Population structure and association mapping of flower-related traits in lotus (Nelumbo Adans.) accessions

Abstract

Association mapping was used to identify candidate molecular markers associated with four flower-related traits-flower color, petal number per flower, flowering time, and flowering duration-in a collection of lotus (Nelumbo Adans.) accessions evaluated over two years. Significant phenotypic variations (P<0.001) were detected for petal number, flowering time, and flowering duration. Use of SSR, SRAP, and AFLP markers identified 423 polymorphic loci in the population. The population structure and principal coordinates analyses based on the three kinds of marker identified three clearly distinct subpopulations, which in general corresponded to Nelumbo nucifera, Nelumbo lutea, and their hybrids. Pairwise linkage disequilibrium (LD) among the 423 markers in the population varied from 0 to 1, with a mean of 0.034. The majority of pairwise LD (96%) were r2<0.2, and less than 1% of LD was r2>0.5, indicating a low level of LD between the markers. Using the general linear model considering population structure, association mapping identified several significant markers (P<0.01) that were correlated with the four flower traits across two years. The contributions of these markers to phenotypic variation ranged from 5% to 16%. Some significant associations were repeated for petal number, flowering time, and flowering duration over the two years, and three markers were associated with both flowering time and flowering duration. Of the three kinds markers used, AFLP was the highest efficient marker in association analysis. Our study suggested that lotus might be amenable to association mapping. Candidate marker loci associated with these traits highlighted regions to be targeted in future association studies, and provided useful information for dissecting genetic variation in traits of interest for lotus improvement.