NSF 2009

Utilizing natural variation in Intermediate wheatgrass to breed for sustainable yield

Key terms: marker-assisted selection, sustainable agriculture, biomass allocation, yield

Introduction: The largest ecosystem type in the world is agriculture, covering >30% of all terrestrial land.⁵ The majority of agro-ecosystems consist of annual monocultures, such as wheat and rice, which have gone through tight genetic bottlenecks and were not selected for their ability to condition soil to prevent erosion or hold water.  High-yielding annuals rely on large amounts of fertilizer, pesticide, and irrigation water inputs that are not sustainable.⁵ The Green Revolution’s high-yielding strategy remains largely effective on prime (rich, non-sloping) farmland, although even this soil continues to be degraded at 10 times the sustainable rate.⁷ Alternatively, perennials not only prevent soil erosion, but also build soil and maintain a protective organic mulch. The perennial intermediate wheatgrass, Thinopyrum intermedium, because it is in a breeding/domestication program at The Land Institute near Salina, Kansas and has the potential to transform global agriculture as a perennial sustainable crop.

The goal of this project is to better understand the physical traits underlying the holistic concept of sustainable yield, including the above and below ground traits important for yield potential and the fertility traits important for seed yield realization. I will investigate whole plant biomass allocation and multi-year grain production across advanced perennial intermediate wheatgrass breeding lines in two locations to determine developmental tradeoffs in above and below-ground growth with multi-year yield. With the advent of genome wide molecular markers, which can be generated for any species with second-generation sequencing technologies, I plan to identify major genetic loci controlling architecture and seed yield. Combining phenotypic characterization with genomic data will allow for direct selection on the genetic basis for yield traits through marker-assisted selection, linking the Genome Revolution to a new Green Revolution.

AIM 1: Quantify potential trade-offs between above-ground and below-ground traits: Below-ground phenomena influence above-ground traits through nutrient-use and water-use efficiencies, stemming from variation in root morphology, architecture, and associated symbiotic and rhizospheric microbial communities.   

Hypothesis: Genetic variation in root system architecture is negatively correlated with above ground tiller height. Root traits include the complexity of branching patterns and is measured as an integration of root length, diameter, and number and angle of branching. Preliminary data suggest that shorter stature genotypes generally possess proportionally fewer structural roots and more fine roots than taller stature genotypes. This observation suggests a tight association between biomass allocation to above- and below-ground structures. What is not known is the consequence under an accelerated breeding program and how these trait trade-offs are maintained through perennial development. 

Methods: I took three soil cores from 30 representative genotypes, spanning above-ground variation for height, number of tillers, yield per stem, fertility, and seed size of the breeding population at The Land Institute. These cores will be analyzed and preliminary data obtained late Fall 2009. I will characterize roots by scanning the washed root systems with a flatbed scanner and analyzing the digitized root systems with WinRHIZO software, which will give me information on traits such as root length, area, branching patterns, surface area, volume, and fractal dimension. I will also determine the amount of mycorrhizal fungi associated with the roots and soil using the marker phospholipid fatty acid (PLFA) 16:1w5c. I will follow the above- and below-ground traits for three years with field plots at Argonne National Laboratory and the Land Institute to validate any preliminary data and to test genotype performance across two sites. Analysis of variance will be used to separate the genotype by environment interaction, analyze within and between plot variations and identify those traits that are phenotypically stable.

AIM 2:  Develop markers for marker-assisted selection: Molecular markers identify sequence variation between individuals to determine relatedness and an underlying genetic map. The map can be used to divide quantitative traits into their individual genetic components, QTL (Quantitative Trait Loci). The use of markers in plant breeding is known as marker-assisted selection, and its implementation has dramatically accelerated the plant breeding process. Populations can be screened at the seedling stage, and plants with desirable gene combinations can be identified early and those with undesirable traits can be eliminated.³ Genome-wide association mapping uses historical recombination events among wild populations to detect population structure and fine map QTL when tens or hundreds of thousands of markers are typed in large populations.^4,6

Hypothesis: I am developing molecular markers and expect to identify large-effect QTL for important quality and domestication traits. Two major QTL for yield/fertility have been identified in perennial ryegrass and are predicted to be present throughout the grass family. Other major domestication QTL for free threshing and seed weight have been shown to follow the 'one gene for one trait' regime. Major variation in such traits has been identified and rare genotypes' frequencies increased in advanced lines of intermediate wheatgrass. 

Methods: I collected almost 800 replicated tissue samples from an advanced selection experiment of 4000 plants at The Land Institute in September 2009, consisting of 96x2 samples representing the high and low extremes of four categories: yield per stem, number of tillers, seed weight, and height. I have a gradient for each of the categories, as individuals at the extreme of one category are intermediate for other categories (e.g. a very tall plant may have an intermediate number of tillers). I will sequence restriction-site associated DNA (RAD) markers to identify 10-100k SNPs (single nucleotide polymorphisms) per individual, simultaneously genotyping the individuals. I will use genome-wide association mapping to determine if the quantitative traits of interest in intermediate wheatgrass are caused by many small-effect or large-effect QTLs.

Broader Impacts: The full cost of agricultural production is not currently considered; millions of dollars are poured into cleaning our waterways and coasts. Nitrogen loads have polluted rivers and caused 'dead zones' in once productive areas. Perennials will require much less fertilization and their root systems should mitigate nutrient leaching. Intermediate wheatgrass is ready to be planted on marginal lands and in pastures to increase soil and forage quality, and the first lines of a perennial wheat could be ready for commercial use within the next 10 years, but only with the incorporation of marker-assisted selection. Land Institute breeders have doubled seed size and increased yield approximately 60% during the past seven years with the use of traditional breeding methods (Lee DeHaan, personal contact).  We have the technology to better understand the genetic basis of the traits being selected, as well as the consequences of selection in an accelerated breeding program.  Intermediate wheatgrass is in its first stages of domestication and could be developed as a model of breeding for sustainable yield.  Changes in plant architecture, fertility, mycorrhizal fungi associations, and microbial communities can be tracked throughout crop development to better understand the physical traits affecting concept of sustainability.

Works cited: 1. VJ Allison et al. In Mycorrhizae 141-61 (2004); 2. IP Armstead et al. New Phyt 178, 3: 559-71 (2008); 3. BCY Collard et al. Ph Trans of Royal Soc Biol Sc 363, 1491: 557-72 (2008); 4. MW Ganal et al., Cur Op Plant Bio 12, 2: 211-17(2009); 5. J. Glover et al., Sci. American August: 82-89 (2007).  6. M  Nordborg et al. Nature 456:720-23 (2008); 7. D Pimental. Env, Dev, and Sustain, 8:119-37 (2006).