Drivers of seasonal dynamics in Ulva spp. associated microbiota and surface metabolome: The interplay between environment and host physiology

Abstract

Marine ecosystems are increasingly affected by climate change and eutrophication, placing considerable stress on macroalgae and potentially disrupting their mutualistic interactions with surface-associated microorganisms, especially bacteria. This study explored the temporal dynamics of epiphytic microbial communities (16S rRNA) and the surface metabolome (LC-MS) of Ulva spp. from southern Brittany (France) over a 17-month period. The data obtained were analyzed through multivariate techniques to uncover correlations between bacterial community structure and surface metabolites. Biochemical and environmental parameters were also examined to identify the factors driving temporal variations in the macroalgal surface landscape. The analysis revealed distinct bacterial communities on Ulva spp. surface, compared to those in surrounding seawater or on rocky substrate. Key environmental factors, including salinity, nitrate concentration, and precipitation, as well as the physiological traits of Ulva spp. (such as uronic acid content, lipid levels, and moisture), were identified as primary drivers of microbiota dynamics. A stable core microbial community, primarily composed of Alphaproteobacteria, Flavobacteriia, Oligoflexia, and Saprospira, was also observed within the Ulva spp. microbiota. Finally, correlations between Amplicon Sequence Variants (ASVs) and surface metabolites underscored the importance of an integrated multi-omics approach to enhance our understanding of the complex dynamics between eukaryotic hosts and their associated microbiota.

Keywords: Epiphytic bacteria; Seasonal drivers; Surface metabolome; Temporal dynamic.

Graphical abstract

Fig. 1

Non-metric multidimensional scaling of Bray-Curtis distance computed from CSS-transformed ASV abundances for the Ulva-associated bacterial community across all stations (S1, S2, S3) and replicates, sampled between February 2021 and February 2022 (Stress: 0.19). Host species were included as explanatory variables using ‘envfit’ with centroid positions displayed.

Fig. 2

Non-metric multidimensional scaling of Bray Curtis distance computed from CSS-transformed ASVs abundances for the bacterial community of Ulva spp. (green), rocky substrate (red) and seawater (blue) sampled between February 2021 and June 2022 (Stress: 0.14).

Fig. 3

Relative abundance of majoritarian bacterial ASVs summarized at the genus taxonomic rank in associated communities of Ulva spp., seawater and rocky substrate.

Fig. 4

Relative abundance of ASVs belonging to the core Ulva spp. bacterial community between February 2021 and June 2022, summarized at the genera taxonomic rank.

Fig. 5

Non-metric multidimensional scaling of Bray Curtis distance computed from CSS-transformed ASVs abundances for the bacterial community in seawater sampled between February 2021 and June 2022 (Stress 0.15). Environmental parameters were added using envfit. The red arrows represent a statistically significant relationship (p> 0.05) measured through the envfit test.

Fig. 6

Non-metric multidimensional scaling of Bray Curtis distance computed from CSS-transformed ASVs abundances for the bacterial community of Ulva spp. sampled between February 2021 and June 2022 (Stress: 0.07). Environmental parameters were added using envfit.The red arrows represent a statistically significant relationship (p> 0.05) measured through the envfit test.

Fig. 7

Non-metric multidimensional scaling of Bray Curtis distance computed from CSS-transformed ASVs abundances for the bacterial community of Ulva spp. sampled between June 2021 and February 2022 (Stress: 0.05). Biochemical parameters were added using envfit.The red arrows represent a statistically significant relationship (p> 0.05) measured through the envfit test.

Fig. 8

Principal component analysis of the seasonal structuring of the Ulva spp. surface metabolome (a); multivariate statistical analysis on discriminant features (orange) and abiotic parameters (b).

Fig. 9

Heatmap of mean normalized and centered seasonal features intensity; p value < 0.05, Kruskall-Wallis test). Abbreviations: DMSP: diméthylsulfoniopropionate; DMPO: 5,5-Dimethyl-1-pyrroline N-oxide; LDAO: Lauryldimethylamine oxide.