Data from: Snowmelt predicts earlier breeding across the latitudinal range of an Arctic nesting seabird, the Little Auk (Alle alle)
Data files
Jun 05, 2026 version files 55.09 KB
-
01_phenology_snowmelt_analysis.R
12.14 KB
-
02_chick_growth_survival_analysis.R
15.54 KB
-
hatching_dates.csv
3.12 KB
-
Polygones.zip
17.80 KB
-
README.md
6.50 KB
Abstract
Climate-driven temporal shifts in seasonal environments are altering some of the environmental cues that organisms use to time reproduction, potentially leading to trophic mismatches across ecosystems. In the Arctic, marine predators must balance conditions at sea with local terrestrial constraints at breeding sites, yet the relative importance of these cues for breeding phenology remains unclear. Here, we used a crevice-nesting High-Arctic planktivorous seabird, the little auk Alle alle, as a model species. Drawing on a unique multi-year dataset from four colonies with distinct climatic regimes, we tested whether breeding onset tracks the timing of snowmelt at breeding sites, a key terrestrial cue determining nest accessibility. We found that snowmelt timing is closely linked with hatching date, with earlier snowmelt enabling earlier access to nesting crevices and advancing hatching across all sites. Importantly, we detected no significant directional temporal trend in snowmelt timing over the study period (2000–2024), suggesting that this relationship reflects interannual variability. Across colonies, later hatching within years was associated with reduced chick growth and survival. However, interannual variation in mean hatching date was linked to chick survival in only one colony, indicating spatial heterogeneity in demographic consequences of breeding phenology.
Future projections indicate that snowmelt timing will advance where little auks breed, potentially advancing breeding timing. However, other ongoing changes—such as borealization of zooplankton communities and the loss of summer sea ice— may alter future fitness consequences of breeding timing. Our study highlights the role of the terrestrial environment in shaping the breeding timing of high-latitude marine birds.
Authors: Martyna Syposz, Øystein Varpe, Sébastien Descamps, Jérôme Fort, David Grémillet, Ann Harding, Dariusz Jakubas, Dorota Kidawa, Nomikos Skyllas, Hallvard Strøm, Tom S.L. Versluijs, Katarzyna Wojczulanis-Jakubas
Journal: Journal of Animal Ecology
DOI: 10.1111/1365-2656.70287
Date of data collection: 2004–2024
Geographic coverage: Svalbard archipelago (Norway) and East Greenland
Files in this dataset
| File | Description |
|---|---|
hatching_dates.csv |
Median, mean and SD of hatching date per colony and year (Supplementary Material 1) |
BearIsland.* |
Shapefile — Bjørnøya colony polygon (6 files) |
Hornsund.* |
Shapefile — Hornsund colony polygon (6 files) |
Isfjorden.* |
Shapefile — Isfjorden colony polygon (6 files) |
EGreenland.* |
Shapefile — Ukaleqarteq colony polygon (6 files) |
01_phenology_snowmelt_analysis.R |
R script — hatching date vs snowmelt and temporal snowmelt trend |
02_chick_growth_survival_analysis.R |
R script — chick growth and survival analyses |
1. hatching_dates.csv
This file corresponds to Supplementary Material 1 (Table S1) in the published manuscript. It contains summary statistics of hatching date for each colony and year, along with the corresponding sample size.
Only years with at least 10 nests monitored with a precision of ±3 days or less are included.
Column descriptions
| Column | Description |
|---|---|
Location |
Name of the little auk breeding colony (see colonies below) |
Year |
Calendar year of data collection |
No. of nests |
Number of nests included (precision ≤ ±3 days) |
Median hatching date |
Median hatching date as calendar date (DD/MM/YYYY) |
Median hatching date (Day of year) |
Median hatching date as day of year (1 = 1 January) |
Mean hatching date (Day of year) |
Mean hatching date as day of year |
SD hatching date (Day of year) |
Standard deviation of hatching date in days |
Study colonies
| Colony name | Coordinates | Country |
|---|---|---|
| Bjørnøya | 74°23'N 19°2'E | Svalbard, Norway |
| Hornsund | 77°1'N 15°32'E | Svalbard, Norway |
| Isfjorden | 78°13'N 15°19'E | Svalbard, Norway |
| Ukaleqarteq (Kap Høegh) | 70°44'N 21°35'W | East Greenland |
Missing data
Not all colonies were monitored every year. Missing years reflect either no fieldwork or insufficient sample size — they do not indicate zero hatching events.
2. Shapefiles — colony polygons (Supplementary Material 3)
These shapefiles define the polygon boundaries used to extract MODIS satellite imagery for snowmelt date estimation, as described in the Methods section and Supplementary Material 3.
Each colony is represented by six files that must be kept together:
| Extension | Content |
|---|---|
.shp |
Geometry |
.dbf |
Attributes |
.prj |
Coordinate reference system |
.shx |
Spatial index |
.cpg |
Character encoding |
.qmd |
QGIS metadata |
Coordinate reference system: WGS84 (EPSG:4326)
File naming from folder Polygones.zip
| Filename prefix | Colony |
|---|---|
BearIsland |
Bjørnøya, Svalbard (74°23'N 19°2'E) |
Hornsund |
Hornsund, Svalbard (77°1'N 15°32'E) |
Isfjorden |
Isfjorden, Svalbard (78°13'N 15°19'E) |
EGreenland |
Ukaleqarteq, East Greenland (70°44'N 21°35'W) |
Note: EGreenland corresponds to the Ukaleqarteq colony throughout the manuscript and in hatching_dates_SM1.csv.
3. R scripts
01_phenology_snowmelt_analysis.R
Analyses the effect of snowmelt date on hatching date (DOY) across four colonies using linear mixed-effects models (LMMs) with colony-centred snowmelt and a snowmelt × colony interaction. Also tests for temporal trends in snowmelt date (DOY ~ year × colony). Produces Figure 2.
Corresponds to Models 1 and 2 in Table 1 of the manuscript.
Input files:
All_Joined.csv— nest-level hatching dates and snowmelt DOY per colony and year*_Snowmelt_NDSI.csv— one file per colony, derived from MODIS NDSI workflow
02_chick_growth_survival_analysis.R
Analyses whether hatching date predicts chick growth rate (g/day, days 1–15) and survival to day 15. Hatching date is decomposed into within-year (individual deviation from annual mean) and between-year (interannual variation) components. Fits binomial GLMMs (survival: Hornsund, Isfjorden) and LMMs (growth: Hornsund, Ukaleqarteq). Produces Figure 3.
Corresponds to Models 3 and 4 in Table 1 of the manuscript.
Input files: colony-specific chick growth and survival files (see script for details).
Note on raw data: Raw input data for both R scripts are not publicly archived and are available from the corresponding author upon reasonable request. File paths in both scripts must be updated to reflect local data locations before running.
Contact
For questions regarding this dataset, please contact: Martyna Syposz syposzmartyna@gmail.com
