Démogénétique environnementale du mildiou de la pomme de terre en Colombie

par Camilo Patarroyo velasquez

Projet de thèse en Biologie

Sous la direction de Stéphane Dupas et de Silvia Restrepo.

Thèses en préparation à université Paris-Saclay en cotutelle avec l'Université des Andes , dans le cadre de École doctorale Sciences du Végétal : du gène à l'écosystème , en partenariat avec Évolution, génomes, comportement et écologie (Gif-sur-Yvette, Essonne ; 2015-....) (laboratoire) et de Faculté des sciences d'Orsay (référent) depuis le 01-09-2019 .


  • Résumé

    Le mildiou de la pomme de terre, en dépit d'être l'une des maladies des plantes les plus étudiées au monde, reste l'une des plus grandes menaces pour la sécurité alimentaire mondiale. Le mildiou de la pomme de terre est la principale maladie affectant les plantations de pommes de terre en Colombie. Cette maladie est causée par l'Oomycete Phytophthora infestans, qui a fait l'objet de nombreuses études. La plupart de ces études ont été réalisées sur des aspects spécifiques de la biologie de ce pathogène, tels que la pathogenèse, la résistance aux fongicides et la reproduction. Cependant, malgré l'étude approfondie de ce pathogène, peu de mesures ont été prises pour combiner les diverses sources d'informations afin de comprendre la dynamique globale de P. infestans.

  • Titre traduit

    Environmental demogenetis of the potato late blight in Colombia


  • Résumé

    Potato late blight, despite being one of the most well studied plant diseases in the world, remains one of the biggest threats to global food security. The potato late blight is the main disease affecting potato plantations in Colombia. This disease is caused by the Oomycete Phytophthora infestans, which has been extensively studied. Most of these studies have been done on specific aspects of this pathogen's biology such as pathogenesis, resistance to fungicides and reproduction. However, despite the extensive study of this pathogen little has been done to combine the diverse sources of information in order to understand the overall dynamics of P. infestans. To address this issue, we plan to use an integrative approach (in this case environmental demogenetics) to be able to integrate the genetic, demographic and epidemiological information collected the by our partner laboratory in Colombia (LAMFU) over the last decade. For this purpose, we will first develop a database to reunite the population genetics and epidemiological information that has been collected on P. infestans in Colombia. In this database we will also include the environmental information made available by local climatic stations and the national weather service (IDEAM). Ideally, the results obtained from future research efforts by our collaborators in Colombia will feed into this database to keep all the information organized and available. In terms of population genetics, the previous work done in Colombia aimed mostly to identify the predominant clonal lines and variants that are found in potato farms, its prevalence among the different potato variants and its resistance to fungicides using microsatellite sequencing. However, little has been done in terms of the integration of the information collected on population genetics and its spatial aspect or the genetic history of the species in Colombia. To address this issue we will use the collected information to identify the main dispersion mechanisms of P. infestans and try to model its spatial dynamics. On the epidemiological side, there have been several studies documenting the dependence and response of this pathogen to fungicide application dynamics and other environmental conditions such as temperature, relative humidity and solar irradiation to name a few. These results were used to develop deterministic epidemiological models that calculate the population growth of P. infestans during a single epidemic for the United States. The development of such models can be divided in two stages, the development of the initial model using the a priori information from previous studies and then the fine-tuning stage using field data. Once adjusted these models have been successfully used to schedule fungicide applications only during periods of elevated risk for late blight appearance reducing the overall use of fungicides in potato farms in the United States. This approach however has a couple major drawbacks. First, because this type of deterministic model uses a large number of parameters its applicability to other localities with different environmental conditions is rather limited. Second, these models do not consider the spatial aspect of the epidemiology. Using the results from the first part and the previous studies on the response of P. infestans we will attempt to develop a generic model for the distribution of this pathogen in Colombia. This model will then be adjusted through Bayesian learning using additional field information that we will collect. During the collection of field data we will also learn the usual cultural practices used by the growers to deal with this kinds of epidemics and include them in the previously mentioned model. Through this approach we expect to overcome the need for experts or large amount of historic information during the adjustment phase of the model development hence making this pipeline applicable for other plant diseases that have not been studied as extensively as the potato late blight. This will also result in an accumulative model, which will be continuously refined as more field information is collected by the partner laboratory in Colombia (LAMFU). And finally, we expect to initiate the creation of a community of the potato growers in Colombia around the sharing of relevant information, as this will provide them with more accurate information about the probability of appearance of the potato late blight in their particular localities.