Data from: Flying without fear: shooting disturbance has little effect on site preferences in a conflict goose species
Data files
Apr 29, 2024 version files 8.49 MB
-
Imp_Grass_Management_Model_Data.csv
-
Overall_Hab_Model_Data.csv
-
README.md
-
Shoot_Field_Disturbance_Model_Data.csv
-
Shoot_Intensity_Model_Data.csv
Abstract
Human-modified landscapes have created opportunities for numerous taxa. Agricultural expansion has proven advantageous for several Arctic-breeding goose species, leading to increased abundance and intensified conflict with farmers. Shooting is frequently implemented as a mitigation strategy to control population and via scaring to alter the spatial distribution of conflict species. However, the efficacy of such regimes in manipulating the fear landscape is not always investigated.
We developed resource selection functions using GPS-tracking data for Greenland barnacle geese (Branta leucopsis) wintering on Islay, Scotland to assess foraging site choice. We assessed overall foraging site preference and evaluated the influence of shooting management on foraging site selection of key habitats.
Barnacle geese selected for improved grassland areas and the likelihood of utilisation varied between these fields according to field-specific management. Protected areas were strongly selected for along with newly reseeded grassland. Field-level exposure to shooting disturbance did not cause a notable change in site selection.
Synthesis and Applications: Our results demonstrate the importance of providing refuges within managed agricultural landscapes to encourage site use and minimise conflict. We highlight how low intensity shooting disturbance may be ineffective in altering winter habitat selection of high-value foraging sites (especially near roosts). If future management aimed to stimulate redistribution higher intensity shooting disturbance along with the spatial and temporal coordination of shooting effort would likely be required to create a stronger perceived gradient of disturbance risk.
README: Data and code associated with McIntosh, A.L.S., Langley, L.P., Hilton, G.M., Shaw, J.M., and Bearhop, S. 2024 "Flying without fear: Shooting disturbance has little effect on site preferences in a conflict goose species" Journal of Applied Ecology
Contact information:
Please contact Aimée McIntosh (a.l.s.mcintosh@exeter.ac.uk) for additional information about the analysis and data.
Project Summary:
This data was collected as part of a PhD funded by the Natural Environment Research Council (NERC) in collaboration with the Wildfowl and Wetlands Trust (WWT) and NatureScot (formally Scottish Natural Heritage)
Description of data and file structure:
R code for each habitat selection model is found in "McIntosh2024_RSF_Model_Code.R"
1. Overall habitat selection:
Data file: "Overall_Hab_Model_Data.csv"
Columns in the file include:
i) Winter <- Winter year
ii) Shooting <- Field-scale classification of shooting management designation
iii) TOD <- Time of day (Day/Night)
iv) ceh_simp_hab <- Habitat classification
v) Case <- Binomial numeric designation of real GPS fixes (1) and pseudoabsences (0)
vi) weight <- Initial weight of 1 assigned to all GPS fixes (real and pseudoabsences)
vii) Used <- Two-level factor classification of real GPS fixes (Used) and pseudoabsences (Avail)
viii) Dist_roost <- Distance to nearest roost (m)
ix) Shoot_Tot <- Field-specific total number of shooting events in a single winter
2. Improved grasssland field selection - Management designation:
Data file: "Imp_Grass_Management_Model_Data.csv"
This file includes only real GPS fixes and pseudoabsences that were previously classified as "Improved grassland"
Columns in the file include:
i) Shooting <- Field-scale classification of shooting management designation (Yes/No)
ii) tod_ <- Time of day (Day/Night)
iii) weight <- Initial weight of 1 assigned to all GPS fixes (real and pseudoabsences)
iii) weight2 <- Weighted data point of 5 assigned to all real GPS fixes and 1 for pseudoabsences
iv) Used <- Two-level factor classification of real GPS fixes (Used) and pseudoabsences (Avail)
v) Case <- Binomial numeric designation of real GPS fixes (1) and pseudoabsences (0)
vi) Winter <- Winter year
3. Improved grassland field selection - Shooting disturbance designation:
Data file: "Shoot_Field_Disturbance_Model_Data.csv"
This file includes only real GPS fixes and pseudoabsences of improved grassland fields that were classified as "NatureScot Shooting"
Columns in the file include:
i) Winter <- Winter year
ii) tod_ <- Time of day (Day/Night)
iii) case_ <- Classification of real GPS fixes (TRUE) and pseudoabsences (FALSE)
iv) Roost_Dist <- Distance to nearest roost (m)
v) Shot_2019 <- Field-specific shooting disturbance classification in 2019 (Zero/No/Yes)
vi) Shot_2020 <- Field-specific shooting disturbance classification in 2020 (Zero/No/Yes)
vii) Case <- Binomial numeric designation of real GPS fixes (1) and pseudoabsences (0)
viii) weight <- Initial weight of 1 assigned to all GPS fixes (real and pseudoabsences)
ix) weight2 <- Weighted data point of 5 assigned to all real GPS fixes and 1 for pseudoabsences
4. Improved grassland field selection - Shooting intensity designation:
Data file: "Shoot_Intensity_Model_Data.csv"
Columns in the file include:
i) Winter <- Winter year
ii) tod_ <- Time of day (Day/Night)
iii) case_ <- Classification of real GPS fixes (TRUE) and pseudoabsences (FALSE)
iv) Dist_Roost <- Distance to nearest roost (m)
v) Shooting <- Field-scale classification of shooting management designation (Zero/Low/High)
vi) Case <- Binomial numeric designation of real GPS fixes (1) and pseudoabsences (0)
vii) weight <- Initial weight of 1 assigned to all GPS fixes (real and pseudoabsences)
viii) weight2 <- Weighted data point of 5 assigned to all real GPS fixes and 1 for pseudoabsences
ix) intensity <- Field-specific classification of shooting intensity per