Meissen, Justin2016-09-192016-09-192016-06https://hdl.handle.net/11299/182161University of Minnesota Ph.D. dissertation. 2016. Major: Conservation Biology. Advisors: Susan Galatowitsch, Meredith Cornett. 1 computer file (PDF); 100 pages.Seed supply limits large-scale restorations, which often rely on seed collection from remnant ecosystems. Overharvesting seed may deplete populations, exacerbate seed limitation, and jeopardize ecosystem integrity, but these risks have not been formally studied. Many life history traits are linked to species’ reliance on seed reproduction, and so may provide a useful framework to address seed harvest risks. I evaluated whether life history traits predict susceptibility to overharvest by comparing tallgrass prairies in Minnesota (USA) harvested at varying frequencies (every year, once per 3-5 years, unharvested). I identified species less likely to occur on frequently harvested sites then tested whether lifespan, clonality and seed production predicted harvest sensitivity. Short-lived, non-clonal species were sensitive to seed harvest while long-lived clonal species were not, suggesting that life history traits provide a means to predict seed harvest risk. To verify the predictive utility of life history traits and determine extinction risks from seed harvest over long restoration timeframes (25 years), I used matrix models of clonal Solidago canadensis and Anemone canadensis and non-clonal Rudbeckia hirta, Packera aureus, Zizia aurea, and Liatris ligulistylis to simulate seed harvest and extinction risk. I simulated 5 scenarios: no harvest; annual harvest at 50%/75% intensity; and triennial harvest at 50%/75% intensity. Non-clonal species were insensitive to triennial and 50% harvest, but susceptible to extinction risks of up to 92% with annual 75% harvest. Clonal species were insensitive to all harvest scenarios. To maintain populations of non-clonal species in the long-term, high intensity annual harvest should be avoided. To demonstrate the risk of overharvest in short-lived, non-clonal species and determine sustainable harvest regimes, I conducted a field experiment varying seed harvest intensity (0, 50%, or 100% seed removed) and management (burned or unburned) for R. hirta populations. I compared seedling recruitment and seed production among treatments, and found that seed production nearly doubled with burning. Moderate intensity harvest with burning allowed high levels of seedling recruitment, but high harvest intensity prevented recruitment, as predicted for a short-lived, non-clonal species. A regime combining moderate intensity harvest with fire management provides seeds while also conserving at-risk seed donor sources.enecological restorationlife historyrisk managementseed harvesttallgrass prairietraitsThesis or Dissertation