A comparison of crop and non-crop plants as sensitive indicator species for regulatory testing.

The effectiveness of regulatory non-target plant testing using crop species to predict the phytotoxicicity of herbicides to non-crop species was evaluated for eleven herbicides. These herbicides were representative of eight chemical classes and six modes of action. Data for non-crop plants from pre-emergence and post-emergence efficacy screening studies were compared with those for the most sensitive crop species defined by regulatory tests conducted to meet US EPA requirements. Testing under pre-emergence conditions for ten compounds indicated that for five of the compounds (K-815910, trifluralin, pyridyloxy A, pyridyloxy B and cyanazine), the most sensitive crop species was more sensitive than all the non-crop species evaluated. For metsulfuron-methyl, chlorimuron-ethyl, hexazinone and bromacil, only one of the non-crop species evaluated was more sensitive than the most sensitive crop species from regulatory tests. Data for the tenth compound, chloroacetamide, showed that four of 32 non-crop species tested in efficacy screens had at least one rate at which greater visual effects were observed than were observed for the most sensitive crop response in a regulatory test. The results of post-emergence exposure comparisons for five of the compounds (pyridyloxy A, cloransulam-methyl, chlorimuron-ethyl, cyanazine and hexazinone) indicated that the most sensitive crop species were more sensitive than all the non-crop species evaluated. Data for pyridyloxy B, metsulfuron-methyl and bromacil indicated that only one of the non-crop species evaluated was more sensitive than the most sensitive crop species. For trifluralin, three of the eight non-crop species were more sensitive than the most sensitive crop species. Data for K-815910 indicated that four of the fourteen non-crop species tested were marginally more sensitive than the most sensitive crop, but were within the same range of sensitivity. These results indicate that the current regulatory test batteries and methods using crop species effectively provide suitable sensitive indicator plants for the eleven diverse herbicides evaluated. This comparison indicates that crop species sensitivity to test substances is likely to be representative of non-crop herbaceous species response, regardless of chemical class, mode of action and magnitude or route of exposure.