The work on plant community image analysis carried out in MAMBO has been used to deploy a new web tool available in Pl@ntNet, supported by investment from the GUARDEN project. 

MAMBO’s contribution focused on methodological innovations in plant community analysis. These advances are enabling researchers to better understand how plant species assemble and interact across different ecosystems. By moving beyond individual species records, MAMBO’s approaches support the interpretation of patterns in diversity, community composition, and ecological processes at the landscape scale. This strengthens the scientific basis for conservation action and habitat management.

Assessing the diversity and density of plant species per square metre is essential for monitoring changes in habitats. However, such monitoring is often time-consuming, which means data isn’t always collected at the required intervals. MAMBO is applying AI image recognition technology to tackle this challenge by (a) identifying all plant species visible in an image and (b) estimating the density of each plant species.

Pl@ntNet, a citizen science platform co-developed in 2009 by a consortium of French research organisations including CIRAD and Inria, underwent major developments as part of GUARDEN. These included a migration to a regionalised global flora, the extension of the platform to handle vegetation images containing multiple species (such as quadrat images), and the development of a tool for batch imports of plant observations. Aligning the platform with Plants of the World Online (POWO), maintained by Kew Gardens, ensured a consistent taxonomic backbone and improved interoperability with other biodiversity platforms.

One of the main benefits for users is the ability to work with regional floras that improve identification accuracy. These regionalised floras have been applied in GUARDEN case studies, including south-western Europe (Spain and France), Eastern Europe (Greece), East Asia (Cyprus), and the western Indian Ocean (Madagascar). In each of these regions, researchers can now select the most relevant flora through maps or GPS, ensuring more reliable identification results and contextualised ecological insights.

The integration of MAMBO’s plant community image analysis into Pl@ntNet contributes to these advances by providing new methods for studying species diversity and density, thereby supporting the monitoring of habitat changes. This innovation supports conservation and habitat monitoring efforts across multiple ecosystems, demonstrating its value in both European and international contexts.