Host range of Pyrenophora semeniperda, a native seed pathogen with potential for biocontrol of Bromus tectorum, in the Intermountain West

Background/Question/Methods The North American fungal seed pathogen Pyrenophora semeniperda has a long evolutionary history with its native grass hosts, but represents a new natural enemy for the invasive annual grass Bromus tectorum (cheatgrass). As part of studies to evaluate this pathogen as a potential biocontrol organism for cheatgrass, we examined its impact on seeds of twenty potential host species, including native and introduced perennial grasses, exotic annual grass weeds, and native perennial dicots. We challenged seeds of each species with conidial inoculum of two pathogen isolates, incubated the inoculated seeds under optimal conditions for the pathogen (20C) for six weeks, and compared germination, infection (as indicated by the presence of fungal stromata), and seed mortality with uninoculated controls. Results/Conclusions Three native dicot species showed both low susceptibility to infection (mean 26%) and low seed mortality (mean 12%), while three exotic annual grass species including cheatgrass were highly susceptible (mean infection 99%) but showed very low mortality (mean <1%), probably because their very fast germination rates (>90% in 2 days) enabled them to escape the pathogen. Native species of the grass genera Achnatherum, Stipa, Sporobolus, and Poa showed intermediate susceptibility (mean 57%), but a majority of infected seeds were killed (mean 38%), probably because their seeds were dormant or slow to germinate. Seeds of species of Agropyron, Pseudoroegneria, Thinopyron, Elytrigia, Pascopyrum, Koeleria, Festuca and Bromus were generally highly susceptible whether native or introduced (mean 97%). Mortality again depended on germination rate. Species whose seeds required 7 days or more to germinate to 50% showed high mortality, while those that reached 50% in 4 days or less were much less likely to be killed. This pathogen potentially has a very wide host range, but species differ substantially in susceptibility to infection. For species that are highly susceptible, a major determinant of the outcome of infection is seed germination rate. Seeds infected but not killed appeared to germinate normally, but further studies are needed to determine whether infection has fitness consequences at the seedling stage. Given the sometimes high mortality on species commonly used in the restoration of cheatgrass-dominated ecosystems, measures must be taken to mitigate the effects on nontarget species if P. semeniperda is to be used for biocontrol. Its effectiveness as a biocontrol depends on the presence of dormant cheatgrass seeds, which, unlike the nondormant seeds of annual grasses used in this experiment, can suffer high mortality under both laboratory and field conditions.