Benjamin BOUREL
Researcher
National Institute for Research in Digital Science and Technology (Inria)
benjamin.bourel@inria.fr
Machine learning
Ecology
Paleo-ecology
Palynology
I am a botanist and palaeo-botanist specialising in the study of pollen, phytoliths (silica concretions produced by plants in and between their cells) and leaves in France, Arabia and East Africa.
My early work focused on the reconstruction of environments, ecosystems and climate. These reconstructions were mainly for the Holocene (present at 0.02 Ma*) and the Plio-Pleistocene (0.02 Ma to 5.3 Ma) where I was interested in questions about the dependence of human and pre-human populations on their environment. I have also worked on older periods through the reconstruction of dinosaur environments in Lozère (France) during the Middle and Lower Jurassic (201.4 to 161.5 Ma).
Over time, my research has gradually included work on the use of statistical analysis, machine learning and deep
learning to solve problems in ecology and palaeoecology. It was in the context of this work at the interface between ecology, palaeontology, statistics and machine learning that I was recruited as a Research at the National Institute for Research in Digital Science and Technology (Inria) in 2023.
* Ma = million d'années
I am working on the taxonomic identification of current and fossil plant remains (mainly pollen, phytoliths and leaves) using automatic image recognition and deep learning. The aim is to help carry out high-resolution monitoring of changes in biodiversity, climate and landscapes in the present and the past using images collected by scientific workflows such as artificial pollen traps, sediment cores and herbariums.
Current or fossil plant remains are signals of vegetation and paleo-vegetation. By studying these signals, it is possible to monitor past and present environmental changes. However, the interpretation of certain signals, such as pollen assemblages, is sometimes difficult due to the complexity of the signal or its degradation. Using statistical analyses, I aim to improve the environmental and climatic interpretation of these complex and/or degraded signals.
I use natural archives (pollens, fossils, etc.) and historical documents (notes, herbariums, archaeological remains, etc.) to take into account both the past and present ecological niches of plants in order to set up models of species distribution. The aim is to better predict the future distribution of plants, to help conserve endangered species or to anticipate the spread of invasive species.
Contact
benjamin.bourel@inria.fr
Antenne Inria de Montpellier
860 Rue Saint-Priest
Bâtiment 5
34090 Montpellier
This site was created with the Nicepage