Effects of glyphosate-based herbicides on embryo-larval development and metamorphosis in the Pacific oyster, Crassostrea gigas.

Pesticides may be involved in oyster summer mortality events, not necessarily as a single causative agent but as an additional stressor. In this context, the present study aimed to assess the toxicity of glyphosate, its by-product, aminomethylphosphonic acid (AMPA) and two commercial formulations, Roundup Express(®) (R(EX)) and Roundup Allées et Terrasses(®) (R(AT)), containing glyphosate as the active ingredient, on the early life stages of the Pacific oyster, Crassostrea gigas. The embryotoxicity of these chemicals were quantified by considering both the rates of abnormalities and the arrested development or types of abnormalities in D-shaped larvae after 48 h exposure. The success of metamorphosis was examined in pediveliger larvae exposed for 24 h. Experiments involving both endpoints included range finding experiments for herbicide concentrations ranging from 0.1 to 100,000 μg L(-1). This range was then narrowed down in order to determine precise EC(50) values. Actual concentrations of the herbicide were determined at the beginning and after 48 h (embryotoxicity) and 24 h (metamorphosis) to evaluate the potential temporal variation in the concentrations. During embryo-larval development, no mortalities were recorded at any of the concentrations of glyphosate and AMPA, whereas no embryos or D-shaped larvae could be observed after exposure to 10,000 μg L(-1) of R(EX) or R(AT). Compared with the controls, no effects on embryo-larval development were recorded between 0.1 and 1000 μg L(-1), regardless of the chemical tested. Above a threshold, which varied according to the chemical used, the gradient of herbicide concentrations correlated with a gradient of severity of abnormality ranging from normal larvae to arrested development (an "old embryo" stage). The EC(50) values were 28,315 and 40,617 μg L(-1) for glyphosate and its metabolite, respectively, but much lowered values of 1133 and 1675 μg L(-1) for R(EX) and R(AT), respectively. Metamorphosis tests also revealed a significant difference between molecules, as the EC(50) values exceeded 100,000 μg L(-1) for glyphosate and AMPA but were as low as 6366 and 6060 μg L(-1) for the commercial formulations, which appeared relatively more toxic. Overall, the embryo-larval development of C. gigas was more sensitive to glyphosate-based herbicides compared to various endpoints studied in regulatory model organisms, and embryos and D-shaped larvae were more sensitive compared to pediveliger larvae.