Data for: Living in fear: How experience shapes caribou responses to predation risk
Cite this dataset
Derguy, Laurie; Leblond, Mathieu; St-Laurent, Martin-Hugues (2024). Data for: Living in fear: How experience shapes caribou responses to predation risk [Dataset]. Dryad. https://doi.org/10.5061/dryad.02v6wwq5c
Abstract
Wild prey can reduce predation risk by avoiding areas used by their predators. As they get older, individuals should be able to fine-tune this avoidance based on their increased experience with predators and cues associated with predation risk. Such learning mechanisms are expected to play a key role in how individuals may cope with risk during their life, particularly in altered landscapes where human disturbances have created habitat conditions distinct from those of the past. We studied the role of experience on the avoidance of risky areas by boreal caribou (Rangifer tarandus caribou) in a system where they are under high predation pressure from grey wolves (Canis lupus) and black bears (Ursus americanus). We described the behavioural responses of caribou to variations in the risk of encountering wolves and bears, investigating whether individuals adjusted their level of predator avoidance with passing monitoring years, a proxy of increasing experience. We used telemetry data collected on 31 wolves and 12 bears to map spatial variations in the risk of encountering predators. We used data from 28 collared female caribou monitored for 3–8 years (mean ± SD: 4.4 ± 2.2) to assess trends in the avoidance of risky areas by caribou with passing years. We observed an increase in the avoidance of areas suitable to wolves with passing years, except during winter and calving when females did not adjust their avoidance of wolves. We also found an increase in the avoidance of areas suitable to bears across all study periods. These results suggest that, in most circumstances, caribou became more efficient at avoiding areas selected by their main predators as they gained experience throughout their life. Provided the avoidance tactics observed in this study are heritable and offer fitness advantages to the individuals that display them, our results suggest that caribou populations living in disturbed environments may have the potential to adapt to changing predation risk. Our findings have encouraging implications for caribou and wildlife conservation in general, as they suggest that experience may allow prey to adjust to novel conditions, including higher predation risk.
README
Databases for final caribou RSFs
*** this README file is the same for the four datasets that are representing different risk periods***
Description of column names:
IDENT: caribou individual number
RISK_PERIOD: biological period of risk of this dataset
presence: GPS location (1) or random point (0); this is the dependent variable of the RSF
habitat: the original habitat categories for caribou that are presented as dummy variables (0 or 1) in columns E to N and that are detailed in Table 1.
Cut_0_5: ≤5-year-old natural disturbances and cutblocks
Cut_6_20: 6–20-year-old cutblocks
Decid_Mixt_Mature: ≥50-year-old mixed and deciduous stands
NatPert_6_20: 6–20-year-old natural disturbances (mostly fires)
OpenLichenWoodlands: Open woodlands often rich in lichens and bare areas
Other: Human infrastructures, water bodies, non-regenerated areas and rare habitat features
OldMatForest: >90-year-old conifer stands
Regeneration: 21–50-year-old regenerating stands originating from natural or anthropogenic disturbances
Wetlands: Bogs, fens, flooded areas, and alder stands
YgMatForest: 50–90-year-old conifer stands
IDYEARPERIOD: caribou individual number combined with the year and the period of risk
ELEV_ORI: elevation in meter
SLOPE_ORI: slope in degrees
MAJOR_EUCL_ORI: Euclidian distance to the closest major (paved) road in meter
MINOR_EUCL_ORI: Euclidian distance to the closest minor (gravel) road in meter
IDENT_YEAR: caribou individual number combined with year
ROADS_EUCL_ORI: Euclidian distance to the closest road (all confounded) in meter
DECAY25: Decay distance to the closest road (all confounded) transformed following Carpenter et al. (2010) equation, with an alpha value of 25
DECAY50: Decay distance to the closest road (all confounded) transformed following Carpenter et al. (2010) equation, with an alpha value of 50
DECAY75: Decay distance to the closest road (all confounded) transformed following Carpenter et al. (2010) equation, with an alpha value of 75
DECAY100: Decay distance to the closest road (all confounded) transformed following Carpenter et al. (2010) equation, with an alpha value of 100
DECAY150: Decay distance to the closest road (all confounded) transformed following Carpenter et al. (2010) equation, with an alpha value of 150
DECAY200: Decay distance to the closest road (all confounded) transformed following Carpenter et al. (2010) equation, with an alpha value of 200
DECAY300: Decay distance to the closest road (all confounded) transformed following Carpenter et al. (2010) equation, with an alpha value of 300
DECAY500: Decay distance to the closest road (all confounded) transformed following Carpenter et al. (2010) equation, with an alpha value of 500
DECAY750: Decay distance to the closest road (all confounded) transformed following Carpenter et al. (2010) equation, with an alpha value of 750
DECAY1000: Decay distance to the closest road (all confounded) transformed following Carpenter et al. (2010) equation, with an alpha value of 1000
MAJ_DECAY25: Decay distance to the closest major road transformed following Carpenter et al. (2010) equation, with an alpha value of 25
MAJ_DECAY50: Decay distance to the closest major road transformed following Carpenter et al. (2010) equation, with an alpha value of 50
MAJ_DECAY75: Decay distance to the closest major road transformed following Carpenter et al. (2010) equation, with an alpha value of 75
MAJ_DECAY100: Decay distance to the closest major road transformed following Carpenter et al. (2010) equation, with an alpha value of 100
MAJ_DECAY150: Decay distance to the closest major road transformed following Carpenter et al. (2010) equation, with an alpha value of 150
MAJ_DECAY200: Decay distance to the closest major road transformed following Carpenter et al. (2010) equation, with an alpha value of 200
MAJ_DECAY300: Decay distance to the closest major road transformed following Carpenter et al. (2010) equation, with an alpha value of 300
MAJ_DECAY500: Decay distance to the closest major road transformed following Carpenter et al. (2010) equation, with an alpha value of 500
MAJ_DECAY750: Decay distance to the closest major road transformed following Carpenter et al. (2010) equation, with an alpha value of 750
MAJ_DECAY1000: Decay distance to the closest major road transformed following Carpenter et al. (2010) equation, with an alpha value of 1000
MIN_DECAY25: Decay distance to the closest minor road transformed following Carpenter et al. (2010) equation, with an alpha value of 25
MIN_DECAY50: Decay distance to the closest minor road transformed following Carpenter et al. (2010) equation, with an alpha value of 50
MIN_DECAY75: Decay distance to the closest minor road transformed following Carpenter et al. (2010) equation, with an alpha value of 75
MIN_DECAY100: Decay distance to the closest minor road transformed following Carpenter et al. (2010) equation, with an alpha value of 100
MIN_DECAY150: Decay distance to the closest minor road transformed following Carpenter et al. (2010) equation, with an alpha value of 150
MIN_DECAY200: Decay distance to the closest minor road transformed following Carpenter et al. (2010) equation, with an alpha value of 200
MIN_DECAY300: Decay distance to the closest minor road transformed following Carpenter et al. (2010) equation, with an alpha value of 300
MIN_DECAY500: Decay distance to the closest minor road transformed following Carpenter et al. (2010) equation, with an alpha value of 500
MIN_DECAY750: Decay distance to the closest minor road transformed following Carpenter et al. (2010) equation, with an alpha value of 750
MIN_DECAY1000: Decay distance to the closest minor road transformed following Carpenter et al. (2010) equation, with an alpha value of 1000
ELEV2_ORI: the square value of elevation (column P)
habwolf: the habitat categories for wolf RSF that are presented as dummy variables (0 or 1) in columns BB to BI and detailed in Table 1.
L_Decid_Mixt_Mature: ≥50-year-old mixed and deciduous stands
L_Other: Human infrastructures, water bodies, non-regenerated areas and rare habitat features
L_OldMatForest: >90-year-old conifer stands, open woodlands often rich in lichens and bare areas
L_Regeneration: 21–50-year-old regenerating stands originating from natural or anthropogenic disturbances
L_Wetlands: Bogs, fens, flooded areas, and alder stands
L_YgMatForest: 50–90-year-old conifer stands
L_Perturb_0_5: ≤5-year-old natural disturbances and cutblocks
L_Perturb_6_20: 6–20-year-old natural disturbances and cutblocks
habbear: the habitat categories for bear RSF that are presented as dummy variables (0 or 1) in columns BK to BO and detailed in Table 1.
O_Mat_Forest: ≥50-year-old mixed or deciduous stands, and 50–90-year-old conifer stands
O_Other: Human infrastructures, water bodies, non-regenerated areas and rare habitat features
O_OldMatForest: >90-year-old conifer stands and wetlands
O_Regeneration: 21–50-year-old regenerating stands originating from natural or anthropogenic disturbances
L_Wetlands: Bogs, fens, flooded areas, and alder stands
O_Perturb_0_20: ≤20-year-old natural disturbances and cutblocks
SEX: Sex of the individual caribou
SLOPE: slope value (culumn Q divided by 10)
ELEV: elevation transformed in kilometer
ELEV2: the square value of elevation in kilometer (column BR)
ROADS_EUCL: Euclidian distance to the closest road (all confounded)
MINOR_EUCL: Euclidian distance to the closest minor (gravel) road in meter
MAJOR_EUCL: Euclidian distance to the closest major (paved) road in meter
YEAR_no: the rank of the year of consecutive telemetry monitoring
YEAR_rsf: the rank of the year of consecutive telemetry monitoring scaled for a start in 2004
RISK_LOUPS: relative probability of occurrence of wolves for this GPS caribou location or caribou random point
RISK_OURS: relative probability of occurrence of bears for this GPS caribou location or caribou random point
Methods
From 2004 to 2011 and from 2017 to 2018, 86 adult caribou were captured and fitted with a GPS collar. In the same study area, we collared and monitored 31 grey wolves from 2005 to 2009 and 12 black bears from 2005 to 2006. We defined four annual periods of risk for caribou (winter, spring, calving, and summer), and built resource selection function models of wolves and bears to describe the probability of occurrence of predators. Thereafter, we built resource selection functions using caribou data and tested interactions between the predator probability of occurrence and the passing years (a proxy of experience) to figure out if - and how - caribou adjusted their selection to predation risk with increasing experience.
Funding
Environment and Climate Change Canada
Ministère de l’Environnement, de la Lutte contre les changements climatiques, de la Faune et des Parcs
Ministère des Transports
Université du Québec à Rimouski
World Wildlife Fund Canada
Natural Sciences and Engineering Research Council, Award: 2016-05196
Natural Sciences and Engineering Research Council, Award: 2022-04307