Harvest and natural predation shape selection for behavioral predictability in male wild turkeys
Data files
Oct 16, 2025 version files 1.01 MB
-
GA_Webb_male_2014_2023_fixes_FINAL.csv
1 MB
-
README.md
1.78 KB
-
survival_data_FINAL.csv
5.03 KB
Abstract
Previous research has shown that traits increasing prey detectability, such as using open areas, higher activity levels, and using areas frequented by hunters, reduce survival rates and are considered risk-taking behaviors. Despite extensive studies on how harvest-induced and natural selection impact average phenotypic expression, the influence of these factors on behavior predictability remains unclear. Using a double hierarchical generalized linear modeling framework, we assessed variation in both average behavioral expression and predictability in male wild turkeys (n = 108) and explored how it related to susceptibility to harvest and predation. Our analysis revealed individual variation in predictability linked to risk-taking and activity (Average speed—CVp = 0.28, 95% CrI = 0.23–0.33), with some males exhibiting more consistent behavior than others. We also identified a behavioral type–predictability syndrome, where riskier individuals were more predictable, and more active individuals were less predictable (Average speed—r = 0.71, 95% CrI = 0.59–0.81). Additionally, individuals became more predictable and occurred closer to risky areas during the hunting season, coinciding with peak mortality for male wild turkeys. Our data suggests harvest-induced and natural selection on both behavioral types and behavioral predictability. Riskier behavior types were more predictable in their behavioral expression and were more likely to be killed by both hunters and predators. We also found evidence that hunters selectively targeted individuals that were more active and unpredictable in their activity patterns, whereas predators preferentially targeted turkeys exhibiting less active, more predictable behaviors. Our findings suggest that adopting a cautious, sit-and-wait tactic may help male wild turkeys detect and avoid hunters, but may increase their vulnerability to predators that use a combination of visual and olfactory cues, along with area-restricted search behaviors, to locate prey.
Dataset DOI: 10.5061/dryad.79cnp5j8p
Description of the data and file structure
├── Data
│ ├── GA_Webb_male_2014_2023_fixes_FINAL.csv <- Behavioral data from both GA/SC used in main analysis
└── survival_data_FINAL.csv <- survival data frame used for fitness analysis
Files and variables
File: survival_data_FINAL.csv
Description: Survival data frame
Variables
- ID2: Transmitter ID for individual wild turkeys
- ID: Transmitter ID for individual wild turkeys
- Site: Study site
- Year: Year of the study
- Age: Age of male wild turkeys (A = adult; J = Juvenile)
- Year_Status: Mortality status of individual wild turkeys during the year they were tracked with GPS-VHF transmitters
- Survival: Binary survival outcome (1= survived, 0 = dead)
File: GA_Webb_male_2014_2023_fixes_FINAL.csv
Description: GPS movement metrics used in the main analysis
Variables
- ID: Transmitter ID for individual turkey.
- GMT_DATE: date (Date the data was collected)
- Site: study site
- Year: year (year the data was collected)
- Stage: Hunting stage
- Open: Distance to open landcover (in meters)
- dist_edge: Distance to edge landcover (in meters)
- Dist_Access: Distance to hunter access areas (in meters)
- total_dist: Total distance traveled (in meters)
- mean_speed: Mean speed (in meters/hour)
- Age: Age of wild turkeys (A = adult; J = Juvenile)
- Status: Daily mortality status
- fixes: number of GPS fixes taken in a day
Code/software
We used R version 4.3.1 for all statistical analyses.
