Thursday, August 6, 2009

PS 66-107: The effect of mowing disturbance on a wild plant-pathogen interaction in a successional, post-agricultural community

Janet A. Morrison, Jacquelyn E. Harth, Lauren J. Frazee, Robert S. Doherty, and Artur Romanchuk. The College of New Jersey

Background/Question/Methods

Plant-pathogen research in non-agricultural species focuses primarily on intact natural communities, yet many host plants live in anthropogenic landscapes. The perennial grass Andropogon virginicus (broomsedge) persists in post-agricultural old-field communities with regular disturbance (e.g. mowing); otherwise it decreases during succession. Of populations from Pennsylvania to South Carolina, 50% were infected by the smut fungus Sporisorium ellisii, which decreases reproduction and increases mortality. In a New Jersey population, disease rates declined from 2003 to 2006 as the community underwent succession without major disturbance. We hypothesize that regularly mown plants are more likely to be diseased, due to spreading of fungal teliospores, reduction of woody competitors, increased host density, and production of wound sites for pathogen entry. We tested this idea by mowing three of five 20x100 m strips in our study site in 2007, and then comparing changes in density, disease, and host size from 2006 to 2008 in the mown and unmown strips. Also, aboveground disturbance may select for a mechanism of pathogen persistence in basal host tissues. We developed DNA primers and a PCR technique to detect S. ellisii DNA within host plants, and tested for its presence in various broomsedge tissues. 

Results/Conclusions

Prior to mowing, in 2006, host density in 480 1m2 plots was not significantly different in the strips assigned to mowing vs. not mowing (3.51 vs. 3.11 plants/m2). After the 2007 mowing, 2008 average density decreased to 1.45 in unmown and 1.41 in mown plots. Average disease prevalence per plot in 2006 was 0.07 in unmown plots and 0.05 in plots slated for mowing; in 2008 it was 0.10 and 0.07, respectively. The decrease in density was not significantly different among treatments, but was significant among strips within treatments (P<0.0001). The increase in prevalence was not significant among or within treatments. Mown plants were 76% as tall as unmown plants (P<0.0001), but had similar numbers of tillers and severity of infection. PCR detection of S. ellisii within plants from the field identified smut DNA only in the root crown of overwintering broomsedge. Broomsedge commonly  inhabits frequently mown landscapes, so a pathogen strategy of localizing to the root crown fosters pathogen fitness. However, a single mowing disturbance was not important for plant or pathogen dynamics, even though it negatively affected plant height. Other factors acting at the scale of approximately 20 m may be more influential.