- Cheatgrass (Bromus tectorum L.) is a widespread exotic weed in the Intermountain sagebrush steppe. An annual grass, it is highly prolific and very competitive with native perennial grass seedlings. A clipping experiment carried out at two cheatgrass-dominated sites (Lincoln Bench and Succor Creek) in eastern Oregon analyzed effects of defoliation on cheatgrass seed production, investigated mechanisms of altered seed production and plant recovery, and considered the potential of defoliation as a cheatgrass control method. Treatments involved hand clipping plants at two heights (tall - 7.6-cm (T) and short - 2.5 cm (S)), two stages of phenological development (boot (B) and purple (P) stages), and two frequencies (once (1) and twice (2)), though purple stage clippings were clipped only once. Treatments were replicated in a randomized complete block design, which included a control with no defoliation. End of season seed production (seeds/m²), plant density (plants/m²), plant seed production (seeds/plant), and tiller production (percentage
of plants with greater than 1 inflorescence) were estimated by sampling plants and litter from each treatment plot at the end of the growing season. Seeds were hand-collected from these samples, counted, and tested for viability. Soil moisture was measured with a TDR device in three randomly selected blocks, and averaged over the season for each treatment. End of season seed production was greater than zero for all treatments at both sites. At Lincoln Bench, all treatments excluding the TB1 treatment produced significantly less seed than the control. At Succor Creek, only the SB2 and SP treatments produced significantly less seed than the control. The SB2 treatment had the lowest seed production at both sites, at 119 and 1243 seeds/m² at Lincoln Bench and Succor Creek, respectively, along with the SP treatment at 1115 seeds/m² at Succor Creek. The response patterns for plant density and seed production of individual plants were similar to that for overall seed production, which suggests that these treatments reduced seed production by increasing plant mortality and reducing plant reproductive ability. Tiller
production increased for the SB1 treatment, which suggests that cheatgrass plants were able to recover from defoliation partly through asynchronous or increased tiller development. There was no significant effect of treatment on seasonal soil moisture. In conclusion, although the SB2 and SP treatments showed the greatest reduction in seed production, plants in these treatments still produced viable seed. Thus, applying a similar defoliation treatment for seedbed preparation with livestock-assuming similar treatment effects- may not be sufficient by itself to reduce cheatgrass to levels low enough to reduce competition in native reseeding projects. Alternatively, defoliation treatments could be intensified, and/or combined with other weed control methods as part of an integrated weed management approach.