Phytotoxicity of glyphosate soil residues re-mobilised by phosphate fertilisation

It has been repeatedly demonstrated that phosphate (P) and the herbicide glyphosate compete for adsorption sites in soils. Surprisingly, the potential consequences of these interactions for plants e.g. re-solubilisation of phytotoxic glyphosate residues in soils by application of P fertilisers or by root-induced mechanisms for P mobilization have not been investigated so far. In model experiments under greenhouse conditions, the potential for glyphosate re-mobilisation by P-fertiliser application was evaluated by bio-indication with soybean (Glycine max L.) cultivated on five contrasting soils with or without glyphosate application at 10-35 days before sowing. Different levels of P-fertilisation (0, 20, 40, 80, 240 mg P kg(-1) soil) were supplied at the date of sowing. Visual symptoms of glyphosate toxicity, plant biomass, intracellular shikimate accumulation as physiological indicator for glyphosate toxicity and the plant nutritional status were determined. On glyphosate-treated soils, P application induced significant plant damage. Expression of damage symptoms declined in the order Arenosol > Acrisol a parts per thousand Ferralsol > Luvisol subsoil > Regosol. On the Arenosol, Ferralsol and Luvisol subsoil plant damage was associated with increased shikimate accumulation in the root tissue. On the Acrisol decline of germination and plant damage in absence of shikimate accumulation indicate toxicity of AMPA (aminomethylphosphonic acid) as the main metabolite of glyphosate in soils. On the Regosol, a growth-stimulating effect of glyphosate soil application (hormesis) was detected. The results suggest that re-mobilisation of glyphosate may represent an additional transfer pathway for glyphosate to non-target plants which is strongly influenced by soil characteristics such as P fixation potential, content of plant-available iron, pH, cation exchange capacity, sand content and soil organic matter.