Description
The recovery of some apex predators has led to concerns for endangered prey that may have developed risky habitat selection tactics during predator-free eras. Environmental heterogeneity affects predator-prey coexistence, but spatial redistribution of prey has rarely been studied. A predator-prey system with white-tailed eagles and common eiders provides a unique opportunity to study the effect of returning predators on an abundant but declining prey population.
Our objective was to investigate how the physical environment affects predator-prey relationships and subsequently the spatial redistribution of the prey population over time, and to perform a large-scale assessment of the population status and distribution of eiders in the North-Eastern Baltic Sea.
Using extensive survey data from the Finnish coast from 1997–2020 on predator and prey breeding numbers, we constructed a spatiotemporal model explaining the distribution of eiders on >3600 islands across highly variable coastal regions. We assessed how the proximity of nesting eagles affected eider abundance, mediated by properties related to physical nest shelter (archipelago type and island forest cover).
Breeding eider numbers decreased on exposed islands particularly near eagle nests, while they increased near eagle nests in the sheltered archipelago. We observed population-scale predator-induced shifts in the breeding distribution, likely reflecting both excess mortality on exposed islands and a shift of the population core to low-risk habitats.
We show that a returning predator can affect the distribution and density of its prey in a habitat-specific manner, which is important to consider in parallel with effects of human-induced ecosystem changes during conservation planning.
Data from the publication:
Living with the enemy: The return of an apex predator is associated with habitat shifts in a common but rapidly declining prey population. Landscape Ecology (2025) https://doi.org/10.1007/s10980-025-02152-7
Description of the data
This data contains numbers of breeding common eiders (Somateria mollissima) on 3648 islands in 1997-2020 along with parameters for the islands. The parameters were standardised to zero mean and unit variance in the analyses, and are presented here in the scaled form. Due to the sensitivity of the spatial data, the scaled latitudinal and longitudinal parameters are rounded to one decimal.
Eider_pairs: Number of eider pairs breeding on the island in the particular year
lnSize: Island size in hectares; lnSize is its natural logarithm
Year: Year (standardised)
Lon: longitude (Euref89), standardised and rounded to one decimal
Lat: latitude (Euref89), standardised and rounded to one decimal
Land: The percentage of land area within a radius of 5 km from the centroid of the island (standardised). Used to describe the archipelago zone. Smaller values (low land percentage) represents outer archipelago, and high values inner archipelago.
WTE:
Categorical variable with three levels describing occupancy of white-tailed eagle (Haliaeetus albicilla, WTE) territories in the vicinity of the island in the previous year.
Closest occupied WTE nest/territory within a radius of 2 km from the centre of the island
Within 2–10 km radius from the island
No occupied territories within 10 km.
Forcov: Median forest cover of the island (%) (standardised)
Island_ID: Unique identifier for each island
Year_f: Year as categorical value
pos2: A distance matrix for the islands based on the coordinates in a geographical grid of 2x2 km squares, used in the spatial exponential covariance structure in the r-package glmmTMB to account for spatial autocorrelation.
Funding for this study was obtained from Victoriastiftelsen, the Swedish Cultural Foundation in Finland, and from Sophie von Julins stiftelse and the Finnish Ministry of Agriculture and Forestry via the Nature and Game Management Trust Finland.
Our objective was to investigate how the physical environment affects predator-prey relationships and subsequently the spatial redistribution of the prey population over time, and to perform a large-scale assessment of the population status and distribution of eiders in the North-Eastern Baltic Sea.
Using extensive survey data from the Finnish coast from 1997–2020 on predator and prey breeding numbers, we constructed a spatiotemporal model explaining the distribution of eiders on >3600 islands across highly variable coastal regions. We assessed how the proximity of nesting eagles affected eider abundance, mediated by properties related to physical nest shelter (archipelago type and island forest cover).
Breeding eider numbers decreased on exposed islands particularly near eagle nests, while they increased near eagle nests in the sheltered archipelago. We observed population-scale predator-induced shifts in the breeding distribution, likely reflecting both excess mortality on exposed islands and a shift of the population core to low-risk habitats.
We show that a returning predator can affect the distribution and density of its prey in a habitat-specific manner, which is important to consider in parallel with effects of human-induced ecosystem changes during conservation planning.
Data from the publication:
Living with the enemy: The return of an apex predator is associated with habitat shifts in a common but rapidly declining prey population. Landscape Ecology (2025) https://doi.org/10.1007/s10980-025-02152-7
Description of the data
This data contains numbers of breeding common eiders (Somateria mollissima) on 3648 islands in 1997-2020 along with parameters for the islands. The parameters were standardised to zero mean and unit variance in the analyses, and are presented here in the scaled form. Due to the sensitivity of the spatial data, the scaled latitudinal and longitudinal parameters are rounded to one decimal.
Eider_pairs: Number of eider pairs breeding on the island in the particular year
lnSize: Island size in hectares; lnSize is its natural logarithm
Year: Year (standardised)
Lon: longitude (Euref89), standardised and rounded to one decimal
Lat: latitude (Euref89), standardised and rounded to one decimal
Land: The percentage of land area within a radius of 5 km from the centroid of the island (standardised). Used to describe the archipelago zone. Smaller values (low land percentage) represents outer archipelago, and high values inner archipelago.
WTE:
Categorical variable with three levels describing occupancy of white-tailed eagle (Haliaeetus albicilla, WTE) territories in the vicinity of the island in the previous year.
Closest occupied WTE nest/territory within a radius of 2 km from the centre of the island
Within 2–10 km radius from the island
No occupied territories within 10 km.
Forcov: Median forest cover of the island (%) (standardised)
Island_ID: Unique identifier for each island
Year_f: Year as categorical value
pos2: A distance matrix for the islands based on the coordinates in a geographical grid of 2x2 km squares, used in the spatial exponential covariance structure in the r-package glmmTMB to account for spatial autocorrelation.
Funding for this study was obtained from Victoriastiftelsen, the Swedish Cultural Foundation in Finland, and from Sophie von Julins stiftelse and the Finnish Ministry of Agriculture and Forestry via the Nature and Game Management Trust Finland.
| Date made available | 23 Jun 2025 |
|---|---|
| Publisher | Zenodo |
Cite this
- DataSetCite