Phytotoxicity of soilborne glyphosate residues is influenced by the method of phosphorus fertiliser application

Abstract

Aims: Glyphosate use has increased in recent decades with the adoption of minimum tillage techniques and the emergence of glyphosate-tolerant crop cultivars. There is evidence that glyphosate residues are common in cropping soils at sowing, but any potential impact of these residues on crop growth is not known. This study aimed to determine concentrations of soilborne glyphosate that are phytotoxic to common crop species and investigate any interactions between glyphosate residues and phosphorus (P) fertiliser. Methods: A pot study examined the impact of soilborne glyphosate residues on the growth of wheat (Triticum aestivum L.) and narrow leaf lupin (Lupinus angustifolius L) seedlings in a sandy soil with a low P buffering capacity in the absence or presence of P fertiliser applied as a liquid to the topsoil. A second pot study investigated whether the phytotoxic effects of soilborne glyphosate residues on lupin growth in the presence of P fertiliser could be mitigated by banding the P fertiliser (as superphosphate) as per standard practice by farmers. Results: Wheat was more tolerant of glyphosate residues than lupin, with glyphosate levels up to 14.8 mg kg−1 having no effect on wheat shoot growth at 48 days after sowing (DAS), while lupin shoot growth at 48 DAS had an ED10 (i.e. a 10% yield reduction) of glyphosate at 6.4 mg kg−1 in the absence of P fertiliser. When P fertiliser was applied, wheat shoot biomass had an ED10 of 5.8 mg glyphosate kg−1 soil while lupins had an ED10 of 3.4 mg glyphosate kg−1 soil. When P fertiliser (superphosphate) was banded at sowing, lupins grown adjacent to band were unaffected by glyphosate whereas the shoot weights of lupins planted above the band were significantly reduced by 68% in soil containing 1.3 mg kg−1 glyphosate and 81% in soil containing 3.4 mg kg−1 glyphosate. Conclusions: Where P fertiliser was applied, growth reductions in lupin seedlings were observed at soil glyphosate concentrations that have been observed in farmers’ fields, suggesting that glyphosate residues may be problematic in sandy soils with low P buffering capacities. Banding of P fertiliser exacerbated, rather than mitigated, the phytotoxicity of glyphosate residues to lupins where P fertiliser was banded directly beneath the seed.