Non-native, annual grasses [e.g., red brome (Bromus rubens)] are prolific throughout the Mojave Desert and have severe consequences for native flora and wildlife habitat. Generally, attempts to control the spread of non-native grasses have been reactive, focusing on already degraded habitats with a prior disturbance history (fire, mining, etc.). However, invasive species management after disturbance is costly and often requires additional measures to re-establish a native plant community. We investigated the effects of proactive herbicide treatments to reduce non-native, annual grass cover and belowground propagules, as well as evaluated the consequences to native vegetation in invaded, but otherwise intact, plant communities. We used a split-plot design to investigate how once- or twice-applied imazapic treatments affected native and non-native vegetative communities and seed banks. We measured vegetation changes in non-native annual grass and annual forb cover in the field and quantified soil seed banks in interspaces between shrubs and under shrub canopies using an emergence seed bank assay, which spanned a drought. Imazapic reduced non-native, annual grass cover in years with near-average precipitation but imazapic effects were negated in drought years. In 2020, once-applied imazapic treatments reduced B. rubens cover by 19.6% relative to controls and reduced the B. rubens seed bank by 42.1% in interspaces and 38.9% under shrub canopies. In 2021, imazapic effects on B. rubens cover were negated by drought; however, the B. rubens seed bank was reduced by treatments; once-applied treatments reduced the B. rubens seed bank in interspaces by 41% and under shrub canopies by 45.3%, and twice-applied treatments reduced the B. rubens seed bank by 33.1% in interspaces and 55.0% under shrub canopies. Bromus rubens seed bank reductions in 2021 may have interacted with drought resulting in reduced B. rubens cover in 2022. We found that herbicide effects on seed banks may persist through drought years even when effects are not seen on vegetation. This suggests timing herbicide treatments during years with adequate precipitation to decrease vegetative and subsequent reproductive growth may lead to longer-term invasive seed bank reductions and hopeful reductions in invasive grass dominance.
