| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
International Rice Research Institute (IRRI), Crop, Soil, and Water Sciences Division, DAPO Box 7777, Metro Manila, Philippines
* Corresponding author (sarah.johnson{at}cgiar.org)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Key Words: CEMR, cation exchange membrane resin • CFG, centrifugation • HPLC, high performance liquid chromatography • OA, organic acids • PVC, polyvinvyl chloride • PT, porous tube • PT-S, porous tube in sand • UV-VIS, ultraviolet–visible • ZTD, zero-tension displacement
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
In the specific case of anaerobic decomposition of organic matter in submerged rice paddies where a wide variety of organic metabolites are produced at low redox potential, the OA produced in largest quantities are fermentation intermediates including acetic, propionic, formic, butyric, and lactic acids (Wang et al., 1967; Gotoh and Onikura, 1971; Chandrasekaran and Yoshida, 1973; Rao and Mikkelsen, 1977b; de Sousa et al., 2002). These OA generally have very short residence time in soil solution, after which they are converted to CO2 or CH4. Although organic matter incorporation can enhance soil fertility and sustainability, these fermentation OA are believed to have adverse effects on growth and yield of rice (Cannell and Lynch, 1984). To isolate effects of potentially toxic OA from other possible growth-inhibiting effects of straw decomposition (such as N immobilization), it is necessary to be able to measure the transient OA in soil solution at various times in the decomposition and plant growth processes.
When sampling soil solution for analysis of its chemical components, the primary task is removal of the solution from the soil without changing its chemical composition. Soil solution sampling methods can be grouped into categories according to the type of force used to separate solution from soil: (i) zero tension, which relies on gravity flow, (ii) medium tension induced by vacuum suction, and (iii) variable tension induced by centrifugation (usually at higher tension than vacuum suction) (Tiensing et al., 2001). Zero-tension sampling can be performed either by removing soil from the field and packing it into a column, known as displacement sampling, or by installing a pan in the field underneath the desired part of the soil profile, known as lysimetry. In either case, solution cannot be collected until the soil above the collection device is saturated and water is flowing through macropores. Medium-tension suction sampling can likewise be performed either on samples that have been removed from the field, by applying suction to the bottom of the displacement sampling column, or on undisturbed samples in the field by installing different types of lysimeters through which soil solution can be extracted by means of a vacuum pump. Centrifugation sampling involves destructive removal of soil cores followed by centrifugation to separate the solution from the soil. This method is classified as "variable tension" because centrifugal force extracts solution from pores of various sizes (various matric potentials) (Tiensing et al., 2001).
When solutions collected in the field by various sampling methods have been compared in terms of chemical composition, tension sampling methods (both suction and centrifugation) have usually demonstrated significantly different concentrations of most measured ions as compared to zero-tension methods. Studies are not consistent, however, regarding which methods result in higher concentrations of specific ions, because those differences apparently depend on the composition of each system's plants and organic matter (Marques et al.,1996; Goyne et al., 2000; Haines et al., 1982; Giesler et al., 1996). When different tension methods have been compared, higher nutrient concentrations have usually been reported for centrifugation than for suction lysimetry (Tiensing et al., 2001; Zabowski and Ugolini, 1990), although in the study comparing two different centrifugation speeds, there was no significant difference between them (Zabowski and Ugolini, 1990). Except in one case in which contamination of the soil solutions from one type of sampling device was demonstrated (Goyne et al., 2000), these differences have been variously attributed to: (i) real differences in the fraction of soil solution sampled at different tensions (Zabowski and Ugolini, 1990; Marques et al., 1996; Tiensing et al., 2001; Haines et al., 1982), and (ii) artificial differences (dilution) caused by preferential flow through larger pores created during installation of zero-tension equipment (Giesler et al., 1996). Higher tension sampling is expected to remove soil water that is more tightly bound to soil particle surfaces (Zabowski and Ugolini, 1990). Since plant roots can also remove some nutrients bound to soil surfaces, tension sampling is considered more representative of soil solution available to plants (Marques et al., 1996). To develop a suitable method of soil solution sampling for transient OA, it was necessary to determine if any of the minor differences between sampling methods might cause artificial differences in OA analysis and if the OA concentration might differ according to sampling tension (i.e., if OA concentrations vary in different fractions of soil solution).
Several analytical techniques have been employed in OA analysis of various types of solutions, including soil solutions. These analytical methods include titrametric analysis (Dilallo and Albertson, 1961), gas chromatography–mass spectrometry (GC-MS) (Lynch, 1980; Fan et al., 1997), ion chromatography (Shen et al., 1996), and HPLC (Lawongsa et al., 1987; Fox and Comerford, 1990). Since the focus of this study was the soil solution sampling method rather than the OA analysis method, all samples were analyzed by HPLC using an ion exclusion column (Fox and Comerford, 1990).
The objective of this study was to develop a suitable method for OA analysis in saturated rice paddies. The study compared the following soil solution sampling methods in terms of OA detection and recovery: (i) zero-tension displacement, (ii) centrifugation, and (iii) suction lysimetry. It further addressed the effect of several sample properties on OA recovery. These sample properties were categorized as: (i) those related to essential variables between sampling methods, including pH, changes in dissolved gas concentration with vacuum-induced tension, and possible adsorption onto sand or soil particles; and as (ii) those related to possible interferences inherent in all methods, such as concentration of salts or soluble Fe and the use of cation exchange membrane resins.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Zero-Tension Displacement Sampling Method
Sampling soil solution via saturated gravity-flow is considered to be a zero-tension sampling method, since no force other than gravity is exerted to remove the solution from the soil matrix (Tiensing et al., 2001). A zero-tension sampling port 3 cm above the base of rectangular plastic pots (51 by 28 by 36 cm) was fabricated as illustrated in Fig. 1
(Port 1). Ten-centimeter glass tubing (0.5-cm inner diameter) was attached to 60-cm silicon tubing. The glass tube was inserted into the pot at 10-cm soil depth through a bored rubber stopper.A 0.5-cm thick glass wool layer (20 by 10 cm) was wrapped around the glass tubing before placing it in a 1- to 2-cm sand layer. With these layers in place, the soil slurry was slowly transferred into the pots. During sampling, the ZTD ports were opened completely and allowed to drip. The first 10-mL drip was discarded. The next 10-mL drip was collected, filtered through a 0.45-µm membrane filter (Millex-HV, PVDF, Millipore, Billerica, MA) and analyzed by HPLC as described below.
|
Porous Tube Sampling Method
This sampling method is considered to be a medium-tension method, with lower tension (approximately 1000 kPa) than centrifugation, due to the use of a vacuum suction system to remove solution from the soil (Alberto et al., 2000; Tiensing et al., 2001; Meijboom and van Nordwijk, 1992). Rhizon soil solution samplers (Eijkelkamp Agrisearch Equipment, Giesbeek, Netherlands) consist of a 13-cm porous tube connected to a 60-cm clear polyvinyl chloride (PVC) tubing with 1-mm internal diameter. The porous tube (pore size 0.1 µm; outer diameter 2.5 mm) was PVC outside and polyethylene inside. The tubes were installed in experimental pots at 10-cm depth through a hole bored in a rubber stopper, which held the protective springs inside the pot as shown in Fig. 1 (Port 3). The first 2 to 3 mL of soil solution was discarded before sample collection into evacuated tubes via syringe. All collection tubes were flushed with N2 for 30 min and then evacuated to 76 cm Hg (approximately 1000 kPa) before use.
Porous Tube in Sand Variation on Porous Tube Sampling Method
This sampling method is not considered to be an alternative method for collecting soil solution samples, but is only included in this set of experiments for the purpose of comparing directly between PT and ZTD methods by placing a PT sample rat 10-cm depth in the same sand layer as the ZTD sampling port (see Fig. 1, Port 2). The method of sample collection from porous tube in sand (PT-S) ports was exactly the same as that described above for normal PT ports.
Organic Acid Analysis by High Performance Liquid Chromatography
Before HPLC analysis, it was necessary to filter samples through a 0.45-µm membrane filter (Millex- HV, PVDF) before injection to avoid clogging of the column with soil particles, which increase column pressure and noise. Organic acid separation was achieved by use of an ion exclusion column (Fox and Comerford, 1990), a Shodex RS-Pak KC-811 column (Waters Corporation, Milford, MA), with 0.0033 M H3PO4 as mobile phase. Organic acid quantification was accomplished using an ultraviolet–visible (UV-VIS) spectrum detector (Waters 2487 Dual Absorbance detector) attached to the column, set at 210 nm. The column temperature was set at 40°C at 0.75 mL min–1 flow rate, with typical column pressure of 4500 to 4800 kPa (650–700 psi). To achieve separation of the longest-retained OA, butyric acid, each sample was run for 20min. Using Jasco-Borwin interface software (v. 1.5) (JMBS Developpements, 1998), it was necessary to manually assign peak names according to retention time and to manually adjust baselines to remove noise from soil solution samples.
Effect of Sampling Method on Organic Acid Recovery from Soil Solution
Using Rice Straw as Organic Acid Source
A 256-kg soil slurry (wet weight, with moisture content of 0.5–0.6 g water g dry soil–1) was divided into two 128-kg portions. The first 128-kg portion was distributed into four pots of 32-kg portions. These served as control pots for the basal concentration of the OA present in the soil samples. The second 128-kg portion was incorporated with 560 g of rice straw (var. IR68) that had been cut into pieces of 5 cm in length. This quantity represented an approximate rate of 10Mg ha–1, calculated on the basis of pot area (0.14 m2). Straw samples were allowed to decompose under flooded conditions at approximately 35°C daytime greenhouse temperature for 3 d before soil solution sampling. Samples were collected at the time of maximum OA production, which was 3 d after straw incorporation (subsequent data not shown), slightly earlier than expected based on previous experiments due to the low initial redox potential of the pre-flooded soil. Samples were taken simultaneously for the three sampling methods used in this part of the experiment: ZTD, PT, and PT-S. Centrifugation was not included because the experimental design would not permit destructive sampling. Three sample replicates were taken from each of the four replicate pots.
Using Reagent Organic Acid Added to Soil as Organic Acid Source
Preliminary tests (data not shown) were used as a basis for determining the amount of reagent acetic acid to add to the soil to produce a measurable peak on the HPLC without creating an unreasonably high concentration: 300 mL glacial acetic acid was added into 128 kg soil slurry, followed by mixing to uniformity before distribution into four replicate pots of 32 kg. The soil was allowed to equilibrate with the added acid for 4 h before simultaneous sampling was done using all four methods: ZTD, PT, PT-S, and CFG. Four sample replicates were taken from each of the four replicate pots.
Systematic Comparison of Variables between Sampling Methods
Effect of Solution pH
A mixture of 1.05 M acetic, 1.31 M butyric, and 2.1 M propionic acids was prepared at three different pH levels from concentrated liquid reagents: pH 2.5 (requiring no adjustment), pH 6.5, and pH 10 (adjusted with NaOH). Initial OA concentrations were selected to produce similar- size peaks of each acid in spite of the fact that different OA responded differently to UV-VIS detection at 210 nm. The OA concentration in all solutions was quantified by HPLC with four replicates.
Effect of Gas Pressure–Induced Tension
A mixture of 1.05 M acetic, 1.31 M butyric, and 2.1 M propionic acids was prepared from concentrated liquid reagents, and then divided into eight 10-mL subsamples in test tubes. Half of the tubes were sealed with a PT sampler inside, followed by transfer of the standard OA solution into an evacuated tube as described above for PT method. The solutions were then transferred under atmospheric pressure to HPLC auto-sampling vials. The other half of the samples were transferred to the HPLC vials at atmospheric pressure, without the intermediate step of suction through the PT sampler. Organic acid concentration in all solutions was quantified by HPLC with four replicates.
Test for Organic Acid Sorption to Sand or Soil
A 0.5-cm layer of air dry sand or soil was placed in a Buchner funnel lined with Whatman No. 40 filter paper. A mixture of 1.05 M acetic, 1.31 M butyric, and 2.1 M propionic acids was prepared from concentrated liquid reagents, of which 250 mL was passed through each funnel. Control samples using filter paper or funnel alone were also used for comparison. The OA concentration in all solutions was quantified by HPLC withfour replicates.
Tests for Possible Interferences across All Sampling Methods
Effect of Fe and NaCl Interference
From preliminary studies, the highest Fe concentration analyzed in soil solutions was 50 mg L–1 (data not shown). To test the effect of excessively high concentrations of Fe on OA recovery, Fe was added at concentrations of 300 and 500 mg L–1 and NaCl at double those concentrations (resulting in NaCl concentrations of 0.09 and 0.18 M) to a solution containing a mixture of 1.05 M acetic, 1.31 M butyric, and 2.1 M propionic acids prepared from concentrated liquid reagents. The OA concentration in all solutions was quantified by HPLC with four replicates.
Effect of Cation Exchange Membrane Resin Strips
To test the effect of cation exchange membrane resin (CEMR) strips (BDH, Inc., Toronto, ON) on the amount of OA measured in soil solutions, four replicate samples were taken from each ZTD port, with CEMR strips added to only two of the four sample collection containers. To test the effect of the CEMR strips on OA recovery in soil-free solution, a mixture of 1.05 M acetic, 1.31 M butyric, and 2.1 M propionic acids was prepared from concentrated liquid reagents and then divided into 10-mL subsamples in eight test tubes. To half of these tubes was added one CEMR strip each, while no strip was added to the other half of the tubes. The tubes were covered, mixed thoroughly, and allowed to stand for 1 h to equilibrate before transfer to vials for HPLC analysis. The OA concentration in all solutions was quantified by HPLC with four replicates.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Systematic Comparison of Variables between Sampling Methods
Variables between sampling methods fall into three categories: (i) differences caused by collection of samples under tension into an evacuated tube (as in PT and PT-S) vs. without tension into a tube filled with N2 at atmospheric pressure (as in ZTD), (ii) differences caused by passing the samples through a buried sand layer before collection, and (iii) differences between collecting samples with a small amount of particulate soil material before filtration into the HPLC vials (ZTD and CFG) and collecting soil solution already filtered during sampling by passing through porous tubing (PT and PT-S). Centrifugation method involved a different type of collection under tension, using centrifugal force rather than gas pressure to produce the tension. This experiment was designed to isolate the effects of gas pressure–induced tension, because of the possibility of an artificial change in chemical composition when dissolved gases were released into the vacuum.
The hypothetical effects of inducing tension in the soil solution sample through changes in gas pressure (i.e., through sample collection into an evacuated tube, as in PT and PT-S) include change in sample pH or change in OA concentration. A change in pH is likely during collection into a vacuum due to the release of dissolved gases, particularly CO2, into the headspace in the tube. As CO2 leaves soil solution, the pH of the solution would be expected to increase (Ponnamperuma, 1972). The change in pH would not affect the concentration of OA in soil solution, but it would affect the percentage protonation of OA, according to the pKa values of each (Table 1). Since the HPLC method used an ion exclusion column, it was possible that an increase in the percentage of OA in ionic form (i.e., an increase in pH) would cause a decrease in the amount of OA detected if the ions passed through the column quickly rather than being retained and separated as they would be in their uncharged protonated form.
|
Since the most important issue to understand was if the percentage of OA in ionic form would affect the measured concentration of OA in solution, this effect was tested by adjusting known concentrations of OA standards to three different pH values (representing nearly 100% protonated uncharged forms at natural pH 2.5, a mixture of uncharged and ionic forms at pH 6–7, and nearly 100% deprotonated ionic forms at pH 10) and comparing the percentage recovery of the OA during HPLC analysis. A comparison of acetic, butyric, and propionic acids at these three different pH levels (Fig. 3A ) indicated that there was no effect of pH on recovery of any of the OA by HPLC. This lack of pH effect is most likely explained by the presence of the mobile phase of 0.0033 M H3PO4 (pH 2.45), which presumably changed the pH at the time of injection into the column of the small injected sample volume (10 µl) until it was below the pKa of all of the acids (see Table 1). It was concluded that although the effect of sampling method on solution pH was unknown, differences in pH would have had no effect on measurement of OA by our HPLC separation and analysis method. This conclusion also implies that variation in soil solution pH (presampling), which could affect OA toxicity to plants (Rao and Mikkelsen, 1977a), would not affect OA separation and quantification by HPLC.
|
|
When solutions of known concentrations of reagent OA were passed through layers of sand or soil in a funnel, OA recovery was approximately 100% for sand and soil (Fig. 3C), indicating that there was no OA sorption on either layer. It was concluded that the presence of soil particulate matter in the samples of some methods had no significant effect on OA recovery, and that sorption of OA to sand was not the cause for the low OA recoveries observed in the two methods, that used a sand layer.
In summary, it was concluded that among the three categories of differences between sampling methods, only the differences related to collection of samples through a layer of sand vs. directly from soil could explain the pattern of high OA recovery with PT and CFG but low recovery with PT-S and ZTD (Fig. 2). Since sorption of OA onto the sand was both theoretically unlikely (McBride, 1994) and unobserved in our experiment (Fig. 3C), and since it was shown that neither changes in pH nor gas pressure during sampling would cause artificial changes in OA concentration (Fig. 3A,B), it was concluded that the observed differences between methods represent real differences in OA concentration in different soil waters (Giesler et al., 1996). Specifically, tension sampling (PT and CFG) removed more of the soil solution that is bound to the soil surface, which contained higher concentration of OA. Zero-tension sampling (ZTD) removed solution that passed through soil macropores into the sand layer around the sampling port during saturated gravity flow. Tension sampling from the buried sand layer (PT-S) removed soil solution with OA concentration similar to zero-tension sampling (ZTD), indicating that the method sampled the same fraction of the soil solution as ZTD. The soil solution that passed through the macropores was the fraction sampled from the sand layer. Sampling with tension in the buried sand layer (PT-S) drew out much less of the soil-particle bound fraction of the soil solution as compared to sampling with tension in the soil (PT).
Tests for Possible Interferences across All Sampling Methods
Regardless of sampling method, the soil solution samples contained many other unknown components in addition to OA, resulting in a large peak at early retention time on the HPLC chromatograms. This first peak presumably consisted of ionic soil solution components, including many different salts, with a significant quantity of Fe, which would have been present in original solution samples as Fe2+ but was oxidized to Fe3+ on exposure to the air during filtration. The early peak had a shorter retention time than oxalic acid, allowing the two peaks to be separated when the early peak was small, as in the OA standard mixture (Fig. 4A). But when large enough, as in soil solution samples, the early peak merged with the oxalic acid peak (Fig. 4B). This observation has two implications: (i) in most soil solution samples, it will be difficult to accurately quantify oxalic acid by ion exclusion HPLC, and (ii) it would be possible to mistakenly identify the large mixed-ion peak in soil solution samples as oxalic acid.
The large early ionic mixture peak in the soil solution samples might be expected to affect OA separation and quantification by HPLC in two ways: (i) the ions making up the early peak might form complexes with OA in solution, causing the OA to become ionic and pass through the column without separation or causing the OA to interact differently with the detector resulting in changes in UV-VIS detector response calibration, or (ii) the high concentration of salt, with or without specific interactions with OA, might be retained in the column long enough to mask later peaks. The effect of the former problem would be a decrease in the measured amount of OA in solution, while the latter problem could artificially increase the OA recovery. If the large early peak contained a significant amount of oxalic acid, it might also be possible for oxalic acid to be retained in the column long enough to mask later peaks, which could artificially increase recovery of other OA with short retention times.
The effects of exaggerated quantities of Fe3+, nonspecifically interactive salts (represented by NaCl), and oxalic acid on the recovery of several reagent OA standards with varying structures and retention times were tested. No significant change in separation and quantification of known amounts of propionic acid was observed (Fig. 5A ). Presumably, since the mobile phase (0.0033 M H3PO4) pH was approximately 2.45, all the OA passing through the column were protonated and Fe complexation was minimized. A small but statistically significant decrease (by ANOVA at the 0.05 level) was observed in the recovery of acetic acid in the presence of oxalic acid (Fig. 5A). Since there was no evidence of peak overlap on the chromatogram, this apparent interference is not expected to be an important problem in future analyses. It was concluded that all of the OA in our experiment, with the exception of oxalic acid because of its overlap with the early ion peak, could be accurately separated and quantified by HPLC, even in the presence of a high concentration of salts. It should be emphasized, however, that the unknown ionic peak present in all soil solution samples changed the shape of the baseline of the chromatograms. To accurately quantify the amount of OA in these samples, it was necessary to use the computer interface software to manually redraw the baselines before using the integrated peak area to determine OA concentration.
|
It was concluded that: (i) the CEMR strips could be used safely without removing OA from solution, and (ii) the CEMR strips were essential for higher recovery of OA from soil solutions. The hypothesis for their usefulness with soil solutions was that the CEMR strips would remove excess Fe from solution so that it would not precipitate out as it oxidized from soluble Fe2+ to insoluble Fe3+. However, our results showed that the presence of high concentrations of Fe3+ in soil-free solution did not affect OA measurement (Fig. 5A), even though the removal of Fe2+ and Fe3+ from soil solution improved OA recovery (Fig. 5B). This apparent contradiction can most likely be explained by the fact that the high concentrations of Fe3+ were added to the soil-free solutions at very low pH (Fig. 5A), minimizing complexation with OA, while the presence of Fe3+ at the natural pH of the soil solution during collection and its subsequent precipitation would be expected to result in some Fe–OA complexation. The high amount of Fe present in soil solution during sample collection decreased OA recovery due to complexation and precipitation of OA with Fe3+. This problem could be counteracted with use of CEMR strips, and Fe remaining in soil solution during HPLC analysis had no effect on the separation or quantification of OA.
Other Variables that Affect Organic Acid Measurement in Soil Solutions
In addition to the variables tested and discussed above, we also observed several other variables that deserve mention. Solution samples must be analyzed by HPLC within approximately 24 h of collection, and stored at 4°C during the intervening hours. It was useful to place samples in ice in the greenhouse during sample collection, followed by immediate transfer to a refrigerator, and to maintain the HPLC autosampler at 4°C during analysis. If samples became too warm, a precipitate (presumably composed partly of Fe3+) formed in the tubes in spite of their lack of direct exposure to oxygen. It is also possible that OA began to decompose when warm. Our HPLC analysis of samples within 24 h after collection followed by a repeat analysis of the same samples 3 d later (after continuous refrigeration at 4°C) revealed that the OA content of the samples had changed completely (data not shown).
Two other variables that significantly affected the amount of OA measured in soil solution were the depth and timing of sample collection (data not shown). Variation due to depth was expected because the production of these OA is redox dependent and redox potential varies with soil depth (Patrick and DeLaune, 1972). Variation due to timing of sample collection was anticipated because OA are formed as intermediates in decomposition processes, implying temporary residence time in soil before conversion to other compounds (Acharya, 1935). Decisions about sampling depth and timing should be made after careful consideration of experimental objectives.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
It is unclear which soil solution fraction would be more relevant to OA toxicity studies with rice seedlings. Since tension-sampled soil solution is reputed to be more plant available (Marques et al., 1996), the higher concentrations of OA obtained by these methods may be more important in determining phytotoxicity. However, it is conceivable that the mechanism of toxicity might be more related to macropore OA concentration, in which case the zero-tension measurements would be more relevant. Most of the early work assessing OA concentrations in soil in relation to plant toxicity used water or dilute H3PO4 acid extractions, rather than soil solution sampling (Gotoh and Onikura, 1971; Rao and Mikkelsen, 1977b; Chandrasekaran and Yoshida, 1973). The design of an experiment to test effects of OA on plant toxicity in the field would require an easily measurable plant parameter indicative of toxicity, which could be correlated to tension and nontension soil solution sampling methods to determine which soil solution fraction is more relevant.
Received for publication March 7, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
= 0.05, and are the same for millimolar and percentage recovery data.
