The effects of sequential herbicide applications on phosphorus cycling and mycorrhization in soybean: A two-year field study

Herbicide use has evolved from single pre-sowing applications to the use of tank mixtures applied at higher rates both pre-emergence and in-crop. Understanding the effects of these practices on non-target organisms is critical for protecting soil functions and regulating pesticide use. This work examines the impact of a commonly used herbicide application scheme in soybean production on non-standardized endpoints related to the phosphorus cycle, under field conditions over two years. Specifically, total arbuscular mycorrhizal colonization, arbuscule and vesicle formation in soybean roots, phosphorus concentration and total content in shoots were evaluated at the vegetative and reproductive stages of the plants. Nine treatments were assessed: a control without herbicides, a full herbicide scheme at label recommended rate comprising preplant, preemergence and early post emergence applications; all seven partial scheme variations; and a worst-case scenario, full scheme at twice the label-recommended rate. Total and arbuscular colonization showed non-linear dose-response to herbicide load (number and rate of active ingredients), not directly related to specific modes of action. Treatments with only the preplant and only the preemergence herbicide application (T2 and T3, respectively) had lower total and arbuscular colonization than the control in the second year at the vegetative stage. The treatment with the preplant and preemergence herbicide applications (T5) had lower total colonization than the control in the first year at the reproductive stage. Vesicle formation was the most sensitive endpoint, with a significant stimulatory response to herbicide-induced stress compared to the control in the reproductive stage of year 1. Phosphorus concentration increased in T3 treatment compared to the control the first year in the reproductive stage. The remaining treatments were not statistically different from the control. The low herbicide levels detected in soil under no-tillage management, with over 95 % of soil covered by straw, could explain the absence of negative effects in the evaluated soil functions under the full sequence of herbicides. Overall, the results suggest mild yet significant counterproductive effects on agroecosystems, and the need for better risk assessment frameworks, pre- and post-registration of pesticides.