Effect of Formulation and Tillage Practice on Volatilization of Atrazine and Alachlor

Conservation tillage practices are being implemented by many farmers to conserve water and soil resources. These practices may modify the soil surface in ways that differentially effect dissipation of pesticide when compared to conventionally tilled fields. We measured volatilization of atrazine [2-chloro-4-ethylamino-6-isopropylamino-s-triazine] and alachlor [2-chloro-2'-6'-diethyl-n-(methoxymethyl) acetanilide] applied as either an experimental starch-encapsulated formulation or as a commercial formulation, containing atrazine as a wettable powder and microencapsulated alachlor, to adjacent no-till and conventionally tilled corn fields in Maryland. Both formulations were applied at the same rate; 1.7 kg ha' for atrazine and 2.8 kg ha~' for alachlor. After 35 d, cumulative volatilization of alachlor from conventionally tilled fields was 14% of that applied for both formulations. Cumulative volatilization of alachlor was less from no-till fields with 9% of the commercial formulation and 4% of the starch-encapsulated formulation being lost. After 35 d, cumulative volatilization of the commercial formulation of atrazine from the conventionally tilled field was 9% of that applied compared with 4% of that applied to the no-till field. Starch encapsulation reduced volatilizaUSDA-ARS, Hydrology Lab., Natural Resources Institute, Beltsville, MD 20705. Received 1 June 1993. "Corresponding author. Published in J. Environ. Qual. 23:292-298 (1994). tion losses of atrazine to <2% of that applied for both tillage practices. Starch encapsulation appears to be a viable formulation modification for reducing volatilization losses of herbicides, especially from no-till fields. I AN EFFORT to conserve soil and water resources, conservation tillage practices are being implemented on a wide scale (Christensen and Norris, 1983). No-till is the most extreme form of conservation tillage in that the soil is never tilled and plant residue is allowed to accumulated on the soil surface (Soil Conservation Society of America, 1982). While no-till conserves water and soil, larger inputs of pesticides are often necessary to maintain favorable crop yields (Christensen and Norris, 1983). In addition, the environmental fete of a pesticide applied to a no-till field may be very different than that of a pesticide applied to a conventionally tilled field (Glotfelty, 1987; Helling, 1987). Volatilization is one pesticide dissipation pathway that may be affected by no-till practices. Pesticides volatilization from soil is controlled by: chemical properties of the pesticide (vapor pressure and solubility), mode of application (surface application vs. incorporated, and formulaWIENHOLD & GISH: FORMULATION AND TILLAGE EFFECT ON VOLATILIZATION 293 tion), and soil properties (soil temperature, soil moisture distribution, and soil organic matter content) (Gloffelty and Schomburg, 1989). With the exception of chemical properties of the pesticide, all of these factors may be quite different when comparisons between no-till fields are made with conventionally tilled fields (Glotfelty, 1987). Since no-till does not allow for any tillage, unless the pesticide is injected into the seed row, pesticides must be surface-applied. A large portion of surface-applied pesticides may be intercepted by the plant residue in no-till fields (Ghadiri et al., 1984). This crop residue will have a much larger surface area and will be rougher than a bare soil surface, resulting in greater volatilization losses (Glotfelty and Schomburg, 1989). Granular formulations and that portion of liquid formulations that penetrates through the plant residue to the underlying soil during application or that is washed from plant residue by precipitation after application will be in a very different microenvironment than formulations applied to conventionally tilled soils. Surface soil under no-till crop residue is usually moister, cooler, and has a greater organic matter content than surface soil under conventional tillage (Thomas and Frye, 1984). Pesticide volatilization is greater from moist soil surfaces than from dry soil surfaces and will be less from cool soil surfaces than from warm soil surfaces (Spencer et al., 1973). Increased soil organic matter content increases pesticide adsorption and decreases volatilization (Spencer and Cliath, 1974). Since surface soil conditions under notillage are very different than those under conventional tillage and these conditions affect pesticide volatilization differentially, the net effect of no-till on pesticide volatilization is largely unknown. Few field studies assessing tillage effects on volatilization losses of pesticides have been conducted (Whang et al., 1993). Volatilization losses of agriculturally applied chemicals can be measured by a number of methods. Disappearance methods involve determination of the mass of chemical lost from soil samples collected over some time interval. Unless sampling is extensive, the uncertainty associated with this method can be large when more than one dissipation pathway exists and when spatial variability at the study site is large. Micrometeorological methods include several techniques in which chemical concentration, windspeed, and temperature gradients are simultaneously measured and used to calculate the chemical vapor flux from the field (Parmele et al., 1972; Harper, 1988; Glotfelty and Schomburg, 1989; Taylor and Spencer, 1990). A potential problem with these methods is the large area needed for each study site (60 to 200 m in diam. plus buffer strips). Soils and meteorological conditions may be quite different at sites separated by these distances; these factors may become confounding variables when comparisons among treatments are being made. Enclosure methods are a third approach for measuring volatilization (Harper, 1988). Enclosures do not require large field areas and are relatively simple; however, care must be used to ensure that modification of soil surface conditions is minimized by the presence of the enclosures. Increased awareness that agriculturally applied chemicals are potential sources for environmental contamination has encourage development of new formulations that may modify the behavior of pesticides. Encapsulation procedures are a formulation modification which modify pesticide behavior. Microencapsulation of chloropropham [isopropyl m-chlorocarbanilate] in nylon capsules reduced volatilization losses five-fold when compared with chloropropham applied as emulsified concentrate (Turner et al., 1978). A method that may modify pesticide behavior and is receiving increased attention involves encapsulating the chemical in a starch matrix (Wing et al., 1987). Schreiber et al. (1987) suggested that starch encapsulation should reduce volatilization losses of herbicides by controlling the rate at which the chemical is released into the soil environment. A greenhouse study compared volatilization losses from moist soils at three temperatures and found that starch encapsulation reduced volatilization losses of atrazine but increased or did not affect volatilization of alachlor (Wienhold et al., 1993). The effect of starch encapsulation on volatilization in the field, where temperatures fluctuate and soil surface moisture conditions vary, has not been studied. We used chambers, similar in design to the agroecosystem chambers of Nash et al. (1977), to measure cumulative volatilization losses of two commonly used agricultural herbicides, atrazine and alachlor. Herbicides were applied as either an experimental starch-encapsulated formulation or as a commercial formulation to adjacent notill and conventionally tilled fields. MATERIALS AND METHODS Volatilization was measured from four 0.25-ha fields, two notill and two conventionally tilled, on the Central Maryland Research and Education Center near Upper Marlboro, MD. The Monmouth sandy loam (clayey, mixed, mesic Typic Hapludul0 present at the site has a pH of 6.4, an organic matter content of 1.1%, and a clay content of 5.6%. Since 1989, all four fields have been planted in corn (Zea mays L.) with a winter cover crop of rye (Secale cereale L.). Corn residue was left on all the fields during the winter; rye was cut in the spring, 1 to 2 wk prior to planting. Crop residues were then incorporated into the soil with a chisel plow on the conventionally tilled fields but remained on the surface of the no-tillage treatment. There is >50% residue cover on the no-till fields (Walter J. Rawls, 1992, personal communication). Atrazine and alachlor was applied as either commercial formulation (Bullet, Monsanto Co., St. Louis, MO 1) or starchencapsulated (Carr et al., 1991). The commercial formulation contained atrazine as a wettable powder and microencapsulated alachior. Starch-encapsulated atrazine contained 11.1% and alachlor contained 10.1% a.i. Starch granules 0.4 to 1.2 mm in diam. were used. Both formulations were applied at the same rate; 1.7 kg ha-1 for atrazine and 2.8 kg ha-1 for alachlor. The starchencapsulated herbicides were broadcast onto the soil surface of two fields, one of each tillage practice. The commercial formulation was sprayed onto the soil surface of the other two fields. Volatilization of atrazine and alachior was measured using 0.25 m3 acrylic chambers that sampled the atmosphere above 0.5 m 2 of each field (Fig. 1). These chambers were designed and constructed to specifications similar to the agroecosystem chambers of Nash et al. (1977), which have been used extensively to evaluI Trade names or company names are included for the benefit of the reader and imply no endorsement or preferential treatment of the product listed by the U.S. Department of Agriculture. 294 I. ENVIRON. QUAL., VOL. 23, MARCH-APRIL 1994 Chamber~ Polyurethane / ~ 1. Schematic drawing of chambers used to measure herbicide volatilization in the field. ated the volatilization behavior of a number of pesticides (Nash, 1983a,b,c; Nash and Gish, 1989; Wienhold et al., 1993). The open-bottom chambers, one per field, were pressed into the soil =10 cm) and air was drawn through each chamber using a vacuum cleaner (Dayton Electric Co., Chicago, IL) attached to a manifold at the exit end of each chamber (Fig. 1). Air was drawn through each chamber at a rate of 31.5 + 0.9 L rain -1 (5.25 L rain-1 plug-~). This flow rate resulted in complete exchange of air within the chamber every 8 min and corresponds to a wind speed of 8 m h-L Air entered the chamber through six evenly-spaced holes (5 cm diam.) present on the front (0.5 by 0.5 m) wall, passed the length of the chamber (1.0 m), exited through six evenly-spaced holes present on the back wall of the chamber. Each air entry and exit hole contained a polyurethane foam plug (5 cm by 5 cm dia.) to trap herbicide present in the vapor phase (Turner and Glotfelty, 1977). Polyurethane foam plugs quantitatively trap pesticides from up to 6.3 × 10~ L of air (Turner and Glotfelty, 1977). Polyurethane foam plugs in entrance holes removed herbicide present in the incoming air. Polyurethane foam plugs in exit holes trapped herbicide which volatilized from the field under the chamber. Chambers were moved often (every 1 to 3 d) to insure that soil surface conditions inside the chambers were representative of those in the field being sampled. Special care was taken to move the chambers as soon after precipitation events as possible. A new location was selected each time the chamber was moved so that no part of the field was repeatedly sampled. Sampling was initiated within 5 min. of herbicide application on June 3, 1992. Polyurethane foam plugs were replaced 1, 2, 5, 8, 12, 16, 21, 28, and 35 d after herbicide application. Polyurethane foam plugs were soxlet-extracted with 150 mL of ethyl acetate for 3 h. The extract was then evaporated to dryness and redissolved in 10 mL of ethyl acetate. Concentrations of atrazinc, and alachlor were quantified using gas chromatography. The effect of the chambers on soil surface conditions was assessed by comparing soil surface temperature and water content for a plot under the chamber to that of an adjacent plot over a 3-d period. Soil temperature was determined by inserting a temperature probe (Type NP penetration probe, OMEGA Engineering, Stamford, CT) into the 0to 3-cm soil layer at two locations in each plot. Soil moisture was determined gravimetrically (Gardner, 1986) by collecting three surface samples (0-3 era) from each plot at five times over the 3-d period. Five surface (0 to 5 cm) soil samples were collected from each field 15 rain after herbicide application and again 8 and 35 d after herbicide application. Soil samples were collected by pressing a soil can (sample area of 38.5 cm2) 5 cm into the soil and removing the soil. Soil samples were pretreated with a phosphate buffer solution containing amylase (Wienhold and Gish, 1991) to facilitate release of the herbicides from starch granules. Following this pretreatment sufficient methanol was added to give a final ratio of methanol/water of 4:1 by volume. Samples were placed on a wrist action shaker for 1 h and suction filtered through glass fiber filter paper. Methanol was removed from the filtrate by rotoevaporation and atrazine and alachlor were isolated from the remaining aqueous olution by solid-phase extraction (Nash, 1990). Concentrations of atrazine and alachlor were quantified using gas chromatography. Operating conditions of the gas chromatograph were: 30 m by 0.32 mm fused silica capillary column coated with 0.26 Ixm SPB-5 (Supelco, Bellefonte, PA); injector temperature of 200 °C, oven temperature of 150 °C and a N-P detector operating at a temperature of 220°C; He carrier gas at 2.5 mL min-1. Trifluralin [ot,~t,~t-trifluro-2,6-dinitro-N,N-dipropyl-p-tolulcYme] was used as an internal standard. RESULTS AND DISCUSSION Air temperature ranged from 7 to 32 °C with an average daily high of 27 °C and an average daily low of 15 °C (Fig. 2). The study site received measurable precipitation on 13 d of the 35 d study, the first rain events occurred 2 and 3 d after herbicide application. Cumulative rainfall over the 35 d study measured 10.6 cm (Fig. 2). Wind speed averaged 350 m h-~ during the 35 d study; however, this average was strongly influenced by several windy days toward the end of the study. Average wind speed during the first 20 d of the study was 7 m h -~. Differences between soil surface temperature inside a chamber of those of an adjacent plot were never more than 1.5 °C (Fig. 3A). Soil surface temperatures inside the chamber were consistently the same or cooler than those outside the chamber. Changes in soil surface water content inside the chamber were similar to those outside the chamber (Fig. 3B). During Day 1, 43 g water kg soil -~ was lost from the surface soil layer inside the chamber compared with 49 g water kg soil-2 from the surface soil layer outside the chamber. Precipitation fell on the site during the evening of Day 1 and the chambers were moved to a new plot. During Days 2 and 3, 39 g water kg soil -~ were lost from the surface soil layer inside the chamber compared with 38 g water kg soil -~ from the surface layer outside the chamber. These results suggest that air flow through the chamber was sufficient to prevent heat buildup within the chamber and that evaporative losses of water from the soil were not greatly modified by the chamber. Variation in the mass of herbicide trapped by the polyo~30 t . Daily Maximum Temperature 9 / I I Daily Minimum Temperature /,I I. I. .. I,I o 0 10 20 30 TIME (days after June 3, 1992) Fig. 2. Daily maximum and minimum temperature, and daily precipitation received at the Central Maryland Research and Education Center near Upper Marlboro, MD. WIENHOLD & GISH: FORMULATION AND TILLAGE EFFECT ON VOLATILIZATION 295