Use of Contour Strip Cropping as a Best Management Practice to Reduce Atrazine Contamination of Surface Water
Abstract:
Atrazine is widely used to control weeds of cereals in the southern USA, but concerns about pesticide contamination of surface waters threatens its continued use. This study was conducted on nine watersheds to compare the effectiveness of contour strip cropping in reducing atrazine, metolachlor and cyanazine in runoff. The dimensions of each watershed was 45 x 140 m (about 0.6 ha). Three cropping systems were evaluated: (1) conventionally tilled field corn with grass filter strips, (2) conventionally tilled field corn with wheat filter strips and (3)conventionally tilled field corn with no filter strips. The three systems were randomly assigned to the watersheds and replicated three times. The soil, a Houston Black clay (Udic Pellustert, fine, montmorillionitic, thermic), ranged in slope from 3 to 5%. Atrazine, metolachlor, and cyanazine were applied according to manufacturers' directions to the areas planted to corn. Runoff was measured and sampled during rainfall events to determine the amount and quality of water leaving each watershed. Instrumentation recorded the total rainfall and intensity during each event. Herbicide concentration in composited runoff water samples and soil samples collected from the filter strip areas were determined using gas chromatography following each runoff event. Results indicated that significant amounts of runoff were retained by the filter strips, thereby reducing herbicide nonpoint source pollution. Contour strip cropping technology should help reduce erosion and prevent pesticide contamination of rivers and lakes.
Background:
Contour filter strips, designed to remove sediment, chemicals and organic material transported in runoff water, were established on nine watersheds at Blackland Research Center, Temple, TX USA, in 1992. The objective was to compare the efficiency of contour strip cropping with grass filter or wheat strips, with conventional cropping and determine if contour filter strips were effective in reducing atrazine non-point source pollution from field corn. Each watershed was 45 x 140 m (0.6 ha). The watersheds designated for filter strips were divided into three equal sections (46 m). Nine meter filter strips were established on the downslope portion of each section resulting in three cropped sections and three filter strips per watershed. Three main treatments were established; Coastal Bermudagrass filter strips, wheat filter strips, and no filter strips. Conventional tillage practices were applied to the corn and wheat cropped areas. The herbicides atrazine (2.24 kg./ha ai), metolachlor (2.8 kg/ha ai) and cyanazine (4.48 kg/ha ai) were applied pre-emergence to the corn crop. Corn and wheat were planted parallel to the catchment terrace. Profile soil samples were collected on 0, 7, 14, 30, and 60 days after planting. Runoff events were monitored and composited water samples were collected during each measurable event. Although excessive runoff can occur anytime in central Texas, it is most prevalent during the spring months just after planting of corn. In central Texas annual total rainfall averages 900 mm and runoff often exceeds 60 mm in May and early June. Soil samples (0-7.5 cm) for herbicide analysis were collected from each of the wheat and coastal Bermudagrass filter strip areas after each runoff event. Herbicides were extracted from the soil and water samples using standard extraction procedures and evaluated with a Hewlett Packard 5890-II gas chromatograph.
Discussion:
Soil profile samples were taken from the cropped area to monitor potential herbicide leaching. The herbicides (atrazine, metolachlor and cyanazine) have not proven to be mobile in the Houston Black clay soil in this study or previous studies. The characteristics of Houston Black Clay (pH=8.0 and above), high soil temperatures and moist soil conditions reduce herbicide half-life to less then 90 days for atrazine and metolacholor and 21 days for cyanazine. During the 1992 and 93 seasons atrazine, metolachlor and cyanazine were detected in soil samples taken from both wheat and Coastal Bermudagrass filter strips following runoff events Significant herbicide concentrations (less then 0.1 ppm) were not detected deeper then 15 cm in the soil profile.
Data collected during preliminary 1992 experiments indicated that 4.5 and 9m foot Coastal Bermudagrass filter strips were intercepting both atrazine and metolachlor. Highest concentrations were recorded in the down-slope filter strip closest to the catchment weir, but hydrological data was not collected in 1992. During the 1993 season atrazine, metolachlor and cyanazine were detected in both wheat and Coastal Bermudagrass filter strips following runoff events. All areas of the filter strips had measurable concentrations of the two herbicides (0.05-0.1 ppm) indicating that the filter strip areas were retaining herbicide contaminated sediments and/or water. Hydrologic data show that runoff volume (l/ha) was reduced over 57% by Coastal Bermudagrass strips and 50% by wheat strips during a May 2, 1993, runoff event. The herbicide concentration in runoff was similar for all nine watersheds. Extrapolation of the runoff data demonstrate that the filter strips reduce herbicide loss in runoff significantly by retaining a significant volume of runoff during two of the three runoff events. The results of the May 9 event were inconclusive, due to flooding which occurred during this event (3.2 inches in less than one hour) permitting runoff into and out of the watershed areas. Damage from flooding on May 9 prevented additional runoff monitoring. Previous studies (1990-92) indicate that herbicide runoff losses occurred during the first 3 or 4 events, with 60-80% of the total loss occurring during the first runoff event. Results from 1993 indicate a similar response, with 50-60% of the total atrazine and metolachlor losses occurring during the first event.
Two major runoff events were recorded during the 1994 cropping season. Both events occurred on May 16, one in the early morning hours (2.4 cm) and the second event (2.5 cm) in the late afternoon. Runoff samples were collected from all nine watersheds following both runoff events. Herbicide concentration present in the runoff was considerably less then previous years which reflects the amount of time elapsed between application and runoff. The afternoon runoff event was the most significant of the two events. The morning rainfall saturated the soil profile resulting in much of the later storm event becoming runoff. Herbicide concentrations were lower in the runoff from both the watersheds containing wheat and grass filter strip areas when compared to corn only watersheds, demonstrating the effectiveness of using contour filter strips to reduce non-point source pollution. Due to the lateness of the 1994 runoff events soil herbicide concentrations were below detection limits. Taking into consideration that contour filter strips reduce the cropped area and total herbicide application by 20% in the watersheds there was a 44-50% atrazine and 30-50% metolachlor NPS reduction observed when using filter strips.
This study is providing information to validate the EPIC water quality farming system model in cooperation with the USDA-ARS, Grassland, Soil and Water Research Laboratory at Temple, Texas.
| Hoffman, D. W., T. J. Gerik, and C. W. Richardson. 1995. Use of contour strip cropping as a best management practice to reduce atrazine contamination of surface water. IAWQ, DIFuse POL’95, 595-96 |
July 13, 1998
Dennis Hoffman, Project Leader
Steve Dagitz, Webmaster