September 30, 2020

Overview of Nitrogen Use Efficiency and Winter Wheat Yield

1. Phenotyping and genotyping elite lines for yield, NUE, and other traits – Years 2 & 3

Breeders have long focused on stable grain yield in crop improvement and it is a vital target trait even with climate change. Direct selection for yield in stress environments improves stress tolerance. Selection techniques for yield have not changed much even with the advent of molecular tools. Nitrogen use efficiency (NUE) is a vital component to yield and end-use quality. Nitrogen fertilizer has become expensive and there are associated environmental concerns. It is important to get as much yield and grain protein per unit of applied N fertilizer as possible. Genetic variation for efficient uptake of soil N and utilization of plant N have been reported.
Objective: Develop more efficient yield and NUE improvement breeding technologies that use high throughput genotyping and phenotyping. This study will assess genetic diversity of yield and NUE in two sets of elite lines: one for the hard red winter wheat region and one for the northern soft red winter wheat region. We will assess each set for yield, grain protein, and advanced phenotyping based on spectrometers for each set in multiple regular and low N environments. All lines will be genotyped using new marker systems. We will quantify genetic variation for yield, yield stability, and NUE and develop whole genome models to use in rapidly improving these traits.
Outputs: Estimate of the genetic diversity and structure of elite populations, develop high-throughput phenotyping for NUE, identify genes for yield and NUE, develop selection models for rapid improvement of these traits.

States involved: NE, OK, KS, CO, OH, KY, VA, MO, MD

2 & 3. Validation of yield and NUE QTL – Years 4 & 5:

Breeders have been trying to identify genes associated with complex traits such as yield for many years with limited success. Most genes are detected in simple experimental populations and have little impact when placed in the complex breeding populations breeders must use to improve yield. This project uses complex populations to identify genes associated with yield, yield stability, and NUE (Objective 1). These genes must be validated.
Objective: Assess the impact of genes identified in objective 1 on yield and NUE in various breeding programs. Breeders in each will select 100 pairs of full sibs that differ for yield and 100 that differ for NUE from their ongoing breeding trials. These will be grown and phenotyped for yield and NUE in appropriate environments. The lines will be genotyped with markers for region of the genome associated with the genes identified in objective 1.
Outputs: Confirm value of the genes so breeders can use markers for the best genes in their yield and NUE breeding efforts.

States involved: Yield genes – NE,OK,KS, MO,OH,KY,MD,VA; NUE genes – NE, OH, MO

4. Validating Yield/Protein response to N – Years 4 & 5

Nitrogen use efficiency (NUE) is an increasing important trait due to rising environmental concerns and N fertilizer prices. Breeding for NUE has been prohibitively expensive as it involves assessing yield of many lines under different N levels. A new approach is emerging that involves a high-throughput phenotyping using canopy spectral reflectance (CSR). It is rapid and cheap so many lines can be assayed. This approach could greatly facilitate breeding for NUE. The CSR approach is experimental and we need to validate that CSR can be used to select lines with good NUE over a range of N application rates.
Objective: Assess the NUE of lines selected to have high and low NUE based on CSR. From each region we will select the six best and six worst NUE lines based on CSR from objective 1. These lines will be grown under five N levels in multiple environments. Yield, grain protein, and CSR data will be obtained.
Output: The results will allow breeders to determine if lines with superior NUE can be selected by using the inexpensive, high-throughput, CSR technology.

States involved: NE, OK, OH, VA

5 & 6. Assessing and implementing Genomic Selection: years 3, 4, 5

Breeders improve complex traits by intermating parents with superior phenotypes and then selecting superior progeny from the matings. This is inherently inefficient for complex traits where parents with the best phenotypes do not always have the best genes. In addition it can take five to ten years to go through one selection cycle: gain per year is very low. Selection based on actual genetic values instead of phenotypes would be vastly superior. Breeders have attempted to assign values to individual genes though this approach has mostly failed for complex traits where most genes have small effects that are difficult to estimate. Animal breeders have successfully used an alternative approach called genomic selection (GS). In GS, a genetic value is assigned to the whole genome of an individual instead of to individual genes. A model that predicts genetic value from molecular marker data is built using genetic data from the whole genome. The model can then be applied to progeny to predict which will be superior. Thus superior progeny can be selected based solely on marker data. Because GS does not use phenotypic data, a breeding cycle may take less than one year so that gain per year can be much greater than with traditional breeding. We need to assess the value of GS in improving yield and NUE.
Objective: Implement GS in yield and NUE breeding efforts. Phenotypic and genetic data from objective 1 will be used to develop GS models. The models will be validated using data from objective 1 and from other independent yield trials. The models will be used to select superior parents that will be intermated. F2 progeny will be developed, genotyped, and superior individual selected using the GS models. These will be intermated and the cycle repeated. We estimate we can accomplish three cycles of GS.
Outputs: Three cycles of GS should be completed producing populations that should have superior yield and NUE. These will be distributed to all breeders so they can derive lines for evaluation in their programs.

States involved: NE,KS,OK,OH,MO,MD,NY

7. Allele-Based Breeding: years 3, 4 & 5

Breeders improve yield and stress tolerance primarily through phenotypic selection of thousands of individuals using data from tens of thousands of plots. Selection is based on performance of an individual. Thus despite enormous testing resources crucial decisions on the value of an individual involves data from a few plots where the individual is grown. The era of inexpensive genotyping mandates that we re-think this strategy. It is now possible to focus on marker-based estimates of the genetic value of alleles versus phenotype-based estimates of the value of individuals. Marker data allows all phenotypic data from all individuals and all plots to be used to determine value of alleles and thus the value individuals with those alleles. In allele-based breeding, replication of alleles in different genetic backgrounds and environments is more important than replication of individuals. Each year breeders test hundreds of new lines in their programs. These existing resources could be exploited to develop allele-based breeding that improves the breeding efficiency of all programs.
Objective: Develop and test a new breeding strategy based on annually determining allele values across breeding programs. Each breeder in a region will place 192 lines that would normally be entering the first year of multi-environment trials in a pool. The lines will be dispersed to breeders with each breeder getting a total of 192 lines to test in their own multi-environment trial. Each line would be phenotyped for yield and CSR and genotyped with 1536 SNPs. In this scheme an allele is evaluated within and between programs, genetic backgrounds, and environments. This will be done for two consecutive years. We will evaluate the consistency of genetic models over programs and years and compared to values estimated from the elite panels evaluated in years 2 & 3.
Outputs: A new breeding strategy that annually evaluates the value of existing and introduced alleles in relevant context, is adaptable to introgression of alleles from the NSGC, uses existing breeding resources, and is adapted to long term gains.

States involved: NE, CO, KS, OK, MO, IL, IN, KY, VA, MD, OH, MI, NY

8. Utilizing high-throughput phenotyping: years 1, 2, 3, 4, 5

A major limitation to improving yield and stress tolerance in crops is obtaining accurate, meaningful phenotypes on thousands of breeding lines. Yield in stress environments is the ultimate measure of tolerance but is expensive to obtain accurately and thus can only be assessed on a limited number of lines. Many physiological parameters have been associated with yield and stress tolerance but are impractical in a breeding program. A promising technology that will be used in this grant is canopy spectral reflectance (CSR). Light hits a crop canopy and some is absorbed and some is reflected. The amount reflected varies by wavelength, genotype, and stress level. Thus assessing CSR can measure the physiological status of different genotypes rapidly and inexpensively. Relevant traits for the grant include biomass (associated with yield), water status (associated with drought tolerance) and nitrogen concentration (associated with NUE). We will use CSR in the basic research objectives of this grant. The more we use CSR the more we will learn about its role in crop improvement. So we want the breeders who receive spectrometers to measure CSR to use them in the regular breeding program to further assess their effectiveness in selection.
Objective: Assess the effectiveness of CSR in crop improvement in standard breeding trials. Seven breeders in this grant will receive spectrometers to measure CSR. Each can use them in selection in any component of their breeding program.
Outputs: We will obtain experience and a broader picture of the role of CSR in crop improvement than can be provided by using CSR in only basic research

9. Utilizing the NSGC: Years 1, 2, 3, 4, 5

Crop improvement requires genetic diversity and variation. Elite populations were primarily formed about 100 years ago from a limited number of ancestors and only small infusions of diversity have occurred since. In objective 1 we will inventory the genetic diversity of our elite pools while also assessing their variation for yield and NUE. Deficiencies will be found and will need to be addressed to improve our long-term gains for important traits. The core of National Small Grains Collection is also being evaluated for its genetic diversity and phenotypic value. That data can be used to identify new sources of genetic diversity in the NSGC that will complement needs of the elite breeding populations.
Objective: Introgress useful diversity from the NSGC into elite breeding populations. Based on phenotypes and genotypes, the most diverse and desirable accessions from the NSGC will be selected and crossed to elite germplasm. About 90 elite x exotic crosses will be made each year. Fifty (25 hard, 25 soft) crosses in the first year will be devoted to creating a nested association mapping population based on one common elite parent per region (Jagger for hard, Branson for soft).
Output: Populations derived from exotic x adapted parents will be developed and distributed. Some of these will be in structures to facilitate gene discovery. All populations will feed into the improved breeding methodologies developed elsewhere in the grant to fuel long term crop improvement.

States involved: NE, OK, KS, OH, MO, MD