96th ESA Annual Meeting (August 7 -- 12, 2011)

COS 43-4 - Cheatgrass effects on native grass seed and seedling fate: Competition, facilitation, and indirect effects of a shared seed bank pathogen

Tuesday, August 9, 2011: 2:30 PM
19A, Austin Convention Center
Julie Beckstead1, Katherine T. Merrill2, Susan E. Meyer3 and Phil S. Allen2, (1)Biology Department, Gonzaga University, Spokane, WA, (2)Department of Plant and Wildlife Science, Brigham Young University, Provo, UT, (3)Shrub Sciences Laboratory, USDA Forest Service, Rocky Mountain Research Station, Provo, UT
Background/Question/Methods

Habitat expansion of the invasive annual grass Bromus tectorum (cheatgrass) has altered the microenvironment of the seed zone, increased competition, and increased the inoculum load of the naturally occurring fungal seed pathogen Pyrenophora semeniperda, leading to potentially important impacts on native seed fate.  We explored seed fate of two native grasses (Pseudoroegneria spicata and Elymus elymoides) planted into high- and low-density B. tectorum patches where disease levels were also manipulated (fungicide, control, and added inoculum).  Parallel precision seeding studies were installed in fall 2009 in at a xeric site in Skull Valley, Utah and at a mesic site near Sprague, Washington. The fate of unemerged seeds was examined at the end of spring 2010 by exhuming seeds, which had been glued individually to toothpicks, and evaluating their status.  

Results/Conclusions

At the xeric site, seedling emergence and survival were higher in high-density cheatgrass patches compared with low-density patches (emergence 37% vs. 27%, survival 72% vs. 44%), indicating facilitation by cheatgrass, probably because high litter improved seed zone water relations. At the mesic site, seedling emergence and survival were lower in the high-cheatgrass patches, indicating a cheatgrass competitive effect (emergence 13% vs. 29%, survival 12% vs. 24%).  Disease incidence on unemerged seeds was higher in high-cheatgrass treatments (32%) than in low-cheatgrass treatments (22%) at both sites, and was much higher overall at the xeric site (46%) than the mesic site (7%). Adding inoculum significantly increased disease levels on unemerged seeds of both species, particularly at the xeric site (40% controls, 64% inoculated).  This translated into significantly reduced seedling emergence with added inoculum primarily for P. spicata at the xeric site. Given the high disease levels in the controls at the xeric site, it is likely that emergence would have been higher if the seed pathogen were absent, but the fungicide treatment designed to test this was unfortunately ineffective.  Emergence was lower overall at the mesic site, where a large fraction of seeds germinated but failed to emerge, most likely due to a hard frost shortly after favorable germination conditions.  We found that the effect of cheatgrass on native grass seed and seedling fate may be directly positive (facilitation), directly negative (competition), or indirectly negative (apparent competition mediated through a shared pathogen).  Competitive effects were seen at the mesic site, whereas direct facilitation was observed at the xeric site, where indirect negative effects via the shared pathogen Pyrenophora semeniperda were also much stronger.