As part of MAMBO’s collaboration with GUARDEN, a hands-on training session took place on 19 June 2025 at the AMAP laboratory in Montpellier, France. Organised by the CIRAD team, the session focused on two innovative digital tools GeoPl@ntNet and Pl@ntNet Plot, developed to support large-scale biodiversity monitoring using AI-powered plant identification.
Led by Pierre Bonnet and Rémi Palard, the training gathered researchers and interns working on plant biodiversity and ecosystem monitoring. Participants were introduced to GeoPl@ntNet, a public web platform that allows users to explore fine-scale biodiversity maps across Europe. With interactive filters based on ecological traits and taxonomic categories, the tool helps identify key conservation areas and floristic hotspots, supporting science-based decision-making.
The session also featured an early demonstration of Pl@ntNet Plot, an experimental tool currently in beta testing. Built on the Pl@ntNet image recognition engine, it is designed to analyse high-resolution vegetation plot images and automatically identify the different plant species present. Participants had the opportunity to test the tools in real time and provide feedback on usability and functionality, contributing to further refinements of the interface and underlying models.
In addition to technical demonstrations, the training sessions highlighted the broader scientific motivations behind these tools: the need for scalable, standardised, and cost-effective approaches to plant community monitoring, especially in light of growing demands for data-driven conservation and restoration strategies under the EU Biodiversity Strategy for 2030.
For those who missed the in-person session, an online version is expected later this autumn. In the meantime, background material, including scientific articles, video tutorials, and challenge datasets from the recent PlantCLEF competition is already available online to support users in getting started.
The resources are:
GeoPl@ntNet paper: “Mapping biodiversity at very-high resolution in Europe”
Pl@ntNet Plot paper: “Adapting a global plant identification model to detect invasive alien plant species in high-resolution road side images”
