Coastal oceanographic connectivity at the global scale: a dataset of pairwise probabilities and travel times derived from biophysical modeling

Abstract

Ocean currents are fundamental drivers of marine biodiversity distribution, mediating the exchange of genetic material and individuals between populations. Their effect ranges from creating barriers that foster isolation to facilitating long-distance dispersal, which is crucial for species expansion and resilience in the face of climate change. Despite the significance of oceanographic connectivity, comprehensive global estimates remain elusive, hindering our understanding of species' dispersal ecology and limiting the development of effective conservation strategies. We present the first dataset of connectivity estimates (including probability of connectivity and travel time) along the world's coastlines. The dataset is derived from Lagrangian simulations of passive dispersal driven by 21 years of ocean current data and can be combined with species' biological traits, including seasonality and duration of planktonic dispersal stages. Alongside, we provide coastalNet, an R package designed to streamline access, analysis, and visualization of connectivity estimates. The dataset provides a new benchmark for research in oceanographic connectivity, enabling a deeper exploration of the complex dynamics of coastal marine ecosystems and informing more effective conservation strategies.

Fig. 1

Global distribution of hexagon-shaped coastal sites (depicted in red; hexagons of 9.85 km side) from where passive Lagrangian particles were released in the simulation of oceanographic connectivity. The zoomed-in areas provide more detailed views of the hexagon-shaped coastal sites.

Fig. 2

Metrics of oceanographic connectivity across a range of planktonic durations. (a) Relationship between the number of connections among sites and propagule duration. (b) Relationship between the average connectivity distance among sites and propagule duration. (c–h) Frequency distribution of connectivity distances for different planktonic durations. The vertical dashed lines depict the distances at which 99% of connectivity events are performed.

Fig. 3

Sensitivity analysis of particle release frequency. (a) Pearson’s correlation coefficient, (b) mean absolute difference in connectivity probability, and (c) Cohen’s d, comparing connectivity matrices derived from simulations with varying numbers of particles released per day (from 1/75 to 24) to a reference simulation with 24 particles released per day (corresponding to one particle released per hour, matching our simulation’s hourly time step). The vertical dashed line and the red dot indicate the original configuration of 1 particle per day. (d) Scatter plot comparing pairwise connectivity probabilities estimated with 24 particles per day versus 1 particle per day.