Master (1/2)/césure internship (three to six months). Start date: January-March 2022.
Institut Sophia Agrobiotech (ISA), 400 route des Chappes, 06903 Sophia Antipolis, France.
Egg parasitoids, i.e. parasitic insect species that lay eggs inside the eggs of other insects, have great applied interest in agriculture: many are biocontrol agents (BCAs). They are released in large numbers to protect crops from pest attacks. They are also remarkable for their minute size (0.5mm or less), qualifying as the smallest insects in the world, and thus featuring extreme morphological and neurological miniaturization. One of the best-known examples is insects of genus Trichogramma, produced and used at industrial scale as an alternative to chemicals in different cropping systems, from maize fields to tomato-producing greenhouses.
In recent years, the experimental study of behavior and movement of insects in the laboratory has greatly benefited from the development of video-analysis and automated video-tracking techniques, in particular Multi-Object Tracking (MOT). This opens new perspectives for the quantitative phenotypic characterization of biocontrol agents, especially regarding the importance of behavior for population level functioning [1,2]. However, a major shortcoming of current tracking methods is the capacity to correctly identify individuals while they are interacting in a group (identity preservation), even without the use of marking techniques. This lack prevents the study of individual differences and personalities in realistic group contexts. IA methods offer the promise to lift this shortcoming, but few tools are available .
Through a collaboration with computer-scientists at Inria, our team is developing an approach using AI for detection and tracking, that exploits cryptic morphological features, and can outperform currently available tools [3,4]. Furthermore, the sex of Trichogramma individuals as well as their species membership could be automatically detected with high accuracy (>90%).
Objectives and Methods.
The goal of this internship is twofold, both methodological and biological. First, test this new tool on existing and new biological data involving groups of Trichogramma insects, and contribute to its development through a close collaboration with the Inria STARS team and another Master intern in computer sciences. Second, apply the tool to address the question of inter-individual differences and personalities in groups of Trichogramma. In particular, our goal is to quantify the extent of behavioral differences within groups, with respect to dominance status/social hierarchies, and to understand functional consequences at the population level (in particular parasitism efficiency).
The work will involve conducting behavioral experiments with groups of Trichogramma insects in the lab, acquiring hi-definition videos of the experiments, and using video-tracking tools to reconstruct individual trajectories and infer behavioral sequences. The part involving video-tracking and the test of IA techniques will be conducted in tandem with the computer-science Master intern based at the nearby Inria team, and will benefit from the co-supervision of a postdoc working on Trichogramma video-tracking and based in our lab. The intern will be responsible for analyzing the resulting data to test for the existence of inter-individual differences in behavior and social dominance within groups of Trichogramma wasps, and to infer their functional consequences for the efficiency of host parasitism and biocontrol.
– Experience of/taste for behavioral ecology, experimenting with insects, image analyses, and/or plant protection
– Taste for interdisciplinary collaborations and team-work
– Good computer skills (running programs in command line, manipulating large data files, data analysis with R)
– Autonomy and rigor
> How to apply
Send email, with CV + motivation letter, to firstname.lastname@example.org
 Lartigue S et al. (2020) Consistent variations in personality traits and their potential for genetic improvement of biocontrol agents: Trichogramma evanescens as a case study. PCI Ecology.  Burte V et al. (2021) Up and to the light: intra- and interspecific variability of photo- and geo-tactic oviposition preferences in genus Trichogramma. PCI Zoology.  Pani V. et al. (2021) TrichTrack: Multi-Object Tracking of Small-Scale Trichogramma Wasps. AVSS Conference Proceedings.  Walter T & Couzin ID (2021). TRex, a fast multi-animal tracking system with markerless identification, and 2D estimation of posture and visual fields. Elife