Shifts in the phytoplankton assemblage induced by environmental changes have significant implications for carbon cycling and marine food webs, but remain poorly constrained across spatiotemporal scales. Here, we investigate the effects of rising sea surface temperatures and increased stratification on the phytoplankton composition and size in Northwestern Mediterranean Sea (2010-2019) using two sediment trap series: one in the oligotrophic Ligurian Sea and the other in the deep convection zone of the Gulf of Lion. We apply deep-learning image analysis to quantify phytoplankton particle fluxes, size distributions, and relative assemblages, focusing on coccolithophores, diatoms, and silicoflagellates. Our results show a general decline of phytoplankton fluxes to the seafloor, mirroring the decrease in vertical mixing in the water column. Both sites show a shift towards phytoplankton species adapted to stratified and nutrient-depleted conditions, although with contrasting patterns. In the Ligurian Sea, deep-dwelling coccolithophore species become dominant, while in the Gulf of Lion, summer-associated siliceous species, including large diatoms and silicoflagellates, show an increase. These contrasted trends, which likely result from differences in nutrient inputs and pH changes in the surface between the two sites, have implications for the efficiency of carbon export pathways at depth. Specifically, the increasing dominance of smaller phytoplankton in the Ligurian Sea leads to a reduction in carbon burial efficiency, while in the Gulf of Lion, the enhanced contribution of larger diatoms may sustain relatively higher export and burial rates in the future.

A document by Chloé Tilliette. Click on the document to view its contents.

Diatom communities preserved in sediment samples are valuable indicators for understanding the past and present dynamics of phytoplankton communities, and their response to environmental changes. These studies are traditionally achieved by counting methods using optical microscopy, a time-consuming process that requires taxonomic expertise. With the advent of automated image acquisition workflows, large image datasets can now be acquired, but require efficient preprocessing methods. Detecting diatom frustules on microscope images is a challenge due to their low relief, diverse shapes, and tendency to aggregate, which prevent the use of traditional thresholding techniques. Deep learning algorithms have the potential to resolve these challenges, more particularly for the task of object detection. Here we explore the use of a Faster R-CNN (Region-based Convolutional Neural Network) model to detect siliceous biominerals, including diatoms, in microscope images of a sediment trap series from the Mediterranean Sea. Our workflow demonstrates promising results, achieving a precision score of 0.72 and a recall score of 0.74 when applied to a test set of Mediterranean diatom images. Our model performance decreases when used to detect fragments of these microfossils; it also decreases when particles are aggregated or when images are out of focus. Microfossil detection remains high when the model is used on a microscope image set of sediments from a different oceanic basin, demonstrating its potential for application in a wide range of contemporary and paleoenvironmental studies. This automated method provides a valuable tool for analysing complex samples, particularly for rare species under-represented in training datasets.