Data from: Navigating danger: How environmental cues and individual traits shape breeding dispersal in the endangered common eider
Data files
Jan 07, 2026 version files 400.71 KB
-
dataset_breeding_dispersal.csv
390.86 KB
-
README.md
9.85 KB
Abstract
Breeding dispersal, or movement between successive breeding sites, is insufficiently understood. A ‘win-stay, lose-switch’ (WSLS) strategy may facilitate predator avoidance, but this strategy may be modulated by individual traits. We studied the drivers of breeding dispersal within and between breeding islands in female common eiders (Somateria mollissima), a ground-nesting site-fidelic sea duck, in the northern Baltic Sea. Rapidly increasing predation led us to expect increased dispersal over time. Using data spanning 21 years (2003-2023), we investigated which individual traits (nest success, age, body condition, boldness, relative head size) and environmental factors (predation risk, breeding density) influenced breeding dispersal. Within-island breeding dispersal distances showed moderate repeatability (r = 0.45), indicating individually consistent dispersal distances across years, but dropped to 0.20 after accounting for significant fixed effects, highlighting the roles of environmental and individual context in dispersal decisions. Consistent with the WSLS strategy, dispersal distances within islands and the likelihood of switching islands increased following predator-induced breeding failure and years with high island-specific nest predation risk. Older females switched islands less frequently, likely due to accumulated experience and age-dependent anti-predator strategies. Island switching was also more likely in years with abundant white-tailed sea eagles (Haliaeetus albicilla), the main predator of adult eiders, and directed toward larger islands offering tree cover, indicating risk perception beyond the immediate nest surroundings. Although philopatry can increase vulnerability to predation, flexible dispersal strategies may facilitate population persistence in this endangered population under intense predation. Future studies should assess the fitness outcomes of dispersal across diverse predation landscapes.
Article title: Navigating danger: how environmental cues and individual traits shape breeding dispersal in the endangered common eider
Journal: Behavioral Ecology
Authors: Ida Hermansson; Mikael von Numers; Kim Jaatinen; Jens Skog; Markus Öst
Corresponding author: Ida Hermansson (ida.hermansson@abo.fi)
Dataset Overview
File: dataset_breeding_dispersal.csv
Dataset underpinning analyses of breeding dispersal in female common eiders across 21 years in the northern Baltic Sea. Variables include individual traits, island attributes and predation risk indices. Used to test win-stay, lose-switch (WSLS) responses and influences of environmental cues on within- and between-island dispersal.
The dataset was used for four analyses: 1) factors explaining breeding dispersal distances within islands, 2) factors explaining breeding dispersal between islands, 3) directionality of movement relative to mainland, 4) repeatability of within-island dispersal distances.
Variable Dictionary
| Variable | Description |
|---|---|
year |
Calendar year of the observation (YYYY). |
island_name |
Island code name where the female bred in the focal year. |
female_id |
Ring ID of the breeding female. |
fid_median_lifetime |
Median lifetime Flight Initiation Distance (FID; meters) across all years the female was measured; proxy for boldness. |
relative_head_size_mean_lifetime |
Mean head size of the female across all measured years, corrected for body size using regression residuals (unitless). Used as a proxy for cognitive ability. |
female_body_condition_index |
Body condition at hatching, calculated as standardized residuals from a linear regression predicting female projected weight at hatching from structural size (radius–ulna length). Higher values indicate better condition relative to body size (see Methods). |
minimum_breeding_experience_years |
Minimum years since first capture (0 = banding year/first recorded breeding; 1 = the following year, etc.). |
island_area |
Area (m²) of the current breeding island. |
island_area_previous_attempt |
Area (m²) of the island used in the previous breeding attempt. |
island_type_previous_attempt_forested1_open0 |
Island type at previous attempt: 1 = forested, 0 = open. |
nest_density_previous_attempt |
Distance-weighted local (within 100 m) nest density index at previous breeding attempt . |
nest_cover_previous_attempt |
Nest cover at previous attempt measured from fish-eye photo (black-pixel proportion). |
dispersal_distance_within_island |
Distance (m) moved within the same island between successive breeding attempts. |
dispersal_between_islands_switched1_stayed0 |
Indicator of between-island dispersal: 1 = changed breeding island; 0 = stayed on same island. |
years_between_observations |
Years elapsed between consecutive observations (to include non-annual breeding). |
early_nest_predation |
Island-specific early nest predation risk (proportion depredated nests to total nesting attempts) during first nest visit in focal year. |
final_nest_predation_previous_attempt |
Island-specific final nest predation risk (proportion of depredated nests to total nesting attempts) in the breeding season of the previous recorded attempt. |
adult_predation_proportion_previous_year |
Adult female predation risk (proportion killed females/breeding attempts) in previous year. |
individual_nest_success_previous_year_succeed1_failed0 |
Individual nest outcome in previous year: 1 = success, 0 = failure. |
eagle_abundance |
Hanko Bird Observatory eagle abundance index in focal year (mean daily count April–June). |
eagle_abundance_previous_attempt |
Eagle abundance index during the year of the female’s previous recorded breeding attempt. |
change_in_distance_to_mainland_meters |
Difference (m) in the distance to mainland between current and previous nest (negative = mainlandward). |
Location
Location: The data was collected at the archipelago off Tvärminne Zoological Station (western Gulf of Finland), Baltic Sea.
Code & Analyses
Statistical analyses were performed in R 4.4.2. See manuscript for specifics.
Loaded packages:
- performance: 1.0+ (model R²)
- MuMIn: 1.48.4 (AICc dredge)
- geepack: 1.3.12 (GEEs)
- rptR: 0.9.23 (repeatability)
Ethics Statement
Handling of animals was reviewed and approved by the National Animal Experiment Board (permit numbers HY-85-2003, ESLH-2009-02969/Ym-23, ESAVI/1697/04.10.03/2012, ESAVI/2831/04.10.07/2015, ESAVI/4053/2018, ESAVI/10022/2021), ringing was done with a requisite ringing license, and our work complied with the regulations of Tvärminne Zoological Station.
File Naming Conventions
Filenames and column names use ASCII, lowercase_with_underscores; indicator variables are coded 0/1 as documented in the variable dictionary. Missing values are represented by empty cells.
