Identifying mechanisms that determine who lives and dies is the first step in developing successful restoration techniques for rare species and endangered habitats. We studied interactions that affect establishment of native plant forbs of conservation concern at the seedling stage to support the theoretical basis for restoration activities in Pacific Northwest prairies. Specifically, we tested the hypothesis that seedling establishment is controlled by 1) competition with or 2) facilitation by existing vegetation and that the interaction is mediated in part by environmental stress. We direct-seeded or planted vegetative plugs of Lupinus oreganus, Castilleja levisecta, Erigeron decumbens, Iris tenax and Sidalcea malviflora ssp. virgata into 20 plots with a range of community compositions in high-stress upland prairies at each of three sites. We counted seedlings and estimated cover of plant functional groups as well as litter, bare soil and disturbance then used linear regression to test for effects of these factors on seedling establishment. We found evidence of indirect facilitation of grass on seedling establishment in the first year: higher accumulations of leaf litter increased seedling numbers at two sites. In the second year, there was evidence of facilitation by live vegetation and litter on seedlings at one site, but no net effect of either competition or facilitation at the other two sites. Overall, we found more evidence for positive interactions than we did for competition. In particular, litter appeared to have a positive effect on seedling establishment of L. oreganus and S. malviflora ssp. virgata. This is contrary to the common perception that litter inhibits plant establishment but supports the theory that facilitation is more common in high stress sites; practitioners should consider seeding into leaf litter at some sites. To support a robust approach to conservation and reintroduction of species with dormant seed, we characterized dormancy types and developed germination protocols for S. malviflora ssp. virgata and I. tenax. S. malviflora ssp. virgata has physical dormancy and may have physiological dormancy. Scarification followed by four weeks of cold moist stratification was effective in initiating germination. I. tenax has morphophysiological dormancy which is overcome by four weeks of warm moist stratification followed by 6-12 weeks of cold stratification. We also conducted a meta-analysis of experiments that tested pre-sowing seed scarification of L. oreganus and conclude that breaking physical dormancy prior to direct seeding does not support higher establishment relative to unscarified seeds in this species. / Graduation date: 2012