Modelling coastal connectivity in the Mediterranean Sea: Similar effects of changes in hydrodynamics and reduction in planktonic duration
Data files
Sep 23, 2025 version files 1.51 GB
-
Cmatrix_decade_PD.zip
115.66 MB
-
Cmatrix_RCP85_2020s.zip
435.26 MB
-
Cmatrix_RCP85_2050s.zip
433.85 MB
-
Cmatrix_RCP85_2090s.zip
435.02 MB
-
Distance_RCP85_2020s.zip
23.99 MB
-
Distance_RCP85_2050s.zip
24.20 MB
-
Distance_RCP85_2090s.zip
24.16 MB
-
polygons.zip
179.99 KB
-
README.md
5.49 KB
-
Temperature.zip
3.94 MB
-
velocity.zip
13.22 MB
Abstract
Climate change can affect the dispersal of marine organisms and the connectivity of their populations through changes in hydrodynamics and the reduction of the planktonic duration of propagules. Few studies have investigated these effects over regional to global scales across a variety of release dates and sites. Here, we use hydrodynamic Lagrangian modelling to investigate these effects, their seasonal variation, and how they are modulated by the planktonic duration of dispersing propagules. Marine coastal organisms with planktonic propagules dispersed passively by ocean currents (e.g., seagrasses, mussels, sea urchins). Simulations of propagule dispersal were carried out using particle tracking model simulations driven by sea current velocity fields derived from the POLCOMS hydrodynamic model, under the Representative Concentration Pathway 8.5. We released propagules over three decades and followed their transport for planktonic durations up to 60 days. We quantified connectivity using three site-level variables (i.e., dispersal distance, coastal retention, and netflow) and a clustering approach. Changes in hydrodynamics and reduction in planktonic duration had similar effects on connectivity, leading to changes of up to 60% of the values of the 2020s. While the largest effects of changes in hydrodynamics were confined to a few areas, the effects of the reduction in planktonic duration were less variable spatially and seasonally. The effects of hydrodynamic changes were larger and more variable in species with longer planktonic durations. Because they had similar magnitudes, both the effects of changes in hydrodynamics and the reduction in planktonic duration should be considered in future studies. The expected changes in connectivity have potential consequences for genetic structure, the effectiveness of protected areas, and biological invasions. The dispersal dataset produced here is openly accessible and will support further research in these critical areas.
Dataset DOI: 10.5061/dryad.w9ghx3g1p
Description of the data and file structure
Results of the particle tracking simulation of propagule dispersal for Mediterranean coastal species, and code to analyse the results and produce the figures of the paper.
Archives: Cmatrix_RCP85_XXXXs.zip and Distance_RCP85_XXXXs.zip
Each .zip archive contains the results of particle tracking simulations for the given decade (2020s, 2050s, 2090s).
Archives named Cmatrix_RCP85_XXXXs.zip contain dispersal matrices.
Archives named Distance_RCP85_XXXXs.zip contain dispersal distances.
Particle tracking simulations were performed by releasing virtual propagules on the first day of each month for each of the ten years of each decade (except for the last two months of the decade), for a total of 118 release dates per decade. For example, for the decade 2020s, the first propagule release was on Jan 1st, 2020, and the last one was on Oct 1st, 2029. For each release date, results were calculated for different planktonic durations, from 5 days to 60 days in increments of 5 days (i.e, a total of 12 planktonic durations). Therefore, each .zip archive contains 118*12=1416 distinct files.
File: Cmatrix_RCP85_2020s.zip
File: Cmatrix_RCP85_2050s.zip
File: Cmatrix_RCP85_2090s.zip
Description: Dispersal probability matrices in RData format. Each file is named with the same pattern: Cmatrix_RCP85_YYYY_MM_KPLD.RData, where YYYY is the year, MM is the month (from 01 to 12), and K is the planktonic duration (from 5 to 60 in increments of 5).
Each file contains an R matrix object named Cmatrix_temp, whose elements Cmatrix_temp[i,j] give the number of propagules released from site j (column) that end up in site i (row) at the end of their planktonic duration. Note that the dispersal matrix is oriented so that Cij is dispersal from j to i, whereas in other datasets, dispersal matrices might be oriented so that Cij is dispersal from i to j. This choice of orientation is arbitrary, but one must always be consistent. From j to i is more common in population dynamics.
Sites are sorted in the same order as in the polygons.shp shapefile.
One hundred propagules were released at each site. To get probability out of the Cmatrix_temp, one must divide each element by 100, i.e.
Cmatrix_temp / 100
File: Distance_RCP85_2020s.zip
File: Distance_RCP85_2050s.zip
File: Distance_RCP85_2090s.zip
Description: Dispersal distance data in.RData format. Each file is named with the same pattern: Distance_RCP85_YYYY_MM_KPLD.RData, where YYYY is the year, MM is the month (from 01 to 12), and K is the planktonic duration (from 5 to 60 in increments of 5).
Each file contains a dataframe with 5055 rows (one per site) and 6 columns:
ID.polygon: the ID of the release site
dist_median: median great circle distance (in km) travelled by the 100 propagules released in that site at the end of their planktonic duration
dist_IQR: interquartile range of the great circle distance (in km) travelled by the 100 propagules released in that site at the end of their planktonic duration
PLD: planktonic duration (days)
year: Year of propagule release
month: Month of propagule release
Note that the PLD, *year, *and month columns do not contain useful information as they are all constant within the same file, but were retained to facilitate binding the dataframes of different files together.
Archive: polygons.zip
Description: The archive contains the shapefile of the 5055 square polygons used as release sites.
Archive: velocity.zip
Description: This .zip archive contains raster files of the zonal (uo) and meridional (vo) components of velocity, and for the resultant (vel) velocity, for each month of each decade. The raster files are named as: uo_surf_01_2020.tif, where "uo" indicates the zonal velocity, 01 indicates the month of January, and 2020 indicates the 2020 decade. Each raster file is the average velocity over the years of the decade, for the given month. These rasters are generated with the 01-Read_oceanographic_data.R script.
Archive: Temperature.zip
Description: This .zip archive contains raster files of temperature for each month of each decade. The raster files are named as: to_surf_01_2020.tif, where 01 indicates the month of January and 2020 indicates the 2020 decade. Each raster file is the average temperature over the years of the decade, for the given month. These raster are generated with the 01-Read_oceanographic_data.R script.
Archive: Cmatrix_decade_PD.zip
Description: This .zip archive contains RData files of dispersal matrices for each decade and planktonic duration. The RData files contain a matrix object analogous to those contained in the Cmatrix_RCP85_2020s.zip archive. These files are generated with the 02-Collate_data.R script.
Code/software
The dispersal matrices and the dispersal distance files are.RData file that can be opened with R.
The polygons shapefile and the velocity raster files can be opened with QGis or with R using the R packages sf or terra.
