Abstract Detail


Iriart, Veronica [1], Baucom, Regina [2], Ashman, Tia-Lynn [3].

Herbicide drift reveals species-level variation in stressor resistance and weakens co-flowering interactions in 25 wild plant species.

Anthropogenic stressors such as herbicide drift threaten the diversity and stability of ecological communities. Drift to nontarget plant communities could affect species differentially due to variation in resistance or tolerance. Moreover, if drift alters flowering phenology, it may affect both pollinator-mediated plant-plant interactions and pollinator food availability. Thus, understanding the effects of herbicide drift on plant communities at the agricultural-ecological interface is an important challenge for biodiversity conservation across kingdoms. Accordingly, we grew 25 species of herbaceous plants spanning 15 families that naturally co-occur near agricultural fields in the southern United States in a common greenhouse environment. We treated the plants as seedlings with either a drift-level dose (~1% of the field application rate) of the widely used herbicide dicamba, which is often linked to drift, or a water and surfactant mixture as a control. We scored species’ initial stressor resistance and long-term tolerance as the percent difference between dicamba and control plants relative to the control in plant size at three-weeks post-treatment and final shoot biomass, respectively, and tested for significance using ANCOVAs with contrasts. We recorded the day of first flower and number of open flowers displayed for all plants, and used a network approach to characterize flowering phenology and assess community level effects of dicamba exposure by comparing network parameters between the drift and control communities. We found significant variation in initial stressor resistance, ranging from high susceptibility to even overcompensation (range = -60% to +59% relative size change), as well as in long-term tolerance (-39% to +34% relative biomass change), and day of first flower (+47 to -11 days until first flower). None of these indices demonstrated significant phylogenetic signal, suggesting species-specific patterns were not constrained by evolutionary history. We also found that dicamba drift weakened average interaction strength between co-flowering species by 24% and altered the composition of co-flowering modules, i.e. the number and membership of subgroups of species potentially interacting. These results demonstrate that even very low levels of herbicide exposure can have significant impacts on plant communities, phenology, and floral resource patterns—thus, community-scale ecological and evolutionary studies on the effects of herbicides on plant-pollinator interactions, particularly sublethal exposures, should be emphasized in future research, especially in the face of increasing agricultural intensification.

1 - University of Pittsburgh, 4249 Fifth and Ruskin Aves, Department of Biological Sciences, Pittsburgh, PA, 15260, United States
2 - University Of Michigan, 4034 BSB, Dept Of EEB, Ann Arbor, MI, 48109, United States
3 - Department Of Biological Sciences, 4249 Fifth Avenue & Ruskin, Pittsburgh, PA, 15260, United States

network analysis
plant tolerance
herbicide drift
Plant Community.

Presentation Type: Oral Paper
Number: ECO5005
Abstract ID:369
Candidate for Awards:Ecological Section Best Graduate Student Paper

Copyright © 2000-2020, Botanical Society of America. All rights reserved