Published online 20 September 2006
Published in Soil Sci Soc Am J 70:1882-1888 (2006)
DOI: 10.2136/sssaj2005.0415
© 2006 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
Soil Chemistry
Trace Element Mobilization in Soils by Glyphosate
K. A. Barrett and
M.B. McBride*
Dep. of Crop and Soil Sciences, Cornell Univ., Ithaca, NY 14853
* Corresponding author (mbm7{at}cornell.edu)
 |
ABSTRACT
|
|---|
Glyphosate, the most widely used herbicide in agriculture, has a strong tendency to sorb on minerals by bonding with surface metals through its metal-coordinating functional groups. However, this same chemical process can potentially mobilize sorbed trace metals by chelation and sorbed anions such as phosphate by displacement. The tendency for glyphosate to mobilize Cu and other elements was tested in soil leaching experiments by applying glyphosate alone or complexed with Cu to mineral and organic soil columns and measuring the concentrations of these elements in the leachates by inductively coupled plasma (ICP) spectrophotometry. Complexation with glyphosate inhibited Cu sorption on the soil columns, rendering this metal more mobile in both the mineral and organic soil. Glyphosate mobility was also enhanced when applied to the organic soil as the Cu complex, but this effect was not detected with the mineral soil. Glyphosate was adsorbed more completely on the organic soil than the mineral soil, suggesting a metal bridging mechanism to explain sorption of this anionic molecule. The application of a commercial Roundup spray to long-contaminated soils containing elevated concentrations of heavy metals and phosphate resulted in significant increases in the leaching of Cu, Zn, Al, Ni, P, Si, and As. No significant increases in elemental leaching were detected in mineral and organic soils with normal background concentrations of heavy metals and P. Batch equilibration studies confirmed the glyphosate-enhanced dissolution of Cu in Cu-contaminated soils, but reductions of free Cu2+ activity were measured only in uncontaminated soils. The results indicate that several elements, particularly Cu, Al, and P, could be mobilized within the thin surface layer of soils receiving a high effective rate of glyphosate during herbicide application.
Abbreviations: AMPA, aminomethylphosphonate • ANOVA, analysis of variance • CBD, citrate–bicarbonate–dithionite • CEC, cation exchange capacity • ICP, inductively coupled plasma spectrometry
|
INTRODUCTION
|
|---|
A CONSIDERABLE AMOUNT OF research has explored the interaction of glyphosate with a range of soil minerals, demonstrating that this herbicide has a high affinity for soil surfaces, including layer silicate surfaces and Fe and Al oxides. However, most published investigations of glyphosate behavior in soil have employed batch adsorption designs (e.g., de Jonge et al., 2001; Morillo et al., 2002; Gimsing et al., 2004b). These studies may overestimate the sorptive capacity of a soil in the field, as batch equilibrations maximize the interaction of the herbicide with the high surface area of water-dispersed soil. In the field, solutes may move rapidly through the profile via macropores, having limited contact with the bulk of the soil. Therefore, the effective adsorptive surface area of soil in the environment may be much less than the actual surface area. Moreover, the disturbance of soil structure leading to disaggregation that occurs during batch studies is likely to artificially enhance glyphosate contact with newly exposed sorption sites.
An additional factor potentially increasing the mobility of glyphosate in soils is the fact that this molecule forms particularly stable complexes with Al3+, Fe3+, and Cu2+ (Motekaitis and Martell, 1985; McBride and Kung, 1989; McBride, 1991). The strong tendency of glyphosate (e.g., Sheals et al., 2001) and to a lesser extent its primary breakdown product aminomethylphosphonate (AMPA) to form chelates with certain metals (Song et al., 1994), may have important implications for the fate and transport of the herbicide in the environment. The potential for degradation (abiotic or biotic) and for adsorption, both critical processes implicated in the attenuation of glyphosate in soil, may be reduced by the formation of metal–glyphosate complexes.
Generally, the mobility of both metal cations and glyphosate is expected to be enhanced in soil by the formation of a complex with less charge (McBride, 1991; Morillo et al., 1997). However, metal–glyphosate interactions may also facilitate immobilization via co-adsorption processes, such as ternary complex formation (Sheals et al., 2003). Decreased mobility has also been suggested with the experimentally observed formation of sparingly soluble glyphosate precipitates of Cu2+, Ni2+, Ca2+, and Mg2+ (Subramaniam and Hoggard, 1988).
With the purpose of improving our understanding of glyphosate behavior in soil, a series of soil column leaching and batch equilibration experiments was undertaken to specifically determine the influence of glyphosate on the potential mobility of various elements in the soil surface layer intercepting applied glyphosate. This research addresses a deficiency in existing studies of glyphosate mobilization of metals in soils, which are mostly based on model mineral systems rather than soils.
|
MATERIALS AND METHODS
|
|---|
Soil Characterization
Several soils variably contaminated by Cu and other metals were collected in New York, Pennsylvania, and Ontario (Canada) and characterized for use in this study. These soils belong to the Hudson series (fine, illitic, mesic Glossaquic Hapludalfs), a silty clay loam sampled from an apple orchard (Hudson A is relatively uncontaminated, whereas Hudson B is highly metal-contaminated from sewage sludge application about 25 yr earlier); Hagerstown series (fine, mixed, semiactive, mesic Typic Hapludalf), a silt loam sampled from a Pennsylvania dairy farm; Carlisle series (Euic, mesic Typic Haplosaprist), an organic soil sampled from a vegetable-growing area of New York state; and Caledon series (coarse- loamy, mesic Ochreptic Hapludalf), uncontaminated and metal-contaminated loamy sands collected near Cambridge, ON, from a long-term sewage-sludge plot experiment, referred to here as Caledon A and B, respectively. Relevant soil properties are presented in Table 1. Soil pH was measured using a pH electrode immersed in a 2/1 distilled water–soil suspension. Organic matter was measured by the Walkley–Black dichromate oxidation method (Allison, 1976). Cation exchange capacity (CEC) (unbuffered) was determined by the method of Hendershot et al. (1993) by BaCl2 displacement of base cations into solution and measurement of base cations by atomic absorption spectrometry.
Total metals and P were determined after digestion of the soils by aqua regia, and measurement of elements in the digests using axial-view ICP spectrometry. Extractable soil Fe was measured by both the citrate–bicarbonate–dithionate (CBD) method for "free" Fe oxides and acid ammonium oxalate method for noncrystalline Fe oxides (Jackson et al., 1986). The Fe dissolved by these methods was measured using flame atomic absorption spectrometry.
Column Study A: Copper Addition Experiment
An initial column experiment was conducted to investigate the implications of complexation between Cu2+ and glyphosate on the mobility of both Cu2+ and glyphosate. The mobility of free glyphosate and Cu2+ separately were compared to that of the Cu2+–glyphosate complex in both mineral and organic soil columns.
Sieved (2 mm), air-dried soils were weighed and placed in plexiglass tubes, each with a porous base comprised of glass wool. The columns were 10 cm in length, with an inside diameter of 1.8 cm. Each column was loosely hand packed with either 20 g of air-dried mineral soil (Hudson A) or 15 g of air-dried organic soil (Carlisle). Pore volumes of the columns were measured at 11.5 mL (Hudson A) and 17 mL (Carlisle). Soil properties are presented in Table 1. A 1:1 molar solution of copper/glyphosate was prepared from reagent-grade glyphosate [N-(phosphonomethyl)glycine, 96%, Aldrich Chemical Co.] and CuSO4 and adjusted to pH 6 with NaOH to maximize complex formation. Treatment solutions, consisting of CuSO4 alone, glyphosate alone, or glyphosate + CuSO4, were added to the columns at an average rate of 8.6 mL h–1 using a peristaltic pump to deliver the treatment solutions at a uniform rate. A 24.5 mL treatment volume in total was added to the air-dried columns, corresponding to 6.49 mg of glyphosate per column, or 324 mg kg–1 air dry soil for the mineral soil (Hudson A) and 433 mg kg–1 air dry soil for the peat (Carlisle). Immediately following treatment, a synthetic acid rain solution (pH 4.1) was applied to the columns at a rate of 8.6 mL h–1. Leachate was collected in 10 mL increments until about 4 to 6 pore volumes were collected, and filtered through fine paper (#42 Whatman) for subsequent ICP spectrometry analysis for total dissolved Cu and P. All column treatments were done in triplicate.
Because total P analysis of leachates by ICP does not distinguish dissolved glyphosate from inorganic phosphate (PO4) already present in the soil and potentially displaced into solution by glyphosate, the column experiments were repeated in duplicate to determine PO4 in leachates by the phosphomolybdate blue colorimetric method (Olsen and Dean, 1965). Preliminary tests showed that this method is completely insensitive to the phosphonate present in glyphosate, and also that essentially all of the dissolved P in leachates from mineral soils not amended with glyphosate is in the form of PO4.
Column Study B: Field-Contaminated Soils
Columns were prepared by lightly packing 200 g of field-moist soil, previously sieved at 2 mm, into Buchner funnels fitted with Whatman #42 filter papers. The resulting columns had a surface area of 68 cm2 and a soil depth of 4 cm. A sprayer was used to distribute 5 mL of Roundup treatment solution evenly over the surface. The Roundup solution consisted of the commercial herbicide formulation (containing polyethylamine surfactant and 25–27% glyphosate, as the isopropylamine salt) diluted in deionized water and was applied at a rate of four times the label-recommended rate (i.e., 25.4 mL m–2). This resulted in each column receiving 220 mg a.i. kg–1. (Preliminary trials with Hudson B had shown that applications lower than 4 times the label rate of Roundup had relatively small effects on soluble Cu, Al, Fe, and P in column leachate). To the controls, 5 mL of deionized water was added in the same manner.
Four soils were selected for this experiment (see Table 1), including organic (Carlisle) and mineral (Hagerstown) soils. Both soils are somewhat elevated in Cu as a result of agricultural practices. In addition, a mineral soil treated with a single high sludge load more than 25 yr prior (Hudson B) and soil from an adjacent site that has not received sludge (Hudson A), were used.
Following a 24-h incubation with the Roundup treatment, up to 100 mL of synthetic acid rain (pH 4.1) was added (in 10-mL aliquots), and a vacuum was applied until 10 mL of leachate was collected and filtered through fine (#42 Whatman) filter paper, for analysis by ICP spectrometry. Both the controls and treatments were completed in triplicate.
To determine the fraction of total leachate P measured by ICP that was in the form of PO4, column study B was repeated in triplicate for the two soils showing the greatest evidence of a Roundup treatment effect. These were the Hagerstown and Hudson B soils, both showing significantly elevated leachate Cu and P as a result of Roundup application. The leachates were in this case analyzed for PO4 using the phosphomolybdate method. In this case, the soils were also treated with pure glyphosate (K salt at pH 6.0) as a separate treatment for comparison with the Roundup treatment, to determine if the surfactant present in the commercial Roundup formulation (which is up to 73% polyethylamine surfactant) affects metal mobilization. The leachates were analyzed for dissolved Cu using flame atomic absorption spectrometry.
Batch Equilibration Study: Copper Activity Measurement
Initial column leaching trials with metal-contaminated soils from the field showed Cu to be the trace metal most likely to be mobilized by glyphosate. Therefore, batch equilibration experiments were conducted with field-contaminated and uncontaminated (control) soils to determine the effects of glyphosate on free Cu2+ ion activity and total dissolved Cu. Initially, 20-g samples of dry soil were weighed into 125-mL Erlenmeyer flasks, and 50 mL of 0.01 M CaCl2 containing glyphosate was added to each flask. This CaCl2 solution contained enough glyphosate reagent to add 0, 10, 20, or 50 mg kg–1 of glyphosate to the soil. The soil suspensions were then stirred continuously up to 24 h, and Cu2+ activity was measured periodically using the Cu ion-selective electrode precalibrated using iminodiacetic acid activity buffers (Sauvé et al., 1995). After 24 h reaction, the supernatants were passed through 0.2-µm membrane filters and analyzed for total dissolved Cu by flame atomic absorption spectrophotometry. All batch experiments were conducted in duplicate.
|
RESULTS
|
|---|
Column Study A
In an effort to examine Cu2+–glyphosate sorption dynamics in a soil matrix, two relatively uncontaminated (low-Cu) soils (Hudson A, Carlisle) were selected for this experiment (see Table 1). Despite high glyphosate loadings, soil Cu was not mobilized by the glyphosate-only treatment in the peat, and mobilized only to a small degree in the mineral soil, as indicated by low Cu concentrations in the leachates (glyphosate-only treatments of Fig. 1
and 2
). The new Cu added to the columns in the Cu-only treatments was completely sorbed and thus not detected in the leachate of either soil (Cu-only treatments of Fig. 1 and 2). In contrast, the Cu2+–glyphosate treatment resulted in high Cu in leachate, detectable in the first 10 mL of leachate collected from the columns (Cu-glyphosate treatments of Fig. 1 and 2). Both the mineral and the organic soil show a similar pattern of Cu release, with a steady decrease approaching the control value after 55 mL (approximately five pore volumes for Hudson A and three pore volumes for Carlisle) of leachate had been collected.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 1. Concentration of Cu in mineral soil (Hudson A) column leachate after soil treatment with CuSO4 alone (Cu), glyphosate alone (Gly), or CuSO4 with glyphosate (Cu+ Gly).
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 2. Concentration of Cu in organic soil (Carlisle) column leachate after soil treatment with CuSO4 alone (Cu), glyphosate alone (Gly), or CuSO4 with glyphosate (Cu+ Gly).
|
|
Despite higher glyphosate and Cu loadings on the Carlisle (due to the lower bulk density of the organic soil), the glyphosate–Cu treatment effect was much more pronounced on the mineral (Hudson A) soil. Approximately 38% of the total Cu added with glyphosate moved through the Hudson A column compared to only 4.5% in the Carlisle.
The leachate P concentrations for the mineral and organic soil columns are shown in Fig. 3
and 4
. For the mineral soil (Hudson A), the substantial cumulative P leached indicates that much of the added glyphosate migrated through the column. Only a small fraction of the P in the mineral soil leachates was PO4, as shown by low dissolved PO4 concentrations (data not shown). Nevertheless, the PO4 concentrations in the leachates were approximately twice as high for the glyphosate and glyphosate–Cu treatments as for the CuSO4 treatment, suggesting that glyphosate is more effective than SO4 in displacing PO4 from soil sorption sites. A mass balance calculation based on leachate concentrations of total dissolved P (corrected for PO4), revealed that about 82 and 74% of the added glyphosate leached through the glyphosate and glyphosate-Cu treated mineral soil columns, respectively.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 3. Concentration of total P and orthophosphate in mineral soil (Hudson A) column leachate after soil treatment with CuSO4 alone (Cu), glyphosate alone (Gly), or CuSO4 with glyphosate (Cu+ Gly).
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 4. Concentration of total P in organic soil (Carlisle) column leachate after soil treatment with CuSO4 alone (Cu), glyphosate alone (Gly), or CuSO4 with glyphosate (Cu+ Gly).
|
|
Leachate Cu concentrations from the Cu-glyphosate–treated mineral soil columns were strongly correlated to leachate P, as shown in Fig. 5
. The close relationship of Cu and P concentrations in the mineral soil leachates strongly suggests the presence of mobile Cu-glyphosate complexes that control Cu solubility and mobility. The fact that the leachate Cu/P mole ratio was below the theoretical value for the 1:1 Cu-glyphosate complex is mainly attributed to the fact that total leachate P included glyphosate not complexed with Cu, as part of the Cu was removed from the complex and sorbed by the soil.
View larger version (10K):
[in this window]
[in a new window]
|
Fig. 5. Correlation of dissolved Cu to dissolved P in the leachates collected from the Hudson A soil column after treatment with Cu-glyphosate (1:1 complex).
|
|
For the glyphosate-treated organic soil column, leachate P concentrations increased in the presence of Cu (Fig. 4), an indication that glyphosate was mobilized by complexation with Cu. Nevertheless, there was no relationship of leachate Cu to leachate P for the Cu-glyphosate treatment, a result that could be due to the greater sorption of Cu (96% of added Cu was sorbed) or glyphosate on this peat soil, or to higher PO4 concentrations in leachates. A much higher percentage of dissolved P in leachate from the organic soil, 50 to 70%, was in the form of PO4 in comparison to the mineral soil. As Cu in the leachate decreased with continued column leaching (Fig. 2), dissolved P tended to increase (Fig. 4). The lack of correlation between leachate Cu and P concentration may largely reflect the fact that most of the Cu added to the organic soil column in the form of the 1:1 glyphosate complex was "stripped" from the complex during migration.
The lower leachate P concentrations for the organic compared to mineral soil (Fig. 3 and 4) indicates that glyphosate was sorbed more completely on the organic soil. In fact, when the high PO4 present in organic soil leachates is subtracted from total dissolved P to estimate the dissolved glyphosate concentration, a mass balance calculation shows that no more than 27 and 34% of the glyphosate could have leached through the glyphosate-treated and Cu-glyphosate-treated organic columns, respectively. However, the dissolved P was comprised in part of natural organophosphorus compounds from the organic soil, as evidenced by the fact that only 68% on average of leachate total P from the columns not amended with glyphosate was found to be PO4. Consequently, these estimates of glyphosate leaching from the organic soil are likely to be high, and glyphosate sorption may exceed 80% of that added.
The glyphosate and Cu-glyphosate treatments increased dissolved PO4 in leachates from the organic soil columns when compared to the Cu-treated columns (data not shown). Glyphosate alone mobilized the highest concentrations of PO4, attributable to the exchange of phosphate anions by the phosphonate group (Gimsing and Borggaard, 2002; Gimsing et al., 2004a, 2004b).
Column Study B
In this experiment, four soils with greatly different properties and levels of metal and P contamination were subjected to Roundup application and subsequently leached to test for the mobilization of elements by glyphosate. Leachate analyses for elements of interest are presented in Table 2. A greater range of elements than those listed in Table 2 were analyzed by ICP, but data are only presented for those elements where the Roundup treatment significantly altered leachate concentrations compared to controls. Due to observed nonequality in variance between trace element concentrations in the treated and control leachate samples (heteroscedasticity), it was necessary to log-transform the data before ANOVA analysis. When greater mean values are accompanied by greater variances, as was observed in this case, logarithmic transformation is considered a useful remedy to allow the statistical analysis of treatment effects (Sokal and Rohlf, 1981). The probable reason for greater variance in trace element concentrations measured in the Roundup-treated column leachates compared to the controls is nonuniformity of glyphosate solution contact with, and flow through, soils during relatively rapid application and subsequent leaching.
Following Roundup application, elevated leachate concentrations of Al, Cu, Ni, Zn, Cd, As, P, and Si were observed in several of the soils tested, with the most significant effects of Roundup observed for the highly metal-contaminated Hudson B, and the heavily-manured Hagerstown soil (Table 2). The largest relative increases in solubility attributable to Roundup application were for Cu and P. In fact, despite considerable variability in dissolved elements from a number of replicate Hudson B soil columns treated with Roundup, dissolved Cu was highly correlated to total dissolved P in collected leachates (data not shown). This result is consistent with Cu being mobilized by formation of a soluble Cu-glyphosate complex.
Copper solubility increased significantly in both Cu-contaminated soils (Hudson B, Hagerstown), but not in the much less contaminated Hudson A or Carlisle. Soluble P increased in leachate from all soils after Roundup application, but the largest relative (and only statistically significant) increase occurred with the Hagerstown soil, which had a high background P level from manure application. Since Roundup contains a surfactant, polyethylamine, which could conceivably have a role in metal mobilization, a comparison of the ability of Roundup and pure K-glyphosate salt (at equal glyphosate application rates) to mobilize Cu was conducted on the Hagerstown and Hudson B soils. There was no significant difference in dissolved Cu in leachates from these two treatments, evidence that glyphosate alone is largely responsible for Cu dissolution in soils.
Roundup treatments had a comparable impact on Hudson B and Hagerstown soils for Ni and Zn mobilization (relative to the control leachate values) despite total Ni and Zn being 3.8 and 18.6 times more elevated, respectively, in the Hudson B soil. This result may be explained by the lability of these elements in each soil or by preemptive sorption and inactivation of glyphosate in Hudson B.
In contrast to the other soils used in this experiment, Hudson B is elevated in total Cd as a result of sewage sludge additions, and Roundup treatment somewhat increased dissolved Cd in Hudson B leachate. Although Cd coordinates to glyphosate in a similar manner to Cu (Ramstedt et al., 2004), with some potential for enhanced mobility (Zhou et al., 2004), it forms a less stable complex than Cu (Motekaitis and Martell, 1985).
No statistically significant trends for Fe, Mn, Mo, and Pb in the leachates as affected by Roundup were evident in any of the soils (data not shown). No mobilization of Pb by Roundup was observed despite the fact that total soil Pb was very elevated in Hudson A and B (200–300 mg kg–1) from historical use of lead arsenate pesticides.
Leachate P increased in some soils after Roundup application, most dramatically in the Hagerstown soil (increased by a factor of 7.5). Analysis of leachate P by ICP did not permit distinction between PO4 displaced into solution by glyphosate sorption and dissolved P in the form of unadsorbed glyphosate. However, measurement of PO4 in leachates revealed that Roundup treatment had no significant effect on leachate PO4 concentrations in the Hagerstown or Hudson B soil, and that PO4 concentrations were much lower than the total dissolved P concentrations in leachates from these Roundup-treated soils. Thus, increased dissolved P in leachates from Roundup-treated soil columns (Table 2) is attributed to dissolved glyphosate.
In addition to P, both As and Si were mobilized by Roundup, presumably via anion displacement at sorption sites (Table 2). As with leachate P, the influence of Roundup treatment on leachate As and Si was most pronounced for the Hagerstown soil; the levels of these elements increased by a factor of 3.3 and 1.7, respectively. Interestingly, As was not mobilized in Hudson A, despite this soil having the highest total As of 67 mg kg–1 (from historical use of lead arsenate pesticides).
In general, the Roundup treatment had very little detectable impact on leachates from the relatively uncontaminated Hudson A and Carlisle soils. However, following treatment, an increase in Al was noted in Hudson A leachate and both P and Si were elevated in leachate collected from the Roundup-treated Carlisle. None of these increases reached significance at the p = 0.05 level, however. The leachate P concentrations were lower in the organic soil than for several of the mineral soils, suggesting relatively effective retention of glyphosate by soil organic matter. In addition, despite the high affinity of glyphosate for Cu, there was no mobilization of Cu in the peat.
Batch Equilibration Studies: Copper Activity Measurement
Batch reaction studies with continuous mixing provide a much greater opportunity for added glyphosate to sorb than column leaching studies; therefore, the mobilization of trace metals would not be expected until relatively high glyphosate loadings in this situation. In fact, preliminary batch adsorption studies on the Hudson B soil showed virtually complete sorption (no measured increase in dissolved P) at glyphosate addition levels approaching 50 mg kg–1. Not surprisingly, then, addition of 10 and 20 mg kg–1 of glyphosate to several mineral and organic soils variably contaminated with Cu showed no tendency for increased Cu solubility or altered Cu2+ activity relative to control soils (data not shown). At the 50 mg kg–1 glyphosate addition rate, however, an increase in Cu solubility and lowered Cu2+ activity was measured in some soils. Figure 6
illustrates the effect of glyphosate on Cu2+ activity for both the uncontaminated and moderately Cu-contaminated Caledon soil (Caledon A and B, respectively). The initial response to added glyphosate was a decrease in Cu2+ activity, an indication that added glyphosate quickly complexed much of the free Cu2+ in soil solution. This response is expected to induce the dissolution of some solid-phase Cu, and ultimately, an increase in soluble Cu. However, the expected change in dissolved Cu occurred only for the Cu-contaminated soil, increasing from 0.040 mg L–1 (initially) to 0.125 mg L–1 after 24-h reaction with glyphosate. Conversely, dissolved Cu remained near the initial value of 0.030 mg L–1 for the uncontaminated Cambridge soil despite 24-h reaction with glyphosate.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 6. Effect of glyphosate addition (at time zero) to Caledon soil, uncontaminated (A) and contaminated by sewage-sludge metals (B), on the measured pCu, where pCu is the negative logarithm of the free Cu2+ activity.
|
|
Further batch equilibrations with field-contaminated soils with even higher concentrations of total Cu (including Hudson B) generally revealed increases in dissolved Cu at the 50 mg kg–1 glyphosate addition rate, but no effect on Cu2+ activity (data not shown). Conversely, in response to the 50 mg kg–1 glyphosate addition, uncontaminated soils showed an initial decrease in Cu2+ activity but little or no change in soluble Cu, consistent with the behavior of the Caledon A soil (see Fig. 6). Thus, it is suggested that glyphosate may solubilize Cu under batch conditions, but only if a large labile "pool" of Cu exists in the soil.
|
DISCUSSION AND CONCLUSIONS
|
|---|
The results of these column leaching and batch equilibration experiments show effects of glyphosate on the mobility of several elements in soils, with the most pronounced effects on Cu and P mobility. These effects are attributable to the ability of glyphosate to form strong chelation complexes with Cu and Al in particular, as well as to compete with PO4 for sorption sites (de Jonge et al., 2001). Glyphosate has been shown to complex and solubilize Cu2+ previously bound to both mineral (McBride, 1991) and organic (Undabeytia et al., 1996) surfaces in pure systems.
The activity of free Cu2+ may be decreased by glyphosate if there is a small reserve of sorbed Cu. In soils contaminated by Cu, the release of sorbed Cu in response to glyphosate addition appears to be rapid enough to prevent an initial lowering of Cu2+ activity. With increasing time after glyphosate application, however, these effects may subside as glyphosate is removed from solution by adsorption on minerals and organic matter. But rapid leaching events following application, such as would occur with heavy rainfall, could allow the mobilization of some metals and phosphorus.
It must be emphasized that the glyphosate loading rates that mobilized soil elements are on the order of 50 mg kg–1 or higher. Typical agronomic application rates, which according to Giesy et al. (2000) currently range from 1.1 to 5.6 kg a.i. ha–1, would only approach these high loadings in the thin soil surface layer that initially intercepts the herbicide spray. Giesy et al. (2000) suggest that soil concentrations of glyphosate rarely exceed 
6 mg kg–1 following agronomic applications. However, occasionally researchers have measured much higher soil glyphosate residues following typical agronomic applications. For example, Feng and Thompson (1990), detected 39.8 mg kg–1 in soil samples and Müller et al. (1981) reported the extraction of 23 mg kg–1. These levels are indicative of uneven deposition of the herbicide, which is probably a common occurrence in the environment. Furthermore, commercially available Roundup (27% a.i.) has label recommendations of 6.38 mL m–2; this rate corresponds to approximately 17 kg a.i. ha–1. In nonagronomic applications, for example in forestry, glyphosate applied aerially may concentrate in the upper organic soil layer and, depending on the topography and weather conditions, impact surface water quality (e.g., via mobilization of Al) by runoff without significant interaction with the underlying mineral soil.
Our results with soils high in organic matter suggest a high affinity of glyphosate for soil organic matter. Since both humus and glyphosate are anionic, the bonding mechanism presumeably involves a "bridging" metal ion such as Fe or Al. Soil organic matter has been shown in a number of studies to have an affinity for glyphosate (Roy et al., 1989; Dousset et al., 2004), and elevated sorption coefficients have been calculated for soils with high organic matter (Nomura and Hilton, 1977; Morillo et al., 2000). Piccolo et al. (1992, 1995) observed considerably higher sorption to a Fe–humic acid complex relative to levels of adsorption measured for cation saturated smectites and soils (Glass, 1987; Piccolo et al., 1994). However, organic carbon appeared to have no significant influence on sorption in some studies (de Jonge et al., 2001; Gimsing et al., 2004a), and Gerritse et al. (1996) reported a negative correlation between organic matter and glyphosate or AMPA sorption in a sandy soil. In the latter study, organic matter was suggested to compete with glyphosate for sorption to clay minerals. This result, which contrasts with most other studies, may be attributable to the exceptionally low organic matter content of the soils used in that study.
Received for publication December 20, 2005.
|
REFERENCES
|
|---|
- Allison, L.E. 1976. Organic carbon. p. 1367–1378. In C.A. Black (ed.) Methods of soil analysis. Part 2. ASA, Madison, WI.
- de Jonge, H., L.W. de Jonge, O.H. Jacobsen, T. Yamaguchi, and P. Moldrup. 2001. Glyphosate sorption in soils of different pH and phosphorus content. Soil Sci. 166:230–238.
- Dousset, S., C. Chauvin, P. Durlet, and M. Thévenot. 2004. Transfer of hexazinone and glyphosate through undisturbed soil columns in soils under Christmas tree cultivation. Chemosphere 57:265–272.[Medline]
- Feng, J.C., and D.G. Thompson. 1990. Fate of glyphosate in a Canadian forest watershed. 2. Persistence in foliage and soil. J. Agric. Food Chem. 38:1118–1125.
- Gerritse, R.D., J. Beltran, and F. Hernandez. 1996. Adsorption of atrazine, simazine, and glyphosate in soils of the Guangara Mound, Western Australia. Aust. J. Soil Res. 34:599–607.[CrossRef]
- Giesy, J.P., S. Dobson, and K.R. Solomon. 2000. Ecotoxicological risk assessment for Roundup® herbicide. Rev. Environ. Contam. Toxicol. 167:35–120.
- Gimsing, A.L., and O.K. Borggaard. 2002. Effect of phosphate on the adsorption of glyphosate on soils, clay minerals and oxides. J. Environ. Anal. Chem. 82:545–552.[CrossRef]
- Gimsing, A.L., O.K. Borggaard, and M. Bang. 2004a. Influence of soil composition on adsorption of glyphosate and phosphate by contrasting Danish surface soils. Eur. J. Soil Sci. 55:183–191.
- Gimsing, A.L., O.K. Borggaard, and P. Sestoft. 2004b. Modeling the kinetics of the competitive adsorption and desorption of glyphosate and phosphate on goethite and gibbsite and in soils. Environ. Sci. Technol. 38:1718–1722.
- Glass, R.L. 1987. Adsorption of glyphosate by soils and clay minerals. J. Agric. Food Chem. 35:497–500.
- Hendershot, W.H., H. Lalande, and M. Duquette. 1993. Ion exchange and exchangeable cations. p. 167–176. In M.R. Carter (ed.) Soil sampling and methods of analysis. Lewis Publishers, Boca Raton, FL.
- Jackson, M.L., C.H. Lim, and L.W. Zelazny. 1986. Oxides, hydroxides and aluminosilicates. p. 101–150. In A. Klute (ed.) Methods of soil analysis. Part 1. Physical and mineralogical methods. 2nd ed. ASA, Madison, WI.
- McBride, M.B. 1991. Electron spin resonance study of copper ion complexation by glyphosate and related ligands. Soil Sci. Soc. Am. J. 55:979–985.[Abstract/Free Full Text]
- McBride, M.B., and K.-H. Kung. 1989. Complexation of glyphosate and related ligands. with iron(III). Soil Sci. Soc. Am. J. 53:1668–1673.[Abstract/Free Full Text]
- Morillo, E., T. Undabeytia, and C. Maqueda. 1997. Adsorption of glyphosate on the clay mineral montmorillonite: Effect of Cu(II) in solution and adsorbed on the mineral. Environ. Sci. Technol. 31:3588–3592.
- Morillo, E., T. Undabeytia, C. Maqueda, and A. Ramos. 2000. Glyphosate adsorption on soils of different characteristics. Influence of copper addition. Chemosphere 40:103–107.
- Morillo, E., T. Undabeytia, C. Maqueda, and A. Ramos. 2002. The effect of dissolved glyphosate upon the sorption of copper by three selected soils. Chemosphere 47:747–752.[Medline]
- Motekaitis, R.J., and A.E. Martell. 1985. Metal chelate formation by N-phosphonomethylglycine and related ligands. J. Coord. Chem. 14:139–149.
- Müller, M.M., C. Rosenberg, H. Siltanen, and T. Wartiovaara. 1981. Fate of glyphosate and its influence on nitrogen-cycling in two Finnish agricultural soils. Bull. Environ. Contam. Toxicol. 27:724–730.[CrossRef][ISI][Medline]
- Nomura, N.S., and H.W. Hilton. 1977. The adsorption and degradation of glyphosate in five Hawaiian soils. Weed Res. 17:113–121.[CrossRef]
- Olsen, S.R., and L.A. Dean. 1965. Phosphorus. p. 1035–1049. In C.A. Black et al. (ed.) Methods of soil analysis. Part 2. ASA, Madison, WI.
- Piccolo, A., G. Celano, and G. Pietramellara. 1992. Adsorption of the herbicide glyphosate on a metal-humic acid complex. Sci. Total Environ. 123:77–82.
- Piccolo, A., G. Celano, N. Arienso, and A. Mirabella. 1994. Adsorption and desorption of glyphosate in some European soils. J. Environ. Sci. Health B29:1105–1115.
- Piccolo, A., L. Gatta, and L. Campanella. 1995. Interaction of glyphosate herbicide with a humic acid and its iron complex. Ann. Chim. 85:31–40.
- Ramstedt, M., C. Norgren, J. Sheals, D. Boström, S. Sjöberg, and P. Persson. 2004. Thermodynamic and spectroscopic studies of cadmium(II)-N-(phosphonomethyl)glycine (PMG) complexes. Inorg. Chem. Acta 357:1185–1192.[CrossRef]
- Roy, D.N., S.K. Konar, S. Banerjee, D.A. Charles, D.G. Thompson, and R. Prasad. 1989. Persistence, movement and degradation of glyphosate in selected Canadian boreal forest soils. J. Agric. Food Chem. 37:437–440.
- Sauvé, S., M.B. McBride, and W.A. Hendershot. 1995. Ion-selective electrode measurements of copper (II) activity in contaminated soils. Arch. Environ. Contam. Toxicol. 29:373–379.[CrossRef][ISI]
- Sheals, J., P. Persson, and B. Hedman. 2001. IR and EXAFS spectroscopic studies of glyphosate protonation and copper(II) complexes of glyphosate in aqueous solution. Inorg. Chem. 40:4302–4309.[CrossRef][ISI][Medline]
- Sheals, J., M. Granström, S. Sjöborg, and P. Persson. 2003. Coadsorption of Cu(II) and glyphosate at the water-goethite (FeOOH) interface: Molecular structures from FTIR and EXAFS measurements. J. Colloid Interface Sci. 262:38–47.[CrossRef][ISI][Medline]
- Sokal, R.R., and F.J. Rohlf. 1981. Biometry. 2nd ed. W.H. Freeman and Co., San Francisco.
- Song, B., D. Chen, M. Bastian, R.B. Martin, and H. Sigel. 1994. Metal-ion-coordinating properties of a viral inhibitor, a pyrophosphate analogue and a herbicide metabolite, a glycine analogue: The solution properties of the potentially five-member chelates derived from phosphonoformic acid and (aminomethyl)phosphonic acid. Helv. Chim. Acta 77:1738–1756.
- Subramaniam, V., and P.E. Hoggard. 1988. Metal complexes of glyphosate. J. Agric. Food Chem. 36:1326–1329.
- Undabeytia, T., M.V. Cheshire, and D. McPhail. 1996. Interaction of the herbicide glyphosate with copper in humic conditions. Chemosphere 32:1245–1250.
- Zhou, D.M., Y.J. Wang, L. Cang, X.Z. Hao, and X.S. Luo. 2004. Adsorption and cosorption of cadmium and glyphosate on two soils with different characteristics. Chemosphere 57:1237–1244.[Medline]
TOP
ABSTRACT
INTRODUCTION
