Association analysis between agronomic traits and AFLP markers in a wide germplasm of proso millet (Panicum miliaceum L.) under normal and salinity stress conditions

Abstract

Background: Proso millet is a highly nutritious cereal considered an essential component of processed foods. It is also recognized with high water-use efficiency as well as short growing seasons. This research was primarily aimed at investigating the genetic diversity among genotypes based on evaluating those important traits proposed in previous researches under both normal and salinity- stress conditions. Use of Amplified fragment length polymorphism (AFLP) molecular markers as well as evaluating the association between markers and the investigated traits under both conditions was also another purpose of this research. Results: According to the phenotypic correlation coefficients, the seed yield had the highest correlation with the forage and biological yields under both conditions. By disintegrating those traits investigated under normal and salinity-stress conditions into principal component analysis, it was found that the first four principal components justified more than 59.94 and 62.48% of the whole variance, respectively. The dendrogram obtained by cluster analysis displayed three groups of genotypes under both normal and salinity- stress conditions. Then, association analyses were conducted on 143 proso millet genotypes and 15 agronomic traits as well as 514 polymorphic AFLP markers (out of 866 created bands) generated by 11 primer combinations (out of the initial 20 primer combinations) EcoRI/MseI. The results obtained by mixed linear model (MLM) indicated that under normal conditions, the M14/E10-45 and M14/E10-60 markers had strong associations with seed yield. A similar trend was also observed for M14/E10-45 and M14/E11-44 markers in relation to forage yield. On the other hand, M14/E10-14, M14/E10-64 markers (for seed yield) and M14/E10-64 marker (for forage yield), had significant and stable association in all environments under salinity-stress conditions. Moreover, a number of markers showed considerable associations and stability under both normal and salinity stress conditions. Conclusions: According to the analysis of phenotypic data, the wide germplasm of Iranian proso millet has significant variation in terms of measured traits. It can be concluded that markers showing strong associations with traits under salinity-stress conditions are suitable candidates to be used in future marker-assisted selection (MAS) studies to improve salinity-resistance genotypes of Panicum miliaceum in arid and semiarid areas.