For overwintering species, individuals’ ability to find refugia from inclement weather and predators likely confers strong fitness benefits. How animals use their environment can be mediated by their personality (e.g., risk-taking), but does personality mediate how overwintering species select refugia? Snow cover is a dynamic winter characteristic that can influence crypsis or provide below-the-snow refugia. We explored how wintering ruffed grouse (Bonasa umbellus) selected snow roosting sites, a behavior that reduces stress and cold exposure. We linked selection for ~700 roosts with survival of 42 grouse, and showed that grouse generally selected deeper snow and warmer areas. Grouse found in shallow snow were less likely to survive winter. However, individuals with personalities for selecting deep snow improved their survival, suggesting that demographic consequences of selecting winter refugia are mediated by differences in personality. Our study provides a crucial, and seldom addressed, link between personality in resource selection and resulting demographic consequences.

Data were collected at Sandhill Wildlife Area in central Wisconsin. Individual Ruffed Grouse were radio-collared and their resource selection and survival were monitored for three winters: 2015-2016, 2016-2017, and 2017-2018. For the resource selection function, "SSFdataset.csv", the dataset consists of GPS coordinates in UTMs and habitat characteristics of used and available Ruffed Grouse roost locations. For the survival analysis,"Survivaldataset.csv", the dataset consists of encounter histories for each individual on a weekly basis throughout the winter months, and covariates for each individual that were used as predictors in the known-fate analysis.

For the step-selection function, predictors shown in "SSFdataset.csv" have not yet been scaled. Environmental data were extracted from raster files in R.

Columns in this dataset are:

• BirdName: a unique identifier for each individual Ruffed Grouse
• Waypoint: The original waypoint used to mark the roost site of a given grouse on a given day. The Waypoint is used as the stratum in this analysis, and is used as an identifier to link used and available locations.
• Easting: UTM Easting (m)
• Northing: UTM Northing (m)
• Depth: Snow depth (cm)
• JDate: modified Julian date where December 1st = 1 and continues numerically through March 31st of each year.
• used: This codes for used and available habitat where 1 = the actual grouse roost location, and 0 = available locations within 308 m of the roost location.
• distedge: The distance in (m) to the forest edge
• minT: Daily minimum temperature (minimum temperature on that day extracted for that location)
• MatureForest: The proportion of habitat classified as Mature Forest within a 100-m buffer of the location
• DenseCover: The proportion of habitat classified as Dense Cover within a 100-m buffer of the location
• Other: The proportion of habitat classfied as Other (mostly open water and marsh) within a 100-m buffer of the location
• Open: The proportion of habitat classified as Open within a 100-m buffer of the location
• ScrubOak: The proportion of habitat classified as Scrub Oak within a 100-m buffer of the location
• spavgT: winter temperature (the average minimum temperature calculated for that location over the course of the winter)
• Year: 1 = 2015-2016, 2 = 2016-2017, 3 = 2017-2018

For the survival analysis, "Survivaldataset.csv", the predictors have not yet been scaled. To add the capture histories to the csv, use the R script "WinterSurvival.R". Descriptions of columns in the dataset are in the R script, and are also listed below.

 Individual covariates (predictor variables) used in known-fate survival analysis:

•  SSEA15M = Mean snow selection above 15 cm ("SSE>15M", or "personality in deep snow selection" in manuscript)
•  SSEB15M = Mean snow selection below 15 cm ("SSE<15M", or "personality in shallow snow selection" in manuscript)
•  WTSEM = Mean winter temperature selection ("WTSEM", or "personality in winter temperature selection" in manuscript)
•  changeWTSE = change in winter temperature selection ("deltaWTSE" or "plasticity in winter temperature selection" in manuscript)
•  changeSSEA15 = change in snow selection above 15 cm ("deltaSSE>15" or "plasticity in deep snow selection" in manuscript)
•  changeSSEB15 = change in snow selection below 15 cm ("deltaSSE<15" or "plasticity in shallow snow selection" in manuscript)
•  WTEXP = median winter temperature that individual was exposed to at used roost sites
•  FEXP = median proportion of mature forest that was present within 100 m of individual's used roost sites ("forest exposure")
•  OEXP = median proportion of open habitat that was present within 100 m of individual's used roost sites ("open habitat exposure")
•  SEXPB15 = median snow depth value at individual's used roost sites when snow depth at roost sites was below 15 cm ("SEXP<15" or "shallow snow exposure")
•  SEXPA15 = median snow depth value at individual's used roost sites when snow depth at roost sites was above 15 cm ("SEXP>15" or "deep snow exposure")
•  year = 1 = 2015-2016; 2 = 2016-2017; 3 = 2017-2018