Management implications

Despite the significant vulnerability to land use conversion, only 9% of grassland climate and land-use strongholds are currently protected in the prairie pothole region of North America (Grand et al., 2019). With the northward shift of suitable bioclimatic space, future conservation planning must take climate change into consideration and protect areas of future suitable bioclimatic space of grassland songbird populations for their long-term persistence. For climate-affected or currently at-risk species, conservation initiatives should focus on currently occupied habitats that are expected to remain suitable under future climate scenarios (Langham et al., 2015). In addition, the conservation targets should not only focus on quantitative measures (e.g., Government of Canada, 2021), but should also consider species responses to climate change to produce a network of connected protected areas that will remain viable over time and facilitate the movement of species. In fact, under all projected climate change scenarios, almost 100% of BAIS and TBLO ranges will reside outside current protected areas, while SPPI will have 70% of its suitable habitat lie outside protected areas under RCP 8.5 in 2080. Therefore, it is imperative to expand the current protected areas network to ensure connectivity, particularly between in situ and ex situ bioclimatic refugia for these species (Carroll & Noss, 2020). Reconfiguring protected area networks by accounting for climate change impacts has been recently reported as a relatively inexpensive adaptation strategy (Lawler et al., 2020).

Additionally, there is currently a mismatch between the conservation status given to a species and its vulnerability to threats. For example, BAIS is projected to lose more than 95% of its current range yet is only assessed as Special Concern by COSEWIC (COSEWIC, 2012). SPPI and TBLO are assessed as Threatened (COSEWIC, 2010, 2016). This study carries a special merit in this regard, as it provides a framework for assessment of climate-integrated conservation status as well as identifies the priority breeding habitats for three important songbirds across Canadian prairies that will require protection now and into the future. We propose this approach for landscape conservation planning and risk assessment for other species with extensive spatial requirements.

Conclusion

Our results further our understanding of the possible impacts of climate change on the distributional range of migratory grassland songbirds in Canada and support other studies that suggest grassland songbirds will expand northward and gain bioclimatically suitable space within the next century. However, BAIS, SPPI, and TBLO are still vulnerable to climate change, especially if native grassland continues to be converted to other land uses and the most severe climate projections are realized. We draw attention to the lack of current protected native grassland habitat that these species depend on. Long-term persistence of grassland species relies on their ability to colonize new bioclimatic space as well as present conservation management of refugia.

Supplementary information

Below is the link to the electronic supplementary material.

ESM 1 (488.6KB, xlsx)

(XLSX 488 KB)

ESM 2 (12.1KB, xlsx)

(XLSX 12.0 KB)

ESM 3 (9.9KB, xlsx)

(XLSX 9.86 KB)

ESM 4 (229.3KB, docx)

(DOCX 229 KB)

ESM 5 (70.4KB, docx)

(DOCX 70.4 KB)

ESM 6 (46.7KB, docx)

(DOCX 46.7 KB)

Acknowledgements

We thank Hanchen Shen of WWF-Canada for initial GIS support, and Kiel Drake of Birds Canada and Jessica Currie of WWF-Canada for their helpful comments on the initial draft.

Authors' contributions

R.S., J.A. and E.G. designed the study. J.A., R.S., C.M. and W.M. analyzed data and interpreted the results. R.S. and C.M. prepared the manuscript. J.S. and E.G. helped with data acquisition and facilitated the analyses. All authors corrected and made suggestions to the text and gave final approval for publication.

Data availability

The data supporting this study’s findings are available from the corresponding author, Rinjan Shrestha ([rshrestha@wwfcanada.org](mailto:rshrestha@wwfcanada.org)) upon request.

Declarations

Ethics approval

All authors have read, understood, and complied as applicable with the statement on “Ethical responsibilities of Authors” as found in the Instructions for Authors.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Akaike, H. (1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control,19, 716–723. [Google Scholar]
  2. Alonso-Crespo, I. M., & Hernández-Agüero, J. A. (2023). Shedding light on trophic interactions: A field experiment on the effect of human population between latitudes on herbivory and predation patterns. Ecology and Evolution,13, Article e10449. 10.1002/ece3.10449 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Anderson, R. C. (2006). Evolution and origin of the central grassland of North America: Climate, fire, and mammalian grazers. Journal of the Torrey Botanical Society,133, 626–647. 10.3159/1095-5674(2006)133[626:EAOOTC]2.0.CO;2 [Google Scholar]
  4. Araújo, M. B., & Pearson, R. G. (2005). Equilibrium of species’ distributions with climate. Ecography,28, 693–695. 10.1111/j.2005.0906-7590.04253.x [Google Scholar]
  5. Ashcroft, M. B. (2010). Identifying refugia from climate change. Journal of Biogeography,37, 1407–1413. 10.1111/j.1365-2699.2010.02300.x [Google Scholar]
  6. Barber, R. A., Ball, S. G., Morris, R. K. A., & Gilbert, F. (2022). Target-group backgrounds prove effective at correcting sampling bias in Maxent models. Diversity and Distributions,28, 128–141. 10.1111/ddi.13442 [Google Scholar]
  7. Bateman, B. L., Wilsey, C., Taylor, L., Wu, J., LeBaron, G. S., & Langham, G. (2020). North American birds require mitigation and adaptation to reduce vulnerability to climate change. Conservation Science and Practice. 10.1111/csp2.242 [Google Scholar]
  8. Bernath-Plaisted, J. S., Correll, M. D., Somershoe, S. G., Dwyer, A. M., Bankert, A., Beh, A., Berlanga, H., Boyle, W. A., Cruz-Romo, J. L., George, T. L., Herkert, J., Koper, N., Macías-Duarte, A., Panjabi, A. O., Ramírez-Flores, O. M., Robinson, B., Ruvalcaba-Ortega, I., Sibbing, J., Strasser, E. H., & VerCauteren, T. (2023a). Review of conservation challenges and possible solutions for grassland birds of the North American Great Plains. Rangeland Ecology & Management,90, 165–185. 10.1016/j.rama.2023.07.002 [Google Scholar]
  9. Bernath-Plaisted, J. S., Ribic, C. A., Hills, W. B., Townsend, P. A., & Zuckerberg, B. (2023b). Microclimate complexity in temperate grasslands: Implications for conservation and management under climate change. Environmental Research Letters. 10.1088/1748-9326/acd4d3 [Google Scholar]
  10. Bird, J. P. (2020). Generation lengths of the world’s birds and their implications for extinction risk. Conservation Biology. 10.1111/cobi.13486 [Google Scholar]
  11. Brambilla, M., Rubolini, D., Appukuttan, O., Calvi, G., Karger, D. N., Kmecl, P., Mihelič, T., Sattler, T., Seaman, B., Teufelbauer, N., Wahl, J., & Celada, C. (2022). Identifying climate refugia for high-elevation Alpine birds under current climate warming predictions. Global Change Biology. 10.1111/gcb.16187 [Google Scholar]
  12. Birds Canada (2018). Data accessed from NatureCounts, a node of the Avian Knowledge Network, Birds Canada. Available: https://www.naturecounts.ca/. Accessed 7, Aug 2025.
  13. Carroll, C., & Noss, R. F. (2020). Rewilding in the face of climate change. Conservation Biology,00(1), 1. 10.1111/cobi.13531 [Google Scholar]
  14. Carroll, C., Roberts, D. R., Michalak, J. L., Lawler, J. J., Nielsen, S. E., Stralberg, D., Hamann, A., Mcrae, B. H., & Wang, T. (2017). Scale-dependent complementarity of climatic velocity and environmental diversity for identifying priority areas for conservation under climate change. Global Change Biology,23, 4508–4520. 10.1111/gcb.13679 [DOI] [PubMed] [Google Scholar]
  15. Convention on Biological Diversity. (2021). First draft of the post-2020 global biodiversity framework. Retrieved from: https://www.unep.org/resources/publication/1st-draft-post-2020-global-biodiversity-framework
  16. Chen, I.-C., Hill, J. K., Ohlemüller, R., Roy, D. B., & Thomas, C. D. (2011). Rapid range shifts of species associated with high levels of climate warming. Science,1979(333), 1024–1026. 10.1126/science.1206432 [Google Scholar]
  17. COSEWIC. (2010). COSEWIC assessment and status report on the Sprague’s Pipit Anthus spragueii in Canada. Committee on the Status of Endangered Wildlife in Canada.
  18. COSEWIC. (2012). COSEWIC assessment and status report on the Baird’s sparrow Ammodramus bairdii in Canada. Committee on the Status of Endangered Wildlife in Canada. Ottawa. [Google Scholar]
  19. COSEWIC. (2016). COSEWIC assessment and status report on the McCown’s longspur Rhynchophanes mccownii in Canada. Committee on the Status of Endangered Wildlife in Canada. Ottawa. [Google Scholar]
  20. Davidson, S.C., Bohrer, G., Gurarie, E., LaPoint, S., Mahoney, P.J., Boelman, N.T., Eitel, J.U.H., Prugh, L.R., Vierling, L.A., Jennewein, J., Grier, E., Couriot, O., Kelly, A.P., Meddens, A.J.H., Oliver, R.Y., Kays, R., Wikelski, M., Aarvak, T., Ackerman, J.T., ..., Hebblewhite, M. (2020). Ecological insights from three decades of animal movement tracking across a changing Arctic. Science, 1979(370), 712–715. 10.1126/science.abb7080
  21. Davis, S. K. (2004). Area sensitivity in grassland passerines: Effects of patch size, patch shape, and vegetation structure on bird abundance and occurrence in southern Saskatchewan. Auk,121, 1130–1145. 10.2307/4090481 [Google Scholar]
  22. Diffenbaugh, N. S., & Field, C. B. (2013). Changes in ecologically critical terrestrial climate conditions. Science,1979(341), 486–492. 10.1126/science.1237123 [Google Scholar]
  23. ECCC & BC (2024). The state of Canada’s birds. Data accessed from NatureCounts, Birds Canada. 10.71842/jqn6-w07
  24. Elith, J., Phillips, S. J., Hastie, T., Dudík, M., Chee, Y. E., & Yates, C. J. (2010). A statistical explanation of MaxEnt for ecologists. Diversity and Distributions,17, 43–57. 10.1111/j.1472-4642.2010.00725.x [Google Scholar]
  25. Elsen, P. R., Monahan, W. B., Dougherty, E. R., & Merenlender, A. M. (2020). Keeping pace with climate change in global terrestrial protected areas. Science Advances,6(25), Article eaay0814. 10.1126/sciadv.aay0814 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Environment and Climate Change Canada. (2021). Canadian protected and conserved areas database. Retrieved from https://www.canada.ca/en/environment-climate-change/services/nationalwildlife-areas/protected-conserved-areas-database.html
  27. Fick, S. E., & Hijmans, R. J. (2017). Worldclim 2: New 1-km spatial resolution climate surfaces for global land areas. International Journal of Climatology,37(12), 4302–4315. 10.1002/joc.5086 [Google Scholar]
  28. Fronzek, S., Carter, T. R., & Luoto, M. (2011). Evaluating sources of uncertainty in modelling the impact of probabilistic climate scenarios on species distributions. Climatic Change,111, 177–197. 10.1007/s10584-011-0137-3 [Google Scholar]
  29. Gahbauer, M. A., Parker, S. R., Wu, J. X., Harpur, C., Bateman, B. L., Whitaker, D. M., Tate, D. P., Taylor, L., & Lepage, D. (2022). Projected changes in bird assemblages due to climate change in a Canadian system of protected areas. PLoS One. 10.1371/journal.pone.0262116 [Google Scholar]
  30. Gauthier, D. A., & Wiken, E. B. (2003). Monitoring the conservation of grassland habitats, Prairie Ecozone, Canada. Environmental Monitoring and Assessment,88, 343–364. 10.1023/A:1025585527169 [DOI] [PubMed] [Google Scholar]
  31. Government of Canada. (2021). Budget 2021: A recovery plan for jobs, growth, and resilience (pp. 149–150). Retrieved from: https://www.canada.ca/en/department-finance/news/2021/04/budget-2021-a-recovery-plan-for-jobs-growth-and-resilience.html
  32. Grand, J., Wilsey, C., Wu, J., & Michel, N. (2019). The future of North American grassland birds: Incorporating persistent and emergent threats into full annual cycle conservation priorities. Conservation Science and Practice,1, Article e20. 10.1111/csp2.20 [Google Scholar]
  33. Groves, C. R., Game, E. T., Anderson, M. G., Cross, M., Enquist, C., Ferdaña, Z., Girvetz, E., Gondor, A., Hall, K. R., Higgins, J., Marshall, R., Popper, K., Schill, S., & Shafer, S. L. (2012). Incorporating climate change into systematic conservation planning. Biodiversity and Conservation,21, 1651–1671. 10.1007/s10531-012-0269-3 [Google Scholar]
  34. Hamann, A., & Wang, T. (2006). Potential effects of climate change on ecosystem. Ecological Society of America,87, 2773–2786. [Google Scholar]
  35. Hébert, R., Lovejoy, S., & Tremblay, B. (2020). An observation-based scaling model for climate sensitivity estimates and global projections to 2100. Climate Dynamics. 10.1007/s00382-020-05521-x [Google Scholar]
  36. Hoekstra, J. M., Boucher, T. M., Ricketts, T. H., & Roberts, C. (2005). Confronting a biome crisis: Global disparities of habitat loss and protection. Ecology Letters,8, 23–29. 10.1111/j.1461-0248.2004.00686.x [Google Scholar]
  37. Jueterbock, A., Smolina, I., Coyer, J. A., & Hoarau, G. (2016). The fate of the Arctic seaweed Fucus distichus under climate change: An ecological niche modeling approach. Ecology and Evolution,6, 1712–1724. 10.1002/ece3.2001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Karger, D. N., Conrad, O., Böhner, J., Kawohl, T., Kreft, H., Soria-Auza, R. W., Zimmermann, N. E., Linder, H. P., & Kessler, M. (2017). Climatologies at high resolution for the earth’s land surface areas (CHELSA V1.2). Scientific Data, 4, 170122. 10.1038/sdata.2017.122
  39. Keil, P., Jetz, W., Belmaker, J., Wilson, A. M., & Unitt, P. (2013). Downscaling of species distribution models: A hierarchical approach. Methods in Ecology and Evolution. 10.1111/j.2041-210X.2012.00264.x [Google Scholar]
  40. Keppel, G., Van Niel, K. P., Wardell-Johnson, G. W., Yates, C. J., Byrne, M., Mucina, L., Schut, A. G. T., Hopper, S. D., & Franklin, S. E. (2012). Refugia: Identifying and understanding safe havens for biodiversity under climate change. Global Ecology and Biogeography,21, 393–404. 10.1111/j.1466-8238.2011.00686.x [Google Scholar]
  41. Kramer-Schadt, S., Niedballa, J., Pilgrim, J. D., Schröder, B., Lindenborn, J., Reinfelder, V., Stillfried, M., Heckmann, I., Scharf, A. K., Augeri, D. M., et al. (2013). The importance of correcting for sampling bias in MaxEnt species distribution models. Diversity and Distributions,19, 1366–1379. [Google Scholar]
  42. Langham, G. M., Schuetz, J. G., Distler, T., Soykan, C. U., & Wilsey, C. (2015). Conservation status of North American birds in the face of future climate change. PLoS One. 10.1371/journal.pone.0135350 [Google Scholar]
  43. Lawler, J. J., Ackerly, D. D., Albano, C. M., Anderson, M. G., Dobrowski, S. Z., Gill, J. L., Heller, N. E., Pressey, R. L., Sanderson, E. W., & Weiss, S. B. (2015). The theory behind, and the challenges of, conserving nature’s stage in a time of rapid change. Conservation Biology,29, 618–629. 10.1111/cobi.12505 [DOI] [PubMed] [Google Scholar]
  44. Lawler, J. J., Rinnan, D. S., Michalak, J. L., Withey, J. C., Randels, C. R., & Possingham, H. P. (2020). Planning for climate change through additions to a national protected area network: Implications for cost and configuration. Philosophical Transactions of the Royal Society B: Biological Sciences. 10.1098/rstb.2019.0117 [Google Scholar]
  45. Liu, C., Newell, G., & White, M. (2016). On the selection of thresholds for predicting species occurrence with presence-only data. Ecology and Evolution,6, 337–348. 10.1002/ece3.1878 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Loarie, S. R., Duffy, P. B., Hamilton, H., Asner, G. P., Field, C. B., & Ackerly, D. D. (2009). The velocity of climate change. Nature,462, 1052–1055. 10.1038/nature08649 [DOI] [PubMed] [Google Scholar]
  47. Maresh Nelson, S. B., Ribic, C. A., Niemuth, N. D., Bernath-Plaisted, J., & Zuckerberg, B. (2023). Sensitivity of North American grassland birds to weather and climate variability. Conservation Biology. 10.1111/cobi.14143 [Google Scholar]
  48. McGinn, S. M. (2010). Weather and climate patterns in Canada’s prairie grasslands. Arthropods of Canadian Grasslands,1, 105–119. [Google Scholar]
  49. McKechnie, A. E., Gerson, A. R., McWhorter, T. J., Smith, E. K., Talbot, W. A., & Wolf, B. O. (2017). Avian thermoregulation in the heat: Evaporative cooling in five Australian passerines reveals within-order biogeographic variation in heat tolerance. Journal of Experimental Biology,220, 2436–2444. 10.1242/jeb.155507 [DOI] [PubMed] [Google Scholar]
  50. McLaughlin, B. C., Ackerly, D. D., Klos, P. Z., Natali, J., Dawson, T. E., & Thompson, S. E. (2017). Hydrologic refugia, plants, and climate change. Global Change Biology. 10.1111/gcb.13629 [Google Scholar]
  51. McMaster, D. G., & Davis, S. K. (2001). An evaluation of Canada’s permanent cover program: Habitat for grassland birds? (Una Evaluación de el Programa de Cubierta Permanente de Canadá: ¿Habitat para Aves de Verbazales?). Journal of Field Ornithology,72, 195–210. [Google Scholar]
  52. Michalak, J. L., Stralberg, D., Cartwright, J. M., & Lawler, J. J. (2020). Combining physical and species-based approaches improves refugia identification. Frontiers in Ecology and the Environment,18, 254–260. 10.1002/fee.2207 [Google Scholar]
  53. Monroe, A. P., Aldridge, C. L., Assal, T. J., Veblen, K. E., Pyke, D. A., & Casazza, M. L. (2017). Patterns in Greater Sage-Grouse population dynamics correspond with public grazing records at broad scales. Ecological Applications,27, 1096–1107. 10.1002/eap.1512 [DOI] [PubMed] [Google Scholar]
  54. Morelli, T. L., Barrows, C. W., Ramirez, A. R., Cartwright, J. M., Ackerly, D. D., Eaves, T. D., Ebersole, J. L., Krawchuk, M. A., Letcher, B. H., Mahalovich, M. F., Meigs, G. W., Michalak, J. L., Millar, C. I., Quiñones, R. M., Stralberg, D., & Thorne, J. H. (2020). Climate-change refugia: Biodiversity in the slow lane. Frontiers in Ecology and the Environment,18, 228–234. 10.1002/fee.2189 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Morelli, T. L., Daly, C., Dobrowski, S. Z., Dulen, D. M., Ebersole, J. L., Jackson, S. T., Lundquist, J. D., Millar, C. I., Maher, S. P., Monahan, W. B., Nydick, K. R., Redmond, K. T., Sawyer, S. C., Stock, S., & Beissinger, S. R. (2016). Managing climate change refugia for climate adaptation. PLoS One. 10.1371/journal.pone.0159909 [Google Scholar]
  56. Nixon, A. E., Fisher, R. J., Stralberg, D., Bayne, E. M., & Farr, D. R. (2016). Projected responses of North American grassland songbirds to climate change and habitat availability at their northern range limits in Alberta, Canada. Avian Conservation and Ecology. 10.5751/ACE-00866-110202 [Google Scholar]
  57. Olimb, S. K., & Robinson, B. (2019). Grass to grain: Probabilistic modeling of agricultural conversion in the North American Great Plains. Ecological Indicators,102, 237–245. 10.1016/j.ecolind.2019.02.042 [Google Scholar]
  58. Olson, D. M., & Dinerstein, E. (2002). The global 200: Priority ecoregions for global conservation. Annals of the Missouri Botanical Garden,89, 199–224. 10.2307/3298564 [Google Scholar]
  59. Pachauri, R. K., Allen, M.R., Barros, V. R., Broome, J., Cramer, W., Christ, R., Church, J. A., Clarke, L., Dahe, Q., & Dasgupta, P. (2014). Climate change 2014: Synthesis report. In R. Pachauri, & L. Meyer (Eds), Contribution of Working Groups I, II and III to the fifth assessment report of the Intergovernmental Panel on Climate Change (p. 151). IPCC.
  60. Partners in Flight. (2021). Partners in flight [WWW Document]. Avian Conservation Assessment Database Scores. URL https://pif.birdconservancy.org/avian-conservation-assessment-database-scores/. Accessed 29 Sept 21.
  61. Paterson, J. E., Bortolotti, L. E., Kowal, P. D., Pidwerbesky, A. J., & Devries, J. H. (2024). Predicting the effects of land cover change on biodiversity in Prairie Canada using species distribution models. Biological Conservation. 10.1016/j.biocon.2024.110754 [Google Scholar]
  62. Phillips, S. J., & Dudík, M. (2008). Modeling of species distributions with Maxent: New extensions and a comprehensive evaluation. Ecography,31, 161–175. [Google Scholar]
  63. Phillips, S. J., Dudík, M., Elith, J., Graham, C. H., Lehmann, A., Leathwick, J., & Ferrier, S. (2009). Sample selection bias and presence-only distribution models: Implications for background and pseudo-absence data. Ecological Applications,19, 181–197. 10.1890/07-2153.1 [DOI] [PubMed] [Google Scholar]
  64. Phillips, S. J., Dudík, M., & Schapire, R. E. (2004). A maximum entropy approach to species distribution modeling. In Proceedings of the Twenty-First International Conference on Machine Learning (pp. 655-662). 10.1145/1015330.1015412
  65. Price, D. T., Alfaro, R. I., Brown, K. J., Flannigan, M. D., Fleming, R. A., Hogg, E. H., Girardin, M. P., Lakusta, T., Johnston, M., McKenney, D. W., Pedlar, J. H., Stratton, T., Sturrock, R. N., Thompson, I. D., Trofymow, J. A., & Venier, L. A. (2013). Anticipating the consequences of climate change for Canada’s boreal forest ecosystems. Environmental Reviews. 10.1139/er-2013-0042 [Google Scholar]
  66. Rapacciuolo, G., Maher, S. P., Schneider, A. C., Hammond, T. T., Jabis, M. D., Walsh, R. E., Iknayan, K. J., Walden, G. K., Oldfather, M. F., Ackerly, D. D., & Beissinger, S. R. (2014). Beyond a warming fingerprint: Individualistic biogeographic responses to heterogeneous climate change in California. Global Change Biology,20, 2841–2855. 10.1111/gcb.12638 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Reddy, S., & Dávalos, L. M. (2003). Geographical sampling bias and its implications for conservation priorities in Africa. Journal of Biogeography,30, 1719–1727. 10.1046/j.1365-2699.2003.00946.x [Google Scholar]
  68. Roberts, C. M., O’Leary, B. C., & Hawkins, J. P. (2020). Climate change mitigation and nature conservation both require higher protected area targets. Philosophical Transactions of the Royal Society B: Biological Sciences,375(1794), Article 20190121. [Google Scholar]
  69. Roberts, D. R., & Hamann, A. (2012). Method selection for species distribution modelling: Are temporally or spatially independent evaluations necessary? Ecography,35, 792–802. 10.1111/j.1600-0587.2011.07147.x [Google Scholar]
  70. Rupp, T., Duffy, P., Leonawicz, M., Lindgren, M., Breen, A., Kurkowski, T., Floyd, A., Bennett, A., & Krutikov, L. (2016). Chapter 2: Climate simulations, land cover, and wildfire. In Zhiliang, Z. & McGuire, A. D (Eds), Baseline and projected future carbon storage and greenhouse-gas fluxes in ecosystems of Alaska. U.S. Geological Professional Paper 1826. 10.3133/pp1826
  71. Sauchyn, D., Davidson, D., Johnston, M., Flannigan, M., Fletcher, A., Isaac, K., Kulshreshtha, S., Kowalczyk, T., Mauro, I., Pittman, J., Reed, M. G., & Schneider, R. (2020). Canada in a changing climate: Regional perspectives report. Government of Canada.
  72. Sauer, J. R., Hines, J. E., Fallon, J. E., Pardieck, K. L., Ziolkowski, D. J., Jr., & Link, W. A. (2014). The North American breeding bird survey, results and analysis 1966–2013. Version,1, 1966–2013. [Google Scholar]
  73. Scheffer, M., Hirota, M., Holmgren, M., Van Nes, E. H., & Chapin, F. S. (2012). Thresholds for boreal biome transitions. Proceedings of the National Academy of Sciences of the United States of America,109, 21384–21389. 10.1073/pnas.1219844110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Schloss, C. A., Nuñez, T. A., & Lawler, J. J. (2012). Dispersal will limit ability of mammals to track climate change in the western hemisphere. Proceedings of the National Academy of Sciences of the United States of America,109, 8606–8611. 10.1073/pnas.1116791109 [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Schneider, R. R., Devito, K., Kettridge, N., & Bayne, E. (2015). Moving beyond bioclimatic envelope models: Integrating upland forest and peatland processes to predict ecosystem transitions under climate change in the western Canadian boreal plain. Ecohydrology,9, 899–908. [Google Scholar]
  76. Schwarz, A. G., & Wein, R. W. (1990). Grassland ecosystems and climate change: Hypotheses for the Northern Grassland Boreal Ecotone. In Proceedings of the Twelfth North American Prairie Conference (pp. 1–5).
  77. Skagen, S. K., & Adams, A. A. Y. (2012). Weather effects on avian breeding performance and implications of climate change. Ecological Applications,22, 1131–1145. 10.1890/11-0291.1 [DOI] [PubMed] [Google Scholar]
  78. Smith, E., O’Neill, J., Gerson, A., McKechnie, A., Wolf, B. (2017). Avian thermoregulation in the heat: Resting metabolism, evaporative cooling and heat tolerance in Sonoran Desert songbirds. Journal of Experimental Biology, 220. 10.1242/jeb.161141
  79. Stewart, R. E. (1975). Breeding birds of North Dakota. Tri-college Center for Environmental Studies.
  80. Stralberg, D., Carroll, C., Pedlar, J. H., Wilsey, C. B., McKenney, D. W., & Nielsen, S. E. (2018). Macrorefugia for North American trees and songbirds: Climatic limiting factors and multi-scale topographic influences. Global Ecology and Biogeography,27, 690–703. 10.1111/geb.12731 [Google Scholar]
  81. Stralberg, D., Matsuoka, S. M., Hamann, A., Bayne, E. M., Sólymos, P., Schmiegelow, F. K. A., Wang, X., Cumming, S. G., & Song, S. J. (2015). Projecting boreal bird responses to climate change: The signal exceeds the noise. Ecological Applications,25, 52–69. 10.1890/13-2289.1 [DOI] [PubMed] [Google Scholar]
  82. Syfert, M. M., Smith, M. J., & Coomes, D. A. (2013). The effects of sampling bias and model complexity on the predictive performance of MaxEnt species distribution models. PLoS One,8(2), Article e55158. 10.1371/journal.pone.0055158 [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Turner, M. G., Calder, W. J., Cumming, G. S., Hughes, T. P., Jentsch, A., LaDeau, S. L., Lenton, T. M., Shuman, B. N., Turetsky, M. R., Ratajczak, Z., Williams, J. W., Williams, A. P., & Carpenter, S. R. (2020). Climate change, ecosystems and abrupt change: Science priorities. Philosophical Transactions of the Royal Society B: Biological Sciences. 10.1098/rstb.2019.0105 [Google Scholar]
  84. Veech, J. A., Pardieck, K. L., & Ziolkowski, D. J. (2017). How well do route survey areas represent landscapes at larger spatial extents? An analysis of land cover composition along breeding bird survey routes. Condor,119, 607–615. 10.1650/condor-17-15.1 [Google Scholar]
  85. Veloz, S. D. (2009). Spatially autocorrelated sampling falsely inflates measures of accuracy for presence-only niche models. Journal of Biogeography,36, 2290–2299. 10.1111/j.1365-2699.2009.02174.x [Google Scholar]
  86. Wang, J. A., Sulla-Menashe, D., Woodcock, C. E., Sonnentag, O., Keeling, R. F., & Friedl, M. A. (2020). Extensive land cover change across Arctic-boreal Northwestern North America from disturbance and climate forcing. Global Change Biology,26, 807–822. 10.1111/gcb.14804 [DOI] [PubMed] [Google Scholar]
  87. AdaptWest Project. (2015). Gridded current and projected climate data for North America at 1km resolution, interpolated using the ClimateNA v5.10 software. (T. Wang et al., 2015). Available at adaptwest.databasin.org.
  88. Warren, D. L., & Seifert, S. N. (2011). Ecological niche modeling in Maxent: The importance of model complexity and the performance of model selection criteria. Ecological Applications,21, 335–342. 10.1890/10-1171.1 [DOI] [PubMed] [Google Scholar]
  89. Warren, D. L., Wright, A. N., Seifert, S. N., & Shaffer, H. B. (2014). Incorporating model complexity and spatial sampling bias into ecological niche models of climate change risks faced by 90 California vertebrate species of concern. Diversity and Distributions,20, 334–343. 10.1111/ddi.12160 [Google Scholar]
  90. Wiens, J. A. (1974). Climatic instability and the “ecological saturation” of bird communities in North American Grasslands. Condor,76, 385–400. 10.2307/1365813 [Google Scholar]
  91. Wiens, J. A., Stralberg, D., Jongsomjit, D., Howell, C. A., & Snyder, M. A. (2009). Niches, models, and climate change: Assessing the assumptions and uncertainties. Proceedings of the National Academy of Sciences of the United States of America,106, 19729–19736. 10.1073/pnas.0901639106 [DOI] [PMC free article] [PubMed] [Google Scholar]
  92. Williams, J. E., & Blois, J. L. (2018). Range shifts in response to past and future climate change: Can climate velocities and species’ dispersal capabilities explain variation in mammalian range shifts? Journal of Biogeography,45, 2175–2189. 10.1111/jbi.13395 [Google Scholar]
  93. Willis, S., Foden, W., Baker, D., Belle, E., Burgess, N., Carr, J., Doswald, N., Garcia, R. A., Hartley, A., Hof, C., Newbold, T., Rahbek, C., Smith, R., Visconti, P., Young, B., Butchart, S. H. M., & Garcia, R. (2015). Integrating climate change vulnerability assessments from species distribution models and trait-based approaches. Biological Conservation. 10.1016/j.biocon.2015.05.001 [Google Scholar]
  94. Wilsey, C., Taylor, L., Bateman, B., Jensen, C., Michel, N., Panjabi, A., & Langham, G. (2019). Climate policy action needed to reduce vulnerability of conservation-reliant grassland birds in North America. Conservation Science and Practice,1, Article e21. 10.1111/csp2.21 [Google Scholar]
  95. With, K.A. (2010). McCown’s longspur (Calcarius mccownii), in: Poole, A., Stettenheim, P., Gill, F. (Eds.), The Birds of North America. Academy of Natural Sciences, Philadelphia, PA, and American Ornithologist’s Union, Washington, D. C., Washington D.C.
  96. World Wildlife Fund. (2021). Plowprint. Washington DC.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ESM 1 (488.6KB, xlsx)

(XLSX 488 KB)

ESM 2 (12.1KB, xlsx)

(XLSX 12.0 KB)

ESM 3 (9.9KB, xlsx)

(XLSX 9.86 KB)

ESM 4 (229.3KB, docx)

(DOCX 229 KB)

ESM 5 (70.4KB, docx)

(DOCX 70.4 KB)

ESM 6 (46.7KB, docx)

(DOCX 46.7 KB)

Data Availability Statement

The data supporting this study’s findings are available from the corresponding author, Rinjan Shrestha ([rshrestha@wwfcanada.org](mailto:rshrestha@wwfcanada.org)) upon request.

Articles from Environmental Monitoring and Assessment are provided here courtesy of Springer

RESOURCES