100 years of plant breeding - what have we done to the seed microbiome?

Humans have domesticated plants for millennia, with cereals first domesticated in the Fertile Crescent of the Middle East around 9000 BCE. The plant has traditionally been considered as a single entity, with selection based on phenotype resulting in modification of the plant genome, and more latterly, understanding of the genome being used to accelerate selection of key phenotypes. However, with the advent of next generation sequencing technology, there has been greater appreciation of the role of the plant microbiome and its metagenome in plant performance. Seed biobanks are critically important for preserving plant diversity which may be lost in the wild due to habitat loss and the changing environment, but we know nothing about the effect of long-term storage on the microbiome. We will test the hypothesis that domestication, breeding and storage alter the composition of the plant microbiome, with impacts on plant performance in terms of yield, resilience and quality. The seed microbiome is the ideal target for this study as this represents the community recruited by the plant and potentially adapted to endophytic existence. We will analyse the bacterial microbiomes of 3 crops with contrasting histories and agricultural uses: 1) Avena sativa (oats): domesticated as a grain crop for human food during the first millennium BC. 2) Lolium perenne: a forage grass for livestock (meat and milk) production that has been bred for ~100 years. 3) Miscanthus: a tall grass recently selected from the wild as a feedstock for bioenergy and industrial products, as well as nutraceuticals (e.g. prebiotics and sweeteners) and represents a novel domestication in progress. The seed biobank at AU holds collections, including wild relatives and breeding lines, of all three species, going back approximately 40 years, 100 years, and 15 years respectively. We will use this resource to: 1. A. Measure seed germination and plant fitness of the three species of different ages in the collection (Y1) and investigate their relationship to both the bacterial composition present (via rapid and cost-effective 16S rDNA profiling) and the climatic data provided for the seed at collection (Y1/2). B. Compare the microbiome of oats and Lolium through the generations of breeding lines (Y1), and correlate the populations present with the extensive trait data available for these lines, including grain and forage quality respectively (Y1/2). C. For Miscanthus, i) compare microbiome composition of wild collected seed with origin of collection and ii) compare these with recently generated hybrids to determine the extent to which the populations are conserved across generations (Y1/2). 2. Subject a recent batch of seed to various conditions to simulate long-term storage (Y1/2), and determine the effect of these treatments on seed germination, microbiome composition, and on subsequent plant performance under optimal conditions and abiotic stresses (Y2/3/4). 3. Isolate bacterial endophytes from oats, Lolium and Miscanthus of different ages and origins (Y1) and conduct plant growth promotion experiments, under optimal conditions and abiotic stresses, to identify beneficial strains (Y2/3/4). We will compare the microbiomes of a historically domesticated grain crop, a more-recently bred forage crop, and a wild species undergoing selection for the first time. We will correlate climatic and other environmental data from the collection site with the seed microbiome, and trace this through generations. Ultimately we will determine whether selection and breeding have inadvertently resulted in reduced diversity within the plant microbiome with reductions to the potential plant benefits, or whether they have resulted in an optimised metagenome. This will provide novel understanding about the impact of the microbiome on plant fitness, with implications for future seed storage and breeding strategies, and for improving sustainability and resilience of UK agriculture to future climates.