Apart from the fact that most information regarding the toxicology of the herbicides is as yet rather fragmentary—a situation which will be considerably improved when reports from the biological laboratories of industry become available to the public—it is important that attention not be confined to the toxicity of the original compounds. Their influence on plant metabolism might, for instance, block synthetic chains, leading to the accumulation of endogenous toxic products in vegetables. In this respect it has been shown that 2,4-D, applied in subtoxic amounts to such plants as sugar beets, increases the concentration of nitrate to twenty times the normal level. Consequently, feeding with the foliage involves a risk of poisoning (99). Attention has also been focused on the increase of cyanide in cyanogenic plants exposed to the action of herbicides (103).
As to the toxicity of metabolites of the herbicides themselves, which are formed in the soil and plants after their distribution, little is known. The problem has been touched in connection with investigations into the breakdown of herbicides by the aid of the micro-organisms in the soil. In particular, the decomposition of the chlorinated phenoxy-acids and the chlorinated aliphatic acids by soil bacteria and fungi has been submitted to extensive studies (vide 11, 57, 58, 59, 60, 93, 121). It appears that the breakdown of these compounds is almost complete, involving dehalogenation (57-60) and hydroxylation of the side-chain (93). Therefore, a persistent deleterious effect on the microflora of the soil need not be feared after use of the great majority of these organic herbicides (38). This is probably due to the fact that in most cultivated soils the populations of bacteria and fungi are so rich that there will always be found at least a few individuals that can adapt to almost any substance presented to them. These will synthesize new specific enzyme systems using the foreign substance as substrate. The resistant strains of micro-organisms will therefore multiply excessively, until the "substrate" has been decomposed (11). In general, the biotransformation of herbicides in soil does not seem to lead to the production of more potent compounds (11, 93, 121). Certain inorganic herbicides, e.g., chlorate, may, however, have a deleterious influence on bacterial nitrification in the soil. By reduction of chlorate, hypochlorite may be produced, which has a very strong inhibitory action on the growth of the soil bacteria (1). Pentachlorphenol may have a harmful effect, especially on the soil fungi; various emulsifying and surface-active agents in the herbicidal preparations may also interfere with the soil flora (38).
The fates of the herbicides in plants have as yet been scarcely elucidated, but intensive studies are being carried on to clarify biotransformation in plants. The explanation of selective phytotoxicity has been founded on these studies. Decisive progress was gained when it was discovered that the herbicidally inactive 4-(2,4-dichlorophenoxy)butyric acid is converted by beta-oxidation to the active 2,4-dichlorophenoxyacetic acid in plants where specific beta-oxidase systems are present. These plants are harmed while species not possessing the specific enzymes are resistant (109). Conversely, the triazine derivative, Simazin, in nonsensitive plants is degradated by the presence of different factors such as catalase, peroxidase, and polyphenols to substances without phytotoxicity (46a). Great difficulties in analytical procedures have to be solved before the fates of herbicides in plants are clarified (92). However, there is reason to expect that a more rational basis will eventually be available, both for the study of the metabolism of herbicides in the animal organism, and for the planning of experimental investigations of the toxicity of the original compounds as well as of their metabolites. A knowledge of persistent metabolites in crops for consumption especially will be of very great importance in the study of long-term actions, including carcinogenic effects (35).