Interest of a Multiparental and Outcrossing Wheat Population for Fine Mapping
Abstract The use of multiparental populations for QTL discovery has been recently highlighted by different theoretical and experimental developments. Here, we explored the interest of French populations using heterogeneous genetic stocks of cultivated wheat, maintained in situ over 12 sites since 1984 with an outcrossing mating system. We studied one of these populations (Le Moulon, 48.4°N, 21°E), derived from 12 cycles of random crosses between 60 founders, selected to maximize genetic diversity. Outcrossing was allowed by the integration of a nuclear male sterility allele (ms1b, Probus donor) in the population. We analyzed 1,000 Single Seed Descent lines (SSD) derived from the 12th generation of cultivation. This population was genotyped using the 9 K i-select SNPs (Single Nucleotide Polymorphisms) array, covering the whole genome. Polymorphism and quality checks resulted in the selection of around 6,500 SNPs. First, the evolution of genetic diversity was explored through the comparison of SSD lines and the inferred initial population. The low population structure and the strong decay in linkage disequilibrium between SSD lines and the inferred initial population confirmed the efficiency of the 12 cycles of the random outcrossing in producing a highly diverse and recombined population. Two years of observations of population earliness under different environments were used to show the complementarity of association genetics, which allowed the detection of already known Vrn major genes, and evolutionary approach, which, lead to the discovery of two new minor effect QTLs.
Keywords Dynamic management • Evolution approach • Recombinant population • Wheat
Dynamic management (DM) aims at maintaining crop genetic diversity through in situ conservation of genetic resources. Genetically diverse populations are grown year after year, in various sites, differing for climate conditions, pathogen pressures and/or agricultural practices (Allard 1988; Henry et al. 1991; Porcher et al. 2004). In France, dynamic management has been experimented on bread wheat (Triticum aestivum L.) since 1984 (Henry et al. 1991), using three gene pools: two selfing populations (based on a pyramidal cross of 16 parents) and one outcrossing population. Samples of each of the three initial populations were sent to 7–12 sites in France and cultivated year after year in the same sites under the same conditions. Thus, these three “meta-populations” evolved over 10 to more than 20 generations without migration or conscious human selection (Enjalbert et al. 2011).
Studies on the selfing populations showed a good maintenance of global diversity at the network level (Raquin et al. 2008), both at phenotypic and molecular levels. A fast evolution of flowering time was observed, both over time and space (Rhoné et al. 2008): all populations flowered later than the initial population; and populations from Northern French sites flowered much later than Southern ones. Association genetics and spatio-temporal shifts in allelic frequencies revealed polymorphisms located in major genes controlling vernalization requirement or photoperiod sensitivity, partially explained climatic adaptation (Rhoné et al. 2008). Therefore, in addition to genetic resource preservation, DM populations can be an appropriate material to detect genes involved in local adaptation (Goldringer et al. 2001).
In the present study, we analyzed one outcrossing DM population, which is characterized by a high number of parents (60 lines), and numerous panmictic generations. In this population, wheat natural selfing habit was turned to an outcrossing mating system, using a recessive male sterility gene (ms1b), and harvesting solely open pollinated male-sterile plants. Taking the parental lines as reference, we studied an evolved population (12th generation) and tested for possible markers selection, trying to link detected markers to the observed evolution in vernalization requirement.