Developing management strategies for invasive plant species requires identifying effective treatment methods (e.g., physical or chemical treatments) and optimally timing their application. For invasive annual plants, effects of treatments on seed production and germinability are paramount to reduce seed banks and subsequent plant establishment. We compared effects of a range of physical and chemical treatments applied to plants containing seeds at different developmental stages at field sites in the Mojave Desert (Nevada and Arizona) for managing the exotic annual forb, African mustard. Effectiveness of physical treatments (hand pulling entire plants, breaking plants, or separating siliques from plants) for reducing African mustard seed size and germination varied by treatment and seed maturity stage at which treatments were applied. With all treated plant material remaining in the field (i.e., not bagged and transported off site), hand pulling or breaking plants resulted in 90 to 100% of developing and developed seeds still able to germinate. Separating siliques from plants, however, resulted in 0% germination of undeveloped and developing seeds. All three tested herbicides (glyphosate, 2,4-D, and metsulfuron) reduced germination to zero or near zero across all seed development stages. Results suggest that physical treatments are most effective at early stages of seed development, whereas several different herbicides are effective across seed developmental stages. Results also highlight the importance of considering the potential for continued seed development and germinability on treated plant material when choosing invasive plant treatment types and timing.

Nonnative, invasive plants are becoming increasingly widespread and abundant throughout the southwestern United States, leading to altered fire regimes and negative effects on native plant communities. Models of potential invasion are pertinent tools for informing regional management. However, most modeling studies have relied on occurrence data, which predict the potential for nonnative establishment only and can overestimate potential risk. We compiled locations of presence and high abundance for two problematic, invasive plants across the southwestern United States: red brome (Bromus rubens L.) and African mustard (Brassica tournefortii Gouan). Using an ensemble of five climate projections and two types of distribution model (MaxEnt and Bioclim), we modeled current and future climatic suitability for establishment of both species. We also used point locations of abundant infestations to model current and future climatic suitability for abundance (i.e., impact niche) of both species. Because interpretations of future ensemble models depend on the threshold used to delineate climatically suitable from unsuitable areas, we applied a low threshold (1 model of 10) and a high threshold (6 or more models of 10). Using the more-conservative high threshold, suitability for Bromus rubens presence expands by 12%, but high abundance contracts by 42%, whereas suitability for Brassica tournefortii presence and high abundance contract by 34% and 56%, respectively. Based on the low threshold (worst-case scenario), suitability for Bromus rubens presence and high abundance are projected to expand by 65% and 64%, respectively, whereas suitability for Brassica tournefortii presence and high abundance expand by 29% and 28%, respectively. The difference between results obtained from the high and low thresholds is indicative of the variability in climate models for this region but can serve as indicators of best- and worst-case scenarios.