Adaptation à la sécheresse et création variétale : le cas de l’arachide en zone sahélienne - Deuxième partie : une approche pluridisciplinaire pour la création variétale

Abstract

Drought is a multiform constraint whose impact on the vegetal metabolism is very variable according to its duration, intensity and phenological stage of the vegetal development where it occurs. Thus, the plant resistance is expressed at different plant organisation levels. The present study was aimed at integrating knowledge generated by experiments carried out in Senegal on groundnut within the framework of a breeding programme geared towards improving groundnut yield under drought conditions. Three studies involved in the breeding work, are presented. The first chapter analyses an incomplete half-diallel cross performed on an original population under recurrent selection for drought adaptation. The study confirmed the weak heritability of yields but concludes that the best predictor of pod-yield was the pod-yield itself. By contrast, the study of the genetic correlations showed that a selection for high haulm-yield could lead to poor pod-maturity under drought constraint. The selection indices were performed and used to estimate genetic gains relative to the main agronomic characters according to selection pressure. The second chapter covers the genetic variability of phenological, agronomic and physiological characters studied in two series of quasi-isogenic early lines. It has indicated that genetic variability was expressed in these lines despite its closeness. Some correlations between yield and physiological parameters, i.e. mainly fluorescence parameters, were significant but not stable across lines and environments showing that groundnut have different drought adaptation strategies according to genetic background and drought pattern. This work was pursued at the molecular level with three reference cvs involving the both recurrent parents of the precedent study. The gene transcript kinetics under drought, obtained using RT-PCR, showed that Phospolipase D and Cysteine protease gene expressions were stimulated by stress in the most susceptible cultivars, whereas their was higher LEA gene expression in the resistant one. These interconnected experiments conducted at different plant organisation levels led to the development of a general methodological model and of new improved genotypes to meet the social demand.