Environmental Effects Of Herbicide Use
As has been suggested above, some substantial benefits can be gained through the use of herbicides to manage unwanted vegetation. Compared with alternative means of weed control, such as mechanically weeding by hand or machine, herbicides are less expensive, often safer (especially in forestry), faster, and sometimes more selective.
However, if herbicides are not used properly, damage may be caused to crop plants, especially if too large a dose is used, or if spraying occurs during a time when the crop species is sensitive to the herbicide. Unintended but economically important damage to crop plants is sometimes a consequence of the inappropriate use of herbicides.
In addition, some important environmental effects are associated with the use of herbicides. These include unintended damage occurring both on the sprayed site, and offsite. For example, by changing the vegetation of treated sites, herbicide use also changes the habitat of animals such as mammals and birds. This is especially true of herbicides use in forestry, because biodiverse, semi-natural habitats are involved. This is an indirect effect of herbicide use, because it does not involve toxicity caused to the animal by the herbicide. Nevertheless, the effects can be severe for some species. In addition, not all of the herbicide sprayed by a tractor or aircraft deposits onto the intended spray area. Often there is drift of herbicide beyond the intended spray site, and unintended, offsite damages may be caused to vegetation. There are also concerns about the toxicity of some herbicides, which may affect people using these chemicals during the course of their occupation (i.e., when spraying pesticides), people indirectly exposed through drift or residues on food, and wildlife. For these and other reasons, there are many negative opinions about the broadcast spraying of herbicides and other pesticides, and this practice is highly controversial.
The intention of any herbicide treatment is to reduce the abundance of weeds to below some economically acceptable threshold, judged on the basis of the amount of damage that can be tolerated to crops. Sometimes, this objective can be attained without causing significant damage to non-target plants. For example, some herbicides can be applied using spot applicators or injectors, which minimize the exposure to non-pest plants and animals. Usually, however, the typical method of herbicide application is some sort of broadcast application, in which a large area is treated all at once, generally by an aircraft or a tractor-drawn apparatus.
An important problem with broadcast applications is that they are non-selective—they affect many plants and animals that are not weeds—the intended target of the treatment. This is especially true of herbicides, because they are toxic to a wide variety of plant species, and not just the weeds. Therefore, the broadcast spraying of herbicides results in broad exposures of non-pest species, which can cause an unintended but substantial mortality of non-target plants. For example, only a few species of plants in any agricultural field or forestry plantation are abundant enough to significantly interfere with the productivity of crop plants. Only these competitive plants are weeds, and these are the only target of a herbicide application. However, there are many other, non-pest species of plants in the field or plantation that do not interfere with the growth of the crop plants, and these are also affected by the herbicide, but not to any benefit in terms of vegetation management. In fact, especially in forestry, the non-target plants may be beneficial, by providing food and habitat for animals, and helping to prevent erosion and leaching of nutrients.
This common non-target effect of broadcast sprays of herbicides and other pesticides is an unfortunate consequence of the use of this non-selective technology to deal with pest problems. So far, effective alternatives to the broadcast use of herbicides have not been discovered for the great majority of weed management problems. However, there are a few examples that demonstrate how research could discover pest-specific methods of controlling weeds that cause little non-target damage. These mostly involve weeds introduced from foreign countries, and that became economically important pests in their new habitats. Several weed species have been successfully controlled biologically, by introducing native herbivores of invasive weeds. For example, the klamath weed (Hypericum perforatum) is a European plant that became a serious pasture weed in North America, but it was specifically controlled by the introduction of two species of herbivorous leaf beetles from its native range. In another case, the prickly pear cactus (Opuntia spp.) became an important weed in Australia after it was introduced there from North America, but it has been successfully controlled by the introduction of a moth whose larvae feed on the cactus. Unfortunately, few weed problems can now be dealt with in these specific ways, and until better methods of control are discovered, herbicides will continue to be used in agriculture, forestry, and for other reasons.
Most herbicides are specifically plant poisons, and are not very toxic to animals. (There are exceptions, however, as is the case with the herbicide paraquat.) However, by inducing large changes in vegetation, herbicides can indirectly affect populations of birds, mammals, insects, and other animals through changes in the nature of their habitat.
For example, studies in Britain suggest that since the 1950s, there have been large changes in the populations of some birds that breed on agricultural land. These changes may be partly caused by the extensive use of herbicides, a practice that has changed the species and abundance of non-crop plants in agroecosystems. This affects the structure of habitats, the availability of nest sites, the food available to granivorous birds, which mostly eat weed seeds, and the food available for birds that eat arthropods, which rely mainly on non-crop plants for nourishment and habitat. During the time that herbicide use was increasing in Britain, there were also other changes in agricultural practices. These include the elimination of hedgerows from many landscapes, changes in cultivation methodologies, new crop species, increases in the use of insecticides and fungicides, and improved methods of seed cleaning, resulting in fewer weed seeds being sown with crop seed. Still, a common opinion of ecologists studying the large declines of birds, such as the gray partridge (Perdix perdix), is that herbicide use has played a central but indirect role by causing habitat changes, especially by decreasing the abundance of weed seeds and arthropods available as food for the birds.
Similarly, the herbicides most commonly used in forestry are not particularly toxic to animals. Their use does however, cause large changes in the habitat available on clear-cuts and plantations, and these might be expected to diminish the suitability of sprayed sites for the many species of song birds, mammals, and other animals that utilize those habitats.
Modern, intensively managed agricultural and forestry systems have an intrinsic reliance on the use of herbicides and other pesticides. Unfortunately, the use of herbicides and other pesticides carries risks to humans through exposure to these potentially toxic chemicals, and to ecosystems through direct toxicity caused to non-target species, and through changes in habitat. Nevertheless, until newer and more pest-specific solutions to weed-management problems are developed, there will be a continued reliance on herbicides in agriculture, forestry, and for other purposes, such as lawn care.
Briggs, S.A. Basic Guide to Pesticides: Their Characteristics and Hazards. Washington, DC: Taylor & Francis, 1992.
Freedman, B. Environmental Ecology. 2nd ed. San Diego: Academic Press, 1995.
Hayes, W.J., and E.R. Laws, eds. Handbook of Pesticide Toxicology. San Diego: Academic Press, 1991.
Pimentel, D., et al. "Environmental and Economic Costs of Pesticide Use." Bioscience 41 (1992): 402-409.