Corinne Vacher

Next‐generation biomonitoring proposes to combine machine‐learning algorithms with environmental DNA data to automate the monitoring of the Earth's major ecosystems. In the present study, we searched for molecular biomarkers of tree water status to develop next‐generation biomonitoring of forest ecosystems. Because phyllosphere microbial communities respond to both tree physiology and climate change, we investigated whether environmental DNA data from tree phyllosphere could be used as molecular biomarkers of tree water status in forest ecosystems. Using an amplicon sequencing approach, we analysed phyllosphere microbial communities of four tree species ( Quercus ilex , Quercus robur , Pinus pinaster and Betula pendula ) in a forest experiment composed of irrigated and non‐irrigated plots. We used these microbial community data to train a machine‐learning algorithm (Random Forest) to classify irrigated and non‐irrigated trees. The Random Forest algorithm detected tree water status from phyllosphere microbial community composition with more than 90% accuracy for oak species, and more than 75% for pine and birch. Phyllosphere fungal communities were more informative than phyllosphere bacterial communities in all tree species. Seven fungal amplicon sequence variants were identified as candidates for the development of molecular biomarkers of water status in oak trees. Altogether, our results show that microbial community data from tree phyllosphere provides information on tree water status in forest ecosystems and could be included in next‐generation biomonitoring programmes that would use in situ, real‐time sequencing of environmental DNA to help monitor the health of European temperate forest ecosystems.

The fungal pathogen Zymoseptoria tritici is the causal agent of Septoria tritici blotch (STB), a major wheat disease in Western Europe. Microorganisms inhabiting wheat leaves might act as beneficial, biocontrol, or facilitating agents that could limit or stimulate the development of Z. tritici. Improving our understanding of microbial communities in the wheat phyllosphere would lead to new insights into STB management. This article provides fungal and bacterial metabarcoding datasets obtained by sampling wheat leaves with and without symptoms caused by Z. tritici. Tissues were sampled from three commercial wheat varieties on three sampling dates during a cropping season. Weeds around wheat fields were sampled as well. In total, more than 450 leaf samples were collected. The pathogen Z. tritici was quantified using quantitative PCR. We provide the raw metabarcoding datasets, the amplicon sequence variant tables obtained after bioinformatic processing, the metadata associated with each sample (sampling date, wheat variety and tissue health condition), a preliminary descriptive analysis of the data, and the code used for bioinformatic and descriptive statistical analysis. [Formula: see text] Copyright © 2022 The Author(s). This is an open access article distributed under the CC BY-NC-ND 4.0 International license .