Mapping the Distribution of an Intertidal Invasive Red Macroalga at a National Scale Using Satellite Remote Sensing.

Remote Sensing
Spectral Library
Invasive Species
Cloud Computing
Satellite
Machine Learning
First Author
France
National Scale
Neural Networks
Bede Ffinian Rowe Davies,Simon Oiry, Mar Roca-Mora, Martin Gade, Valérie Stiger-Pouvreau, Anne-Laure Barillé, Anthony Le Bris, Pierre Gernez & Laurent Barillé
Author

Bede Ffinian Rowe Davies;Simon Oiry; Mar Roca-Mora; Martin Gade; Valérie Stiger-Pouvreau; Anne-Laure Barillé; Anthony Le Bris; Pierre Gernez; Laurent Barillé

Published

November 6, 2026

Link here: Davies et al., 2026

Abstract

Invasive species have been ranked among the greatest causes of biodiversity loss, specifically the loss of already threatened or endangered species, with coastal and estuarine systems often bearing the brunt of initial introductions and subsequent invasions. This is likely due to high levels of global transport, alongside other anthropogenic activities within these systems. Macroalgae species make up a large proportion (40 %) of Invasive Alien Species in the European marine environment. Yet, monitoring of these macroalgal species is lacking: how they spread, change seasonally and affect local ecosystems is poorly understood. This lack of spatio-temporal information is prevalent even for species widespread across multiple countries and coastlines, such as the intertidal red macroalga Gracilaria vermiculophylla (formerly Agarophyton vermiculophyllum). Here we present the methods and results of using open-access Copernicus Sentinel-2 satellite imagery to map this invasive red macroalga across the Atlantic Coast of France for the summer of 2024. We initially improved training data of a published intertidal habitat classifier and validated the results using independently sourced data. By doing so, we achieved high levels of classification accuracy at predicting this red macroalga (93.2% global accuracy and 0.82 F1 score). We found that, while there were no evident latitudinal patterns, G. vermiculophylla was most prevalent within transitional (i.e. estuarine) intertidal areas when compared to more open coastal areas, where some hotspots were still identified. Furthermore, while it has been suggested that the initial invasion of G. vermiculophylla was mediated through aquaculture activities, the presence of aquaculture showed no clear relationship to the proportion of the red macroalga covering the intertidal area. Moreover, seasonal dynamics of extent for this red macroalga revealed difference between two sites, Belon Estuary in Brittany and Bonne Anse in the Gironde Estuary, known to have consistent G. vermiculophylla cover. This work provides an open-access map of G. vermiculophylla across the whole Atlantic Coast of France, as well as illustrating a framework that could be employed to allow rapid monitoring of invasive macroalgae. Monitoring and understanding spatio-temporal dynamics is an essential step in managing both invasive species and the ecosystems that they influence.