Aim To determine seed retention rates on vehicles as a function of distance driven, road surface, weather condition and seed location on the vehicle undercarriage. Location Montana, United States. Methods Metal plates were covered with a seed‐soil slurry, dried and attached to different locations underneath a vehicle. The vehicle was then driven on paved and unpaved roads under both wet and dry conditions. Plates were removed from the vehicle at seven distances between 4 and 256 km. The number of seeds remaining was determined. Four general models were assessed to explain observed seed retention. Results Under dry conditions, seed retention rates were high on both unpaved and paved roads, with 86–99% of the seeds remaining at 256 km. Under wet conditions, lower rates of seed retention were observed for both road surfaces: 0.3–80% of seeds were retained at 256 km on paved wet roads and 50–96% of seeds were retained at 256 km on unpaved wet roads. Plate location had a significant effect on seed retention under certain road surfaces and conditions, with loss generally being highest from the wheel wells. Of the statistical models compared, a double exponential model explained the most variation in seed retention. Main conclusions Vehicles act as vectors of long‐distance dispersal. Seed adhered to vehicles can be retained for hundreds of kilometres under dry conditions. When wet conditions occur, a greater proportion of seeds will be dispersed shorter distances. Consequently, vehicle seed dispersal has implications for plant invasions and species migration rates, and those concerned with prevention and control of non‐native plant invasions should consider vehicle seed transport when developing management strategies and plans.

The invasive annual grass Bromus tectorum (cheatgrass) forms a positive feedback with fire in some areas of western North America's sagebrush biome by increasing fire frequency and size, which then increases B. tectorum abundance post-fire and dramatically alters ecosystem structure and processes. However, this positive response to fire is not consistent across the sagebrush steppe. Here, we ask whether different climate conditions across the sagebrush biome can explain B. tectorum's variable response to fire. We found that climate variables differed significantly between 18 sites where B. tectorum does and does not respond positively to fire. A positive response was most likely in areas with higher annual temperatures and lower summer precipitation. We then chose a climatically intermediate site, with intact sage-brush vegetation, to evaluate whether a positive feedback had formed between B. tectorum and fire. A chronosequence of recent fires (1-15 years) at the site created a natural replicated experiment to assess abundance of B. tectorum and native plants. B. tectorum cover did not differ between burned and unburned plots but native grass cover was higher in recently burned plots. Therefore, we found no evidence for a positive feedback between B. tectorum and fire at the study site. Our results suggest that formation of a positive B. tectorum-fire feedback depends on climate; however , other drivers such as disturbance and native plant cover are likely to further influence local responses of B. tectorum. The dependence of B. tectorum's response to fire on climate suggests that climate change may expand B. tectorum's role as a transformative invasive species within the sage-brush biome. Author contributions: Kimberley Taylor conceived of/designed study; performed research; analyzed data; and wrote paper. Tyler Brummer performed research and wrote paper. Lisa J. Rew conceived of/designed study and wrote paper. Matt Lavin contributed new methods/models and wrote paper. Bruce D. Maxwell conceived of/designed study; performed research; and wrote paper. Results for the ten climate variables that were most correlated with PCO 1 are shown in declining order of importance. temp. Temperature, precip. precipitation, min. minimum, max. maximum.

No abstract

Bromus tectorum can transform ecosystems causing negative impacts on the ecological and economic values of sagebrush steppe of the western USA. Although our knowledge of the drivers of the regional distribution of B. tectorum has improved, we have yet to determine the relative importance of climate and local factors causing B. tectorum abundance and impact. To address this, we sampled 555 sites distributed geographically and ecologically throughout the sagebrush steppe. We recorded the canopy cover of B. tectorum, as well as local substrate and vegetation characteristics. Boosted regression tree modeling revealed that climate strongly limits the transformative ability of B. tectorum to a portion of the sagebrush steppe with dry summers (that is, July precipitation <10 mm and the driest annual quarter associated with a mean temperature >15°C) and low native grass canopy cover. This portion includes the Bonneville, Columbia, Lahontan, and lower Snake River basins. These areas are likely to require extreme efforts to reverse B. tectorum transformation. Our predictions, using future climate conditions, suggest that the transformative ability of B. tectorum may not expand geographically and could remain within the same climatically suitable basins. We found B. tectorum in locally disturbed areas within or adjacent to all of our sample sites, but not necessarily within sagebrush steppe vegetation. Conversion of the sagebrush steppe by B. tectorum, therefore, is more likely to occur outside the confines of its current climatically optimal region because of site-specific disturbances, including invasive species control efforts and sagebrush steppe mismanagement, rather than climate change.