Characterization and stability of transition metal complexes of chestnut (Castanea sativa L.) leaf litter

The water- and acid-insoluble fractions of a chestnut ( Castanea sativa L.) leaf litter sample and their complexes with Cu(II), Fe(III), and Mn(II) prepared in the laboratory were characterized by major elemental analysis, total Cu, Fe, and Mn content, infrared (IR), and electron spin resonance (ESR) spectroscopy. The IR spectra revealed a broad typology of functional groups (particularly carboxyls) in the solid litter, whereas the ESR spectra showed the existence of indigenous organic free radical species, inner-sphere Fe³⁺ complexes, and outer-sphere Mn²⁺ complexes. The litter exhibited a high residual binding capacity for Cu, Fe, and Mn in chemical forms of differing stability against water leaching and proton exchange. The ESR spectra of the metal complexes prepared in the laboratory indicated that Fe³⁺ and Cu²⁺ formed highly water-stable, inner-sphere complexes, whereas Mn²⁺ formed water-labile, outer-sphere complexes. Oxygen ligands of the litter were involved in metal complexation in all cases. The litter showed the highest affinity for Cu²⁺, followed by Fe³⁺ and Mn²⁺, when it was reacted with a single metal, whereas it complexed Fe³⁺ preferentially in the presence of both Cu²⁺ and Fe³⁺. Only a limited portion of the metal ions retained at the pH of distilled water remained bound in stable forms by the litter when the pH was lowered. Thus, variations of pH in forest soils will significantly affect micronutrient metal content and mobility in leaf litter.     

Effect of Oxidizing Power of Roots on Iodine Uptake by Rice Plants

Although iodine is harmful to plants, rice plants ( Oryza sativa L.) absorbed iodine more selectively than bromine. To explain this selective absorption, the authors proposed the following hypothesis based on the fact that the standard redox potential for (I₂+ 2e = 2I⁻) is lower than that for (Br₂+ 2e = 2Br⁻) and (Fe³⁺+ e = Fe²⁺), and the roots of rice plants are able to oxidize ferrous ion (Fe²⁺) into ferric ion (Fe³⁺), namely rice plants oxidize iodide ion (I⁻) to form molecular iodine (I₂) via the oxidizing power of their roots, and absorb the molecular iodine formed more selectively than iodide ion. Bromine, by contrast, is absorbed by rice plants only in the form of ion (Br⁻). According to this hypothesis, there should be a significant correlation between the oxidizing power of the rice roots and the amount of iodine absorbed. Therefore, the relationship between the oxidizing power of the roots and the concentration of iodine absorbed was studied in a water culture using 8 varieties of rice plants. Rice seedlings, 14 d after germination, were cultured in a solution containing 1 mg L⁻¹ each of iodide and bromide ions for 3 d. The oxidizing power of the rice roots was evaluated based on the amount of 1-naphthylamine oxidized by the roots. A significant correlation (0.78, n = 16, 0.1% significant level) was found between the oxidizing power and the concentration of iodine absorbed by the roots. However, no relationship was found between the oxidizing power of the roots and the amount of bromine absorbed.     

Correlations between extractable Na, K, Mg, Ca, P & N from fresh and dried samples of two Aquults

Significant increases in extractable ions resulted from air-drying and grinding samples of two infertile Aquults. Effects of the sample preparation differed markedly between ions and between the two soils. Regression equations were calculated to predict extractable ions in dried, ground samples from extractable ions in fresh, unground samples and the relationships were compared between the two soil series. Regressions were significantly different between soils for extractable PO³₄, Mg⁺⁺, and K⁺, but not for Ca⁺⁺ and Na⁺. Extractable NH ⁺₄ and NO-₃ in fresh, unground samples were not correlated with those in air-dry, ground samples of either soil. Differences in response to preparation between soil types appeared to be related to the oxidative status of these soils in the field, wherein constituents of more poorly-drained soils may be less stable to the oxidizing conditions of air-drying and grinding. Such complexities suggest that effects of sample preparation should be considered when interpreting soil nutrient data for studies of forest nutrient cycling and forest soil fertility.     

Accumulation of pyrite and Fe-rich carbonate and phosphate minerals in a lowland moor area

Pyrite and iron-rich carbonate and phosphate minerals are found in large quantities in an inland moor in western Jutland, Denmark. Sediments with up to 10.4% pyrite (FeS₂) occur well-separated from sediments with iron-rich carbonate and phosphate minerals both reaching up to ∼6–7%. Analysis of the ground water showed that water enters the moor from different sources. Water of high alkalinity (up to 2.05 meq l⁻¹) as well as water rich in sulphate (up to 79 mg l⁻¹), sulphide (up to 1.75 mg l⁻¹) and Fe²⁺ (up to 22.4 mg l⁻¹) were sampled from the upper ground water flowing into the moor. On entering the moor processes within moor sediments control authigenic mineralogy, resulting in segregated mineral precipitation.     

Pollution of carbonate soils in a Mediterranean climate due to a tailings spill

The retention walls in a pond containing the residues from the pyrite mine of Aznalcóllar (southern Spain) broke open on 25 April 1998, spilling approximately 6 × 10⁶ m³ of polluted water and toxic tailings, which affected some 55 km². Drying and aeration of the tailings resulted in oxidation, forming an acidic solution with high pollutant contents, the effects of which were studied in a calcareous soil. The infiltration of this solution markedly affected only the first 12 mm of the soil, where strong acidification caused the weathering of the carbonates, and where the fine mineral particles were hydrolysed. The SO₄²⁻ ions in the acidic solution precipitated almost entirely at this depth, forming gypsum, hydroxysulphates and complex sulphates. The Fe³⁺ ions also precipitated there, mainly in amorphous or poorly crystallized forms, adsorbing to As, Sb, Tl and Pb dissolved in the acidic solution. The Al³⁺ ions, though partly precipitating in the acidic layer, accumulated mostly where the soil pH exceeded 5.5 (12–14 mm in depth). They did so primarily as amorphous or poorly crystallized forms, adsorbing to Cu dissolved in the acidic solution. The Zn²⁺ and Cd²⁺ ions accumulated mainly at pH > 7.0 (19–21 mm in depth), being adsorbed chiefly by clay mineral. After 15 months, only the first 20 mm of the soil were acidified by the oxidation of the tailings and most of the pollutants did not penetrate deeper than 100 mm. Consequently, the speed of the cleanup of the toxic spill is not as important as a thorough removal of tailings together with the upper 10 cm of the soil.     

Retarded leaching of nitrate measured in monolith lysimeters in south-east Nigeria

At Onne in South-east Nigeria, drainage water was collected from four monolith lysimeters and analysed for nitrate. The lysimeters contained an acid sandy loam. At the start of the first rainy season two lysimeters received urea labelled with ¹⁵NO₃^– and two received no nitrogen fertilizer; all four were uncropped in the first year.
The peak concentrations of ¹⁵NO₃^– and of unlabelled (soil) NO₃^– were found after 2.5 pore volumes of water had passed through the lysimeters. Using the same soil in the laboratory after fine sieving, the peak concentration of tritiated water was found at 1 pore volume whereas nitrate leaching was retarded. The pattern of nitrate leaching was well described by miscible and immiscible models which included an adsorption coefficient for nitrate. Over the 2 years 81.4% of the ¹⁵N added at the start of the first rainy season was recovered in the drainage water.     

Phosphorus and nitrogen turnover and risk of waterborne phosphorus emissions in crop rotations on a clay soil in southwest Sweden

Abstract. Nutrient losses from arable land are important contributors to eutrophication of surface waters, and phosphorus (P) and nitrogen (N) usually act together to regulate production of Cyanobacteria. Concentrations and losses of both nutrients in drainage water from pipe drains were studied and compared in 15 crop rotations on a clay soil in southwest Sweden. Special emphasis was placed on P and it was possible to evaluate critical components of the crop rotations by flow-proportional water sampling. Total P concentrations in drainage water were generally small (0.04–0.18 mg L⁻¹), but during two wetter years out of six, high P concentrations were measured following certain management practices, including ploughing-in lucerne ( Medicago sativa L.) and fertilizing in advance without incorporation into the soil to meet the needs of several subsequent crops. This resulted in average flow-weighted concentrations of total P between 0.3 and 0.7 mg L⁻¹. In crop rotations containing green manures, green fallow or leguminous leys, there was also a risk for increased P losses after these crops were ploughed in. The losses increased in the order: cash crops < dairy with grass < dairy with lucerne < monoculture with barley < organic farming with cattle slurry < stockless organic farming with green manure. P balances varied between −9 and +8 kg P ha⁻¹ and N balances between +4 and +35 kg N ha⁻¹. The balances were not related to actual leaching losses. Phosphorus losses in drainage from set-aside were 67–82% of those from cash crops grown in ploughed and P-fertilized soil at the same site, indicating a high background P loss from this clay soil.     

Effects of the nitrification inhibitor dicyandiamide on potassium, magnesium and calcium leaching in grazed grassland

Abstract. Leaching of calcium (Ca), potassium (K) and magnesium (Mg) from urine patches in grazed grassland represents a significant loss of valuable nutrients. We studied the effect on cation loss of treating the soil with a nitrification inhibitor, dicyandiamide (DCD), which was used to reduce nitrate loss by leaching. The soil was a free-draining Lismore stony silt loam (Udic Haplustept loamy skeletal) and the pasture was a mixture of perennial ryegrass ( Lolium perenne ) and white clover ( Trifolium repens ). The treatment of the soil with DCD reduced Ca²⁺ leaching by the equivalent of 50%, from 213 to 107 kg Ca ha⁻¹ yr⁻¹ on a field scale. Potassium leaching was reduced by 65%, from 48 to 17 kg K ha⁻¹ yr⁻¹. Magnesium leaching was reduced by 52%, from 17 to 8 kg Mg ha⁻¹ yr⁻¹. We postulate that the reduced leaching loss of these cations was due to the decreased leaching loss of nitrate under the urine patches, and follows from their reduced requirement as counter ions in the drainage water. The treatment of grazed grassland with DCD thus not only decreases nitrate leaching and nitrous oxide emissions as reported previously, but also decreases the leaching loss of cation nutrients such as Ca²⁺, K⁺ and Mg²⁺.     

A constant potential titration method for studying the kinetics of Cu2+ desorption from soil and clay minerals

The kinetics of Cu²⁺ desorption from a contaminated Malaysian padi soil, English Supreme kaolinite and Wyoming bentonite are studied using a constant potential titration method. This technique involves maintaining a constant Cu²⁺-sensitive electrode potential corresponding to a Cu²⁺ ion activity below that which would naturally be supported by the soil in suspension. This is achieved by the controlled addition of complexing agent. Desorption occurs in response to the artificially reduced solution activity and the rate of the desorption reaction is revealed by the rate of addition of the complexing agent.
The desorption results were fitted to three established kinetic equations, all of which showed reasonable agreement with the data. In particular, a two-constant rate equation demonstrated some consistency between different materials. Both first-order and 'diffusion' kinetic equations appeared to apply within specific titration conditions. Each equation is discussed in terms of its implied desorption mechanism and the inconclusive nature of such mechanisms for short term (< 1 hour) desorption reactions is illustrated.     

Acid inputs from the atmosphere in the United Kingdom

Abstract. Inputs of acidity to the ground arise through two distinct routes: wet deposition which includes all acidity deposited in rain and snow and dry deposition, the direct sorption of SO₂, NO₂ or HNO₃ gases by vegetation or soil surfaces. The acidity from dry deposition of SO₂ and NO₂ is created during the oxidation of deposited SO₂ and NO₂ to SO²₄ and NO₃⁻ respectively. The areas of Britain experiencing the largest wet deposition of acidity are the high rainfall areas of the west and north, in particular the west central highlands of Scotland, Galloway and Cumbria where inputs exceed 1 kp H⁺ ha⁻¹ annually. Wet deposited acidity in the east coast regions of Britain is in the range 0.3–0.6 kg H⁺ ha⁻¹ a⁻¹. Monitoring data for rainfall acidity at rural sites throughout northern Britain show a decline in deposited acidity of about 50% during the last six years. Dry deposition is largest in the industrial midlands and southeast England and in the central lowlands of Scotland, where concentrations of SO₂ are largest. In these regions the dry deposition of SO₂ following oxidation may lead to acid inputs approaching 3 kg H⁺ ha⁻¹ a⁻¹ and greatly exceeding wet deposition.     

THE OXIDATION OF IRON SULPHIDES IN SOILS IN RELATION TO THE FORMATION OF ACID SULPHATE SOILS, AND OF OCHRE DEPOSITS IN FIELD DRAINS

Aerating pyritic soils causes acidification and the forrnation of acid sulphate soils, or cat-clay. The Oxidation of pyrite in soils is associated with the deposition in tile drains of a form of ochre quite distinct from that formed by the action of filamentous iron bacteria. Pyrite-derived ochre results from the action of Thiobacillus ferrooxidans, which, below pH 3.5–4.0, catalyses the Oxidation of Fe²⁺ and pyrite. In soils less acid than c. pH 4, pyrite oxidizes relatively slowly by chemical reactions to Fe²⁺ and SO²₄^?. Under these conditions iron enters the drains as Fe²⁺ and is there oxidized by T. ferrooicidans and deposited as hydrated ferric oxide. Once the soil becomes acid enough for T. ferrooxidans to multiply, the rate at which pyrite oxidizes increases several-fold, and at c. pH 3 iron appears in the drainage water in the ferric form. Liming seems to decrease the rate of Oxidation.     

Fixation of radiocaesium traces in a weathering sequence mica → vermiculite → hydroxy interlayered vermiculite

Radiocaesium fixation in soils is reported to occur on frayed edge sites of micaceous minerals. The weathering of mica in acid soils may therefore influence the Cs⁺ fixation process and thereby the mobility of the radiopollutant. We produced a laboratory weathering model biotite → trioctahedral vermiculite → oxidized vermiculite → hydroxy interlayered vermiculite (HIV) and quantified the Cs⁺ fixation of each mineral both in a fixed K⁺–Ca²⁺ background and in acid conditions. The transformation process was achieved through K depletion by Na-tetraphenylboron, oxidation with Br₂ and Al-intercalation using NaOH and AlCl₃. In a constant K⁺–Ca²⁺ background, vermiculite fixed 92–95% of the initial ¹³⁷Cs⁺ contamination while biotite and HIV fixed only 18–33%. In acid conditions, the interlayer occupancy by either potassium (biotite) or hydroxy-Al groups (HIV) strongly limited Cs⁺ fixation to 1–4% of the initial ¹³⁷Cs⁺ contamination. Cs⁺ fixation occurred on vermiculitic sites associated with micaceous wedge zones. Though both oxidized and trioctahedral vermiculites fixed similar Cs⁺ amounts in a constant K⁺–Ca²⁺ background (92–95%), the oxidized vermiculite retained much more radiocaesium in acid conditions (78–84% against 54–59%), because of its dioctahedral character.     

Effect of redox conditions on phosphate exchangeability and iron forms in a soil amended with ferrous iron

Application of iron (Fe) -rich amendments to soils has been proposed as a means of decreasing phosphorus (P) losses from soils. However, anoxic conditions following soil saturation are known to increase Fe and P solubility in soils, thus cancelling out the potential benefits. Our aim was to evaluate the effects of continuous oxic, continuous anoxic and alternating anoxic/oxic conditions on P exchangeability and Fe forms in soil amended with Ca(OH)₂ and FeSO₄. We incubated amended and unamended soils under these conditions for 8 weeks and measured Fe forms and P exchangeability. Under oxic conditions, addition of Ca(OH)₂ and FeSO₄ resulted in a strong decrease in P exchangeability and an increase in oxalate-extractable Fe. Mössbauer analyses suggested that an unidentified Fe oxide (D1oxide) with a strong sorbing capacity for P was precipitated. Under continuously anoxic conditions, P exchangeability and oxalate-extractable Fe increased with or without the amendments. Mössbauer analyses suggested that there was a partial dissolution of the D1oxide phase, precipitation of another unidentified Fe oxide (S3) and a reduction of structural Fe³⁺ in phyllosilicate, thereby increasing soil negative charge. These transformations resulted in a strong increase in rapidly exchangeable P. Alternating anoxic and oxic periods induced the dissolution and precipitation of iron oxides and the increase and decrease in P exchangeability. Implications of the results for limiting P losses from grassland soils are discussed.     

Seasonal variation of pyrite oxidation rates in colliery spoil

Abstract. Oxidation rates of pyrite in colliery spoil were measured under both field and laboratory conditions. Meld oxidation rates varied through the year, depending primarily upon temperature. Rates of acid release of 7–15 μmoles H⁺/day were measured in field lysimeters in the period May to November. Little oxidation of the pyrite occurred between November and May; the rates in the summer months were approximately 5–10 times those during die winter. The rate of oxidation in the summer was limited by the solubility of amorphous iron oxides. Slow oxidation during the winter is probably related to the inactivity of Thiobacillus ferrooxidans at low temperatures. The rates of acid production in the laboratory in the temperature range 0–18°C were similar to those in the field. Materials inhibiting pyrite oxidation should be added when oxidation rates are slow, so that they are not overwhelmed by large amounts of acid.     

Temporal trends in bromide release following rewetting of a naturally drained gully mire

Abstract. Recently, there has been interest in the occurrence of bromide (Br^-) in natural waters since it has been demonstrated that Br^-, in association with humic substances in raw waters, is readily incorporated into haloacetic acids in the form of organically bound bromine (Br) during water chlorination. We report results of the effects of experimentally rewetting a naturally drained gully mire on the hydrochemistry of Br^-, iron (Fe) and dissolved organic carbon (DOC) in the peat water. Results obtained over a three year period showed that rewetting substantially increased the concentrations of these solutes in the pore water, with peak values of 1 mg dm^-3 (Br^-), > 60 mg dm^-3 (Fe) and > 300 mg dm^-3 (DOC) detected in some samples after rewetting, compared with typical values < 0.05 mg dm^-3 (Br^-), < 1 mg dm^-3 (Fe) and < 15 mg dm^-3 (DOC) under the drained conditions. Bromide, Fe and DOC release were highly seasonal, with the largest concentrations observed in late-summer to autumn. However, whereas seasonal peak concentrations of Fe and DOC have since remained at these higher levels, seasonal peak concentrations of Br^- were progressively attenuated over time, suggesting the latter phenomenon is a flush effect, with no longer-term consequences for water quality.     

Vertical Distribution of Physico-Chemical Properties and Number of Sulfur-Oxidizing Bacteria in the Buried Layer of Soil Profiles with Marine-Reduced Sulfur Compounds

We observed the presence of reduced sulfur compounds in the buried soil layer of a paddy field on Sado Island, Niigata Prefecture. We sampled the paddy field soil from 0 to 300 cm depth and analyzed the physico-chemical properties of the soil and the numbers of sulfur-oxidizing bacteria and iron-oxidizing bacteria in order to elucidate both the sulfur-oxidizing mechanism and the function of sulfur-oxidizing bacteria in the subsoil. Based on the physico-chemical properties of the soil, layers 4 and 5, which were located below 1 m in depth, were found to be potential acid sulfate soils and to be under semi-anaerobic conditions. However, the concentrations of water-soluble sulfate ions in layers 4 and 5 (88.2 to 444 mg S kg⁻¹) were higher than those in layers 1 and 3 (16.1 and 8.29 mg S kg⁻¹, respectively) and a significant number of sulfur-oxidizing bacteria (10^2–6 MPN g⁻¹) was detected in layer 4. These results suggested that the oxidation of reduced sulfur compounds by sulfur-oxidizing bacteria had occurred in layer 4. Since no iron-oxidizing bacteria were detected in any layers, and it was reported that sulfur-oxidizing bacteria such as Acidithiobacillus thiooxidans could not oxidize pyrite directly, it was considered that the oxidation of the reduced sulfur compounds in layer 4 occurred through the following processes. At first, reduced sulfur compounds such as pyrite were oxidized chemically by ferric ions to intermediary sulfur compounds such as thiosulfate ions. Subsequently, sulfur-oxidizing bacteria in layer 4 oxidized these intermediary sulfur compounds to sulfate ions. However, it was considered that the oxidation rate of the reduced sulfur compounds in layer 4 was far slower than would occur under aerobic conditions.     

Leaching and balances of phosphorus and other nutrients in lysimeters after application of organic manures or fertilizers

Abstract. A long-term lysimeter experiment with undisturbed monoliths studied leaching behaviour and balances of phosphorus (P), potassium (K) and nitrogen (N) during a seven year crop rotation on four types of soil receiving inorganic fertilizers, manure and grass compost respectively. It was shown that application of manure did not lead to any direct change in nutrient leaching, unlike the application of fertilizers to soils of normal fertility. However, soil type considerably affected the nutrient concentrations in the drainage water.
Manure applied in amounts equal to the maximum animal density allowed by Swedish legislation slightly oversupplied P and N (0.5–3.5 and 18–38 kg ha⁻¹ y⁻¹ respectively) compared to the crop requirement and leaching losses for most of the soils. The relationship between lactate-soluble P in the topsoil and the concentrations of dissolved P in the drainage water was very strong. However the strength of this relationship was dependent on just one or two soils. P losses from a fertile sandy soil were large (1–11 kg ha⁻¹ y⁻¹) throughout the crop rotation and average crop removal (13 kg ha⁻¹ y⁻¹) plus the leaching losses were not balanced (average deficit 3–6 kg ha⁻¹ y⁻¹) by the addition of fertilizer, manure or grass compost. No decreasing trend was found in the P losses during seven years. However, the K deficit (average 26 kg ha⁻¹ y⁻¹) led to a significant reduction in the leaching trend from this soil. The other soils that had a smaller K deficit showed no significant reduction in the leaching of K.     

Protective role of proline against salt stress is partially related to the improvement of water status and peroxidase enzyme activity in cucumber

A salt-sensitive cucumber cultivar "Jinchun No. 2" ( Cucumis sativus L.) was used to investigate the role of proline in alleviating salt stress in cucumber. Proline was applied twice (day 0 and day 4 after salt treatment) as a foliar spray, with a volume of 25 mL per plant at each time. Plant dry weight, leaf relative water content, proline, malondialdehyde (MDA), Na⁺, K⁺ and Cl⁻ contents, as well as superoxide dismutase (SOD), peroxidase (POD), catalase (CAT) and ascorbate peroxidase (APX) activities in the plants were determined at day 8 after salt treatment. The results showed that 100 mmol L^–1 NaCl stress significantly decreased plant dry weight, leaf relative water and K⁺ contents, and increased leaf MDA, Na⁺ and Cl⁻ contents and SOD, POD, CAT and APX activities. However, leaf proline accumulation was not affected by salinity. The exogenous application of proline significantly alleviated the growth inhibition of plants induced by NaCl, and was accompanied by higher leaf relative water content and POD activity, higher proline and Cl⁻ contents, and lower MDA content and SOD activity. However, there was no significant difference in Na⁺ and K⁺ contents or in CAT and APX activities between proline-treated and untreated plants under salt stress. Taken together, these results suggested that the foliar application of proline was an effective way to improve the salt tolerance of cucumber. The enhanced salt tolerance could be partially attributed to the improved water status and peroxidase enzyme activity in the leaf.     

Potassium budgets in grassland systems as affected by nitrogen and drainage

Abstract. The main inputs, outputs and transfers of potassium (K) in soils and swards under typical south west England conditions were determined during 1999/00 and 2000/01 to establish soil and field gate K budgets under different fertilizer nitrogen (N) (0 and 280 kg ha⁻¹ yr⁻¹) and drainage (undrained and drained) treatments. Plots receiving fertilizer N also received farmyard manure (FYM). Potassium soil budgets ranged, on average for the two years, from −5 (+N, drained) to +9 (no N and undrained) kg K ha⁻¹ yr⁻¹ and field gate budgets from +23 (+N, drained) to +89 (+N, undrained). The main inputs and outputs to the soil K budgets were fertilizer application (65%) and plant uptake (93%). Animals had a minor effect on K export but a major impact on K recycling. Nitrogen fertilizer application and drainage increased K uptake by the grass and, with it, the efficiency of K used. It also depleted easily available soil K, which could be associated with smaller K losses by leaching.     

Characterization of Treatment Processes and Mechanisms of COD, Phosphorus and Nitrogen Removal in a Multi-Soil-Layering System

Characteristics of the treatment processes inside a MSL system were investigated by using a laboratory-scale MSL system, which was set up in a D 10 × W 50 × H 73 cm acrylic box enclosing "soil mixture blocks" alternating with permeable zeolite layers. For the study of the treatment processes inside the system, wastewater, with mean concentrations (mg L⁻¹) of COD: 70, T-N: 12, T-P: 0.9, was introduced into the system at a loading rate of 1,000 L m⁻² d⁻¹. Treatment processes in the MSL system were different for the COD, P and N pollutants. Eighty percent of COD was removed in the 1st soil layer among the 6 layers, and the removal rate increased as water moved down and finally reached 90% in the last layer of the system. Phosphorus concentration was lower under the soil mixture layers than under the permeable layers, presumably because P was adsorbed mainly by soil and mixed iron particles. The P concentration in water gradually decreased in the lower layers of the system. The concentration of PO₄^3--P was generally lower in the aerated MSL system than in the non-aerated one. NH₄⁺-N was adsorbed and nitrified in the upper part of the system. The NO₃⁻-N concentration was lower in water under the soil mixture layers than under the permeable layers, indicating that denitrification mainly occurred in the soil mixture layers.