| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
G.W. Andrews Forestry Sciences Laboratory, 520 DeVall Drive, Auburn, AL 36849
* Corresponding author (michajl{at}auburn.edu)
Received for publication May 21, 2002.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Abbreviations: DAT, days after treatment • SM, sulfometuron
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Use of herbicides in forest management is already associated with increased public concerns over human health and safety, and forest and aquatic ecosystem impacts. Any additional uses will intensify those concerns, and the public demands proof of safety. However, it is not possible under tenets of science to "prove" anything. In the face of this conundrum, How can we be sure of the safety of technology when our science is unable to prove the safety?, we must find alternative methods of making decisions about new scientific developments and the role they will play in society. Currently, the most widely accepted alternative is risk assessment. Risk assessment recognizes and incorporates uncertainty into the process of evaluation and allows for the management of uncertainty in decision making. Risk assessment utilizes a tiered approach in which each tier may generate additional researchable questions and point to potential problems. At the conclusion of each tier more research may be required and more information provided. Risk assessments require toxicological data and knowledge of the fate of herbicides in the environment.
The active ingredient of the herbicide Oust is sulfometuron (SM). Sulfometuron is a member of a new class of herbicides that inhibit acetolactate synthase (ALS, also referred to as AHAS, acetohydroxy acid synthase, two acceptable names for the same enzyme system). Sulfometuron is herbicidally active at very low concentrations, and is registered for use in forest sites at rates up to 420 g ha-1. Solubility of SM in water increases with increasing pH, but is relatively low at pHs typically observed in pine forests of the southern USA (10 mg L-1 at 25°C and pH 5.5; DuPont Chemical Company, 1983). Sulfometuron is stable to hydrolysis at pH 7.0 but unstable in acid aqueous solution with a hydrolytic half-life of approximately 19.8 d at pH 5 and 4.2 d at pH 2 (temperature 25°C; DuPont Chemical Company, 1983). Under acidic conditions, SM hydrolyzes to 1,2-benzisothiozol-3-one, 2,3-dihydro 1,1-dioxide (saccharin), methyl 2-(aminosulfonyl)benzoate, and CO2 (Harvey et al., 1985; Anderson and Dulka, 1985; Brown, 1990). Degradation is enhanced in the presence of ultraviolet light and similar products are formed. Aqueous photolysis is rapid with an 80 to 90% decrease in SM at the end of two weeks, even at pH 8.6 (Harvey et al., 1985). Sulfometuron, with its low solubility, low vapor pressure (5.5 x 10-16 mm Hg at 25°C; Beyer et al., 1987), and low partition coefficient (Kd = 0.29, Dickens and Wehtje, 1986; Kow = 0.31, Harvey et al., 1985), would be available for transport in storm runoff unless its movement from or degradation in the upper few centimeters of soil is rapid. The chemistry of sulfonyl urea herbicides has been reviewed by Beyer et al. (1987), Brown (1990), and Hay (1990). Breakdown of SM through metabolism and by chemical degradative pathways significantly decreases SM persistence in the environment. Sulfometuron persistence in the environment is summarized in Table 1. Breakdown also occurs in soils with a more complex mixture of breakdown products formed, which include those observed in hydrolysis and photolysis (Hay, 1990). The instability of SM in acid aqueous solution may reduce its potential to move off-site and downward through the soil profile in the acid soils of southern U.S. pine forests.
|
Michael and Neary (1993) summarized the state of knowledge of offsite movement of forestry herbicides. The fate of sulfonylurea herbicides has been reviewed (Blair and Martin, 1988). Neary and Michael (1989) monitored a 4-ha watershed on deep sandy soil in Florida treated with SM at the rate of 0.42 kg a.i. ha-1, and found offsite movement for only 7 DAT. The maximum concentration observed in water (7 µg L-1) followed 54 mm of precipitation that began 1 DAT and continued to 3 DAT. There are no other published data concerning offsite movement of SM in forest ecosystems. A study with metsulfuron (another sulfonylurea acetolactate synthase inhibitor with similar chemical structure) found a maximum of 8 µg L-1 in two surface water samples collected 9 DAT following an application of 84 g ha-1 of metsulfuron applied as the herbicide Escort (DuPont Chemical Company, Wilmington, DE). Metsulfuron was not detected in 229 additional surface water samples collected during the two months following application to this poorly drained loamy sand in the Florida flatwoods near Gainesville, FL (Michael et al., 1991).
A few studies have reported offsite movement of SM from agricultural applications. Hubbard et al. (1989) reported that offsite movement of SM exceeded 1000 µg L-1 when plots treated with 0.6 kg a.i. ha-1 received extremely high-intensity precipitation (125 mm h-1). GLEAMS (Leonard et al., 1987) model predictions did not compare well with actual observations, because GLEAMS is a long-term simulation model, not an event-based model, and is therefore not capable of predicting single-event concentrations of offsite movement (Hubbard et al., 1989). Wauchope et al. (1990) reported SM concentrations in runoff following application of 0.4 kg ha-1 to agricultural sites. Simulated rainfall (69 mm h-1, total < 30 mm) within 24 h of herbicide application produced runoff concentrations up to 700 µg L-1. They also reported that grass-covered plots yielded the same total offsite movement as observed from bareground plots, indicating that most offsite movement occurred in overland flow as a result of the high-intensity rainfall simulation. Concentrations of SM in runoff from the agricultural applications and simulated rainfall of Hubbard et al. (1989) and Wauchope et al. (1990) are 142- and 100-fold (respectively) greater than values reported by Neary and Michael (1989) for natural rainfall on treated forest sites.
Adverse environmental impacts are a function of herbicide concentration, exposure time, and species susceptibility. Available acute toxicology data for SM (rat [Rattus norvegicus], mallard duck [Anas platyrhynchos], and bobwhite quail [Colinus virginianus] LD50 > 5 g kg-1; bluegill sunfish [Lepomis macrochirus] and rainbow trout [Oncorhynchus mykiss] 96-h LC50 > 12.5 mg L-1, where LD50 and LC50 are the dose or concentration that causes the death of 50% of a group of test subjects; Beyer et al., 1987) indicates it is almost nontoxic to mammals, birds, and fish, and there is no evidence for mutagenic or teratogenic properties (Beyer et al., 1987; DuPont Chemical Company, 1983). No adverse effects were observed on embryo hatch or larval survival and growth of fathead minnows (Pimephales promelas) from long-term exposure to 1.2 mg L-1 (DuPont Chemical Company, 1983). The impacts of SM on some aquatic plant species have also been studied. Growth of Cladophora spp. was not reduced by exposure to 5 mg L-1 of SM (USDA Agricultural Research Service, 1982). Hydrilla [Hydrilla verticillata (L. f.) Royle], a submerged aquatic plant, was not affected by exposure to 50 µg L-1 of SM, but exposure to 100 µg L-1 resulted in a significant reduction in growth (USDA Agricultural Research Service, 1982). Byl et al. (1993) found that hydrilla exposed for 5 d to 10 µg L-1 of SM exhibited significant increases in peroxidase activity and decreased growth, but these changes were not significant after 3 d of continuous exposure. Hartnett et al. (1987) found wild-type green alga Chlamydomonas reinhardtii completely resistant to 144 µg L-1 SM, but growth inhibition was observed at 364 µg L-1 following two weeks of continuous exposure. Wild-type chronically exposed to 364 µg L-1 developed resistant strains that were unaffected by up to 36.44 mg L-1. Clearly, susceptibility of aquatic species varies just as it does for terrestrial species, but because of generally shorter generation times than exist for terrestrial plants, recovery options exist for some aquatic species if environmental exposures reach the extremely high levels tested in laboratories. Rooted or floating aquatic macrophytes should be more sensitive to SM than other aquatic organisms due to the similarity of their metabolism to terrestrial plants. Growth of Eurasian water-milfoil (Myriophyllum spicatum L.) was reduced at water concentrations of 1 µg L-1 (USDA Agricultural Research Service, 1982).
There is little information available on environmental fate and aquatic ecosystem contamination from SM use in forest management and this lack of information hinders adequate risk analysis. This study was conducted to determine the environmental fate of SM and aquatic organism exposure on two forestry sites in Mississippi following operational application.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Soils for all three watersheds (Table 2) are predominately Wilcox silty clay loam to clay loam (very-fine, smectitic, thermic Chromic Dystrudert), acidic, and somewhat poorly drained.
|
Treatments
The watersheds were mechanically site-prepared by shearing, windrowing, discing, and bedding across the slope in June 1984 followed by planting of loblolly pine in January and treatment with SM in April 1985 following an operational approach to pine regeneration. The disking and bedding are accomplished long before planting to permit adequate time for settling of the soil in the beds. Sulfometuron is normally used on these sites as a preemergence to early postemergence treatment for the control of herbaceous weeds, but application rates typically range from 0.075 to 0.26 kg a.i. ha-1.
One watershed was treated with Oust, a commercial formulation of dispersible granules containing 75% SM (sprayed). The sprayed watershed is 253 ha, of which only the lower 148 ha (58%) was sprayed. The second watershed is 247 ha, of which the lower 44 ha were treated with an experimental pelleted formulation containing 0.9% SM (henceforth called the pelleted formulation). The target treatment rate for both SM formulations was 0.42 kg a.i. ha-1. The control site is a smaller watershed in which 33 ha had a cultural history similar to the two treated watersheds.
Sulfometuron was applied aerially over the sprayed site early postemergence at the target rate of 0.42 kg a.i. ha-1 (9–10 Apr. 1985). The spray was applied without surfactant in a total spray volume (water) of 112 L ha-1 from a Hughes H500D helicopter (McDonnell-Douglas Corp., St. Louis, MO) fitted with a 9-m Through-Valve boom. The pelleted formulation was applied at 0.37 kg a.i. ha-1 (18 Apr. 1985) by a Hughes H500D helicopter equipped with an Isolair Series 2600 applicator–spreader (Bell Helicopter Textron, Hurst, TX). A 15-m untreated streamside management zone (SMZ) was maintained on each side of the streams throughout their lengths where water was visible at the time of SMZ establishment.
Sample Collection
Samples of water, sediment, plants, litter, and soil were collected from both watersheds, transported to the laboratory, and frozen until analyzed. Samples collected from the untreated control watershed were used mainly in analytical methods development, freezer storage stability studies, and for quality control. The sample collection schedule for all matrices was approximately -1, 0, 1, 3, 7, 14, 30, 45, 60, 90, 120, and 180 DAT. Water was intensively sampled by prearranged timed sequence and stream flow according to precipitation events. Phosphate buffer sufficient to buffer water samples to pH 7.0 was added to each water bottle before samples were collected to prevent hydrolysis of SM.
Water
The watersheds were instrumented at 0.5-m rectangular gauging weirs with pressure transducers for water level recording and automatic water samplers capable of collecting discrete volume samples up to 1 L. Water samples were also collected by grab sampling. Time sequence samples were collected every 6 h (500 mL) and composited by the autosampler into 12-h samples for the first 37 DAT. Another automatic sampler was set to initiate sampling with an increase in stream stage without compositing. Grab samples were taken 3, 5, 9, 15, 21, 27, 35, 43, 50, 63, 92, and 125 DAT at 0, 30, 60, and 150 m below the gauging weirs. All water samples were immediately frozen and maintained frozen until thawed for analysis.
Sediment
Sediment samples were separated from water samples by decantation after the water samples had been thawed for analysis. The decanted water sample was filtered and the residue combined with the sediment sample. Suspended sediment samples were composited by week with bedload sediment collected weekly at the weir for total sediment analysis. Sediment samples were air dried, weighed, and frozen until analyzed.
Plants
Plant tissues were collected according to the predetermined sampling schedule (with minor variations due to weather, etc.) from grasses, pokeweed, blackberry, oak sprouts, and pine seedlings at 10 randomly located positions on each watershed and composited by species. The entire aboveground portion of each plant was collected when possible. Several branch tips from each of 10 plants were collected from species that were too large for collection of the entire plant to be feasible. All plant sampling was discontinued on the sprayed watershed with the 27 DAT sample and on the pelleted watershed at 60 DAT, because plants remaining on the sites were inadequate to provide sufficient material for analysis.
Litter
Litter samples were collected from three 900-cm2 areas randomly selected at ridge, midslope, and toe-slope positions. These samples were composited by slope position and frozen until analyzed.
Soil
Soil samples were collected in 46-cm-deep (18-in) cores from bare ground and from under litter at three slope positions (upper, middle, and lower) by driving 5-cm-i.d., Schedule 40 PVC pipe into the ground and extracting the core. Three cores were collected from each slope location. All cores were frozen until analyzed. Immediately before analysis, the frozen cores were cut into 15-cm sections representing depth from the soil surface. Samples were composited by depth for each soil cover condition, slope position, and sampling date. Composited samples were subsampled for SM analysis with a 14-channel proportional splitter, and a subsample was taken for oven-dry weight determinations.
Analytical Methods
Sulfometuron was determined by high performance liquid chromatography in water by the method described by Wells and Michael (1987), in plants and litter by the method of Zahnow (1985a), and in sediment and soil by the method of Zahnow (1985b). Limits of quantitation (LOQ, a signal to noise ratio of 5) were 1 µg L-1 for water, 50 µg kg-1 for plants and litter, and 20 µg kg-1 for soil and sediment. Limits of detection (LOD, a signal to noise ratio of 2) were established at 0.2 µg L-1 for water, 10 µg kg-1 for plants and litter, and 2 µg kg-1 for soil. All values falling between the LOQ and LOD were identified as trace quantities. Average values were computed substituting half of the LOD for all samples containing trace or nondetectable amounts. Quality control (QC) samples consisted of blank and fortified samples from appropriate matrices. Materials for these samples were collected from the control watershed or as pretreatment samples from the treated watershed. Samples for QC were interspersed among actual samples for extraction and analysis. Analytical standards were included as every fourth sample during routine analysis.
Freezer storage stability studies were conducted to determine whether any degradation occurred during sample storage. Samples of water, soil, and plants collected from the study areas were fortified with analytical standard (water was buffered to pH 7), frozen simultaneously with the field samples, and analyzed periodically.
Sulfometuron dissipation was evaluated on the basis of half-life. Half-life (t1/2) was calculated from the slope of the best fit line:
 |
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
Sulfometuron Dissipation and Half-Life
Sulfometuron residues were higher on all plant tissue, litter, and surface soil samples from the sprayed watershed than observed in the pellet-treated watershed during the first few days following application (Table 4). Residue concentrations increased significantly in plant tissues and soil under litter on the watershed treated with the pellet formulation after sufficient precipitation had fallen (5 DAT) to activate the pellet. Following the accumulation of peak concentrations, SM dissipated rapidly from all matrices. Interception alone by vegetation, litter, and soil frequently does not result in the highest concentrations observed in these matrices, especially on sprayed sites. Additional uptake or mass transfers among the matrices may result in higher concentrations during the first several days following application. Mass transfer, presumably from mechanical causes, was observed in this study between 0 and 1 DAT. Significant loses of SM from sprayed plant tissues and litter are evident (Table 4) in the absence of any precipitation. This disappearance cannot be explained by photolysis, which occurs only in the aqueous state. A trace amount of precipitation (3 mm, Table 5) occurred 2 DAT and plant residue concentrations decreased by approximately 50% during this time. This is in agreement with the findings of Michael et al. (1992), who reported that up to 100% of applied herbicide could be washed from leaf surfaces by just 3 mm of precipitation 1 h after application and approximately 30% when the 3 mm fell within 48 h of application.
|
|
Sulfometuron half-life was calculated for bare-ground soil (33 d), litter (9 d), pine foliage (6 d), and composited plants (4 d) from the sprayed watershed. The very short half-lives for plant and litter reflect the inordinately large impact of high values on regression analysis, and mechanical transfers and washoff of large portions of the applied herbicide just a few days after application. The half-life in soil was affected by these same conditions, but the impact was to attenuate the half-life because the soil was the ultimate receptor of herbicide transfers due to mechanical effects and washoff.
Calculation of half-life normally starts with day of application, but in the case of soil in this study, the maximum concentration did not occur until 7 DAT. Using all data from bare soil on the sprayed site in this study, soil half-life was determined to be 33 d. However, SM was not washed from foliage and into the soil until the first storm event 5 DAT. Using the residue data for samples collected after this precipitation event resulted in a half-life of 14 d. The latter half-life is very close to the hydrolytic half-life for pH 5.0 (19.8 d at 25°C; DuPont Chemical Company, 1983). Hydrolysis was probably the most important route of dissipation in this study.
Plant Residues
Sulfometuron was detected in composited plant tissue samples with a minimum detection limit of 0.050 mg kg-1. Dissipation of herbicide on competing vegetation occurred rapidly. Initial residue concentrations for all species averaged about 24 mg kg-1. The highest SM concentrations were observed on sprayed plants with the largest horizontal leaf intercept area (Table 6). By 27 DAT, SM had dissipated to near nondetectable levels for most sprayed species, but continued to increase in vegetation on the pellet-treated site. Loblolly pine still contained low levels of herbicide residue that may represent additional uptake from the soil or storage at inactive sites. Plants from pellet-treated sites showed little uptake from soil. Any plants that had emerged at the time of treatment and perennial species had small amounts of SM as the result of root uptake; however, species that germinated after treatment were killed and therefore did not emerge to be collected as plant samples.
|
Sulfometuron residues in mineral soil samples under litter peaked 14 DAT at a concentration of 0.397 mg kg-1 and dissipated to nondetectable levels by 60 DAT. The peak concentration measured 14 DAT corresponds to a precipitation event occurring 13 DAT. During this event additional SM was transferred from the litter layer to soil beneath. As with the bare-ground sample, SM residues were detected in soil at a 15- to 30-cm depth, but not in samples below 30 cm at any time during the study. Residue concentrations were highly variable, and the dissipation attenuated, presumably by the litter layer, which held a reservoir of SM for release through time into the upper level of soil. Because of the impact of the litter layer, half-life was not calculated for these data.
Precipitation and Streamflow Residue ConcentratIons
The first 40 d after treatment contained five storm events on the sprayed watershed and four storm events on the pelleted watershed (Table 5) that individually yielded at least 12 mm of precipitation. Because of the poor drainage on these sites, each storm event resulted in a sluggish rise and relatively rapid fall of stream stage, but base flow continued for several days at a very low rate. After 40 DAT streamflow became discontinuous and sampling was restricted to isolated pools on the treated watersheds.
A total of 200 water samples were collected and analyzed for SM from streamflow on the sprayed watershed and 175 on the pelleted watershed. Water from the pelleted watershed contained SM more frequently and at higher concentrations than those of the sprayed watershed (Fig. 1) . Only 12.5% of water samples from the sprayed watershed contained quantifiable residues of SM compared with 70.9% of the samples from the pelleted watershed. The remaining samples contained either trace amounts or SM was not detectable in them. The pelleted formulation was an experimental formulation that has been dropped from consideration for registration.
|
Time sequence samples were collected every 6 h and composited to 12-h samples for the first 37 DAT. I calculated 24-h average concentrations (Fig. 1) by averaging the two 12-h composites. While the preponderance of samples with nondetectable residues of SM occurred on the sprayed watershed after the third major storm event, a few occurred just after application. For example, the 24-h average residue level of 12.3 µg L-1 for 1 DAT (Fig. 1) is the result of averaging the 24.6 found in the composite sample collected at midnight and 0600 h with the nondetectable residue in the composited sample collected at 1200 and 1800 h.
Flow sequence samples were collected periodically as a function of stream stage, but not more frequently than every 15 min and were not composited. While the occurrence of SM residues was correlated with storm events, peak concentrations occurred on the pelleted watershed more than 3 h after peak discharge when antecedent soil moisture was low and more than 1 h after peak discharge when antecedent soil moisture was high (Fig. 2) . Michael et al. (1999) reported water samples collected from watersheds aerially sprayed with hexazinone had peak herbicide concentrations at peak discharge, probably due to overland flow conditions. Pelleted hexazinone applications resulted in peak herbicide concentrations several hours after peak discharge. This latter finding is attributable to the fact that the pelleted form requires rain to break down the pellet and during this process, most of the herbicide percolates into the soil and is moved via subsurface flow. Because of the small number of samples that contained SM from the sprayed watershed, it is impossible to determine whether peak discharge occurred simultaneously with peak SM concentrations.
|
The presentation of herbicide residues in runoff in a line graph of positive and measurable concentrations by previous authors, even when sampling periods varied from 7 to 11 d, has contributed to development of the concept that stream contamination is a continuous phenomenon. Modelers also contribute to this concept by displaying their predictions in line graph form. Perhaps the most serious aspect of this continuum concept of stream contamination is the resultant testing of herbicides in 24-, 48-, or 96-h static tests for toxic impacts on aquatic organisms. Toxicological data derived from such tests are almost impossible to interpret in light of a pattern of discrete and discontinuous streamflow contamination that results in short-term exposures for organisms within the affected habitat.
Sediment Residues
A total of five sediment samples from 3, 4, 5, 6, and 7 weeks after treatment were analyzed for SM. Suspended sediment samples did not contain SM residues at the minimum detection limit of 0.02 mg kg-1. Lack of SM residues associated with sediment is a result of its low Kd and indicates that movement was principally in the dissolved state.
Environmental Consequences
The use of herbicides in forest management has resulted in concern over the nontarget impacts that may result from either their movement through the treated site, or from movement off-site. Two issues of special concern are the potential for impacts on aquatic ecosystems, and the potential for impacts on microbial populations encountered during movement through the soil column.
Herbicides are specifically designed to affect plant growth processes such as cell division, photosynthesis, and respiration. Concentrations of SM, which impact alga species like Cladophora, are 62 and 160 times higher than observed on the pelleted and sprayed watersheds, respectively (USDA Agricultural Research Service, 1982). Other aquatic species may be more susceptible. Growth of hydrilla is reduced by prolonged exposure to 100 µg L-1 SM, but not by prolonged exposure to 50 µg L-1, while Eurasian water-milfoil, an invasive submerged aquatic plant with adverse impacts on U.S. and Canadian aquatic ecosystems, exposed to 1 µg L-1 suffers reduced growth.
Herbicides can also affect nonplant organisms when concentrations are extremely high. For example, SM can kill fish when water concentrations exceed 12.5 mg L-1 for 96 h continuously, but at lower concentrations has no observable impact on fish. Fathead minnow embryo hatch and larval survival and growth are not affected by concentrations of 1.2 mg L-1 for extended periods of time. The maximum observed SM concentrations in this intensive monitoring study were 0.089 mg L-1, a component of the 24-h average for 9 DAT on the pelleted watershed and 0.034 mg L-1, a component of the 24-h average 5 DAT on the sprayed watershed (Fig. 1). These concentrations, which lasted for 12 h, are respectively 140 and 360 times lower than the concentrations known to kill bluegill fish, and 13 and 34 times lower than the concentrations known to have no effect on embryo hatch or larval survival and growth for fathead minnows.
Catastrophic events caused by fire (Rinne 1996), flooding (Gray and Fisher, 1981; Minckley and Meffe, 1987), drought (Deacon and Metcalf, 1961), and grazing (Rinne, 1990), which decimate fish and macroinvertebrate populations resulting in complete extirpation, may occur frequently in headwater streams. With but a few exceptions, aquatic ecosystems in general and fish populations in particular have a remarkable ability to recover following such catastrophes provided habitat is not damaged or altered (Olmsted and Cloutman, 1974). Peterson and Bayley (1993) simulated a catastrophic fish kill by poisoning fish in 45- to 113-m-long sections of 18 streams with rotenone. Recolonization began almost immediately, and recovery to 90% of original abundance occurred in four to 11 d, while recovery to 70% of original species composition required only 2.5 to 6 d. Olmsted and Cloutman (1974) reported on fish recolonization following a fish kill caused by pollution from an unknown pesticide. Complete extirpation of fish was observed over a 1.6-km reach of Mud Creek in Washington County, Arkansas on 17 May 1971. Recolonization of Mud Creek (normally 1 to 8 m in width with pools up to 1.5 m depth) began 6 d later with the migration of fish from downstream. Most species originally present had returned within three months, and repopulation was nearly complete after one year.
Macroinvertebrates are also remarkably resistant to catastrophic events. Gray and Fisher (1981) studied macroinvertebrate recolonization in Sycamore Creek following flooding and bottom scouring, which resulted in elimination of 80 to 90% of benthic invertebrates. Recolonization was principally by aerial pathways and nearly 66% of total taxa recolonized in nine weeks. Thus, recolonization can be very rapid, particularly when the affected area is relatively small. Given that recolonization can occur rapidly under such catastrophic conditions, it is reasonable to conclude that similar or more rapid rates of recolonization may occur under considerably lower levels of impact. Such is the case with SM in the watersheds in this study. Residue concentrations observed were of short duration and up to several orders of magnitude less than those shown to cause species mortality under conditions of long exposure. Thus, the potential for long-term adverse impacts on aquatic ecosystem functioning in this treated watershed is very low due to the low SM concentrations observed, and the documented ability of fishes and macroinvertebrates to rapidly recolonize following catastrophic extirpation. Neither are downstream impacts likely, because of the dilution of already low concentrations in streamflow. Sulfometuron was detected at trace levels (<1 µg L-1) or not at all 150 m downstream of the treated watersheds after 18 DAT on the pelleted watershed and 9 DAT on the sprayed watershed.
Clearly, it is difficult to assess the potential impacts of SM on all aquatic organisms from the available toxicological data and the levels of contamination observed in this study. Contributing factors to this difficulty include variability in species susceptibility, cumulative effects, short generation times potentially leading to development of resistance, and possible interactions with many other environmental elements. Hartnett et al. (1987) reported that Chlamydomonas reinhardtii was completely resistant to 0.144 mg L-1 SM but growth was inhibited at 0.36 mg L-1. Wild-type exposed to 0.36 mg L-1 developed resistant strains, which were unaffected by up to 36 mg L-1. Exposure to sublethal concentrations may, in some species, result in development of strains that are highly resistant.
Movement of SM through soil resulted in exposure of microbial organisms in the surface 15 cm to a maximum concentration of 0.397 mg kg-1. There is little direct evidence of the potential impact of such low concentrations on soil microorganisms. In the few cases studied, concentrations of SM greater than 1 mg kg-1 are usually required to produce any measurable impact on microbial populations, and those impacts are not always negative. Xing and Whitman (1987) studied 28 strains of methanogenic bacteria exposed to 200 mg L-1. Of these 28 strains, 7 (two species) were unaffected, while an additional 14 strains were slightly inhibited. The remaining 7 strains showed some growth inhibition. Van Dyk and LaRossa (1986) found several strains of Salmonella typhimurium resistant to 51 mg L-1. LaRossa and Smulski (1984) found Escherichia coli and Salmonella typhumurium resistant to sulfometuron methyl concentrations ranging from 0.36 mg L-1 to 36 mg L-1. Rank (1986) reported Saccharomyces cerevisiae resistant to concentrations up to 50 mg L-1. In these studies, exposures sufficient to cause some measurable impact were generally 90 to 500 times greater than observed in the sprayed watershed from the high application rate used in this study. This finding is in agreement with previous findings of herbicide interactions with soil microorganisms discussed in several reviews (Martin, 1963; Audus, 1964; Greaves et al., 1976) and more recently by Rhodes et al. (1980), Greaves et al. (1981), Repova (1985), and Tu (1993). Martin (1963) concluded that quantities of herbicides or insecticides applied to soils in agricultural situations are not sufficient to significantly depress growth of specific soil microbes. Audus (1964) also concluded that populations of microflora are not seriously damaged by herbicides, but that observations of numbers alone may not be sufficient to demonstrate impacts on specific elements in the microflora. Greaves et al. (1976) concluded that no herbicide had been found that had a long-term harmful effect on microbial activity at concentrations arising from normal agricultural usage. A survey of the literature since that time supports the position that soil microorganisms are generally resistant to the low-concentration short-term exposures, which occur from use of herbicides at labeled rates, but the relative degree of resistance is highly variable. Short-term impacts lasting a few days or weeks may be detected in some studies. Tu (1993) studied impacts of 10 herbicides on bacterial and fungal populations in soil treated with 10 mg kg-1. Populations of both groups of organisms either decreased or were not different from the control group during the week after treatment, but by two weeks after treatment, the only statistically significant differences among treatments were a few cases in which populations were greater than observed in the controls. Soil microbial respiration was significantly increased over controls for all herbicide treatments. There seems to be no evidence, direct or indirect, that herbicide use in forestry in general, or specifically SM use, adversely affects soil microbial populations to the extent that one could expect any impact on present or future soil productivity.
While there is much we do not know about SM, this research indicates that on forest sites like those in the southern USA, where soil and water conditions are acidic, sometimes extremely so, SM is short-lived. The half-lives calculated from this research are in line with those reported from Russia, Canada, and around the USA (Table 1). The principal routes of degradation appear to be hydrolysis and photolysis. However, where soil and water conditions are neutral or basic, rates of hydrolysis (but not photolysis) decrease and aqueous solubility increases (Hay, 1990). Under these conditions, longer half-lives and increased potential for off-site movement would probably be observed. Adverse impacts on watersheds in the southern USA are unlikely.
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Related articles in JEQ:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
