| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
MTT Agrifood Research Finland, FIN-31600 Jokioinen, Finland
* Corresponding author (risto.uusitalo{at}mtt.fi)
Received for publication February 5, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: Al-QS, quartz sand with synthetic Al-(hydr)oxide coating • BD, bicarbonate–dithionite • BD-Pi, inorganic bicarbonate–dithionite-extractable phosphorus determined in the original sediment phosphorus speciation procedure of Psenner et al. (1984) • BD-PP, redox-sensitive particulate phosphorus • BD-Pt, phosphorus determined after bicarbonate–dithionite extraction, filtration, and digestion of the filtrate • DRP, dissolved (molybdate) reactive phosphorus • Eh, normal electrode potential • Fe-QS, quartz sand with synthetic Fe-(hydr)oxide coating • PP, particulate phosphorus • QS, quartz sand • TDP, total dissolved phosphorus • TP, total phosphorus • TSS, total suspended solids
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As a result of decomposing phytoplankton, limnic vegetation, or organic wastes, water near the bottom of a water body may periodically become depleted in oxygen. If organic matter continues to decompose during a period of oxygen shortage, Fe(III), among other reducible compounds, serves as an electron acceptor, changing the oxidation state and breaking up the Fe-P associations. The solubilized redox-sensitive P may be brought to the productive zone by diffusive, advective, and resuspension fluxes. The delayed reduction-induced P flux from eroded soil material may then speed up eutrophication in the same manner as the external loading of DRP or desorbable PP.
On the basis of extractions of P and Fe from sediments collected from the Canadian and Portuguese continental margins and the Chesapeake Bay in the USA, Anschutz et al. (1998) estimated that 20 to 30% of the sedimented P had been remobilized and exported to the overlying water, mainly due to the reductive dissolution of poorly crystalline iron oxides. In the Swiss lake of Sempach (Hupfer et al., 1995), settling material became enriched in reductant-soluble P due to sorption of dissolved P from the hypolimnion. However, this P pool acted only as a transient P sink and largely vanished during sediment diagenesis; Ca-P associations (HCl-extractable P) were then primarily responsible for permanent P burial in the sediment, although they accounted for only 5 to 7% of the total phosphorus (TP) in the settling particles (Hupfer et al., 1995).
These findings emphasize the need to take the redox-sensitive P pool into account when considering management strategies for controlling P transport from agricultural fields to surface waters, especially in fine-textured catchments. It is fine-textured soils that produce runoff in which the PP pool dominates the DRP pool (Turtola, 1999; Uusitalo et al., 2000). Hence, in areas where PP makes a significant contribution to the annual P loading, only after assessing the risk of reduction-induced solubilization of P from eroded soil material can we discuss erosion control and the reduction of DRP losses as alternative strategies for mitigating eutrophication of agriculturally affected waterbodies.
For PP in runoff water samples, a wide range of P speciation techniques, commonly applied to soils and sediments, may be used to selectively remove P from different compounds (Sonzogni et al., 1982), but only when a sufficient amount of suspended sediment is collected. The collection of sediment for extractions adds not only to the work load but also to the cost of large surveys of runoff waters. What is needed, then, for the estimation of PP release from eroded soil matter is techniques that can be applied to runoff waters without sediment preconcentration; this has been an underlying objective of our recent research efforts. In this paper, we report the results of a preliminary investigation of a technique for the estimation of PP release in reduced conditions, based on bicarbonate–dithionite (BD) extraction, a modification of the reductive extraction step of the sediment P speciation presented by Psenner et al. (1984).
|
MATERIAL AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
First, 0.100 mg of dithionite (Na2S2O4) and 0.025 g of bicarbonate (NaHCO3) were added to 40 mL of runoff water. The changes in the redox state were then monitored, and the pH level and solution P, Fe, and Al concentrations (without replicates) were determined. For redox and pH determinations, a 40-mL subsample was measured into a decanter glass and the salts were added. The sample was stirred slowly with a magnetic stirrer throughout the experiment. A pH electrode (Model 8155; Orion Research, Beverly, MA) and a platinum redox electrode (Orion Model 9778) were inserted into the decanter glass through holes made in the covering laboratory film. The operation of the redox electrode between each of the samples was checked with solutions made by mixing quinhydrone and pH buffers (Patrick et al., 1996). The readings obtained with the platinum redox electrode used were afterward converted to correspond to the normal hydrogen electrode potential (Eh) values at the laboratory temperature by adding 200 mV (Chateau, 1954).
The changes in the concentrations of soluble P, Fe, and Al due to chemical reduction were monitored by measuring 40-mL subsamples of runoff into 50-mL-capacity centrifuge tubes; these subsamples were independent of those used for Eh and pH measurement. After addition of the salts, the tubes were immediately capped and placed on an orbital shaker adjusted to a rotation speed of 120 rpm. After 5, 15, 30, and 60 min, one subsample of each runoff water sample was taken from the shaker and filtered through a 0.2-µm Nuclepore polycarbonate filter (Whatman International, Maidstone, UK). For the determination of soluble Fe and Al, a 5-mL aliquot of the filtrate was acidified by adding 5 mL of 1.2 M HCl to prevent Fe and Al from precipitating, and Fe and Al concentrations of the acidified filtrate were measured with an IRIS Advantage inductively coupled plasma–atomic emission spectrometer (ICP–AES) (Thermo Jarrel Ash, Franklin, MA).
For the determination of total dissolved phosphorus (TDP), an aliquot of the filtrate was digested with peroxodisulfate and sulfuric acid in an autoclave (120°C, 100 kPa, 30 min; as in TP analysis of water samples) to oxidize the excess dithionite. The digestion step was employed here because in preliminary tests major interferences due to dithionite were observed in the spectrometric P measurement unless the samples were digested. An alternative pretreatment is given in Psenner et al. (1984). The soluble P determined from the digested BD extracts was by Psenner et al. (1984) referred as BD-Pt; therefore, in Experiment 3 when BD was added to all of the subsamples, BD-Pt is used instead of TDP. In this experiment, the subsample representing time zero (Fig. 1) was the original sample without BD additions.
|
As for the iron-oxide coatings synthesized, the Munsell color for the quartz sand with synthetic Fe-(hydr)oxide coating (Fe-QS), 5YR 5/8 (air-dry sample), corresponded to that typical for (
-FeOOH) lepidocrocite (Scheinost and Schwertmann, 1999). However, compared with the formation of oxides in soils and natural waters, the chemical environment during the synthesis of Al- and Fe-QS was rather simplistic, and there is a possibility that the synthesized minerals differed by their physical and chemical properties from those found in nature (Mayer and Jarrell, 2000).
In addition to the QSs, four mineral samples consisting mainly of calcium phosphates (Table 1) were tested: an apatite sample from Siilinjärvi (eastern Finland), a biogenic phosphate sample from Morocco, and two finely ground rock phosphate samples of unknown origin and mineralogy. For background information, subsamples of the calcium-phosphate minerals were digested with aqua regia and HF in a microwave digestor (Lamothe et al., 1986) and the digest was analyzed for Ca, P, Fe, and Mn with an ICP–AES. Median particle size was measured with an LS 200 particle sizer (Beckman Coulter Ltd., Fullerton, CA) without sample pretreatment.
|
Experiment 3: Determination of Redox-Sensitive Phosphorus in Runoff Samples
To test the method when applied to water samples with a wide range of TSS concentrations, 79 runoff water samples from the Jokioinen field (a very fine, mixed Typic Cryaquept) were chemically reduced, mostly without replicates. Automated flow-proportional sampling of surface and subsurface runoff was arranged by tipping buckets by conducting 0.1% of the total flow to 23-L high-density polyethene containers. For laboratory analyses, the containers were sampled every one to three days, depending on flow volume. After collection from the field, the samples were stored at 4°C in 0.5- or 1-L polyethene bottles in the darkness until analyzed. In May 2000, 19 samples (for this study selected to represent variable PP concentrations) that were collected during spring and autumn of 1998, were analyzed for P by the BD extraction. Most of the samples (60) were, however, sampled during three dates: 10 Dec. 2000 and 20 and 30 Apr. 2001, and analyzed in June 2001. All 16 drainage plots and four surface runoff plots of the Jokioinen field were sampled during these three samplings.
The BD addition differed from the procedure used in the first two experiments in that bicarbonate and dithionite were added as solutions: each 40-mL runoff sample contained 1 mL of both chemicals, and the final bicarbonate and dithionite concentrations were the same as in the earlier experiments. This modification was made to speed up the analysis (weighing small amounts of salts for each sample was omitted) after it was found not to affect the final Eh and pH of the BD-treated samples. Dithionite was dissolved in deionized water (1.00 g of Na2S2O4 in 10 mL) by gently stirring the solution until the salt crystals were not visible. Due to the immediate onset of dithionite decomposition in an aerobic environment, this solution had to be prepared for each extraction series separately and used straightaway (only four samples were extracted simultaneously, as the number of filtering units available dictated the size of the extraction series). Bicarbonate solution (1.25 g of NaHCO3 in 50 mL) was prepared for daily use.
The final extraction procedure for the runoff waters was as follows: 40 mL of runoff water was pipetted into a 50-mL-capacity centrifuge tube, and 1 mL of 0.298 M NaHCO3 and 1 mL of 0.574 M Na2S2O4 solutions were added in this order. After dithionite addition, the centrifuge tube was immediately capped and placed on an orbital shaker adjusted to 120 rpm. After 15 min, the sample was removed from the shaker and filtered through a 0.2-µm filter. For colorimetric determination of BD-Pt, the filtrate was digested as in TP analysis. The amount of particulate P solubilized by the reduced conditions (BD-PP) was calculated by subtracting TDP in the original sample (without BD extraction) from that of the BD-treated sample (i.e., BD-Pt). A duplicate of one randomly selected sample (a subsample independent of the actual P analysis) was included in each extraction series for redox and pH control. In all cases, the Eh value of the control sample was lower than -200 mV, and the pH value was 6.5 or higher.
To assess the reproducibility of the BD-Pt data, 10 of the runoff samples were reanalyzed four more times, all extractions being performed on separate days. The minimum detectable value (detection limit, DL) of the method was deduced from the data on seven blank samples (deionized water extracted in the same manner as the runoff samples) and the following equation:
 |
The effect of the long sample storage period in the results of the BD extraction was examined by dividing the samples used in Experiment 3 into two classes: (i) those analyzed 8 to 9 wk and (ii) those analyzed 6 to 25 mo after collection. Because the BD-extractable portion of P in the sample should logically increase with PP, linear equations were fitted for BD-PP vs. PP relationships for the two groups and, after ensuring that the regression residuals were normally distributed, the hypothesis of equal slope values in the PP range of 0 to 1 mg L-1 was tested with the t test (e.g., Johnson, 1994).
The evaporation residue of a 40- to 80-mL subsample was used as a measure of TSS. The phosphorus concentrations were measured by the method of Murphy and Riley (1962) after the following pretreatments: filtration (0.2 µm) for DRP, digestion by peroxodisulfate and sulfuric acid in an autoclave (120°C, 100 kPa, 30 min) for TP, and the same digestion of a filtered (0.2 µm) subsample for TDP. The concentration of PP was taken as the difference between TP and TDP. In the samples studied the DRP and TDP concentrations were almost identical (results not shown) and hence DRP could also have been used in this calculation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Compared with the initial TDP concentration of 0.05 to 0.12 mg L-1, a three- to sevenfold rise in TDP was observed 5 min after dithionite addition (Fig. 1). The greatest TDP concentrations were typically measured 15 min after addition of the BD reagents, and in one sample the concentration of TDP clearly decreased from 15 min to 30 and 60 min reaction times. Soluble Fe acted similarly to TDP (Fig. 1).
In a preliminary experiment (results not shown), the addition of dithionite without simultaneous addition of bicarbonate caused a different reaction pattern from that shown in Fig. 1, that is, a distinct rise in Eh and a steep decline in pH between 5- and 15-min reaction times. The pH values measured were less than 4, and soluble Al increased from less than 0.1 to 1.2–2.6 mg L-1. Clearly, the addition of dithionite alone and the following decrease in pH resulted in dissolution not only of Fe-P associations, but also of other runoff sediment components. Here, we took the increased Al solubility as an indication of instability of minerals at the low pH, even though solubilization of Al does not necessarily coincide with breakup of Al-P associations. The solubility of Al remained low, less than 0.1 mg Al L-1 in all samples (not shown), when dithionite was added after bicarbonate. Adding bicarbonate in conjunction with dithionite also seemed to suppress Fe solubility slightly: 15 min after BD addition, the soluble Fe concentration was 70% of that in the samples treated with dithionite only.
Experiment 2: Release of Phosphorus from Inorganic Aluminum-, Iron-, and Calcium-Phosphorus Associations
Bicarbonate–dithionite seemed to release fresh, probably amorphous, Fe-P associations (P from Fe-coated QS) effectively, but P release from the newly synthesized Al compound studied (Al-coated QS) was insignificant due to BD addition (Table 2), as was also solubility of Al in the BD-treated runoff samples discussed above. As for calcium phosphates, BD solution extracted more P than did bicarbonate solution or deionized water from all the Ca-P minerals studied, probably due to clearly lower pH measured for all mineral samples after the BD treatment (Table 3). Relative to the total P content of the calcium-phosphate minerals, the amount of P solubilized was, however, only 0.2% at most.
|
|
Experiment 3: Extraction of Redox-Sensitive Phosphorus from Turbid Field Runoff Samples
In the field runoff waters, BD-PP responded clearly to the changes in PP concentration (Fig. 2)
. The TP concentration in these samples ranged from 0.10 to 2.62 mg L-1, the concentration of DRP from 0.019 to 0.067 mg L-1, and that of TSS from 0.21 to 3.35 g L-1. Of the TP in these samples, 73 to 99% was PP, values typical of runoff waters from the clayey soils of southern Finland (Pietiläinen and Rekolainen, 1991; Turtola and Paajanen, 1995; Uusitalo et al., 2000). In the runoff samples that contained the most PP and TSS, relatively less P was extracted by BD from a mass unit of PP (or TSS). The BD-PP concentration responded in a linear manner up to about 1.0 mg PP L-1 in runoff. For this PP range (less than 1.0 mg PP L-1; not taking the two samples having BD-PP well above the trendline in Fig. 2 into account), a linear equation (BD-PP = 0.298 x PP + 0.024) had a slightly higher r2 value (0.94) than a curvilinear one (BD-PP = 0.303 x PP0.797, r2 = 0.92). However, over the whole concentration range studied, the response was curvilinear (Fig. 2).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Second, owing to factors affecting the kinetics of P solubilization, our extraction time was shorter and the temperature of the extraction lower than in the procedure of Psenner et al. (1984). The extraction time of 1 h did not result in the release of any additional Fe or P compared with that solubilized in 15 min of shaking. Because we performed the extraction in normal atmosphere, extending the shaking time might well have an adverse effect on soluble P due to reoxidation of Fe2+ and the formation of sparsely soluble Fe-P precipitates. Mayer and Jarrell (2000) observed that the oxidation of Fe(II) (added as FeCl2 at a rate of 5.6 to 27.9 mg Fe L-1, hence, comparable with the concentrations measured in Experiment 1; Fig. 1) at pH 7 completed in two hours when air was the source of oxygen for the oxidation of Fe(II).
Extraction at several different temperatures was tested by Psenner et al. (1984). In contrast to BD extractions performed at 60 and 85°C, the risk of thermally labile P in cells of algae and bacteria being included in the estimate of Fe-associated redox-sensitive P was equally low at extraction temperatures 40 and 20°C. Psenner et al. (1984) preferred extraction at 40°C because 90% of the Fe solubilized at 85°C was still extracted at 40°C, but only about 70% at 20°C. Later, Hupfer et al. (1995) extracted BD-P from settling material and lake sediments at room temperature, instead of at 40°C as in the original method. They found that the changes in BD-P and BD-Fe during early sediment diagenesis were closely coupled. By scanning electron microscope and X-ray spectroscopy, they also detected P and Fe in the same (particulate) structures having atomic Fe to P ratios identical to the BD extracts. Thereby, to us BD extraction at 20°C seemed to be a reasonably well-justified alternative.
Our modification further differs from the original BD extraction step of Psenner et al. (1984) in that we cannot obtain a separate estimate of reduction-sensitive organic P, because the P concentration is measured on digested filtrates only. Psenner at al. (1984) separated inorganic BD extractable P (BD-Pi) from the sum of inorganic and organic BD-extractable P (BD-Pt). In their work, BD-Pi was determined after aeration of the sample into which EDTA (ethylene-diamine-tetraacetic acid) was added to complex Fe and Mn, and BD-Pt after digestion with concentrated sulfuric acid and hydrogen peroxide. Redox-sensitive organic P was then assessed as the difference between BD-Pt and BD-Pi. Even for lake sediments (e.g., Loch Leven, Scotland, with 7 to 10% organic C; Farmer et al., 1994) BD-Pi has, however, been reported to be dominant (84%) in BD-Pt (see also Gonsiorczyk et al., 1998). Because the field runoff waters studied here probably contain clearly less organic C than lake sediments typically do (preliminary unpublished data from the Jokioinen field suggest that runoff sediment contains less than 3% organic C, that is, amounts comparable with those in Ap horizon of the field), we assumed that the organic P pool contributed relatively little to BD-Pt.
With the modified method, we recorded high P recovery from pure Fe hydroxides (Fe-QS) and low P recovery from pure Al hydroxides (Al-QS) after addition of bicarbonate and dithionite. Compared with extractions with bicarbonate and deionized water, additional P was solubilized from Ca-P minerals when dithionite was added. In relation to the total amount of P in the Ca-P minerals, the amount of P solubilized was, however, insignificant. Provided that the relative effect on dissolution of fine-sized calcium phosphates is as small as suggested by the BD extraction of the apatite sample, our results suggest specific release of P from poorly crystallized Fe-P associations. The high selectivity of the extraction for poorly crystallized Fe hydroxides (not all Fe-P associations) was earlier demonstrated by Psenner et al. (1984). Even though they used greater BD concentrations than we did, they were only able to extract about 25% of the P contained in a FePO4 mineral. The authors also showed that the percentage of extractable P from hydrated AlPO4, as well as from apatite and ATP (pure compounds), was very low.
As for the applicability for runoff analysis, the detection limit of the modified method was small enough to allow the method to be used in P extraction from turbid runoff samples. Also, P solubilization from the eroded sediment responded approximately linearly to PP up to a relatively high PP concentration; in the Jokioinen field, turbidity at a PP level of about 1 mg L-1 corresponds to that in the spring or autumn peak runoff periods (Turtola, 1999). The similar PP vs. BD-PP slopes for the samples stored for 8 to 9 wk and for those stored for 6 to 25 mo suggest that, for runoff samples containing relatively much inorganic PP, BD-extractable P fraction remains reasonably stable for long periods of time. For unfiltered runoff from the Jokioinen field, Turtola (1999) only found small changes in DRP and TP concentrations during a 12-wk storage at 4°C in the dark; the samples had thus already prior to sampling (flow-weighted sampling arranged as in the present study) reached an equilibrium that was maintained during the three-month period.
In a catchment scale study, Pacini and Gächter (1999) found that an increase in the discharge volume was accompanied by a decline in the P content of suspended stream sediment. At the same time the dominant PP fraction shifted from BD-PP at a low flow rate to the more firmly bound NaOH-PP at peak flow. They related these observations to the changes in particle size distribution, geochemistry, and sediment sources in different discharge phases. The apparent curving of the BD-PP/PP relationship (Fig. 2) at the high PP concentrations in our study thus resembles that described by Pacini and Gächter (1999). This pattern could be explained by the higher relative share of "amorphous mineral material," probably including easily reducible Fe-P associations, in the fine clay (particles less than 0.001 mm in diameter) fraction of the Finnish mineral soils (Sippola, 1974), as compared with the coarse clay fraction (0.001- to 0.002-mm particles). Although the above mentioned is a more likely option, methodological reasons might also explain the curving observed in the BD-PP/PP relationship. We hypothesize that this type of a pattern could result when P initially released during extraction reacts with Al-oxides and reoxidizing Fe(II) during shaking and filtration; these possible sinks for soluble P are more abundant in the runoff samples richer in suspended soil.
In half of the samples, reproducibility of the results of the BD-Pt determinations was good, with coefficients of variation (CVs) less than 10%. For some, however, reproducibility was rather poor, as shown by CVs greater than 20%. There seemed to be no clear tendency to larger CVs in either high- or low-TSS samples, and thus there does not appear to be a simple answer to the question of the probable cause of the occasionally large variation. Moreover, our results are similar to those of Psenner et al. (1984) and Farmer et al. (1994): for example, for BD-Pt of three sediment samples, Psenner et al. (1984) reported CVs of 2, 14, and 17%. As is usually the case with weak extractants, highly repeatable results may be hard to obtain with BD procedures. It is, however, possible that contact of the dithionite solution with air (during pipetting, shaking, and filtration) is a major source of variation in the modified procedure. We attempted to restrict excessive oxidation of dithionite before the digestion step by applying gentle stirring and keeping the filtration time as short as possible; we did not, however, really control the procedure, for instance, by doing the work in an anoxic atmosphere.
Further experiments are needed to test the extraction procedure presented here, if it is to be used for the quantitative determination of redox-sensitive P. In addition to the question of whether the work should be done in an anoxic atmosphere, the possible changes in the P pools during shorter storage periods are of practical importance. The trends in Fe and P solubilization in the BD extractions appear similar to those of redox-sensitive metals and P reported for studies on flooded soils (Lefroy et al., 1993), microcosms with microbial mats (Joye et al., 1996) or sediments (Gunnars and Blomqvist, 1997), and to those measured in water columns (Kneebone and Hering, 2000); however, whether the microbial redox reactions result in a similar pattern in unconcentrated runoff samples was not verified here. In addition, further studies including runoff samples from several soils, P status, and the abundance of different P forms of which differ from that of the Jokioinen field, are required; this is especially stressed due to the lack of relevant reference materials.
Finally, we would like to comment on the potential importance of redox-sensitive PP transport as a source of the eutrophying P load from the Jokioinen field. Assuming that the eroded sediment will end up in a water body with a tendency to develop anoxia, and that the equation presented in Fig. 2 is appropriate for estimating P release before and during early sediment diagenesis, the reduction-induced P release might be as high as 2.3 times the measured DRP loss from the field. This result was obtained by using the measured P losses and the equation of Fig. 2 to recalculate the data presented by Turtola and Paajanen (1995) for the period 1987–1993. In this hypothetical case, a strategy to reduce DRP losses that would not affect erosion at the same time would probably prove to be ineffective. Because the loss of PP from the Jokioinen soil overwhelms that of DRP, PP desorbable in an aerobic environment may be a significant potential source of P for the primary producers (Uusitalo et al., 2000). Taking redox-sensitive P into account, too, further stresses the role of eroded soil as a carrier of potentially bioavailable P from clayey soils.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Ekholm, H. Rita, H. Pitkanen, P. Rantanen, J. Pekkarinen, and U. Munster Algal-available Phosphorus Entering the Gulf of Finland as Estimated by Algal Assays and Chemical Analyses J. Environ. Qual., October 29, 2009; 38(6): 2322 - 2333. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. He, C. W. Honeycutt, T. S. Griffin, B. J. Cade-Menun, P. J. Pellechia, and Z. Dou Phosphorus Forms in Conventional and Organic Dairy Manure Identified by Solution and Solid State P-31 NMR Spectroscopy J. Environ. Qual., July 23, 2009; 38(5): 1909 - 1918. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Liikanen, M. Puustinen, J. Koskiaho, T. Vaisanen, P. Martikainen, and H. Hartikainen Phosphorus Removal in a Wetland Constructed on Former Arable Land J. Environ. Qual., May 1, 2004; 33(3): 1124 - 1132. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Uusitalo, E. Turtola, M. Puustinen, M. Paasonen-Kivekas, and J. Uusi-Kamppa Contribution of Particulate Phosphorus to Runoff Phosphorus Bioavailability J. Environ. Qual., November 1, 2003; 32(6): 2007 - 2016. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
