Identifying novel variation in grain yield, biomass and n-use efficiency and its physiological and genetic basis in wheat

Wheat (Triticum aestivum L) is the most widely grown crop in the world and accounts for approximately 31% of global cereal consumption every year (FAO, 2012). Furthermore, wheat is the most traded product worldwide (Dixon et al. 2009). In UK, wheat is the largest arable crop, it is grown on about 1.8 M ha p.a. For wheat productivity and grain quality, nitrogen (N) is the most important fertilizer applied, and modern wheat culitvars are generally selected under high N supply. However, N fertilizer represents a significant cost of production and may have negative environmental impacts. N-use efficiency (grain DM yield / soil and fertilizer N available; NUE) can be divided into two main components: N-uptake efficiency (N uptake / soil and fertilizer N available; NUPE) and N-utilization. efficiency (grain DM yield / NUP; NUTE). The current study aims to identify novel variation for grain yield, above-ground dry matter (AGDM). NUE and NUE components in a panel of AE Watkins landraces compared to UK check cultivars in the glasshouse and field trials and in two Single Seed Descent (SSD) populations derived from the crosses between UK spring wheat Paragon and Watkins landrace line 481 (PxW481) and between Paragon and Watkins landrace line 141 (PxW141) in field experiments and to investigate the physiological and genetic bases of this variation. Two glasshouse experiments (2011/12 and 2012/13) examining 17 genotypes (comprising five UK check cultivars and 12 Watkins landraces), and one field experiment (2011/12) examining a panel of 10 genotypes (comprising three UK cultivars and seven of the 12 Watkins landraces examined in the field) were carried out. In addition, three field experiments examining two SSD populations PxW481 (79 lines) (in 2011/12 and 2012/13) and PxW141 (97 lines) (in 2012/13) under two N treatments (HN=200 kg N/ha and LN=0 kg N/ha) were carried out. The glasshouse experiments were located at the University of Nottingham, School of Biosciences, Sutton Bonington Campus, UK and field experiments were located at the University of Nottingham experimental station at Bunny Park, Nottinghamshire, UK. The genetic maps were presently constructed using 201 and 491 markers, including single nucleotide polymorphism (SNPs) and microsatellites (SSR) for PxW481 and PxW141, respectively. Grain yield (GY), yield components, AGDM, NUPE, NUTE, and physiological traits including leaf light-saturated photosynthesis rate (Amax), flag leaf chlorophyll content (SPAD) and the spectral reflectance index, Normalized Difference Vegetation Index (NDVI, indicative of canopy size, field experiments only) were measured in the experiments. Genetic variation in grain yield was more closely associated with AGDM than with harvest index (HI) in the glasshouse and in the field under both HN and LN. For biomass production, the results showed that Paragon, Rialto, Xi19 and landrace W141 had higher AGDM than all other genotypes. In summary, UK check cultivars generally had higher grain yield (GY) compared to Watkins landraces and higher AGDM in the glasshouse under HN conditions. Similar results were found in the field under HN; however, under LN, the majority of Watkins lines maintained GY and AGDM than the UK cultivars compared to HN conditions. With regard to N uptake at harvest, the Watkins landraces W040, W141 and W579 had similar NUP to Paragon and higher than Oakley in the glasshouse. In the field, in general, Watkins lines maintained relatively better NUPE under LN than the UK check cultivars.. A negative association amongst genotypes was found between NUTE and grain N concentration (%) in the glasshouse and in the field under HN and LN treatments, with the strongest association observed under LN conditions. Variation for light-saturated leaf photosynthesis rate (Amax) was also observed amongst Watkins lines. In the glasshouse experiments, the UK cultivars, and landraces W141 and W352, generally had higher pre-anthesis Amax than other Watkins landrace lines. In the field experiment, landrace lines W352 and W141 had higher pre-anthesis Amax than Paragon under LN, whereas lines, W223 and W034 had higher values under HN. Genetic variation in GY and AGDM were overall better associated with post-athesis flag-leaf Amax than with pre-anthesis flag-leaf Amax in the field, and the associations with post-anthesis Amax tended to be stronger under HN rather than under LN conditions. With regard to the two SSD mapping populations, both N and genotype had a significant effect on GY, AGDM, NUPE and NUTE for PxW481 in all three experiments. For the PxW481 population, no SSD line had either higher GY or AGDM than Paragon under HN, while under LN, 13 lines had higher GY and two lines had higher AGDM than Paragon (P< 0.05). For the PxW141 population, three lines had higher grain yield under HN and three lines had higher GY under LN than Paragon. Five and 10 SSD lines under HN and LN, respectively, had higher AGDM than Paragon. The transgressive segregation for GY was due to either AGDM or HI, depending on the line. In both populations under both N treatments, genetic variation in NUE was more strongly associated with NUPE than with NUTE under HN and LN; however, the response of NUE to N availability was better associated with NUTE. Averaging years for the PxW481 population under HN, no SSD line had higher NUPE than Paragon, while under LN L481-76 had higher NUPE than Paragon. For PxW141, no SSD line had higher NUPE than Paragon either under HN or under LN conditions. Seventeen SSD lines of the PxW481 population had significantly higher NUTE than Paragon under HN and 10 lines higher under LN conditions. For PxW141 under HN, L141-95 and L141-94 had higher NUTE than Paragon under HN, while under LN 12 SSD lines had higher values. In both the PxW481 and PxW141 populations, genotype and N treatment had an effect on NDVI (canopy size) for all measurements pre and post-anthesis. Genetic variation in yield, biomass and NUPE were more closely associated with NDVI measured at around anthesis than at other measurement dates. In addition, the correlation amongst genotypes was stronger under LN than under HN conditions. Genetic variation for flag-leaf senescence parameters was also found in both mapping populations, and association with grain yield amongst genotypes was stronger under LN conditions than under HN conditions in both populations. QTL analysis identified two major QTLs for grain yield on 6AL and 5A under HN and one major QTL on 2B under LN conditions in the PxW481 population. The QTL on 6AL co-located with a QTL for NUPE, while the one on 5A co-located with a putative QTL for AGDM. Simonds et al. (2014) reported a major stable QTL in the same region of 6AL in the Spark x Rialto DH population across several environments. Under LN conditions, the QTL on 2B co-located with a QTL for plant height. For AGDM, a major QTL under both HN and LN was identified on chr 4B in the PxW481 population. This QTL co-located with a QTL for NUPE under both HN and LN conditions. In addition, a major QTL under HN for NUPE on 3BL (PxW481) co-located with QTLs for GY and AGDM. Another major QTL on 2A (PxW141), co-located with flag-leaf stay-green traits. This is consistent with the findings of Bogard et al. (2011) who reported two QTLs for NUPE under LN on 2A, associated with a delay in onset of flag-leaf senescence. With regard to NUTE, two major QTLs on 3B, co-locating with QTL for GY and HI, and on 1B, co-locating with a QTL for N grain concentration, were found for the PxW481 population under LN. In addition, three QTLs on 3A, co-locating with a QTL for TGW, and on 6D and on 7B, co-locating with QTLs for TGW, were identified