JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 1 May 2008
Published in J Environ Qual 37:830-838 (2008)
DOI: 10.2134/jeq2007.0256
© 2008 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Laitinen, P.
Right arrow Articles by Hartikainen, H.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Laitinen, P.
Right arrow Articles by Hartikainen, H.
Agricola
Right arrow Articles by Laitinen, P.
Right arrow Articles by Hartikainen, H.
Related Collections
Right arrow Agricultural Pesticides
Right arrow Phosphorus
Right arrow Soil Analysis

TECHNICAL REPORTS

Organic Compounds in the Environment

Effects of Soil Phosphorus Status on Environmental Risk Assessment of Glyphosate and Glufosinate-Ammonium

Pirkko Laitinena, Katri Siimesb, Sari Rämöa,*, Lauri Jauhiainena, Liisa Eronenc, Seija Oinonena and Helinä Hartikainend

a MTT Agrifood Research Finland, FI-31600 Jokioinen, Finland
b SYKE Finnish Environment Inst., P.O. Box 140, FI-00251 Helsinki, Finland
c SBRC Sugar Beet Research Centre, Korvenkyläntie 201, FI-25170 Kotalato, Finland
d Dep. of Applied Chemistry and Microbiology, Box 27, FI-00014 Univ. of Helsinki, Finland

* Corresponding author (sari.ramo{at}mtt.fi)

Received for publication May 23, 2007.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Material and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
The increased use of herbicides poses a risk to the aquatic environment. Easy and economical methods are needed to identify the fields where specific environment protection measures are needed. Phosphorus (P) and organophosphorus herbicides compete for the same adsorption sites in soil. In this study the relationship between P obtained in routine Finnish agronomic tests (acid ammonium acetate [PAC]) and adsorption of glyphosate and glufosinate-ammonium was investigated to determine whether PAC values could be used in the risk assessment. The adsorption of glyphosate ((N-(phosphonomethyl)glycine) and glufosinate-ammonium (2-amino-4-(hydroxymethylphosphinyl)butanoic acid) was studied in a clay and a sandy loam soil enriched with increasing amounts of P added as potassium dihydrogen phosphate. Desorption was also determined for some P-enriched soil samples. The adsorption of both herbicides diminished with increasing PAC value. The correlations between Freundlich adsorption coefficients obtained in the adsorption tests and PAC were nonlinear but significant (r > 0.98) in both soils. The exponential models of the relationship between soil PAC values and glyphosate adsorption were found to fit well to an independent Finnish soil data set (P < 0.1 for glyphosate and P < 0.01 for glufosinate-ammonium). The desorption results showed that glufosinate-ammonium sorption is not inversely related to soil P status, and the high correlation coefficients obtained in the test of the model were thus artifacts caused by an abnormal concentration of exchangeable potassium in soil. The solved equations are a useful tool in assessing the leaching risks of glyphosate, but their use for glufosinate-ammonium is questionable.

Abbreviations: FMOC-Cl, 9-fluorenylmethyl-clororoformiate • KF, Freundlich adsorption coefficient • PAC, acid ammonium acetate extractable phosphorus


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Material and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
GLYPHOSATE (N-(phosphonomethyl)glycine) and glufosinate-ammonium (2-amino-4-(hydroxymethylphosphinyl)butanoic acid) are broad-spectrum organophosphate herbicides. Glyphosate has been one of the world's most widely used herbicides since it came on the market in 1974. In Finland, glyphosate accounted for 50.4% and glufosinate-ammonium for 0.13% of herbicide active ingredients sold in 2005 (Evira, 2006). In recent years, the use of glyphosate in cereal cultivation has changed. Instead of a post-harvest spraying every second or third year, an annual spring or autumn application has become a more common practice.

On the basis of the chemical structure, with the active phosphonate group at the end of the molecule (Fig. 1 ), glyphosate is able to form an inner-sphere complex on the Al and Fe oxide surfaces in soil. This reaction pattern, being similar to that of phosphate, means that glyphosate competes with phosphate for the same sorption sites. Sprankle et al. (1975a and 1975b) were the first to produce evidence that phosphate can diminish glyphosate adsorption. This finding was confirmed in later studies (de Jonge et al., 2001; Gimsing and Borggaard 2001, 2002).


Figure 1
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 1. Chemical structures of glyphosate and glufosinate-ammonium.

 
In conformity with phosphate chemistry, oxidic material is found to be superior to clay silicates in glyphosate sorption (Nomura and Hilton, 1977; Gimsing and Borggaard, 2001, 2002). Furthermore, de Jonge et al. (2001) noticed that liming enhanced the sorption of glyphosate in a very acid coarse sandy soil through the formation of reactive aluminum and iron oxides. Other soil factors, such as organic carbon, clay content, and mineralogy of the clay fraction have been found to have no effect on glyphosate and phosphate adsorption (Gimsing et al., 2004).

There are several studies on the desorption of glyphosate from soil material. Hance (1976) reported that the phosphonic moiety of adsorbed glyphosate could be displaced by phosphate. Gimsing and Borggaard (2001, 2002) showed that phosphate displaced absorbed glyphosate from goethite, but glyphosate was not able to displace phosphate. In a recent study, Gimsing et al. (2004) noticed that phosphate and glyphosate are able to some extent to displace each other at the sorption sites, the amount of glyphosate desorbed by phosphate being large in soils high in iron and aluminum oxides. These dissimilar outcomes can be explained by differences in saturation degrees of sorption sites. For example, the results of de Jonge et al. (2001) show that an increase in phosphorus (P) status rendered the glyphosate sorption more reversible. Similarly, in the studies by Sørensen et al. (2006) and Mamy and Barriuso (2007), desorption of glyphosate was small, whereas its adsorption tendency was high.

The sorption of glufosinate-ammonium has been studied less than that of glyphosate. The glufosinate-ammonium molecule is larger, and, in contrast to glyphosate, its P atom is bound to two carbon atoms (Fig. 1). These factors reduce the sorption ability of glufosinate-ammonium. Dorn et al. (1992) reported that adsorption of glufosinate-ammonium depends for the most part on the amount and type of clay minerals. Tomlin (2000) has reported that the sorption of glufosinate-ammonium is low, especially in soils with a low clay content. Due to the positive charge, glufosinate-ammonium can bind to soil cation exchange sites that are more abundant in clay soils than in coarser soils. On the basis of its low Freundlich adsorption coefficient (KF) values, glufosinate-ammonium is usually classified as a mobile compound. This property was also shown in 12 Finnish soils where KF values varied from 0.6 to 11.7 L kg–1 (Autio et al., 2004). There are no published studies of glufosinate-ammonium desorption. Nevertheless, the low sorption coefficients and the expected reversible sorption mechanism to cation exchange sites indicate high desorption potential. The influence of soil P status on glufosinate-ammonium sorption had not been previously studied.

The retention of organics on soil particles retards their mineralization and causes build-up of chemicals in soil. Thus, the intensive use of herbicides, especially of those of slow degradation, may lead to their accumulation in soil and increase the risk of water pollution. To assess this risk, easy and economical methods based on the chemical behavior of pesticides are needed. The agri-environmental program in Finland regulates the determination of P status of fields at regular intervals. Thus, a large amount of data on P values obtained with acid ammonium acetate extraction (PAC) used in routine soil testing are available. The aim of the present study was to determine whether it is possible to assess the accumulation tendency and leaching risk of pesticides on the basis of these data. The study was undertaken to compare the retention tendency of glyphosate and glufosinate-ammonium and its dependency on the P status of soils.


   Material and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Material and Methods
Results
 Discussion
 Conclusions
 REFERENCES
 
Experimental Soils
About 50-L samples of sandy loam and clay soil from Southwest Finland were collected from the plow layer (0–30 cm) of two fields before onset of a field study on the persistence and transport of glyphosate and glufosinate-ammonium and conventional sugar beet herbicides (see Laitinen et al., 2006). Twenty subsamples from both fields were combined, and the soil material was homogenized. The properties of the soil samples are given in Table 1 . The clay soil was most likely a Vertic Cambisol, and the sandy loam soil was a Dystric Regosol. The same soil material was previously included in a herbicide adsorption study of Autio et al. (2004).


View this table:
[in this window]
[in a new window]

  		Table 1. Soil properties.

 
Soil texture was determined by the pipette method of Elonen (1971), and organic carbon content was determined according to Graham (1948). pH was measured in an aqueous suspension at a soil to water ratio of 1:2.5 (Lakanen and Erviö, 1971). Iron (Fe) and aluminum (Al) oxides were extracted with ammonium oxalate (0.5 mol L–1; pH 3.3) (Niskanen, 1989) and determined by inductively coupled plasma optical emission spectrometry (Thermo Jarrell Ash Corp., Franklin, MA). Phosphorus reserves were characterized by using a modified Chang and Jackson's sequential fractionation procedure (Hartikainen, 1979), where NH4Cl is used to remove soluble P, and NH4F is assumed to extract P bound by Al oxides, NaOH P bound by Fe oxides, and H2SO4 apatitic or Ca-bound P. Agronomic P status was determined with acid ammonium acetate extraction (Vuorinen and Mäkitie, 1955) used in routine advisory soil testing in Finland since the 1950s. For practical reasons, air-dried soil material was sampled for this analysis on volume basis, and the results are given in mg L–1 of soil. This procedure speeds up the routine laboratory work and makes the plant-available P reserves obtained for soils of contrasting texture (peat vs. mineral) comparable.

Phosphorus Additions
For the herbicide sorption tests, a set of soil samples was enriched with KH2PO4 (J.T. Baker, Phillipsburg, NJ) solutions of increasing P concentrations (0–15.5 g P L–1) to give additional P levels of 100 to 1200 mg kg–1 in clay and 100 to 3000 mg kg–1 in sandy loam soil (Table 2 ). Either 75 mL or 100 mL of water were added in air-dried sieved sand or clay soil sample (425 g) and were kept moist overnight or over the weekend. A P solution (80 mL) was added. A suitable P concentration was made by mixing different volumes of water and 0.2 mol L–1 or 0.5 mol L–1 KH2PO4 solution. Sandy loam P.1100 and P.1400 were made by adding 82 mL and 98 mL of 0.2 mol L–1 KH2PO4 solution, and clay P.3000 was made by adding 91 mL of 0.5 mol L–1 KH2PO4 solution. After addition of the enrichment solution, the moisture content of clay soil was 30%, and the moisture content of sandy loam 27%. The soils were incubated for 2 wk at room temperature. After incubation, they were air-dried and ground to pass through a 2-mm sieve. Homogenized soils were stored at an ambient temperature of 25°C for P analysis and sorption tests of the herbicides.


View this table:
[in this window]
[in a new window]

  		Table 2. Freundlich adsorption coefficient (KF) and exponent (1/n) with SD for glyphosate and glufosinate-ammonium at different levels of acid ammonium acetate–extractable P (PAC) (n = 4).

 
Sorption Tests with Pesticides
The soil samples adjusted to represent different P levels were subjected to batch equilibrium sorption tests performed according to the Organization for Economic Cooperation and Development Guideline for Testing of Chemicals 106: "Adsorption/Desorption" (OECD, 1981). The new version of the test guideline (OECD, 2000) was not available when we started the first laboratory experiments. The soil to solution ratio was 1:5, and the tests were performed at a standard laboratory temperature (25 ± 2°C). Stock standard solutions (approximately 100 µg mL–1) of glyphosate and glufosinate-ammonium (Dr. Ehrenstorfer GmbH, Augsburg, Germany) were prepared with high-performance liquid chromatography–grade water (Ultra-pure water system; Millipore, Bedford, MA) in polypropylene volumetric bottles. The adsorption test analyses were performed with two different concentrations of the active substance, being 10 and 2 mg L–1 for glyphosate and 5 and 1 mg L–1 for glufosinate-ammonium. The highest concentrations were selected to correspond to those used in practical cultivation. Moreover, two additional glufosinate-ammonium concentrations, 8 and 2 mg L–1, were used for some soil samples because these concentrations were included in the desorption study. Glyphosate and glufosinate-ammonium were derivatized with 9-fluorenylmethyl-clororoformiate (FMOC-Cl) (Fluka/Riedel-de Haën, Buchs, Swizerland) and determined by a high-performance liquid chromatograph (Hewlett-Packard series II 1090; Hewlett Packard GmbH, Waldbronn, Germany) equipped with a fluorescence detector (Hewlett-Packard 1046A programmable fluorescence detector) as their FMOC derivatives (Sancho et al., 1996). A slight modification of the guideline was necessary: Water was used instead of CaCl2 as the background solution. Preliminary tests had shown that reaction failed when Cl– solution was used due to recrystallization of FMOC-Cl (Autio et al., 2004).

For identification and quantitation of the active substances, the following analytical columns were used. For glyphosate and glufosinate-ammonium, a Hamilton PRP-1 (Hamilton, Reno, NV) or Supelco guard column (Supelco, Bellefonte, PA) was connected in series with a HP Hypersil APS column with KH2PO4 and HPCL-grade acetonitrile (J.T. Baker) as eluents. The exitation wavelength of the fluorescence detector was 263 nm, and the emission wavelength was 317 nm for the analysis of both herbicides. Two control samples (50 mL of test solution) and one blank soil sample (10 g of soil and 50 mL of water) were handled in the same way as the test samples. The control samples have two purposes in the test: (i) A test compound should be stable enough during the test, and (ii) only weak adsorption is allowed to occur in the test tubes. Any interference signals from soil were controlled by using the blank sample. The recovery of the tests was 88 to 104% of glyphosate and 91 to 101% of glufosinate-ammonium. The analytical data were used to calculate the Freundlich adsorption coefficients KF and 1/n according to Eq. [1].

Formula 1[1]
where S is the sorbed amount (mg kg–1), C is the concentration in equilibrium solution (mg L–1), KF is the Freundlich adsorption coefficient (L kg–1), and 1/n (-) is a correction term needed to describe the nonlinearity of the sorption isotherm. The percentage of adsorbed herbicide was calculated according to Eq. [2].

Formula 2[2]
where A is adsorbed herbicide (%), Ce is the concentration of supernatant (i.e., the non-adsorbed herbicide in solution), and Cc is the concentration of herbicide in control solution.

Herbicide Desorption Tests
Desorption tests were performed with sandy loam samples adjusted to three P levels. For economic reasons, for clay soil only subsamples not enriched with P were used. The test was continued as a desorption test by adding the same volume of water as equilibrium solution was removed in the of adsorption test. Then the suspension was reshaken and centrifuged, and the supernatant was analyzed. The desorption procedure was then repeated. The supernatants were analyzed separately in the same way as in the adsorption test. The desorption percentile was calculated according to Eq. [3].

Formula 3[3]
where D is desorbed herbicide (%), Cdes1 is the concentration in the first desorption extraction (mg L–1), Cdes2 is the concentration in the second desorption extraction, V is the volume of solution recovered from the adsorption test (ml), V0 is the volume of the test solution in the adsorption test (ml), Ce is the herbicide concentration in the supernatant in the adsorption test (mg L–1), and X is the amount of herbicide adsorbed in the soil (mg).

Developing and Testing of Model
The results of the adsorption tests and P analyses were used to model the KF values as a function of soil PAC. Several nonlinear models of the SAS/NLIN procedure (SAS Institute Inc. 1999) were applied. The developed models were then tested with an independent data set obtained in a previous study (Autio et al., 2004). The equations developed for clay soil were applied to clay soils, and the equations developed for sandy loam soil were used for all other soils. Calculated and observed KF values were compared with each other, and their correlation coefficients were calculated.


   Results
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Material and Methods
 Results
Discussion
 Conclusions
 REFERENCES
 
Soil Phosphorus Status
According to the Finnish agri-environmental classification, the initial P status of the two soils was fair. After P enrichment PAC values covered the whole classification range. The PAC limits were as follows: between "fair" and "satisfactory" P status is 8 mg L–1, between "satisfactory" and "good " P status is 14 mg L–1, between "good" and "high" P status is 23 mg L–1, and between "high" and "excessive" P status is 40 mg PAC L–1 in clay soil. These values are slightly higher for coarse soils.

Statistical analysis showed that the increase of PAC value was linear when the addition of P was less than 1000 mg kg–1. The increase in PAC value induced by each addition of 100 mg kg–1 of P was 2.0 mg L–1 (SE, 0.17) higher in clay than in sandy loam soil. This difference was statistically significant (P < 0.001). In sandy loam soil, pH tended to increase with the amount of P added. In the clay soil, the impact of P additions on soil pH was less regular (Table 2).

Herbicide Adsorption
Selected herbicide adsorption isotherms for soils of different P levels are depicted in Fig. 2 . The order of the isotherms, the highest curve corresponding to the lowest P status, demonstrates decreasing sorption with increasing soil P level. Adsorption coefficients at different P addition levels are given in Table 2. Glyphosate adsorption was about four times higher than that of glufosinate-ammonium (P < 0.001). When comparing the results at about the same increase in the P test level (4.1–152 mg L–1 in clay and 3.3–156 mg L–1 in sandy loam soil), the decrease was more marked in sandy loam than in clay soil. The KF value for glyphosate decreased from 99 to 41 L kg–1 in clay soil and from 99 to 13 L kg–1 in sandy loam. Expressed in percentages, the adsorbed amount of glyphosate decreased from 99 to 86% in clay soil and from 89 to 65% in sandy loam. The corresponding changes in glufosinate-ammonium KF values were 2.8 to 1.2 L kg–1 in clay soil and 1.3 to 0.5 L kg–1 in sandy loam, and the adsorbed amounts decreased from 28 to 12% and from 16 to 7%, respectively.


Figure 2
View larger version (38K):
[in this window]
[in a new window]

  		Fig. 2. Glyphosate (A and B) and glufosinate-ammonium (C and D) adsorption isotherms at different agronomic phosphorus levels from fair to excessive. n = 4. Note the different scales.

 
Herbicide Desorption
The results of glyphosate desorption tests in Table 3 show that desorption increases when the herbicide solution concentration increases. In other words, the relative proportion of bound residues decreased. This concentration effect became more pronounced at the higher PAC levels. The effect of test solution concentration on herbicide desorption seemed to be even more marked in the case of glufosinate-ammonium (Table 4 ). In contrast to glyphosate, the desorbed amounts of glufosinate-ammonium decreased when the PAC levels increased. However, the effect of test solution concentration increased with PAC. The proportions of adsorbed and desorbed herbicides are shown in Fig. 3 as percentages of added amounts. The desorbed proportion stands for reversibly bound herbicides. With increasing PAC values, the irreversibly bound fraction of glyphosate slightly decreased, whereas for glufosinate-ammonium it remained about the same. The reversible fraction of glyphosate tended to increase, but that of glufosinate-ammonium decreased drastically.


View this table:
[in this window]
[in a new window]

  		Table 3. Adsorption and desorption of glyphosate at different levels of acid ammonium acetate–extractable phosphorus (PAC) and test solution.

 

View this table:
[in this window]
[in a new window]

  		Table 4. Adsorption and desorption of glufosinate-ammonium at different levels of acid ammonium acetate extractable phosphorus (PAC) and test solution.

 

Figure 3
View larger version (29K):
[in this window]
[in a new window]

  		Fig. 3. Sorption percentiles of applied glyphosate and glufosinate-ammonium in phosphorus-adjusted sandy loam soil. The herbicide addition levels used were 2 mg L–1 for glyphosate and 1 mg L–1 for glufosinate-ammonium.

 
Model Fitting
Graphical examination of KF as a function of PAC revealed that the relationship was not linear in the glyphosate data. Spherical and exponential nonlinear models with and without an intercept were therefore fitted to the data. These four models were compared by graphical methods (e.g., plotting the data and the fitted curve to the same figure), and the nature of the relationship was found to follow the exponential model. However, a simple exponential model did not have a good fit in the small values of the PAC. The problem was solved fitting the following double exponential model (Eq. [4]), which was suitable for both soil types with slightly different parameter values (Table 5 , Fig. 3). Estimates for the parameter β were higher in clay than in sandy loam (P < 0.001). No statistically significant difference was found in the slope parameter, {alpha}. The difference between the models of the two soil types was statistically significant (P = 0.03).

Formula 4[4]
where β and (100 – β) represent the magnitude of the first and second parts of the model, and {alpha}1 and {alpha}2 are slope parameters for the first and second parts.


View this table:
[in this window]
[in a new window]

  		Table 5. Parameter values for exponential Equations [4] and [5].

 
Glufosinate-ammonium data were modeled in the same way. However, a simple exponential model (Eq. [5]) was selected for both soil types (Fig. 4 ).

Formula 5[5]
where β is the curve's highest value, and {alpha} is a slope parameter. The parameter values are shown in Table 5. Estimates for parameter β were higher in clay than in sandy loam (P < 0.001). No statistically significant difference was found in the slope parameter, {alpha}. The determination coefficients of the equations were the same for both herbicides (0.99 for clay and 0.98 for sandy loam soil).


Figure 4
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 4. Observed KF (see Table 2) of glyphosate (A) and glufosinate-ammonium (B) versus soil acid ammonium acetate–extractable phosphorus (PAC), and fitted models.

 
The observed KF values of the independent data set and KF values calculated using the developed models (Eq. [4] and [5]) are shown in Table 6 . The correlation coefficients were 0.47 for glyphosate (P < 0.1) and 0.64 for glufosinate-ammonium (P < 0.01). There were some great differences between calculated and measured values, especially when the soil type was clay. This means that the designed models are not universal and that their ability to predict is poor in certain situations.


View this table:
[in this window]
[in a new window]

  		Table 6. Testing statistical Equations [4] and [5] for independent data sets: Calculated and measured sorption coefficient (KF).

 

   Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Material and Methods
 Results
 Discussion
Conclusions
 REFERENCES
 
Effect of Soil Phosphorus Status on Adsorption
The herbicide adsorption decreased with increasing P status (Fig. 2). The nonlinear relationship is attributable to two parallel factors decreasing the retention of glyphosate: (i) The competitive phosphate anions occupy the oxide surfaces, and, at the same time, (ii) they render the surface charge of the oxides more negative, thereby increasing repulsion. The KF value diminished with increasing PAC value more drastically in sandy loam than in clay soil. This difference between the soil samples can be explained by their dissimilar sorption properties. Phosphate additions required to raise the PAC values to a given level were much higher in the sandy loam than in clay soil. This finding is consistent with the higher Fe and Al oxide concentration of the sandy loam soil (Table 1), which is evident if the metals are expressed on a molar basis (in total 261 mmol kg–1 in sandy loam and 216 mmol kg–1 in clay). In terms of adsorption reactions, this means that in the sandy loam at a given P addition level there were still more sorption sites available for glyphosate. According to the review by Vereecken (2005), in addition to P status, the amount of iron and aluminum oxides, phosphonic moiety and pH control the adsorption, whereas soil organic matter seems to have an indirect effect. In our study, the impact of oxides was not evident when the soils were adjusted to the same PAC level.

The adsorption of glufosinate-ammonium was lower than that of glyphosate. Furthermore, in contrast to glyphosate, the soil types differed decisively in that the KF value of glufosinate-ammonium was much higher in the clay soil than in the sandy loam (Table 2). This is in accordance with the concept that the retention of this herbicide takes place at the cation exchange sites whose number is lower in the coarse-textured soils. However, the impact of the P enrichment on the adsorption of glufosinate-ammonium was the opposite of that expected. Theoretically, the P adsorption renders the surface charge more negative when the phosphate anion forms an inner sphere complex with the oxide surface (e.g., Hingston et al., 1974). As a result of this reaction, the cation exchange capacity increases. Thus, the P enrichment can be expected to enhance the adsorption of this cationic herbicide. The opposite outcome obtained in our adsorption test may be due to K+ cations added in high quantities in the P enrichment treatments. It is likely that they efficiently competed for the cation exchange sites and interfered with the adsorption of glufosinate-ammonium.

Effect of Soil Phosphorus Status on Desorption
The relative amount of glyphosate desorbed being very small (only 6%) in clay soil (Fig. 3) indicates that the sorption was irreversible. The bound residues of glyphosate amounted to 89 to 71% of applied herbicide (Fig. 3). This result is consistent with the results of Aamad and Jacobsen (2001), who found it to be about 10% in clay soil. Sørensen et al. (2006) showed that desorption of glyphosate was small when the adsorption tendency was high. In the sandy loam, an increase in P status enhanced the glyphosate desorption. This means that the competition of phosphate for sorption sites weakens the binding strength, rendering the adsorption more reversible, as concluded earlier in several studies (e.g., de Jonge et al., 2001; Gimsing et al., 2004).

In contrast to glyphosate, the relative amount of bound residues of glufosinate-ammonium was very low (16–20% of added herbicide) (Fig. 3) and seemed not to be affected by the soil P status. Also, the desorption behavior of glufosinate-ammonium seemed to differ from that of glyphosate, decreasing with increasing soil P status (Tables 3 and 4). This response indicating increasing adsorption tendency disagrees with the outcome of the adsorption test. This discrepancy can be explained by the desorption process, including the solution replacement steps, which diminish the salt concentration in the solution phase. Thus, the effect of K+ cations competing with glufosinate-ammonium for the cation exchange sites decreases, enhancing the adsorption ability of the herbicide. Further studies are needed to test this hypothesis.

Phosphorus Test Values in Risk Assessment
The PAC values obtained in the laboratory tests corresponded to typical values in Finnish cultivated soils, varying from 3 to 30 mg L–1 (mean, 13–14 mg L–1) (Mäntylahti, 2003). However, an order of magnitude higher PAC values are found in some fields. The developed statistical models described well the relationship between soil PAC values and sorption coefficients of glyphosate and glufosinate-ammonium in the experimental data set. Validation of the models was difficult because the independent data available were rather small and heterogeneous, including several soil types and soil layers. It is possible that the equations are not useful for all soil types. However, the models explained some of the variation existing in the validation data; therefore, the models are useful in assessing the influence of soil P status on adsorption. However, desorption results show that the underlying sorption mechanism of glufosinate-ammonium is not affected by soil P status. Thus, the use of the model for glufosinate-ammonium is questionable. The desorption study revealed that the sorption of glufosinate-ammonium was probably affected by the increase in cation exchange capacity caused by P enrichment rather than by the PAC value per se.

Because KF values are used for the classification of pesticide mobility and leaching risks, we compared our results with Finnish mobility limits (Nikunen et al., 2000). The exponential models (Equations [4] and [5]) were used for calculating PAC values corresponding to the mobility limit classes (Table 7 ). The mobility of glyphosate varied from immobile to low, being slight when the P status was at the level typical of Finnish soils (range, 3–30 mg L–1) according to Mäntylahti (2003). However, PAC values higher than 30 mg P L–1 are common (e.g., in sugar beet fields) (unpublished result of the Sugar Beet Research Centre, Finland, Eronen, 2007). In 46,600 soil samples tested, the mean values for mineral, clay, and organic soils were 37, 16, and 28 mg PAC L–1, respectively (average, 32 mg P L–1). At a given PAC value, the mobility of glufosinate-ammonium was higher than that of glyphosate. In the soils of a typical PAC level, it was medium or high. The mobility risk of glufosinate-ammonium was shown to be higher in coarse-textured soil than in clay soil.


View this table:
[in this window]
[in a new window]

  		Table 7. Mobility classification of glyphosate and glufosinate-ammonium at different levels of acid ammonium acetate–extractable phosphorus (PAC).

 
The strong adsorption and low desorption to soil suggested that the risk of glyphosate leaching through the soil matrix is minor when the soil P status is low. On the other hand, the strong adsorption reduces glyphosate bioavailability (degradation) and could lead to accumulation in soil if glyphosate application is prolonged.


   Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Material and Methods
 Results
 Discussion
 Conclusions
REFERENCES
 
On the basis of the KF values obtained in the adsorption tests, glyphosate can be considered to be a compound that is fairly immobile or of low mobility. Our results show that its adsorption decreases with increasing soil PAC value. They indicate that the risk of glyphosate leaching into water courses is higher in fields with high reserves of adsorbed P. Nevertheless, the change of soil P status is a very slow process, and therefore the glyphosate leaching potential can be expected to be more dependent on soil characteristics than on fertilization intensity. As for glufosinate-ammonium, the KF values indicated that mobility varies from medium to very high. The results of adsorption and desorption tests revealed that the retention of this herbicide was affected primarily by factors other than PAC value.

Although glyphosate is not considered to be a very toxic herbicide, its intensive use makes it one of the herbicides of most environmental concern. This study gives information that can be used in the field-specific risk assessment of herbicides. The exponential models of the relationship between soil PAC values and glyphosate were found to fit well to Finnish data sets. The results of the desorption study indicated that the equations are useful in assessing the risks caused by glyphosate, but are questionable in the case of glufosinate-ammonium.


   ACKNOWLEDGMENTS
 
This study was funded by the Ministry of Agriculture and Forestry of Finland, MTT Agrifood Research Finland, and SBRC. We thank Helvi Heinonen-Tanski and Sirpa Kurppa for their valuable suggestions concerning the study and improving the manuscript. We thank the laboratory personnel at SBRC and MTT Agrifood Research Finland for their skillful analysis and the data from the results of the analysis. We thank Sevastiana Ruusamo for revising the English manuscript. Finally, we are grateful for the anonymous reviewers of our manuscript for their comments, which improved the article significantly.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Material and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Material and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Laitinen, P.
Right arrow Articles by Hartikainen, H.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Laitinen, P.
Right arrow Articles by Hartikainen, H.
Agricola
Right arrow Articles by Laitinen, P.
Right arrow Articles by Hartikainen, H.
Related Collections
Right arrow Agricultural Pesticides
Right arrow Phosphorus
Right arrow Soil Analysis

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