Weed management: a need for ecological approaches - Ecological Management of Agricultural Weeds - page 15

the 12362 samples tested; maximum contaminant levels for alachlor and atra-
zine were exceeded in 1.1% and 0.1% of the samples, respectively. Two large-
scale, multistate investigations of herbicides in American wells and springs
detected at least one of seven targeted compounds in 35% to 40% of the sites
sampled, although maximum contaminant or health advisory levels were
exceeded at fewer than 0.1% of the sites (Barbash et al ., 1999).
Although concentrations of individual herbicides in American groundwa-
ter rarely exceed existing regulatory standards, important concerns remain
concerning health risks.Detection of one herbicide in groundwater at an indi-
vidual site is often accompanied by the detection of others (Barbash et al .,
1999), but little is known about the health-related impacts of exposure to
multiple herbicides, or to herbicides in combination with nitrates, which are
also common water contaminants.Breakdown products of herbicides are gen-
erally found in well water more frequently and at higher concentrations than
the corresponding parent compounds (Kolpin, Thurman & Goolsby, 1996),
but little is known about their possible effects on human health.Health-based
standards for breakdown product concentrations in groundwater generally
do not exist.

Herbicide drift
Herbicides can contaminate off-target sites by moving in air as well as
in water.Generally,herbicide drift from tractor-mounted sprayers is about 5%
to 10% of the material applied,with most off-site deposition occurring within
20m of field edges (Freemark & Boutin, 1995). However, depending on
meteorological conditions, application equipment, and physical characteris-
tics of herbicide products,spray drift concentrations of 0.02% to 2% of applica-
tion rates may occur at distances as great as 400m from application sites
(Fletcher et al ., 1996).
The implications of aerial movement of herbicides are especially proble-
matic for highly phytotoxic chemicals, such as sulfonylurea and imidiazoli-
none compounds. Although these compounds may have low mammalian
toxicity, their drift onto nontarget crops and wild land areas, even at low con-
centrations, may greatly alter plant performance, particularly reproduction.
Fletcher et al . (1996) found that flower and seed production by rapeseed,
soybean, sunflower, and Polygonum persicaria could be reduced by exposure to
chlorsulfuron at rates from 0.1% to 0.8% of those recommended for field
applications to cereal crops. For certain combinations of plant species, chlor-
sulfuron rates, and application times, reproductive damage occurred even
when effects on vegetative growth were minimal. For example, chlorsulfuron
treatment of rapeseed (at 9.2 10 5 kg a.i. ha 1 ) and soybean (at 1.8 10 4 kg