Data from: Limited source-sink connections shape south western Pacific coral reef resilience under current and future warming
Data files
Apr 01, 2026 version files 82.80 KB
-
Dataset_S1_Reef_Classifications.xlsx
59.57 KB
-
README.md
3.72 KB
-
stochastic_replicates.R
19.51 KB
Abstract
This script implements a stochastic Lagrangian particle tracking model to simulate coral larval dispersal across 850 reefs in the southwestern Pacific (17–32°S, 149–168°E), encompassing the southern Great Barrier Reef, Lord Howe Island, New Caledonia, and the Coral Sea. Larval trajectories are driven by daily oceanographic current data from BRAN2020 (u, v velocity fields), with survival probabilities modulated by sea surface temperature and salinity using Gaussian penalty functions. A stochastic random walk component (factor = 0.2) is added to velocity fields to account for sub-grid turbulent diffusion. Larvae are tracked over species-specific pelagic larval durations, with settlement occurring when a larva passes within one kilometre of a reef following a pre-competency period. The model outputs daily larval trajectories and a reef-to-reef connectivity matrix quantifying successful dispersal events between source and sink reefs across the 13-year simulation period (2011–2024).
Authors
- E. Dehont¹*, S. Choukroun², A. Grech³, K.M. Quigley⁴·⁵*
¹ Centre for Fisheries Ecosystem Research, School of Fisheries, Fisheries and Marine Institute, Memorial University, St. John's, NL, Canada
² Centre of Tropical Water and Aquatic Ecosystem Research, James Cook University, Townsville, Queensland 4811, Australia
³ College of Science and Engineering, James Cook University, Townsville, Queensland 4811, Australia
⁴ Minderoo Foundation, Perth, Western Australia, Australia
⁵ James Cook University, Townsville, Queensland 4811, Australia
Associated manuscript: Limited source-sink connections shape south western Pacific coral reef resilience under current and future warming — Journal of Applied Ecology
DOI: to be added upon publication
Dataset description
This dataset contains the R code and reef classification data used to simulate coral larval dispersal across 850 reefs in the southwestern Pacific (2011–2024) using a stochastic Lagrangian particle tracking model. Data support analyses of source-sink dynamics, reef connectivity networks, and the effects of climate warming on larval dispersal.
File list
| Filename | Description |
|---|---|
stochastic_replicates.R |
R script implementing the stochastic Lagrangian particle tracking model used to simulate coral larval dispersal |
Dataset_S1_Reef_Classifications.xlsx |
Coordinates, names, eco-evolutionary classifications and metadata for all 850 reefs included in the study |
Variable descriptions
Dataset_S1_Reef_Classifications.xlsx
| Variable | Description | Units |
|---|---|---|
lat |
Reef latitude | Decimal degrees |
lon |
Reef longitude | Decimal degrees |
Reef_name |
Reef name | — |
Reef_class |
Eco-evolutionary classification (recovery, resistance, avoidance, additional) | — |
location |
Geographic location / region | — |
Reef_Number |
Unique reef identifier | — |
Source |
Source study for reef classification | — |
Zone |
Spawning zone (Northern or Southern) | — |
Methods summary
Larval dispersal was simulated using daily BRAN2020 oceanographic data (0.1° resolution). Five larvae were released per reef per year on November 20th (northern reefs) and January 20th (southern reefs). Survival probability was modulated by temperature and salinity using Gaussian penalty functions. A stochastic random walk (factor = 0.2) was applied to velocity fields. Settlement occurred within 1 km of a reef after a pre-competency period over a specified pelagic larval duration.
Software
R version 4.5.1 — key packages: terra, ncdf4, sf, geosphere, dplyr