Chrom und Quecksilber in einem seit 80 Jahren mit städtischem Abwasser berieselten Boden

Chromium and mercury in a soil after 80 years of treatment with urban sewage water . A soil (tab. 1) which is since 80 years under irrigation with sewage water contains 112 ppm Cr and 1.5 ppm Hg (Fig. 1). The sewage water presently contains 0.07 ppm chromium and 0.002 ppm mercury. In order to study the migration and leaching of chromium in the soil experiments were carried out with ⁵¹Cr as CrO₄^2- under aerobic and ⁵¹Cr als Cr³⁺ under anaerobic conditions in undisturbed soil columns. In all cases chromium was adsorbed in the layer 0–5 cm (Fig. 2 and 3). Similar experiments with ²⁰³Hg as HgCl₂ (Fig. 4) showed that mercury ist adsorbed near the surface. These experiments confirmed our results of ground water analyses, neither chromium nor mercury contaminated the ground water.     

Distribution and Concentration of Pb and Cd in Two Adjacent Luvisols under Grassland and Forest in Denmark

The effect of land use (forest and grassland) on the deposition and mobility of Pb and Cd has been studied by extraction with 1M HCl, 1M NH₄Ac/HAc and 1M NH₄Ac. The soils were described pedologically and the main physical and chemical parameters were determined. The Pb concentration in the A horizon is higher in the profile under forest than in the profile under grassland. We assign this difference to the filter action of the forest. The Cd concentration is highest in the grassland profile because the forest soil is more degraded and Cd is leached to a higher degree in this soil. Both metals are more mobile (easier to extract) in the forest because the forest soil is most acidic, and because a greater part is bound to organic compounds (humic and fulvic acids) which are more abundant in the forest than in the grassland. The concentrations of Pb and Cd in the A horizons are 26 ppm and 0.1 ppm under the forest and 9 ppm and 0.2 ppm under grassland, respectively.     

重庆地区紫色土和紫色泥岩物质组成与微结构研究

根据紫色土与紫色泥岩的X-射线矿物鉴定、粒度分析、电镜鉴定及野外观察,研究了紫色土与紫色泥岩的矿物组成、粒度组成及其易发生侵蚀的原因。结果表明,紫色土与紫色泥岩矿物成分与含量差异很小,均以石英和长石为主,黏土矿物中蒙脱石含量较其他矿物多;粒度成分也非常接近,以粉砂和黏粒为主。紫色土以碎屑粒状孔隙充填式胶结微结构为主,另有少量粒状接触胶结微结构和粒状块状胶结微结构,胶结物以黏土为主,仅含有3%左右的Fe2O3化学胶结物;紫色泥岩微结构的突出特点是微型片结构和微型块状结构占主要地位。蒙脱石的较高含量指示紫色泥岩与紫色土为膨胀类岩土。紫色泥岩与紫色土在湿润气候条件下的频繁胀缩是它们易于发生物理风化和易受侵蚀的主要原因,微型块状结构、微型片状结构和密集微型裂隙也是其抗蚀性差的重要原因。紫色泥岩的微型块状与微型片状结构和密集微型裂隙主要是在成岩过程中产生的,是原始沉积形成的大量球形、椭球形黏土形态受到成岩过程中挤压作用的结果。     

Mercury contamination in the Carson River,Nevada: A preliminary study of the impact of mining wastes

From 1860 to 1890, approximately 7 100 metric tons of metallic mercury (Hg) were released into the Carson River-Lahontan Reservoir watershed as a by-product of Comstock Lode silver (Ag) and gold (Au) ore refining. Present-day Hg contamination is most severe in mine tailings, where total Hg concentrations can exceed several hundred µg/g. Hg-laden tailings were also dumped directly into the Carson River, and were subsequently transported downstream into the Lahontan Reservoir and the Stillwater Wildlife Management Area. The Hg, Ag and Au contents of sediments from the Carson River and the Lahontan Reservoir are well above local background levels, and both Ag and Au contents are positively correlated to Hg. Thus, tailings-derived Hg has been redistributed throughout the entire Carson-Lahontan watershed over the last century. Total Hg concentrations in water samples from the Carson River at seven localities show that: 1) elevated (e.g., >20 ng/L) Hg levels in Carson River waters first appear downstream from accumulations of mill tailings, 2) total Hg concentrations in unfiltered and filtered water from the Carson River increase downstream (i.e. away from the tailings piles), and 3) Hg concentrations in both the Carson River (downstream from the tailings piles) and the Lahontan Reservoir are among the highest known worldwide (100 to 1000 ng/L). Filtered water samples from the Carson-Lahontan system also have high Hg contents (up to 113 ng/L), and suggest that the >0.4 µm particle fraction constitutes over 60% of the total water-borne Hg.     

Evaluation of Soil Tests for Plant‐available Mercury in a Soil–Crop Rotation System

A reliable soil test is needed for estimating mercury (Hg) availability to crop plants. In this study, four extraction procedures including 0.1 M hydrochloric acid (HCl), 1 M ammonium acetate (NH₄OAc) (pH 7.0), 0.005 M diethylenetriaminepentaacetic acid (DTPA), and 0.1 M calcium chloride (CaCl₂) (pH5.0) were compared for their adequacy in predicting soil Hg availability to crop plants of a rice–cabbage–radish rotation system. The amounts of Hg extracted by each of the four procedures increased with increasing equilibrium time. The optimal time required for extraction of soil Hg was approximately 30 min, though it varied slightly among the four extractants. The amounts of Hg extracted decreased with increasing soil/solution ratio, and a soil/solution ratio of 1:5 appeared to be adequate for soil Hg availability tests. The amounts of Hg extracted increased in the order of NH₄OAc < CaCl₂ < DTPA < HCl in silty loam soil (SLS) soil, and the order was NH₄OAc < CaCl₂ ≈ DTPA < HCl in yellowish red soil (YRS) soil. Significant positive correlations among the four extractants were obtained in SLS soil. In contrast, the correlations were poor in YRS soil, especially for HCl. There were significant correlations between concentrations of Hg in edible tissue of three plants and the amounts of soil Hg extractable to the four extractants for soil–rice system and soil–radish system, but not for soil–Chinese cabbage system. The 0.1M HCl extraction overall provided the best estimation of soil‐available Hg and could be used to predict phytoavailability of Hg in soil–crop systems.     

Modelling long-term phosphorus leaching and changes in phosphorus fertility in excessively fertilized acid sandy soils

The sound management of agricultural soils that are heavily loaded with phosphorus (P) involves minimizing the losses of P responsible for eutrophication of surface waters, while ensuring enough P for crops. This paper describes a simple model to examine the compatibility of these two objectives in acid sandy soils in a temperate humid climate. The model is based on several assumptions regarding reversible and irreversible P sorption by P-reactive soil compounds (mainly poorly crystalline Fe and Al oxides) and release of P to water (water-P test). Model inputs are amount of P leached, P uptake by crops, and contents of poorly crystalline Fe and Al oxides in soil. The model predicts that reducing the amount of leached P to what is environmentally acceptable (e.g. 0.44 kg P ha^–1 year^–1, equivalent to 1 kg P₂O₅ ha^–1 year^–1) results in the long run in available soil P test values below target concentrations for optimum crop growth. When the amount of leached P is set to a fixed value the model predicts that soils with large contents of Fe and Al oxides can maintain the initial soil P test values for longer periods than other soils. The content in available P decreases if fertilizer P is applied to the soil at a rate equal to P uptake by crops. These results stress the difficulties involved in trying to make agricultural and environmental needs compatible in acid sandy soils.     

Effect of silver from cloud seeding on rumen microbial function

Hazards of Ag from weather modification to micro-organisms of the rumen were evaluated in vitro and in vivo. Silver iodide complexes at a concentration of 38 ppm of Ag in dry alfalfa substrate did not affect substrate digestion in vitro. Inhibition of rumen microbial digestion was noted with 100 or more ppm of Ag, as Ag NO₃, in an in vitro fermentation system. A chronic intake of 1 ppm of Ag in dry-matter intake did not produce gradual accumulation of Ag in the rumens of two goats. Silver passed from a goat rumen with a half time of 1.3 days. In two in vivo trials, neither suspensions of AgI complexes simulating 1 ppm of Ag in dry food intake nor solutions of AgNO₃ simulating 100 ppm of Ag in dry food intake inhibited rumen digestion rates. An abundance of chloride ions, proteins, bacteria and other organic matter precludes persistence of Ag ions in the rumen. Based on these results, Ag in forages in areas where clouds are seeded with AgI complexes, including areas near AgI generators, where forages rarely contain more than 1 ppm of Ag in dry matter, will not affect rumen microbial digestion.     

Ability of sorption–desorption experiments to predict potassium leaching from calcareous soils

When potassium (K⁺) fertilizers are applied to the soil, K⁺ is subject to displacement through the soil profile. Leaching can play an important role in agricultural K⁺ losses that can decrease groundwater quality. To avoid overfertilization, estimation of K⁺ leaching from soil is important. The ability of the soils to retain K⁺ against leaching varies according to the adsorption coefficient of the soils. The aim of this study was to relate the K⁺ leaching from a wide range of calcareous soils to the values obtained from a sorption–desorption experiment. The soil columns were leached with 10 mM CaCl₂ solution and the leachate was analyzed for K⁺. The breakthrough curves for K⁺ were different, and the amounts of K⁺ leached varied considerably between different soils. In these calcareous soils where crops are irrigated with water containing significant concentrations of Ca²⁺ and other cations, large amounts of K⁺ will be leached. Cumulative K⁺ leached after five pore volumes leaching with 10 mM CaCl₂ was significantly (r = 0.776, p < 0.01) related to the equilibrium K⁺ concentration. The results of this study enabled us in many cases to estimate the K⁺ leaching from soil without conducting column experiments, minimizing the laborotary work.     

Building upon the Conceptual Model for Soil Mercury Flux: Evidence of a Link Between Moisture Evaporation and Hg Evasion

Parameters known to influence mercury (Hg) release from soils include substrate and air Hg concentration, light, atmospheric oxidants, temperature, and soil moisture. However, for low Hg-containing soils, the influence of these parameters has been shown to vary across space and time. Here, we expand upon previous work by investigating whether soil–water evaporative loss, which integrates the influence of multiple parameters, could be applied for predicting Hg flux from soil with low Hg concentrations when bare and planted. To investigate our hypothesis, Hg flux was measured from three soil types (<100 ng Hg g^?1). When these soils were saturated, flux was suppressed. Soil moisture evaporative stage was used to partition the parameters most important for controlling Hg flux as the soils dried. Classification and regression tree (CART) analyses showed that soil moisture was the most important parameter predicting Hg flux. Results also showed an important predictor for Hg flux was whether actual evaporation (E _a) was equal to potential evaporation (E _p) or E _a < E _p. Depending on evaporative stage, the parameters with the next highest correlation to Hg flux were light, temperature, and soil moisture evaporation rate. The presence of vegetation also influenced flux with lower Hg flux when the plants were transpiring. Results indicate for those developing models that estimate Hg flux from low Hg-containing soils, soil moisture and evaporative stage are useful tools for predicting flux.

     

Interaction of solid gold with mercury in ambient air

A number of experiments aimed at developing methods for speciated measurements of the concentrations of airborne Hg⁰, Hg^11a and Hg^11b, with sampling by Au traps and H_2O-bubblers have been performed. The results have led to some unexpected conclusions regarding the interaction of atmospheric components, including the Hg forms mentioned, and a Au surface. It is now believed that the sampling technique in question may give artifacts due to redox processes occurring in the sampling equipment during collection. These limitations are of major interest since at present Au collectors are most frequently applied in the determination of Hg in ambient air.     

SOME EFFECTS OF PARTIAL STERILIZATION ON MINERAL NITROGEN IN A LIGHT SOIL

Field plots on loamy sand with and without the soil fumigant ‘D-D’ (dichloropropane-dichloropropene mixture) were sampled at 0–12, 12–24, and 24–64 cm depths during May 1966–8. All plots received 125 kg/ha nitrogen fertilizer early in March each year and were cropped with sugar beet. Fumigation increased the total amount of mineral (NH₄⁺+ NO₃^?) nitrogen to a depth of 64 cm by 67 kg/ha on average, largely by increasing NH₄⁺-nitrogen. In unfumigated plots some mineral nitrogen leached below 12 cm every year but there was at least 125 kg/ha in the surface 64 cm in 1966 and 1968; during the wet spring of 1967 much leached below 64 cm. Sequential sampling in 1968 confirmed that some of the fertilizer nitrogen moved down the profile during spring. Plots partially sterilized either one or two years previously and not again, contained the same quantity of mineral nitrogen as plots that had never been fumigated. Fumigation increased the mineral nitrogen in plots by a similar amount whether or not they had been fumigated previously. Thus, on this soil, partial sterilization with ‘D-D’ seemed to have little residual effect on mineral nitrogen.     

Nitrous oxide emissions under different soil and land management conditions

Nitrous oxide (N₂O) emissions of three different soils – a rendzina on cryoturbed soil, a hydromorphic leached brown soil and a superficial soil on a calcareous plateau – were measured using the chamber method. Each site included four types of land management: bare soil, seeded unfertilized soil, a suboptimally fertilized rapeseed crop and an overfertilized rapeseed crop. Fluxes varied from –1g to 100g N₂O-nitrogen ha^–1 day^–1. The highest rates of N₂O emissions were measured during spring on the hydromorphic leached brown soil which had been fertilized with nitrogen (N); the total emissions during a 5-month period exceeded 3500gNha^–1. Significant fluxes were also observed during the summer. Very marked effects of soil type and management were observed. Two factors – the soil hydraulic behaviour and the ability of the microbial population to reduce N₂O – appear to be essential in determining emissions of N₂O by soils. In fact, the hydromorphic leached brown soil showed the highest emissions, despite having the lowest denitrification potential because of its water-filled pore space and low N₂O reductase activity. Soil management also appears to affect both soil nitrate content and N₂O emissions. Received: 4 April 1997     

Evaluation of efficient soil test methods for Zn and their critical vales in salt‐affects soils for rice

Abstract

Soil pot culture experiment was conducted on 22 soils of Balewal‐Phaguwala‐Narike (BPN) and 24 soils of Isri‐Langrian‐Narike (ILN) associations using rice (PR 106) as test crop at 0 and 7.5 ppm Zn levels. Chelating extractants 0.005M DTPA, 0.01M EDTA‐(NH₄)₂CO₃ and 0.05M EDTA, extracted more soil Zn than double‐acid and were significantly correlated with each other as well as with soil pH and clay in BPN and only with clay in ILN soil association. Soil CaCO₃ governed the double‐acid extractable Zn in these soils. Dry matter yield and Zn uptake by rice significantly increased with 7.5 ppm Zn application. The response was higher in ILN than BPN soil association, The DTPA method gave the highest correlation with Bray's yield and Zn uptake (r =0.72 and 0.55) followed by 0.05M EDTA (r ‐ 0.75 and 0.61) or EDTA‐(NH₄)₂CO₃ (r =0.70 and 0.61). The predictability of rice yield improved from 18–27 to 27–35, 32–43, 34–44 and 51–55 percent as a result of stepwise inclusion of pH, CaCO₃, organic carbon (OC) and clay respectively in the regression equation alongwith Zn extracted by chelating agents.

The critical levels of DTPA, EDTA‐(NH₄)₂CO₃ and EDTA extractable Zn significantly differed in the two associations and were 0.69, 0.82 and 1.24 ppm in BPN and O.BC, 1.09 and 1.42 ppm in ILN soil association. Soil properties further affected the critical levels. This for DTPA available Zn was 0.80 and 1.03 ppm in soil containing less and greater than 2% CaCO₃, 1.03 and 0.80 ppm in soils containing less and greater than 0.25% OC. These values for EDTA‐(NH₄)₂CO₃ available Zn were 1.09 and 0.91 ppm Zn in soils containing less and greater than 15% clay suggesting that critical levels of Zn for each category of soil properties should be considered while making recommendations of Zn fertilization of crops.,     

Toxicity interaction of metals (Ag, Cu, Hg, Zn) to urease and dehydrogenase activities in soils

A study of the individual and mixture toxicities of the trace metals Ag, Cu, Hg, Zn to the soil enzymes dehydrogenase and urease was undertaken. An agricultural soil and a sandy forest soil were spiked with metal salts, individually and in combinations. The anion additions to the various treatments and controls were normalized for added anions using Ca salts. The soils were then left to equilibrate and leached to reduce the excess metals and anions. Total and dissolved metal concentrations were measured concurrently in order to consider the effect of soil chemistry on the enzyme activities. Dose-response relationships for total soil metals and soil solution metals were estimated for each metal separately following a log-logistic curve fitting. Ag and Hg were the most efficient metals to reduce soil enzyme activities. The Bliss independence model was used to predict the toxicity of metal combinations. The enzyme responses in relation to the total soil metal combinations were synergistic for the agricultural soil and antagonistic for the forest soil; possibly as a result of a higher organic matter content and higher pH in the latter soil. Enzyme activities expressed in relation to the dissolved metal concentrations were more variable than against the total metal contents and consequently we observed both synergistic and antagonistic interactions.     

Rhizodeposition: Its contribution to microbial growth and carbon and nitrogen turnover within the rhizosphere

A greenhouse rhizobox experiment was carried out to investigate the fate and turnover of ¹³C‐ and ¹⁵N‐labeled rhizodeposits within a rhizosphere gradient from 0–8 mm distance to the roots of wheat. Rhizosphere soil layers from 0–1, 1–2, 2–3, 3–4, 4–6, and 6–8 mm distance to separated roots were investigated in an incubation experiment (42 d, 15°C) for changes in total C and N and that derived from rhizodeposition in total soil, in soil microbial biomass, and in the 0.05 M K₂SO₄–extractable soil fraction. CO₂‐C respiration in total and that derived from rhizodeposition were measured from the incubated rhizosphere soil samples. Rhizodeposition C was detected in rhizosphere soil up to 4–6 mm distance from the separated roots. Rhizodeposition N was only detected in the rhizosphere soils up to 3–4 mm distance from the roots. Microbial biomass C and N was increased with increasing proximity to the separated roots. Beside ¹³C and ¹⁵N derived from rhizodeposits, unlabeled soil C and N (native SOM) were incorporated into the growing microbial biomass towards the roots, indicating a distinct acceleration of soil organic matter (SOM) decomposition and N immobilization into the growing microbial biomass, even under the competition of plant growth. During the soil incubation, microbial biomass C and N decreased in all samples. Any decrease in microbial biomass C and N in the incubated rhizosphere soil layers is attributed mainly to a decrease of unlabeled (native) C and N, whereas the main portion of previously incorporated rhizodeposition C and N during the plant growth period remained immobilized in the microbial biomass during the incubation. Mineralization of native SOM C and N was enhanced within the entire investigated rhizosphere gradient. The results indicate complex interactions between substrate input derived from rhizodeposition, microbial growth, and accelerated C and N turnover, including the decomposition of native SOM (i.e., rhizosphere priming effects) at a high spatial resolution from the roots.     

Degradation and Leaching of Simazine in Arctic Sandy Soils

Abstract

Degradation and leaching of ¹⁴C-labelled simazine in coarse sandy soils at 15 + 1°C were investigated using radiometric and mass-spectrometric methods. During 6 months incubation approx. 4–7% of the applied ¹⁴C-simazine was evolved as ¹⁴CO₂. 4–9% of the simazine still remained in the soil. Addition of hen manure or acidification by addition of peat did not clearly influence the rate of degradation of simazine, whereas mechanical treatment significantly increased its degradation. In a nitrogen atmosphere the rate of degradation of simazine was reduced.

9–15% of the simazine or its radioactive metabolites leached through a 33 cm sandy moraine soil column (diameter 6 cm) in ca. 1770 mm of precipitation over a 4 month period, and 2% was leached from a fine sand soil under the same conditions.     

14C标记秸秆碳素在老成土和变性土中的转化

Hg vertial transference in soil-water system was studied by analyzing Hg vertical ditribution in soil column after adding Hg and one of the two leacheates,deionzied water or acid rain,into soil column.The results indicated that Hg was hardly transferable in puple soil.About 86%-88% of the total soil Hg was distributed in the top layer (0-2cm) and to Hg was detected in the leakage when the purple soil column was leached by deionized water and simulated acid rain.But Hg was more movalbe in yellow soil with only about 20%-22% of the total soil Hg distributed in the top layer (0-2cm),and about 17%-25% washed out from the soil column by deionized water and simulted acid rain,Incremant in soil bulk density colud reduce Hg leaching,thus the more the Hg kept in soil,the less the Hg leached into underground water,Deionized water and acid rain almost played the same role in leaching Hg.Bentioint was most effecient in preventing Hg from vertcal transferring in the soil coulumn.     

DISTRIBUTION OF FORMS OF NITROGEN IN A PODZOLIC SOIL PROFILE FROM GARPENBERG, CENTRAL SWEDEN

The chemical nature and distribution patterns of the forms of N in a podzolic soil profile from central Sweden, developed under a stand of Norway spruce, were studied. Total N, native fixed NH₄, acid hydrolysable-N, and the amounts of ammonia, hexosamine, and amino-acid N in the hydrolysate were determined. From 17 to 27 per cent of the N was insoluble in 6N HC1, the highest percentage being in the A₂ horizon. Amino-acids in the acid hydrolysates decreased from 50 per cent of the total N content of the soil in the humus layer (A₀) to 24 per cent in the B horizon. Amino-acid composition varied little in samples from different horizons. Hexosamine-N was 11–14 per cent of the total soil N, tending to increase with depth. Approximately 15 per cent of the total soil N was found in the soil hydrolysate as NH₄. Values for native fixed NH₄ extracted by N HF: N HCl were 34–80 ppm but were reduced to 10–17 ppm when corrected for the NH₄ released by N HCl only. The figures thus obtained are considerably lower than those reported by other workers.     

MODELLING LEACHING OF INORGANIC Hg(II) IN A SCANDINAVIAN IRON-HUMUS PODZOL - VALIDATION AND LONG-TERM LEACHING UNDER VARIOUS DEPOSITION RATES

Increasing mercury contents are reported from freshwater systems and fish in northern Europe and North America. Mercury input from soils is a major source with the leaching being affected by increased atmospheric mercury deposition compared to pre-industrial times and by other environmental conditions such as acid rain. The results of a mathematical model-calculation of vertical inorganic Hg(II) leaching in a Scandinavian iron-humus podzol under different atmospheric input rates of mercury are presented. Leaching under background rain conditions was calculated to be considerably stronger than under acid rain conditions. Increasing fractions of deposited soluble or solute atmospheric mercury were leached from the O _f(h)-horizon with decreasing soil content of soluble mercury under acid rain conditions; this effect was less pronounced under background rain conditions. The steady state concentrations of soluble mercury of the upper soil horizons were calculated and compared with the actual concentrations of total (= soluble + insoluble mercury) and extractable (= estimate of soluble) mercury measured in these horizons. The results indicate that even if the deposition of airborne mercury to soil is strongly reduced, the total mercury content of the soil decreases only slowly. It may take decades or even centuries before a new steady state concentration of total mercury is established in the soil. The decrease of the mercury concentration in the O _f(h)-horizon is probably largely dependent on the turnover of organic matter, binding most of the deposited airborne mercury in an insoluble form. Hence, present day mercury leaching is likely to be dominated by mercury deposited during former times and temporarily retained in an insoluble form in the organic matter.     

Soil texture effect on nitrate leaching in soil percolates

Abstract

Nitrate nitrogen (NO₃‐N), which is an essential source of nitrogen (N) for plant growth, is now also considered a potential pollutant by the Environmental Protection Agency (EPA). This is because excess applied amounts of NO₃‐N can move into streams by run‐off and into ground water by leaching, thereby becoming an environmental hazard. Soils have varied retentive properties depending on their texture, organic matter content, and cation exchange capacity (CEC). The purpose of this study was to determine the effect of soil texture on NO₃‐N retention to reduce NO₃‐N contamination in the environment. A sand, 85:15 sand:peat Greensmix, a loamy sand, and sandy clay loam soils were placed in 2×3 inch metal cylinders and soaked in a 240 ppm solution of NO₃‐N for seven days to saturate the soil with NO₃ ions. The columns were leached with water to collect 10 soil percolate samples of 50 mL each until a total volume of 500 mL was collected. Nitrate‐N was measured in each 50‐mL aliquot by automated colorimetry. The results showed that soil texture affected the retention of N03‐N in the sand, which adsorbed the least amount of NO₃‐N at 119 ppm, followed by the Greensmix at 125 ppm, loamy sand at 149 ppm, and sandy clay loam at 173 ppm. More NO₃‐N was released in the first 50 mL of the sand percolate at 63% followed by the Greensmix, loamy sand, and sandy clay loam at 58,46, and 37% NO₃‐N released, respectively. Soils with more silt, clay, and organic matter retained more NO₃‐N than the straight sand. Therefore, a straight sand would be the poorest of soil types since NO₃‐N retention was low.