Identification of fusarium resistance within UK oat breeding lines (PhD)

The challenge

Fungal Fusarium species can infect cereals pre- or post-harvest, resulting in the contamination of harvested grains with mycotoxins. HT2+T2 mycotoxins have been identified at high levels in UK oat grains at harvest as a result of F langsethiae infection. The EU has introduced indicative levels of HT2 and T2 in cereals and cereal products for human consumption. Differences in the susceptibility of oats to Fusarium infection are genetic rather than cultural (e.g. sowing date) or morphological (e.g. height). As a result, the most achievable and economically viable long-term strategy is the breeding of oat genotypes with good resistance to Fusarium langsethiae. Studies are, however, hampered by the inability to artificially inoculate oats. This project aimed to develop an inoculation method for the infection of oats with F langsethiae and identify genetic markers for resistance/susceptibility to the pathogen. The outputs from this project could facilitate efforts of plant breeders in their efforts by using genetic marker assisted selection of new oat varieties with enhanced and stable expression of these traits.

Project abstract

HT2+T2 mycotoxins are produced by Fusarium fungal species. The EU is drafting legislation for maximum limits of these mycotoxins in cereals and cereal products intended for human consumption. Until the discovery of high concentrations of HT2+T2 in UK oats, oats were considered largely resistant to fusarium infection. Fusarium langsethiae is the main producer of both mycotoxins in UK oats. The infection of oats by F. langsethiae is symptomless and its epidemiology is unknown. Opoku et al. (2013) suggested a life cycle for the fungus. Increased growth of the fungus on emerged plant heads and the pathogen’s DNA being almost undetectable prior to anthesis are crucial aspects.

To effectively assess control measures, reliable artificial infection of oats with F. langsethiae is desirable. Successful artificial inoculation was achieved under glass by using spore suspensions and bagging plants (to increase humidity) at various growth stages. Inoculation made after the emergence of the panicle but before anthesis resulted in higher infection levels. Although it is possible reliably infect oats under glass, there is no evidence that this provides a reliable mimic of natural infection in the field. Furthermore, the ranking of the genotypes, in terms of resistance, did not match that of the naturally infected field-grown plots. In-field artificial inoculation using misting systems (to induce high humidity) failed to achieve higher infection levels than unmisted/uninoculated plots.

Cultivars have varying resistances to HT2+T2 accumulation, with ranking relatively consistent across years. This work further clarified the resistance imparted on oats by the parental origin of quantitative trait loci (QTL). QTL, designated Chr6D, Chr4A, Chr4D and Chr4C, have been identified previously as being associated with F. langsethiae DNA and HT2 +T2 concentration in harvested grains. Near isogenic lines (NILs), developed from a mapping population derived from crossing Tardis (a taller, earlier cultivar) and Buffalo (a semi-dwarf, later cultivar), were used to examine these QTL. Buffalo is the more susceptible of the two cultivars to F. langsethiae.

Introgression of the Buffalo-derived Chr6D into the Tardis background resulted in a shorter plant with panicles only partially emerged from the flag leaf boot. The opposite introgression led to plants taller than either parent line. Introgression of the Tardis Chr4D into the Buffalo background resulted in a later plant when sown in autumn, the effect was close to tenfold when sown in spring. The introgression of the Buffalo Chr4D into the Tardis background caused the resultant plant to be earlier in autumn-sown plots and four times as much so in spring-sown plots. Through comparison of the NIL with original parent lines, reductions in HT2+T2 concentrations were seen when Tardis Chr6D and Chr4D alleles were introgressed into the Buffalo background genome. The impact of Chr6D was consistent across all experiments, while the impact of Chr4D was dependant on sowing season. Chr4D had a weaker effect compared to Chr6D, but introgression of the Buffalo alleles into the Tardis background resulted in a reduction of HT2+T2 in autumn-sown plots. Introgression of Tardis-derived Chr4A into Buffalo had no impact on the HT2+T2 concentration, and introgression of Buffalo-derived Chr4A into Tardis had inconsistent effects across years. The ranking of the NIL population after artificial inoculation under glass did not match that of the naturally infected field grown plots.

Plant height and panicle extrusion were correlated to one another, and evidence is presented that either or both could influence plant susceptibility to F. langsethiae infection. Dissection of naturally infected panicles and quantification of F. langsethiae DNA concentration at the spikelet level demonstrated the independent nature of the infection in each spikelet. This finding reinforces previous work that oats have high type II resistance to fusarium infection. Window-pane analysis of summarised environmental variables, which utilised the NIL field experiments over four years, demonstrated that warm dry conditions post panicle emergence are conducive to higher HT2+T2 concentrations in harvested oats. Grid sampling of the experimental field (at large and small scale) did not identify any consistent patterns in infection but showed high spatial heterogeneity.