Damage and recovery from drift of synthetic-auxin herbicide dicamba depends on concentration and varies among floral, vegetative, and lifetime traits in rapid cycling Brassica rapa

Highlights: • Herbicide drift is a major threat to biodiversity and ecosystem services. • Damage by synthetic-auxin herbicide drift varies among functional plant traits. • Floral traits are more sensitive to drift, but modularity may allow full recovery. • Flowering time and female fitness but no other lifetime traits were affected by drift. • Variation in damage and recovery, and fitness could cascade to ecological interactions. Herbicides can drift from intended plants onto non-target species. It remains unclear how drift impacts plant functional traits that are important for fitness. To address this gap, we conducted an experiment where fast cycling Brassica rapa plants were exposed to one of three drift concentrations (0.5%, 1%, 10%) of synthetic-auxin dicamba. We evaluated damage to and capacity of floral and vegetative traits to recover as well as lifetime fitness by comparing treated plants to controls. Response to dicamba exposure was concentration-dependent across all traits but varied with trait type. At 0.5% dicamba, three out of five floral traits were affected, while at 1% dicamba, four floral traits and one out of two vegetative traits were negatively impacted. At 10% dicamba all floral and vegetative traits were stunted. Overall, floral traits were more responsive to all dicamba drift concentrations than vegetative traits and displayed a wide range of variation ranging from no response (e.g., pistil length) to up to 84% reduction (ovule number). However, despite floral traits were more affected across the dicamba drift concentrations they were also more likely to recover than the vegetative traits. There was also variation among lifetime traits; the onset of flowering was delayed, and reproductive fitness was negatively affected in a concentration-dependent manner, but the final biomass and total flower production were not affected. Altogether, we show substantial variation across plant traits in their response to dicamba and conclude that accounting for this variation is essential to understand the full impact of herbicide drift on plants and the ecological interactions these traits mediate.