Trichogramma for plant protection: Pangenomics, Traits, and establIshment Capacities – TriPTIC

WP1 « From genome to species » will provide 1) accurate genetic characterisation of strains using RAD markers 2) easy-to-use identification tools for other WPs, ecologists and agricultural practitioners (by assessing the power of classical markers or choosing some markers among the RAD loci) and 3) insights into the evolutionary history of the group (inference of phylogenetic trees). WP2 « From microbiome to phenotype » will use high-throughput sequencing to identify the diversity of symbiotic bacteria infecting the strains and that may impact their phenotype or reproduction. WP3 « From individual phenotype to species » will provide a comprehensive assessment of key individual phenotypes by using state-of-the-art image analysis and computer vision tools (“ethomics” approaches). It will generate data on 1) fundamental life-history traits, their variability and adaptive potential, 2) host-seeking and movement strategies related to host attack and ecological specialization, and 3) the strength of species boundaries. WP4 « From phenotype to population dynamics » will investigate the demographic and genetic processes at play during population establishment and early expansion using laboratory microcosms. It will also investigate how heterogeneous selective pressures along an invasion front affect the spatial and temporal dynamics of species assemblages, and test if functional diversity increases the performance of a community in a complex agro-ecosystem. WP5 « From collection to transfer » will be in charge of sampling, strain maintenance and knowledge transfer towards stakeholders. It will contribute to the creation of a perennial bio-repository (Biological Resource Center) that will store strains and side-products (e.g. vouchers, DNA extracts) and aggregate knowledge. Resources accumulated by the Biological Resource Center are planned to be used for applied and academic purposes.

During the first part of the project we have,
WP1 : 1) validated a Whole Genome Amplification method to obtain 120,000 RAD markers from a single Trichogramma specimen 2) constructed a first RADseq library on 13 species of Trichogramma 3) constructed and published a bioinformatic pipeline to make the processing of RAD-seq data easier and allow rapid and automated exploration of parameters for phylogenetic inference
WP2 : 1) set up a 2-step PCR protocol for the amplification and simultaneous sequencing of COI (barcode) and 16S (microbiome) genes, 2) developed a bio-informatic pipeline for the fast and accurate analysis of raw data 3) tested our protocol on 5 strains of Trichogramma.
WP3 : 1) developed and implemented as a user-friendly ImageJ plugin (CODICOUNT) that automatically quantifies parasitism rates. The method has been tested on c.a. 50 strains of T. cacoecieae, T. brassiceae and T. chilonis, 2) set up experimental and analytical pipelines to automatically quantify: i) the phototactic and geotactic preferences of Trichogramma, ii) the movement speed and overall diffusion rate of Trichogramma, iii) the parasitic performance of Trichogramma as a function of the spatial distribution of host eggs.
WP4 : 1) characterized development time and sex-ratio in 55 strains available from the first sampling season, 2) characterized local population dynamics on a subset of 30 strains representative of the sampled diversity.
WP5 : 1) initiated a citizen-science project at national level to collect strains. We also sampled in about 60 sites in multiple French departments, 2) barcoded all strains using one or more genes, 3) conducted crossing experiments and behavioural observations to investigate the reproductive compatibility of strains with shared haplotypes, 4) developed a database with photos of key morphological features for the world species of Trichogramma, 5) have submitted a project for the accreditation of the Biological Resource Center, which has been accepted.

The critical phase of the project was sampling. During the first part of the project efforts have thus been made to collect strains on the field (collaborative network, sampling on our own etc.). Today, more than 200 strains are stored in the Biological Resource Center. Inferences about the genomic, phenomic and population dynamics of each strain initiated on the strains collected at the beginning of the project will now be generalized to all the collected strains. The preliminary results confirmed that the taxonomy of the group is unclear, that lab and wild type strains have different phenotypes (e.g. different strategies to exploit host eggs) and that the collected strains exhibit different genotypes and population dynamics.

- Paper published : Cruaud A.*, Gautier M.* (eq. contributors), Rossi J.-P, Rasplus J.-Y, Gouzy J. (2016). RADIS: Analysis of RAD-seq data for InterSpecific phylogeny. Bioinformatics 32, 3027–3028
- Software publicly available : RADIS (http://www1.montpellier.inra.fr/CBGP/software/RADIS/)
- Paper in preparation : Perez G., Burte V., Baron O., Calcagno V. CODICOUNT : une méthode d'analyse d'image pour déterminer automatiquement le nombre d'oeufs et les taux de parasitisme lors de l'élevage de Trichogramma sp. Les Cahiers Techniques de l'INRA.
- Accreditation of the Biological Resource Center EP-coll (http://www.spe.inra.fr/Toutes-les-actualites/EP-Coll)

France has recently adopted the Ecophyto 2018 plan, which recommends the development of effective biological control programs against pests as alternatives to pesticides. It is acknowledged that successful and safe biological control depends on 1) accurate genetic and phenotypic characterization of the strains of natural enemies released; 2) strong knowledge of their life-history traits, strategies of host exploitation and population dynamics; 3) good understanding of the processes determining their successful establishment. Nevertheless, such studies are rarely, if ever, conducted, which hampers our ability to implement efficient programs against pests. During this project, joining our complementary skills, we will tackle these topics, thereby addressing key questions in evolutionary biology about speciation, hybridization, adaptation, specialisation, and interaction dynamics.
We will focus on four species complexes of the genus Trichogramma (Hymenoptera), one of the most studied and commercialized groups to control pests of major economic importance in Europe. Nevertheless Trichogramma are frequently released without taxonomic, genetic, phenotypic or behaviour characterization and in-depth risk assessment study. This practice must be improved to comply with a recent decree adopted in France stipulating that macro-organisms must be precisely characterized before released, to ensure traceability and prevent unintended effects. Trichogramma species complexes are poorly understood and entities with weak reproductive barriers may exist. Strains display significant inter- and intra-population variations in biological traits that are influenced by environmental factors, and several traits important for biocontrol success have been poorly documented (e.g. movement and dispersal capacities). Infections by symbionts can influence individual fitness, reproductive isolation and ecological specialization, but there is little data on most symbionts associated with Trichogramma species. Finally their evolutionary history is poorly known. Using novel and powerful methods, we will:
1) Investigate species limits, phenotype biocontrol traits and quantify gene flow and evolutionary potential within the species complexes. We will characterize sampled strains and those already commercialized, using high-throughput methods of genotyping (RAD-sequencing), phenotyping (high-throughput ethomics), and endosymbiont screening, still rarely used, though informative and powerful.
2) Study the impact at the population level of the individually-characterized traits
3) Analyse the evolutionary processes at play when populations are introduced into a novel and complex habitat, either alone or in pluri-specific combinations.
Understanding the selective pressure driving Trichogramma evolution (e.g. how strategies of host exploitation have evolved) and developing knowledge on phenotypes, hybridization potential of released strains with each others and local fauna or on the relative influence of competition and facilitation for parasitoid communities organization should allow us to i) better select appropriate strains for targeted biocontrol, ii) maximise the chances of safely introducing agents into a new habitat and iii) better anticipate their impact and modify control strategy if needed. Biological control often involves public-private partnerships. These collaborations are essential as (i) agents are produced and sold by private firms and (ii) identification and evaluation of biocontrol candidates require scientific skills. This is why we included a work package dedicated to the maintenance of characterized strains and knowledge transfer towards stake-holders (ecologists, agricultural practitioners, private companies) through a web-interfaced database. This WP will be organized around the recently established Biological Resource Center devoted to biological control using macro-organisms and hosted by ISA.