Context

Variability in temperature and snow cover are characteristics of high-latitude environments that impose significant pressures on overwintering species. To cope with increased energetic demands and decreased resources, species occupying seasonal environments often seek out refugia that buffer them from inclement conditions. Ruffed grouse (Bonasa umbellus) roosting in the thermally stable microhabitat beneath deep snow are buffered from negative effects of cold temperatures on physiological stress (glucocorticoid hormone levels).

Objective

Despite physiological advantages of accessing warmer refugia during winter, it is unknown how land cover and winter climate promote the occurrence of such refugia over space and time. Analogous to the landscape of fear, which mediates how prey navigate spatial variation in predation risk, mapping a landscape of stress, or stress-scape, may identify hotspots where metabolic challenges persist.

Methods

We assayed droppings for fecal corticosterone metabolites (FCMs) collected from radio-tagged ruffed grouse over three winters and developed a spatial model of FCM concentrations across the extent of our study area, thus quantifying a stress-scape.

Results

FCMs increased with shallower snow depths, less dense snow, colder ambient temperatures, and more open habitat. However, despite considerable spatiotemporal variation in snow depth, snow density, and temperature, the regions across the landscape where grouse had elevated FCM levels were consistent and predictable across years.

Conclusions

Stress-scapes offer a new tool for understanding and quantifying indirect effects of stressors and can identify areas of the landscape where there may be consistent hotspots of stress that are the result of multiple ecological and environmental challenges.  

Data were collected at Sandhill Wildlife Area in central Wisconsin. Individual ruffed grouse (Bonasa umbellus) were radio-collared, and their roost behavior was monitored and fecal samples were collected for three winters: 2015-2016, 2016-2017, and 2017-2018. Fecal samples were processed to measure corticosterone concentrations (ng/g) in the lab. We used maps of snow and temperature characteristics that we prepared on a daily or biweekly basis during the study period, and extracted snow and temperature data, as well as land cover data, from the locations where fecal samples were collected to create a generalized additive mixed model that describes how winter weather and land cover influence ruffed grouse physiological stress (i.e., corticosterone concentrations, ng/g). The dataset "Stress-scape Dataset.csv" consists of multiple observations of each individual grouse over the course of three winter seasons. We include variables that were included in the final model. For each observation, we show the UTM location of the roost site (i.e. where the fecal sample was collected), the year, the individual bird ID, the roost type, the time of day that the sample was collected, and snow, temperature, and land cover variables that we extracted from previously created maps (see manuscript for additional details).

Columns in this dataset are:

BirdName: a unique identifier for each individual ruffed grouse

Easting: UTM Easting (m)

Northing: UTM Northing (m)

ng.g2: concentration of fecal corticosterone metabolites (ng/g), log-transformed

year: 1 = 2015-2016, 2 = 2016-2017, 3 = 2017-2018

RoostType: SNOW = snow bowl or snow burrow; OTHER = all other roost types (see manuscript for further details)

STime: the time of day the fecal sample was collected, measured as minutes from midnight, scaled

Sroostsnow: snow depth at the roost site (cm), scaled

Sroostden: snow sink depth at the roost site (cm), scaled. Snow sink depth is an index of snow density. Larger values indicate more powdery snow.

SNorthing: scaled version of Northing

SEasting: scaled version of Easting

SminT: minimum daily temperature (°C), scaled

SMFa: the area of the largest patch of Mature Forest within 309 m of each roost site, scaled

SDCa: the area of the largest patch of Dense Cover within 309 m of each roost site, scaled

SOPa: the area of the largest patch of Open habitat within 309 m of each roost site, scaled

SyrminTvar: annual variance of minimum temperature, scaled