A toolkit to rapidly modify root systems through single plant selection

Background The incorporation of root traits into elite germplasm is typically a slow process. Thus, innovative approaches are required to accelerate research and pre-breeding programs targeting root traits to improve yield stability in different environments and soil types. Marker-assisted selection (MAS) can help to speed up the process by selecting key genes or quantitative trait loci (QTL) associated with root traits. However, this approach is limited due to the complex genetic control of root traits and the limited number of well-characterised large effect QTL. Coupling MAS with phenotyping could increase the reliability of selection. Here we present a useful framework to rapidly modify root traits in elite germplasm. In this wheat exemplar, a single plant selection (SPS) approach combined three main elements: phenotypic selection (in this case for seminal root angle); MAS using KASP markers (targeting a root biomass QTL); and speed breeding to accelerate each cycle. Results To develop a SPS approach that integrates non-destructive screening for seminal root angle and root biomass, two initial experiments were conducted. Firstly, we demonstrated that transplanting wheat seedlings from clear pots (for seminal root angle assessment) into sand pots (for root biomass assessment) did not impact the ability to differentiate genotypes with high and low root biomass. Secondly, we demonstrated that visual scores for root biomass were correlated with root dry weight (r = 0.72), indicating that single plants could be evaluated for root biomass in a non-destructive manner. To highlight the potential of the approach, we applied SPS in a backcrossing program which integrated MAS and speed breeding for the purpose of rapidly modifying the root system of elite bread wheat line Borlaug100. Bi-directional selection for root angle in segregating generations successfully shifted the mean root angle by 30° in the subsequent generation (P ≤ 0.05). Within 18 months, BC2F4:F5 introgression lines were developed that displayed a full range of root configurations, while retaining similar above-ground traits to the recurrent parent. Notably, the seminal root angle displayed by introgression lines varied more than 30° compared to the recurrent parent, resulting in lines with both narrow and wide root angles, and high and low root biomass phenotypes. Conclusion The SPS approach enables researchers and plant breeders to rapidly manipulate root traits of future crop varieties, which could help improve productivity in the face of increasing environmental fluctuations. The newly developed elite wheat lines with modified root traits provide valuable materials to study the value of different root systems to support yield in different environments and soil types.

A Toolkit to Rapidly Modify Root Systems Through Single Plant Selection

Background: The incorporation of root traits into elite germplasm is typically a slow process. Thus, innovative approaches are required to accelerate research and pre-breeding programs targeting root traits to improve yield stability in different environments and soil types. Marker-assisted selection (MAS) can help to speed up the process by selecting key genes or quantitative trait loci (QTL) associated with root traits. However, this approach is limited due to the complex genetic control of root traits and the limited number of well-characterised large effect QTL. Coupling MAS with phenotyping could increase the reliability of selection. Here we present a useful framework to rapidly modify root traits in elite germplasm. In this wheat exemplar, a single plant selection (SPS) approach combined three main elements: phenotypic selection (in this case for seminal root angle); MAS using KASP markers (targeting a root biomass QTL); and speed breeding to accelerate each cycle.Results: To develop a SPS approach that integrates non-destructive screening for seminal root angle and root biomass, two initial experiments were conducted. Firstly, we demonstrated that transplanting wheat seedlings from clear pots (for seminal root angle assessment) into sand pots (for root biomass assessment) did not impact the ability to differentiate genotypes with high and low root biomass. Secondly, we demonstrated that visual scores for root biomass were correlated with root dry weight (r = 0.73), indicating that single plants could be evaluated for root biomass in a non-destructive manner. To highlight the potential of the approach, we applied SPS in a backcrossing program which integrated MAS and speed breeding for the purpose of rapidly modifying the root system of elite bread wheat line Borlaug100. Bi-directional selection for root angle in segregating generations successfully shifted the mean root angle by 30o in the subsequent generation (P ≤ 0.05). Within 18 months, BC2F4:F5 introgression lines were developed that displayed a full range of root configurations, while retaining similar above-ground traits to the recurrent parent. Notably, the seminal root angle displayed by introgression lines varied more than 30° compared to the recurrent parent, resulting in lines with both narrow and wide root angles, and high and low root biomass phenotypes.Conclusion: The SPS approach enables researchers and plant breeders to rapidly manipulate root traits of future crop varieties, which could help improve productivity in the face of increasing environmental fluctuations. The newly developed elite wheat lines with modified root traits provide valuable materials to study the value of different root systems to support yield in different environments and soil types.

Variation in root system architecture among the founder parents of two 8-way MAGIC wheat populations for selection in breeding

Root system architecture (RSA) is a target for breeding because of the interest to develop crops with roots that use nutrients and water more effectively. Breeding for RSA requires phenotypic diversity in populations amenable to QTL identification to provide markers for large breeding programs. This study examined the variation for root traits across the parents of two multi-parent advanced generation inter-cross (MAGIC) wheat populations from NIAB and CSIRO for 16 days in an upgraded version of the non-invasive, germination paper-based phenotyping platform, GrowScreen-PaGe. Across all parents, total root length varied up to 1.90 fold, root biomass 2.25 fold and seminal root angle 1.16 fold. The CSIRO parents grew faster, exhibited slightly wider seminal root angle and produced larger root systems compared to NIAB parents. Lateral root lengths, leaf lengths and biomass contrasted most between fastest (Robigus - NIAB and AC Barrie - CSIRO) and slowest growing parents (Rialto - NIAB and G204 Xiaoyan54 - CSIRO). Lengths of lateral and total root, and leaf number and length had moderate to high heritability (0.30-0.67) and repeatability. Lengths of lateral roots and leaves are good targets for enhancing wheat crop establishment, a critical stage for crop productivity.

Chemical and physical influence of sodic soils on the coleoptile length and root growth angle of wheat genotypes

Background and Aims High exchangeable sodium percentage (ESP) and bulk density of sodic soils can reduce seedling emergence. This study examined variation in seedling coleoptile length and seminal root angle of wheat (Triticum aestivum. L) genotypes to determine whether these traits vary between genotypes that differ in their tolerance to sodic soils. Methods Wheat genotypes were grown in three different experiments. First, four wheat genotypes were grown using soils of three ESPs (4, 10 and 17 %) and secondly in soils of three different bulk densities (1.2, 1.4 and 1.5 g cm–3) and ESP 10 %. Thirdly, seedling coleoptile length and seminal root angle were determined for 16 genotypes grown in a soil of ESP 10 % and bulk density 1.2 g cm–2. Seminal root angle and coleoptile length measurements from the current study were compared with seedling emergence rate and force measured previously. Key Results The seedling coleoptile length of all genotypes decreased with increasing soil ESP and bulk density, but with no significant differences between genotypes. In contrast, seminal root angles differed significantly between genotypes, but were not significantly affected by ESP or bulk density. There was an inverse relationship between the seminal root angle of the 16 genotypes and seedling emergence rate (R2 = 0.89) and also between seminal root angle and seedling emergence force (R2 = 0.61). Conclusions Lack of significant variation in coleoptile length between genotypes suggests that this may not be a suitable characteristic to identify wheat tolerance to sodic conditions. However, a narrower seminal root angle was correlated with rate and force of seedling emergence, traits likely to improve establishment. The mechanism underlying this correlation is not yet clear. Genotypes with a narrow root angle had greater root depth. One possible mechanism might be that genotypes with narrow root angles were able to take up more soil moisture at depth, leading to a higher proportion of seedling emergence.

Assessing the importance of subsoil constraints to yield of wheat and its implications for yield improvement

Many of the soils in the Australian cereal belt have subsoils with chemical and physical properties that restrict root growth, which limits water use and yield. On alkaline sodic soils salinity, high pH, high available boron (B), deficiencies of zinc (Zn) and manganese (Mn) and high soil strength occur commonly and aluminium (Al) toxicity restricts root growth on acid soils. While the effects of individual subsoil constraints have been studied there is some debate about the relative importance to yield of the different soil stresses across the region. To address this issue yield variation among a set of 52 varieties of bread wheat was analysed using yield data from 233 trials conducted over 12 years. The trials were conducted in all mainland States but the majority were in South Australia and Western Australia. Each variety was characterised for its response to high B, high pH, Al toxicity, salinity, deficiencies in Zn and Mn and resistance to root lesion nematode (Pratylenchus neglectus), root growth through strong soil, seminal root angle, carbon isotope discrimination (CID) and maturity. This data was then used to examine the contribution of each trait to the genetic variation in yield at each of the 233 trials. The contribution of a specific trait to the genetic variation in yield at each site was used to infer the importance of a particular constraint to yield at that site. Of the traits linked to soil constraints, salinity tolerance, (measured by Na+ exclusion) was most often associated with genetic variation in grain yield (34% of all experiments), followed by tolerance to high Al (26%) and B tolerance (21%). Tolerance to low Zn and Mn were not consistently associated with yield variation. However, maturity was the trait that was most frequently associated with yield variation (51% of experiments), although the relative importance of early and late flowering varied among the States. Yield variation was largely associated with early flowering in Western Australia and the relative importance of late flowering increased as trials moved eastward into South Australia, Victoria and New South Wales. Narrow, rather than wide, seminal root angle was more commonly associated with high yield (25% of sites) and there was little evidence of any regional pattern in the importance of root angle. CID was important in 18% of trials with a low CID being most commonly associated with high yields. The yield advantage at sites where a trait contributed significantly to yield variation ranged from ~15% for Na+ exclusion and B tolerance to 4% for tolerance to high pH. The analysis has provided an assessment of the relative importance of a range of traits associated with adaptation to environments where subsoil constraints are likely to affect yield and has indicated patterns in the importance and effects of these traits that may be linked to regional variation in rainfall and soils.