Comments on Tests of Herbicide Mixtures

NE of the important areas of current herbicide research is the investigation of herbicide mixtures. It is a common field experience that use of a single material-even of a broad-spectrum herbicide-exerts a selective pressure on the field population because of differential effectiveness on the various weed species. There is some evidence that a part of this derives from purely mechanical considerations (e.g., the spraying system in use may not reach plants of different growth habit equally well), but in any case the weed control program often appears to decrease in effectiveness after several seasons. In addition to favoring the survival of particular species, the non-eradicatory application of even a highly efficient chemical apparently aids in the establishment of populations of individuals which are physiologically most tolerant of the herbicide used. It will probably be granted, then, that the use of mixtures of toxicants will provide more effective control of populations of mixed weed species, and may also reduce the numbers of individuals and species exceptionally tolerant of the herbicide program, especially when the toxicants kill by affecting different physiological processes. Two points of particular interest in tests of herbicide mixtures are (1) whether or not there is any interaction in the responses to the herbicides which would make their combination unexpectedly effective, i.e., synergistic, and (2) how to determine the expected effects. In respect to the first point, and to take the simplest case, the relative effectiveness of two chemically similar toxicants may readily depend on the relative affinities for both the site of toxicity and the site of some mechanism of detoxification. These affinities and the ratios of the toxicants used in the mixture will condition the competition at each site, and if the less toxic material spares the destruction of the more toxic material, synergism will be obtained. A complication arises in respect to the second point, the determination of the expected effect. This can only occasionally be established by simple summation of the responses to the chemical used separately. When one chemical has already accounted for certain individuals, the other may appear either more or less effective than when used alone, depending, among other things, on the level of kill being tested and on the degree to which susceptibility to the two toxicants is related. (Susceptibility to two quite dissimilar toxicants may be related. If, for instance, one affects permeability and the other respiration, they both might be most effective against weeds in full sunlight, the one because of a connection with transpiration, and the other because of a connection with temperature.) Consequently, demonstration of a real interaction-of synergism or antagonism-between toxicants may not be a simple matter.