Data from: Habitat conditions during winter explain movement among subpopulations of a declining migratory bird
Data files
Oct 14, 2025 version files 12.67 KB
-
covariate_data.csv
1.05 KB
-
monthly_ch.csv
8.34 KB
-
README.md
3.28 KB
Abstract
Understanding animal movement patterns among subpopulations is crucial for identifying spatiotemporal patterns in animal abundance. Quantifying such movement rates enables a better understanding of population dynamics and how animals decide to move between habitats throughout their range. However, detailed assessments of animal movements among subpopulations are often difficult to obtain with typical capture-mark-recapture methods, limiting our ability to incorporate movement information into conservation planning. We used six years of high-resolution Global Positioning System (GPS) tracking data in a Bayesian multistate model to quantify habitat drivers of monthly intra- (i.e., from one month to the next within a winter) and inter-winter (i.e., from the last month of one winter to the first month of the following winter) movements among eight different subpopulations of a declining migratory bird, the Greenland white-fronted goose (Anser albifrons flavirostris). We found that while Greenland white-fronted geese were highly philopatric to geographically distinct wintering subpopulations, individuals made intra- and inter-winter movements based on locally available foraging habitat. These decisions changed within and among winters; geese were more likely to make intra-winter movements to areas with fewer croplands and boglands, potentially in response to local food depletion, and more likely to make inter-winter movements to areas with more boglands and “greener” grasslands. We demonstrated a framework for using high-frequency GPS tracking data to discover movement patterns and test hypotheses about environmental drivers of movements, which could be linked with population dynamics and applied to other species of concern for an improved understanding of metapopulations across space and time. Implementing habitat management strategies that optimize foraging conditions throughout winter, including rewetting degraded peatlands to provide bogland food plants, provision of cereal stubbles in early winter and high-quality grasslands throughout winter may therefore help improve conservation outcomes for Greenland white-fronted geese.
All analyses were completed using R. See manuscript for details.
Please contact the corresponding author (A. Schindler) with questions about this data package or to seek potential collaborations using these data.
Description
Model code and associated data files to quantify habitat drivers of both intra- and inter-winter movements among Greenland white-fronted goose (Anser albifrons flavirostris) wintering subpopulation groups.
Code files
winter_multistate_model.R: Multistate model codenimble_restart_functions.R: Functions used inwinter_multistate_model.Rto continue model run from a new session (for improving model convergence)
Data files (See manuscript for further details)
monthly_ch.csv: Monthly multistate capture histories for all Greenland white-fronted geese captured during the study- rows: individual Greenland white-fronted goose capture histories
- columns: month/year (i.e., 2018_1 = first month [October] of winter 2018/2019)
- states: 1 = observed at Lough Swilly, 2 = observed at Sheskinmore,
3 = observed at Connemara, 4 = observed at Lough Iron,
5 = observed at Wexford, 6 = observed at Islay,
7 = observed at West Freugh, 8 = observed at Loch Ken,
9 = observed elsewhere, 10 = not observed
covariate_data.csv: Covariate data used in the multistate model to quantify the effects of habitat on movement probabilitiesstate: 1 = observed at Lough Swilly, 2 = observed at Sheskinmore,
3 = observed at Connemara, 4 = observed at Lough Iron,
5 = observed at Wexford, 6 = observed at Islay,
7 = observed at West Freugh, 8 = observed at Loch Kenprop_farm: Average proportion of land within the 95% minimum convex polygons associated with the corresponding state managed through a farm plan.prop_farm_scaled: Average proportion of land (scaled) within the 95% minimum convex polygons associated with the corresponding state managed through a farm plan.prop_grass: Average proportion of land within the 95% minimum convex polygons associated with the corresponding state classified as grassland.prop_grass_scaled: Average proportion of land (scaled) within the 95% minimum convex polygons associated with the corresponding state classified as grassland.prop_ag: Average proportion of land within the 95% minimum convex polygons associated with the corresponding state classified as agriculture.prop_ag_scaled: Average proportion of land (scaled) within the 95% minimum convex polygons associated with the corresponding state classified as agriculture.prop_bog: Average proportion of land within the 95% minimum convex polygons associated with the corresponding state classified as bog.prop_bog_scaled: Average proportion of land (scaled) within the 95% minimum convex polygons associated with the corresponding state classified as bog.max_evi: Maximum enhanced vegetation index value for all land classified as grass within each minimum convex polygon.evi_scaled: Maximum enhanced vegetation index value (scaled) for all land classified as grass within each minimum convex polygon.
