Genome-wide association study and drought tolerance evaluation of a winter wheat association mapping panel

Drought is one of the most important environmental challenges farmers face around the globe, with water stress the main cause for yield loss. Therefore, the objectives of the first part of this study were to 1) evaluate a hard winter wheat association-mapping panel (HWWAMP) in multiple environments differing for soil moisture for agronomic and drought tolerance related traits; 2) determine the relationship between yield and other agronomic and physiological traits; and 3) identify QTL involved in drought tolerance through association analysis. The HWWAMP consists of 299 entries (cultivars and breeding lines) adapted to the U.S. Great Plains region. The panel was characterized using a high-density 90 000 single nucleotide polymorphism (SNP) genotyping platform. Field evaluations for the HWWAMP were conducted in two sites in a side-by-side experiment under two soil moisture regimes in two years (2011-2012 at Greeley, CO and 2012-2013 at Fort Collins, CO). In addition, a replicated confirmation study was conducted in two sites (Greeley and Fort Collins in 2013-2014) to evaluate performance in field trials of a subset of 50 entries.;At Greeley 2011-2012, genotypes differed significantly for grain yield (GY) in both well-watered (WW) and water-stressed (WS) trials (P<0.001). Genotypes also differed significantly (P<0.001) for total biomass (TBM), biomass grain weight (BGW), harvest index (HI), plant height (Ht), relative water content (RWC), and carbon isotope discrimination (CID) in the WW trial and differed significantly (P<0.001) for BGW, HI, Ht, canopy temperature at late heading stage (Tclh), and RWC in the WS trial. At Fort Collins 2012-2013, genotypes differed significantly (P<0.001) for GY in both WW and WS trials. Moreover, genotypes differed significantly (P<0.001) for BGW, HI, Ht, RWC, and canopy temperature at booting stage (Tcbs) in the WW trial and for TBM, BGW, HI, Ht, RWC and Tcbs in the WS trial.;Markers with MAF <5 % and SNPs with ≥10 % of calls missing were removed to produce a set of 16,052 filtered SNPs. These 16,052 SNPs were used by the software package GAPIT in R to perform the genome-wide association study (GWAS). A kinship matrix was estimated in the rrBLUP package for R and was incorporated in the analysis. Several models were considered in the analysis with principal components and kinship (P+K) to control for kinship and structure. A total of 331 significant (P<0.001) marker-trait associations (MTA) was detected in one or more environment for 10 measured or calculated traits (GY, TBM, BGM, HI, Ht, Tcbs, Tclh, canopy temperature at grain filling stage (Tcgf), CID, and drought susceptibility index (DSI)). For the five main traits that were measured in all four environments, the highest number of MTA was recorded for HI (58) followed closely by GY (57), while the lowest number of MTA was recorded for BGW (18). The MTA for HI and GY were spread along 14 and 12 chromosomal regions, respectively in four environments. Amongst the three different Tc measurements analyzed for genome-wide association study, canopy temperature at late heading stage (Tclh) had the highest number of detected MTA (50). Carbon isotope discrimination was measured in the Greeley 2011-2012 WW trial, where the number of detected MTA was 29. Multi-trait chromosome regions were detected on chromosomes 4A and 4D associated with GY and CID, which may be useful in marker-assisted selection, following proper validation. In the confirmation study at Greeley 2013-2014, genotypes differed significantly for GY under WW conditions only (P<0.001) and for GY under both WW and WS conditions at Fort Collins 2013-2014 ( P<0.001).;The development of a deep and extensive root system is a drought adaptation mechanism to allow water and nutrient extraction from the soil profile. We conducted two studies to investigate the variation in root architecture and related physiological and morphological traits in winter wheat under drought stress. The first study evaluated 30 entries primarily from Colorado, and the second study included 30 entries from seven Great Plains states. Entries were evaluated in a greenhouse in 2012 and 2013 in 1 m x 10 cm plastic tubes filled with a fritted clay medium. Drought stress was imposed by withholding water after the emergence of the fourth leaf. After three weeks without watering, above ground biomass was harvested and roots were separated from the growing medium, washed, scanned, and digitally analyzed. Colorado entries differed significantly (P<0.05) for estimated transpiration, above ground biomass, average root diameter, total root length for bottom, middle, and top sections, and root length in most diameter classes. Great Plains entries differed significantly (P<0.05) for above ground biomass, stomatal conductance, water use efficiency, total root length, and root length for several diameter classes. Total root length adjusted for above ground plant size of Colorado entries ranged from 5212 to 7279 cm and average diameter ranged from 0.33 to 0.40 mm. Total root length correlated positively ( P<0.05) with leaf elongation rate and RWC for Colorado entries and total root length correlated negatively with average root diameter for entries of both studies. No significant differences were observed for any root trait between entries with and without Rht semi-dwarf alleles. The variation in root traits among Colorado and Great Plains winter wheat entries can be exploited in breeding programs to help develop plants with the best adapted root systems to withstand drought stress.;Because it is very important to test root performance of entries in the actual environment where they grow, a soil coring study was conducted at multiple environments to directly quantify variation in root traits. The objectives of this study were 1) to determine the variation in root architecture traits among US Great Plains winter wheat germplasm in the field, under water-stressed field conditions, 2) to examine correlations among the evaluated root traits and yield, canopy temperature, CID, plant height and harvest index, and 3) to examine correlations between evaluated root traits in the field and greenhouse. The study was conducted in three location-year environments, each with a different set of entries, all part of the HWWAMP. In all three environments, soil cores were collected from the WS treatment using a 1 m high, 5 cm diameter truck- or tractor-mounted hydraulic soil probe at three depths. Entries at Greeley 2011-2012 WS trial differed significantly for TL, TLTS, TLMS, AD, average diameter in the middle root section (ADMS), and average diameter in the bottom root section (ADBS) (P<0.001). Total root length correlated negatively with Tclh (r=-0.51, P<0.05). The 25 entries at Fort Collins 2012-2013 WS trial differed significantly for TL, TLTS, TLMS, TLBS, AD, and average diameter in the top root section ( P<0.01). There was significant (P<0.05) negative correlation between Tcbs and three root traits, TLTS, TLMS, and TL ( r=-0.42, -0.24, and -0.36, respectively). The 12 entries at Fort Collins 2013-2014 WS trial differed significantly for TL and TLMS ( P<0.01). There was a significant negative correlation between TL and Tcvg (r=-0.23, P<0.05) and a significant negative correlation between TLMS and canopy temperature at grain filling stage (r=-0.40, P<0.05). Root traits collected from this study should be regarded as a useful resource to gain insights about wheat adaptation to water stress in the field.;Our results demonstrate the relative importance of several physiological and morphological traits for drought tolerance evaluation in wheat. The most important traits, which showed significant association with grain yield, include root length, carbon isotope discrimination, and canopy temperature. The important chromosomes that comprised QTL for yield and drought tolerance traits in this study are chromosomes 4A, 4D, 2D, 3B, 3A, 2A, 2B, and 1D. Insights gained from this research will help aid our understanding of drought tolerance mechanism of winter wheat and help us define the morphological and physiological traits that define productivity under drought stress.