Simulated herbicide drift alters native plant flowering phenology

Abstract

Data for herbicide effects on plant flowering are needed to determine potential impacts on plant reproduction. Thus, flowering phenology was determined for up to 12 weeks after herbicide treatment for native Willamette Valley plants growing in small plots on two Oregon State University experimental farms. Six perennial species were evaluated: Camassia leichtlinii (CALE), Elymus glaucus (ELGL), Eriophyllum lanatum (ERLA), Festuca idahoensis subsp. roemeri (FEID), Iris tenax (IRTE), and Prunella vulgaris var. lanceolata (PRVU). Effects of glyphosate and dicamba, alone and in combination, were determined using simulated drift rates of 0.1 or 0.2 x field application rates (FAR) of 1119 g ha⁻¹ active ingredient (a.i.) (830 g ha⁻¹ acid glyphosate) for glyphosate and 560 g ha⁻¹ a.i. for dicamba. Flowering phenology was evaluated as stage of development on a scale from no buds (converted to 0), buds (1), pre-flowering (2), flowering (3), post-flowering (4), to mature seeds (5) before herbicide treatment and for 12 weeks after treatment. Flowering response to herbicides varied by species and farm; but, in general, dicamba and glyphosate resulted in earlier flowering stages (delayed or not full flowering) for the dicot ERLA, and to a lesser extent, PRVU; and glyphosate resulted in earlier flowering stages for the monocot IRTE. Based on these data, the concentration of herbicide affecting flowering stage was 0.1 x FAR. Once flowering stage was inhibited by dicamba and glyphosate, plants generally did not recover to full flowering. This study provided evidence that common herbicides can affect flowering phenology of native plants with implications for seed production.