RESEARCH OBJECTIVES AND ACTIVITIES
Behavioral biomarkers offer unique insights into how contaminants impact individual fitness, revealing physiological changes and adaptive responses. Despite the sensitivity of behavioral assays, they are rarely included in pesticide regulation tests. Characterizing dose-response curves for biomarkers is crucial for identifying ecotoxicological parameters such as No Effect Concentration (NEC) and Effect Concentration 50 (EC50), thus providing important information for environmental risk assessment and risk mitigation strategies. Advances in computer vision now enable tracking body-part coordination allowing for automatic distinction between different behavioral categories (e.g. locomotion, grooming, social interactions). Since many pesticides can disrupt the neuro-endocrine system, these technologies have the potential to help differentiate exposed vs. non-exposed individuals.
Your objective will be to compare behavioral endpoints from classical methods for quantifying the activity of carabid beetles (distance moved and percent of arena explored) with pose-tracking approaches that train AI models to detect the motion of different body-parts. You will have access to an already-collected dataset of carabid exposed to a pesticide gradient for comparing the outcomes of both methods. In a second phase you will test whether pose-tracking allows for detecting toxicity parameters (NEC and EC50) at lower concentration compared to classical behavioral indicators.
WORKING ENVIRONMENT
You will be co-supervised by Raphaël Royauté and Colette Bertrand at UMR EcoSys and hosted at the Campus Agro-Paris Saclay in Palaiseau within the Soil & Tox team. You will have close interactions with the soil fauna technical staff for carabid collection in the field and maintenance under laboratory conditions.
TRAINING AND SKILLS REQUIRED
– Recommended training: PhD in Behahioral Ecology, Ecotoxicology or Ecology,
– Knowledge required: Proficiency with machine learning and / or deep learning techniques applied to video analyses of behavioral data. Very good Python and R knowledge. Good capacity to handle complex and high-dimensionality datasets
– Appreciated experience: Experience with arthropod behavioral assays
– Must have completed the PhD no later than fall 2022 (position subject to 3-years post-PhD rule as per French regulations)
Contract conditions
– Unit: UMR EcoSys
– AgroParisTech Campus, 91120, Palaiseau, France
– Type of contract: Postdoctoral contract
– Duration of the contract: 12 months
– Starting date: November 2025
– Remuneration: 3135€ gross salary
How to apply
Send a cover letter and a CV to raphael.royaute@inrae.fr and colette.bertrand@inrae.fr. Applications will be reviewed immediately until the position is filled. Deadline for applying: October 15th
Commentaires récents