| |||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||
Executive Summary Extensive use of the herbicides dicamba and picloram in Wyoming, along with the large volumes of irrigation water used in some areas, has created a concern for the potential contamination of surface and ground water by these herbicides. Persistence and mobility of dicamba and picloram were investigated in batch adsorption, soil column, and field studies of both agricultural and rangeland soils. The objectives of this study were to characterize soil chemical and physical properties that affect herbicide transport, examine herbicide sorption, estimate degradation rate constants and model herbicide movement.
Soil chemical and physical properties that influence pesticide fate and mobility were examined in soils and substrata within three areas of Wyoming. Deep core incremental sampling was employed for pesticide analysis and was used as material for sorption studies to determine the potential extent of pesticide movement. Soil/substrata chemical and physical properties were also used in conjunction with results from the sorption studies to predict pesticide transport.
Essentially no sorption of dicamba was detected in laboratory studies; however, picloram sorption was found to be greater in soils containing increasing organic carbon contents. In saturated column (5.90, 2.96, and 0.82 kg ha-1 dicamba and 1.85, 0.97 and 0.47 kg ha-1 picloram) and unsaturated column (2.76 and 1.00 kg ha-1 dicamba and picloram, respectively) experiments, both herbicides and a Br tracer were displaced through soils using distilled water applied daily (60 ml d-1). Herbicide and tracer breakthrough curves were obtained from the column experiment. Degradation rate constants were calculated using both a simple recovery fraction technique and by matching LEACHP-generated breakthrough curves to experimental data. For the two columns receiving intermediate application rates, anaerobic picloram dissipation was more rapid (t1/2 = 19 d) than for aerobic conditions (t1/2= 87 d). The rate of dicamba dissipation was approximately the same under aerobic and anaerobic conditions (t1/2 = 15 and 17 d in the saturated and unsaturated columns, respectively). Picloram and dicamba dissipation was more rapid at the lowest application rates, with t1/2 of 13 and 10 days. Both herbicides were found to be highly mobile, with the mobility of picloram increasing at the higher pore-water velocities.
Soil collected at the Torrington experiment station prior to herbicide application, did not, for the most part, contain any dicamba. Concentrations below the detection limit (8 ug/kg) were found at the 3 conventional tillage sites at all depths. The highest dicamba content was found at the 60-90 cm depth for some no-till-injection and chisel-broadcast sites with different tillage and fertilizer treatments. All samples from no-till sites with fertilizer injection treatment contained trace amounts of dicamba at all depths. Results also suggest dicamba leached to depths of 60-90 cm (trace content) and deeper after 80 days. The highest persistence of dicamba was in the no-till sites with fertilizer injection treatment.
The results of picloram concentrations in soils from the Sundance area (Crook County) were variable with respect to ifs movement and degradation rate. An initial application rate of 0.25 lb a.i./ac was equivalent to 250 ppb of picloram in the top 15-20 inches of soil. In some sites, picloram was present in the top 15 cm of soil at 8.5 mg/kg one to six months after application of 1 lb a.i./ac. Fourteen months after an application of 1 lb a.i./ac, picloram content was highest at the 40-100 cm depth. Results indicate that picloram residues were limited to the top 100 cm with the highest concentration in the upper 40 cm. At one site, 10% of the applied picloram remained after 34 months.
Field studies were performed at an irrigated pasture site equipped with 64 soil water
extractors installed at four depths. Samplers extract solution from 15, 30, 60 and 90 cm
depth, with four replicated per plot at each depth. Two different herbicide application rates
were applied to a total four plot. The highest application rates for picloram and dicamba
were 2.9 kg ha-1 and 9.4 kg ha
Mean herbicide concentrations as a function of time were determined because of
the spatial variability in contaminant movement that was evident both between neighboring
samplers and plots. For plots, with lower herbicide application rates, herbicide
disappearance was relatively quick, especially for dicamba. Dicamba concentrations
approached the detection limit (0.0015 pm) in 96 days at the depth of 15 cm and in 57-89
days at the depth of 30 cm. Maximum concentration at the depth of 60 cm (0.01-0.02 pm)
was reached in 29 days and was below the detection limit in 43 days after application.
Dicamba was not detected at the 90 cm depth during the entire experiment.
Picloram remained in the profile longer, but did not penetrate into the vadose zone
as deeply as dicamba into the vadose zone. Picloram was not detected in any solution
samples collected from 60 and 90 cm samplers. However, picloram was detected on 327th
day at the depth of 60 cm. This is consistent with column study results in which picloram
was adsorbed by all the soil materials.
For plots with higher herbicide application rates, dicamba content within the profile
decreased more rapidly than picloram. Concentration peaks of picloram and dicamba
diminished and spread with increasing depth. Picloram concentration peaks moved
throughout the profile at a slower rate than dicamba peaks. The highest measured
concentration of picloram and dicamba was reached 15 days after application at the depth
of 15 cm; the peak reached a depth of 30 cm two weeks later and a depth of 60 cm six
weeks later. Picloram content at the 90 cm depth was still increasing 327 days after
herbicide application.
Modeling pesticide movement in a vadoze zone proved to be a useful research tool.
LEACHP provided modeling parameters of solute movement in repacked soil columns.
However, comparing model-predicted and field contaminant movement was more difficult
due to spatial variability. Hydraulic conductivity was found to be highly variable, and would
have the greatest effect on contaminant movement.
Additional studies are being conducted on the field site used for this project. Ground
water monitoring wells have been installed at 10 locations throughout the research site.
Both dicamba (=9.4 kg ha-1) and Br- (225 kg ha-1) have been applied to plots 1 and 2, and
a Cl-tracer (50,000 mg L-1) added to one of the ground water monitoring wells. Results of
this additional study will provide information on pesticide and tracer characteristics in both
vadose zone and ground water environments.
Water Resources Publications List
Water Resources Data System Library |
Water Resources Data System Homepage
| |||||||||||||||||||||||||||||||||||
