VARIATION IN GENERAL RESISTANCE TO THE FUNGAL PATHOGEN PYRENOPHORA SEMENIPERDA IS ECOLOGICALLY SIGNIFICANT FOR BROMUS TECTORUM SEED POPULATIONS
Allen, P.S.1, Meyer, S.E.2, Beckstead, J.3 and Clement, S.2
1Department of Plant and Wildlife Sciences, Brigham Young University, Provo, Utah, USA
2US Forest Service Rocky Mountain Research Station Shrub Sciences Laboratory, Provo, Utah, USA
3Department of Biology, Gonzaga University, Spokane, Washington, USA
Contact: Phil S. Allen, firstname.lastname@example.org
Bromus tectorum is a selfing winter annual grass that is highly invasive in the western United States. The generalist seed bank pathogen Pyrenophora semeniperda can kill B. tectorum seeds in the seed bank, with highest in situ mortality typically observed with dormant seeds and where autumn precipitation is unreliable. Inoculum augmentation in contrasting habitats resulted in near-complete mortality at sites where P. semeniperda was less abundant, but higher seed survival where the pathogen naturally occurred at high levels. This suggested selection on genetic variation in general resistance to pathogen attack. We tested two hypotheses. First, host genotypes from sites where high seed mortality coincides with occasional stand failure (i.e., necessitating stand recovery from the seed bank) will be more resistant to the fungus. Second, seeds in the carryover seed bank are more likely to belong to resistant lines than seeds produced on current-year plants from the transient seed bank. Bromus tectorum SNP haplotypes from contrasting habitats were subjected to laboratory experiments in which dormant seeds were inoculated with low levels of the fungus. General resistance was greatest in haplotypes from salt desert sites where the pathogen is abundant and lowest in haplotypes from montane or warm desert sites with lower levels of disease. Seeds from current-year plants and from the carry-over seed bank were used to produce 100 lines of each group from two cold desert sites, one with high disease levels and a history of stand failure and one with lower levels of seed bank diseases. These 400 lines were SNP-genotyped to identify host lineages overrepresented in either the current stand or the seed bank. Seeds of eight lines of each lineage were included in laboratory inoculation tests with dormant seeds inoculated at low levels. For the site with low field disease levels, lineages from current-year and seedbank seed populations did not differ in pathogen resistance. However, seedbank lineages from the die-off-prone site were significantly more resistant to the pathogen than lineages from current-year seeds at this site. These results confirm both hypotheses, and indicate that selection on genetic variation in general resistance has significant ecological consequences in this pathosystem.