Search

Structure liming aims to improve soil structure (i.e., the spatial arrangement of particles and pores) and its stability against external and internal forces. Effects of lime application on soil structure have received considerable interest, but only a few studies have investigated effects on macro- and mesopore networks. We used X-ray computed tomography to image macropore networks (o ≥ 0.3 mm) in soil columns and mesopores (o ≥ 0.01 mm) in soil aggregates from three field sites with (silty) clay soils after the application of structure lime (3.1 t ha−1 or 5 t ha−1 of CaO equivalent). Segmented X-ray images were used to quantify soil porosity and pore size distributions as well as to analyse pore architecture and connectivity metrics. In addition, we investigated the amount of readily dispersible soil particles. Our results demonstrate that structure liming affected both, macropore networks and amounts of readily dispersible soil to different degrees, depending on the field site. Significant changes in macropore networks and amounts of readily dispersible soil after lime application were found for one of the three field sites, while only some indications for similar changes were observed at the other two sites. Overall, structure liming tended to decrease soil macroporosity and shift pore size distribution from larger (e>1.0 mm) and medium sized macropores (e0.3–1.0 mm) towards smaller macropores (e0.1–0.3 mm). Furthermore, liming tended to decrease the critical and average pore diameters, while increasing the surface fractal dimension and specific surface area of macropore network. Structure liming also reduced the amounts of readily dispersible soil particles. We did not find any changes in mesopore network properties within soil aggregates or biopore networks in columns and aggregates. The effects of lime on macropore networks remain elusive, but may be caused by the formation of hydrate phases and carbonates which occupy pore space.

With time, different soil-forming processes such as weathering, plant growth, accumulation of organic matter, and cultivation are likely to affect phosphorus (P) speciation. In this study, the depth distribution of P species was investigated for an agricultural clay soil, Lanna, Sweden. Small amounts of apatite-P was demonstrated in the topsoil whereas the speciation of Pat 70-100 cm depth consisted of approximately 86% apatite according to P K-edge XANES (X-ray absorption near-edge structure) spectroscopy. Because there were only minor differences in bulk mineralogy and texture, these variations in P speciation were interpreted as the result of apatite weathering of the topsoil. Speciation modeling on soil extracts supported this idea: hydroxyapatite was not thermodynamically stable in the top 50 cm of the soil. Apatite was enriched in the bulk soil relative to the clay fraction, as expected during apatite dissolution. Combined results from batch experiments, XANES spectroscopy and X-ray diffraction suggested chemical transformations of the topsoil as a result from accumulation of organic matter and airing from tillage followed by enhanced weathering of apatite, amphiboles, clay minerals, and iron oxides. This caused the formation of poorly crystalline secondary iron and aluminum (hydr)oxides in the topsoil, which retained part of the released P from apatite. Other P was incorporated into organic forms. Furthermore, the results also showed that short-term acidification below the current pH value (below 5.5 in the topsoil and 7.2 in the deeper subsoil) caused significant solubilization of P. This is attributed to two different mechanisms: the instability of Al-containing sorbents (e.g. Al hydroxides) at low pH (in the topsoil), and the acid-mediated dissolution of apatite (the subsoil). (C) 2016 Elsevier B.V. All rights reserved.

Risk assessment for elevated leaching losses of phosphorus (P) from agricultural land is commonly based on indices, since such losses are highly episodic and difficult to predict. Here a flashiness index (FI) representing changes in daily water flow from drainage systems was estimated from measured discharge (agrohydrological years 2004–2013) after reconstruction of subsurface drainage systems in 16 fields on a former swine farm. The fields were analysed for ammonium lactate-extractable soil P (P-AL), clay, carbon and other soil parameters in 2004. Transport of total P (TP), dissolved reactive P (DRP) and unreactive P (UP) was estimated from concentrations in composite water samples taken flow-proportionally up to 20 times per year. On average, 2.20 kg TP ha−1 yr−1 was leached, with 27% in DRP form, from the entire farm. FI was significantly negatively correlated (Pearson correlation coefficient p < 0.05) to mean yearly discharge from each field. Stepwise regression demonstrated that FI index was the most important single explanatory parameter for flow-proportional yearly mean concentration of unreactive P losses (UP) from each field, with a coefficient of determination (r2) of 0.67. The corresponding concentration of dissolved reactive P (DRP) was significantly positively correlated (p = 0.015) to soil P-AL and FI. A regression model for TP leaching losses based on FI, P-AL and yearly discharge (Q) from 11 of the fields over nine years (r2 = 0.67, p = 0.002) was validated against TP leaching from the remaining five fields (32% of farm area). Root mean square error (RMSE) was 0.43, which represented 20% of measured leaching (mean 2.14 kg TP ha−1 yr−1). For individual years, RMSE for different fields was 37–80% of measured TP leaching (0.8–3.7 kg TP ha−1 yr−1). The FI index could be used together with soil P test to predict P leaching from individual fields of a drained farm.

Due to rapid growth of the horse industry in the EU, manure production from horse farms has become substantial. Horse manure is a valuable resource in terms of plant nutrients that should be utilised primarily as a fertiliser and soil conditioner on agricultural soils, but in an environmentally friendly manner. This study estimated potential retention and leaching losses of phosphorus (P), nitrogen (N) and carbon (C) from two mineral (Nantuna loamy sand and Wiad loam) and one organic (Hidinge peat) agricultural topsoil amended with two levels of composted horse manure, 18 and 36 megagram (Mg) ha − 1 , compared with non-amended control soil. Soil columns (20 cm long, 18.8 cm diameter) were exposed to three consecutive simulated rainfall events with an intensity of 8 mm h − 1 and a pore volume of 100 mm each time, which resulted in about 280 mm total discharge. Results showed that the background losses from Hidinge peat were several-fold higher (8 kg P, 122 kg N, 235 kg C ha − 1 ) than from Nantuna loamy sand (0.05 kg P, 2 kg N, 12 kg C ha − 1 ) or Wiad loam (1.6 kg P, 24 kg N, 19 kg C ha − 1 ). On applying composted horse manure the leachate concentrations of P, N and C significantly increased for the mineral soils, but only at the high application rate of 36 Mg ha − 1 . Unexpectedly, P, N and C concentrations in leachate from organic soil decreased after manure addition, indicating greater potential for retention than leaching. The increased amounts of P, N and C lost from Nantuna loamy sand and Wiad loam corresponded to less than 10% of the amount added via composted manure, indicating high potential for retention. We concluded that composted horse manure can improve soil nutrient status, as potential nutrient retention was higher than leaching from manure-amended soils, and that information on background nutrient content in soil may be more important than data on nutrient load applied through amendments when assessing retention and/or leaching potential in manure-amended soil.

Seven constructed wetlands (0.05–0.69 ha), situated in agricultural catchments (22–267 ha) in the south of Sweden, were studied for two years with two aims: to (i) quantify their function as sinks for particles and phosphorus (P) lost from the catchments, and (ii) investigate to what degree catchment and wetland characteristics and modeled loads (using hydrochemical catchment models) could be used to explain differences in retention between the wetlands. The wetland areas ranged from 0.04 to 0.8% of the respective catchment area, and they were situated in areas dominated by fine-textured soils with relatively high P losses and the main proportion of P transported in particulate form. Net P and particle retention were estimated during two years from annual accumulation of particles on sedimentation plates (40 × 40 cm) on the bottom of the wetlands. There was an annual net retention of particles and P, but with a large variation (for particles 13–108 t ha−1 yr−1 and for P 11–175 kg ha−1 yr−1), both between wetlands and between years. The difference between the two years was larger than the difference in mean P retention between the seven wetlands. There was a positive relationship between P and particle retention and three catchment factors, i.e. P status (P-AL) of agricultural soils, average slope in the catchments and the livestock density, and a negative relationship with the agricultural soil clay content. In addition, there was a positive relationship with the wetland length:width ratio. Contrary to expectations, neither the modeled hydraulic load nor P load was significantly correlated with the measured particle and P retention. There was also a positive relationship between P concentration in the sediment and soil P status in the catchment. The results imply that considerable errors are introduced when down-scaling modeled regional nutrient losses to estimate the P loads to small wetlands in agriculturally dominated catchments. A more qualitative approach, using catchment characteristics for identification of hot-spot fields, may be equally good to identify suitable locations for constructed wetlands to reduce diffuse P loads.

Of the EU countries, Sweden utilizes the highest proportion (10%) of its total agricultural land for horses. Horse paddocks commonly hold horses on a limited space, in the present study at a rate of 5–14 livestock units ha−1. Thus these paddocks receive significant amounts of phosphorus (P) and nitrogen (N) through feed residues and deposition of feces and urine, which can lead to nutrient build-up in the soil. This study examined soil P and N status in different parts of horse paddocks (feeding, grazing and excretion areas) and compared it with that in adjacent, unmanaged reference soils. The paddock areas were then categorized with respect to environmental risk using the threshold concentrations of plant-available P extracted with ammonium acetate lactate solution at pH 3.75 (P-AL) and total N set by the Swedish Environmental Protection Agency. In total, seven horse farms, covering different grazing densities and soil textures representative of Swedish horse paddocks, were examined. The results showed that concentrations of water-soluble P (WSP), P-AL, and total N were highest in feeding and excretion areas within the paddocks. Weighted concentrations of soil P for the whole paddocks amounted to 2.9–10.5 mg WSP kg−1 and 35–224 mg P-AL kg−1, and were higher than in the corresponding reference fields (0.8–4.9 mg WSP kg−1 and 17–102 mg P-AL kg−1). The WSP concentration in the paddocks was strongly correlated with horse density (R2 = 0.80***1, n = 13) and P-AL with years of paddock management (R2 = 0.78***, n = 13). Total organic C was significantly correlated with P-AL, total P and total N in the feeding and excretion areas. The degree of soil P saturation percentage (DPS%) were important soil parameters determining WSP concentration in the paddocks (R2 = 0.63***, n = 110), whereas total P concentration in soils were determining P-AL concentrations (R2 = 0.82***, n = 112). According to Swedish environmental guidelines, two of the seven farms studied posed a high risk and three a moderate risk of extensive P leaching losses, but the risk of extensive N losses was moderate for all farms studied. As regards the specific sections of the paddock, the feeding and excretion areas had the highest risk of P leaching losses. Thus paddock soils can be high-risk areas for P leaching comprising about 3.85% of the total high-risk land area in Sweden.

Catch crops are a potential option to reduce phosphorus (P) losses, but little is known about their establishment success and capacity to retain P on clayey soils in regions with short autumns, e.g. Sweden. This study screened biomass production and P retention by eight catch crop species: the perennials chicory (Cichorium intybus L), cocksfoot (Dactylis glomerata L), perennial ryegrass (Lolium perenne L.) and red clover (Trifolium pratense L) and the annuals phacelia (Phacelia tanacetifolia L), white mustard (Sinapis alba L), oilseed radish (Raphanus sativus L oleiformis) and white radish (R. longipinnatus). The catch crops were grown at six field sites, where the perennial species were under-sown with barley and the annual species were after-sown following barley harvest. Biomass production, P content in above-ground and below-ground plant parts and content of available P in the soil were determined in autumn and survival rate of the catch crops in the following spring. Biomass production and P retention in autumn both differed significantly between species (p < 0.0001), and were greatly affected by site-specific conditions and time of sowing, which differed between experiments. Growth of catch crops can also be suppressed by low precipitation. Content of P in roots varied substantially between species, a factor which must be considered in species comparisons. The under-sown species produced more or equivalent amounts of biomass, retained more or equivalent amounts of P in autumn and survived better over winter than the after-sown species. Thus under-sown catch crops generally seem more suitable as catch crops for P.

High application rates of poultry ( Gallus gallus domesticus ) manure can impair drainage water quality through enhanced leaching of phosphorus (P) and nitrogen (N). In two years with application rates of broiler chicken manure corresponding to 99 and 79 kg P ha −1 year −1 to a tile-drained field with loamy soil in SW Sweden, mean concentrations of dissolved reactive P (DRP) were significantly higher ( p 3 -N) in peak, base and intermediate flows occurred in the 8-year period following manure application. Mean NO 3 -N concentration in peak flows decreased approximately linearly, by 5.6 μg L −1 , for every mm of total water discharge. Thus poultry manure should not be applied in repeated high loads in order to avoid the risk of enhanced P leaching losses during subsequent peak flows. Furthermore, N leaching losses during peak and base flow conditions can persist for longer than a 5-year crop rotation and were shown here to recede at a slower relative rate than P.

Constructed free water surface wetlands (CWs) are used to reduce nutrient and sediment loads to receiving waters in highly impacted catchments by e.g., agricultural land use. In this paper, we evaluated the retention effectiveness of a Swedish CW in two consecutive hydrological years, 3-4 years after construction respectively. We compared nutrient loads based on concentrations from flow-proportional sampling (C-fp) and turbidity and nitrate concentrations measured with optical sensors (C-s). CW's retention was estimated based on differences between inlet and outlet concentrations and flow (Q) at both the inlet and outlet (2Q), or only at the inlet (1Q). In the first year (2012/2013), with a cold winter (mean topsoil temperature -0.2 degrees C), nitrate-nitrogen (NO3N) retention was 32% (C-fp and C-s, 2Q). In the second year, with a mild winter (mean topsoil temp 1.8 degrees C) and less water runoff, the corresponding values were 37% (C-fp, 2Q) and 39% (C-s, 2Q). Nitrate-nitrogen retention was significantly correlated to water residence time and temperature, and was most effective relative to the load (80%) in summer and least effective (40%) in winter. Quantitatively, however, summer NO3N retention contributed only 7% (2Q) or 8% (1Q) of yearly NO3N mass retention. Particulate phosphorus (PP) concentrations were significantly correlated with suspended solids (SS) concentrations at both inlet and outlet. Seasonal PP retention (C-fp, 2Q) was related to particle residence time estimated from turbidity measurements by sensors, and was less effective in the cold winter (3%) than in the mild winter (32%) (C-fp, 2Q). Yearly retention (2Q) as a mean of the two years was: SS 40%, total P 36%, PP 34%, dissolved reactive P 30%, total N 56%, NO3N 35%, organic N 75%, and organic C 30%. Overall, the wetland satisfactorily removed nutrients from agricultural drainage water. However, longer-term studies over a range of flow and temperature conditions are needed to evaluate climate conditions and hydrological residence time as key factors in nutrient removal efficiency.

Of the EU countries, Sweden utilizes the highest proportion (10%) of its total agricultural land for horses. Horse paddocks commonly hold horses on a limited space, in the present study at a rate of 5–14 livestock units ha−1. Thus these paddocks receive significant amounts of phosphorus (P) and nitrogen (N) through feed residues and deposition of faeces and urine, which can lead to nutrient build-up in the soil. This study examined soil P and N status in different parts of horse paddocks (feeding, grazing and excretion areas) and compared it with that in adjacent, unmanaged reference soils. The paddock areas were then categorized with respect to environmental risk using the threshold concentrations of plant-available P extracted with ammonium acetate lactate solution at pH 3.75 (P-AL) and total N set by the Swedish Environmental Protection Agency. In total, seven horse farms, covering different grazing densities and soil textures representative of Swedish horse paddocks, were examined. The results showed that concentrations of water-soluble P (WSP), P-AL, and total N were highest in feeding and excretion areas within the paddocks. Weighted concentrations of soil P for the whole paddocks amounted to 2.9–10.5 mg WSP (kg−1) and 35–224 mg P-AL (kg−1), and were higher than in the corresponding reference fields (0.8–4.9 mg WSP (kg−1) and 17–102 mg P-AL (kg−1)). The WSP concentration in the paddocks was strongly correlated with horse density (R2 = 0.80*** 1 , n = 13) and P-AL with years of paddock management (R2 = 0.78***, n = 13). Total organic C was significantly correlated with P-AL, total P and total N in the feeding and excretion areas. The degree of soil P saturation (DPS) concentrations was important soil parameters determining WSP concentration in the paddocks (R2 = 0.63***, n = 110), whereas total P concentration in soils was determining P-AL concentrations (R2 = 0.82***, n = 112). According to Swedish environmental guidelines, two of the seven farms studied posed a high risk and three a moderate risk of extensive P leaching losses, but the risk of extensive N losses was moderate for all farms studied. As regards the specific sections of the paddock, the feeding and excretion areas had the highest risk of P leaching losses. Thus paddock soils can be high-risk areas for P leaching comprising about 3.85% of the total high-risk land area in Sweden.

In 1988–2009, efforts involving manure regulations, subsidies and advisory work were made to reduce losses of nitrogen (N) and phosphorus (P) from arable land to water in Sweden, especially southern Sweden. To evaluate ensuing long-term trends, nutrients in total and dissolved form were monitored in drainage water from 13 fields under ordinary cultivation. The Mann-Kendall test, adjusted for biweekly sampling, revealed significant decreasing trends (magnitude 3% per year) in nitrate–nitrogen (NO 3 –N) concentrations in two fields. One of these had received no manure and had been subjected to delayed tillage after fallow since 1993, while the other had been frequently cropped with a catch crop combined with spring tillage since 1999. An additional decreasing trend in dissolved reactive P (DRP) was observed over the years from this and another field, both with unfertilised fallow grown as an interior buffer on an area at the field outlet, one in 1999–2010 and the other since 2001. By contrast, there was an increasing trend in DRP in a field with accelerating P loads that included almost yearly manure additions to crops following potatoes and sugar beet. However, the general DRP level in the drainage water was low (4–13 μg L −1 ), probably as a result of high P sorption capacity of the topsoil and subsoil. No significant trends in total P or particulate P were found for any of the fields. The importance of long-term monitoring of nutrients in water leaving agricultural fields in order to estimate trends in nutrient losses is discussed.

Summary A refined version of a conditional phosphorus risk index (PRI) for P losses to waters was developed based on monitoring and analyses of PRI factors from an agricultural catchment in Sweden. The catchment has a hummocky landscape of heavy glacial till overlying moraine and an overall balanced soil P level. Single P source factors and combinations of factors were tested and discussed together with water movement and water management factors important for catchments dominated by drained clay soils. An empirical relationship was established (Pearson correlation coefficient 0.861, p 2 ), measured in a weak calcium chloride solution, and iron (Fe-AL) aluminium (Al-AL) and phosphorus (P-AL) in soil extract with acid ammonium lactate. Differing relationships were found for a field that had not received any manure in the last 15 years and a field that had received chicken litter very recently. In addition, a general relationship (Pearson correlation coefficient 0.839, p 2 and the degree of phosphorus saturation in lactate extract (DPS-AL). One exception was a single field, representing 7% of agricultural land in the catchment, that had been treated with glyphosate shortly before soil sampling. Saturated hydraulic conductivity (SHC) in heavy clay in contact with the moraine base (at 1 m depth) was on average 0.06 m day −1 . In clay not in contact with moraine, SHC was significantly lower (mean 0.007 m day −1 ). A reduction in the present tile drain spacing (from 14–16 m to 11 m) is theoretically required to maintain satisfactory water discharge and groundwater level. Up to 10% of the arable land was estimated to be a potential source area for P, based on different indices. Parts of a few fields close to farm buildings (1% of total arable land) were identified as essential P source areas, with high DPS-AL values and low PSI-CaCl 2 values throughout the soil profile. A further 2% of arable land was identified as potential important transport areas, based on visible surface water rills or frequent water-ponded conditions. Fields comprising 10% of the total arable land in the catchment should be re-drained in the near future to improve water infiltration and avoid unnecessary channelised water flow. The need for an improved PRI for erosion and water transport is discussed.

Concentrations of dissolved reactive phosphorus (C DRP ) and particle-bound phosphorus (C PP ) were investigated in soil columns taken from different parts of three selected agricultural fields. The columns (mini-lysimeters 20 cm in diameter, 20 cm high) were exposed to on average 64 ± 6 mm of simulated rainfall. Total phosphorus concentrations (C TP ) in water percolating from the lysimeters were similar to the C TP observed in drainage water leaving the fields, but the C DRP /C TP ratio was significantly ( p DRP (up to 0.4 mg L − 1 ) measured in the percolate may have partly derived from decomposition of accumulated senescing vegetation and litter material, since C DRP in percolating water was related to total organic carbon concentration (TOC) in the shallow topsoil (0–5 cm) and to phosphorus concentration in soil extract of ammonium lactate (P-AL) from the same soil layer. Another easily identified factor clearly related to C DRP in percolate from the lysimeters was the degree of phosphorus saturation (DPS) in the same soil P-AL extract. The factors giving best prediction of C PP and leaching were the amount of percolate passing through the soil columns and total phosphorus concentration in the topsoil determined after oxidation with nitric acid (TP-HNO 3 ). One field under grass ley, including the slope and depression close to the field outlet, comprised 24% of total field area. In the other two fields these areas (11 and 17% of total area respectively) were under long-term permanent green fallow as an internal buffer zone (IBZ). The lysimeter studies indicated that these areas had a higher risk of DRP losses than the rest of the fields. In addition, the P content of the particles in water from the depression was unusually high. Aggregate stability, measured indirectly as readily dispersible clay and expressed as nephelometric turbidity units (NTU), was significantly stronger at the lowest point of one IBZ (27–35 NTU) than in stubbled parts of the same field (62 NTU), and the turbidity of water percolating through the topsoil was lower (5–27 NTU).

Adapting regional eutrophication targets for surface waters - influence of the EU water framework directive national policy and climate change

Diffuse phosphorus losses from arable land contribute significantly to eutrophication in many regions. To enable the most efficient measures for reducing P losses to be identified, it is important to quantify the contributions from different sources and the effects of various scenarios aimed at reducing these losses. To account for transport to tile-drains through macropores, we extended the field-scale simulation model ICECREAM with descriptions based on mixing layer and dual porosity concepts. For losses of particulate phosphorus (PP) to tile-drains, a new description of particle generation and detachment was included. The model was applied on a 6-year field experiment on a structured clay soil in south-west Sweden and simulation results were compared with measured water flow and losses of suspended particles (SP), PP and dissolved reactive phosphorus (DRP) to tile-drains. After calibration of parameters related to macropore flow and transport, and to particle generation and detachment, the model could describe the extremely episodic losses of SP, PP and DRP reasonably accurately. However, some short-term fluctuations were not captured (model efficiency during the validation period −2.1 for PP and 0.43 for DRP). Thus, further development and testing are required before the model can be used with confidence for predictive applications or management purposes. Based on this application, we suggest that direct leaching from non-decomposed plant material and infiltration in partly frozen soil be included in the ICECREAM model. We also suggest that the controls on the sorption distribution coefficient for partitioning between DRP and labile phosphorus (PL) and the effects of using the ‘storage routing’ equation for calculation of soil water content instead of the Richards equation be further investigated.

Phosphorus (P) binding to minerals and ion exchange capacity in different clay fractions were examined for a non-calcareous soil in southwest Sweden. The soil had received pig slurry during three decades, 2 kg lower than the recent maximum load of 22 kg P ha − 1 year − 1 as regulated by livestock density legislation. The topsoil was found to contain 33% clay by weight. Illite was the predominant clay mineral and constituted 13% of total soil. Vermiculite (10%), K-feldspar (14%) and plagioclase (21%) also constituted significant proportions of the mineralogical matrix. Within the most fine-grained clay fraction, 50% of which was less than 0.1 μm in particle size, illite and vermiculite dominated totally, 50 and 23% respectively. In fine-grained (FG), most fine-grained (MFG) and colloidal fractions, there were strong relationships (Pearson correlation coefficient 0.98–1.00) between calcium (Ca) and P. There was a low molar ratio Ca:P in added manure and the presence of Ca–P complexes in the fine soil fractions was indicated. In contrast, in the coarse soil fraction (> 2 μm), there was a clear relationship (Pearson coefficient 0.97) between P and iron oxide (Fe 2 O 3 ) and between P and aluminium oxide (Al 2 O 3 ) throughout the soil profile. Thus even for non-calcareous soils, formation of Ca–P complexes should be taken into account with regards to losses of colloidal P to drainage water.

Within the Swedish Monitoring Programme for Agriculture, 27 small agricultural catchments were investigated for between 9 and 14 years (6 years for one catchment), starting in the late 1980s, to determine the impact of agricultural activities on discharge water. The catchments represent major variations in climate, soil types and farming in Sweden and hence long-term average load of total nitrogen (N) and total phosphorus (P) at stream outlets varied widely among the catchments, from 2 to 41 and from 0.1 to 0.9 kg ha −1 yr −1 , respectively. Catchments with large N loads were characterised by several factors favourable for leaching, such as mild winter climate, large precipitation, sandy or organic soils and high animal density in combination with intensive cropping systems with a low percentage of ley. Large P loads were mainly related to factors such as clay and clay loam soils, medium to high precipitation and large proportions of annual crops in the catchments. Flow-normalised time series of monthly loads of nitrate N, phosphate P and particulate P (for 24 catchments) were tested for trends. Significant downward trends were revealed for nitrate N, phosphate P and particulate P for seven, eight and three catchments, respectively, whereas, upward trends were revealed for nitrate N and particulate P for one and two catchments, respectively. Downward nitrate N trends for two catchments were correlated to smaller amounts of applied manure, especially during autumn, and to a decrease in the area of spring cereals and spring rape.

A case study was carried out in central Sweden on local initiatives for a river basin in accordance with the European Water Framework Directive. The catchment has erosion problems and the recipient waters suffer from eutrophication. Collective learning and successively more collaborative measures were introduced. According to stakeholders, the following measures were considered feasible and ranked in order of importance: improved soil fertility (improved organic concentration) > measures at or in the stream > measures in the lake itself > improvements to single household wastewater treatment systems. Personal contacts and dialogue at an early stage between stakeholders in order to motivate actions were found to be necessary for progress in the work. Local demonstration areas and information were found to be important in the collaborative learning process according to interviews with the farmers. They indicated more interest and willingness to change behaviour than was demonstrated from real actions during 2 years when several economic subsides were introduced. Sustained financial backing is crucial and funding must be secured to implement the actions decided upon. Implementation required improved contact between local authority employees, politicians, householders and farmers within the catchment area, and in addition, introduced social values as a factor defining the environmental goal to be reached for the lake.

A study of erosion and phosphorus problems was carried out in central Sweden in an area with very visible erosion and negative offsite effects on water quality. The Massingsboan catchment is dominated by agriculture and the recipient Lake Brunnsjon suffers from eutrophication. Local demonstration areas were established and local personal contacts to motivate action and local subsidies were financed for 2 years. Measures identified and/or implemented to reduce phosphorous load on the lake were grouped into: (1) improved soil fertility and minimising of soil erosion; (2) measures at or in the stream; (3) measures in the lake itself; (4) improvements of single household wastewater treatment systems. Soil management that improved aggregate stability was demonstrated to significantly reduce soil erosion and loss of particulate-bound phosphorus. In addition, high soil fertility was generally associated with good aggregate stability. Local experiments in the stream and in the lake were less successful and more research is needed. Maximum phosphorus concentration for water loading the lake without causing excessive eutrophication was calculated to be similar to the background phosphorus concentration from arable land (0.05 mg l−1).

The aim of this paper is to collate information on potential measures to reduce phosphorus (P) losses, with particular reference to Sweden. In Sweden, three dominant types of soils or soil profiles are at risk of suffering from high P losses through different pathways: clay soils with a naturally poor structure, usually situated on the large central plains, are prone to lose P through drain tiles; silty soils, usually situated on hilly terrain, are prone to erosion; and sandy soils with a low P sorption capacity in both topsoil and subsoil and that have had heavy applications of fertilisers, are prone to high leaching losses. Fields with such soils and landscape position with connectivity to surface waters may lose more than 0.5 kg P ha year−1 and often more than 1 kg P ha year−1. Losses of soluble reactive P (SRP) are commonly 30–50% of the total P losses. Improved soil tillage together with other measures aimed at improving soil structure are suggested to have the best potential to reduce losses, with short-term and long-term effects, for the clayey and silty soil profiles. For the silty soils, keeping the surface densely vegetated in winter and liming backfill of the drainage systems are suggested to be important ways to reduce the losses of particulate P (PartP). For sandy soils prone to P leaching, appropriate application of fertiliser and improved contact between fertiliser and the soil may be one of the few effective ways to reduce SRP losses.

As a part of nation-wide monitoring of the impact of agriculture on the aquatic system, phosphorus (P) transport was measured from subsurface drained fields in Sweden. The measured transports was a mixture of P from surface, subsurface and groundwater contribution. Fourteen of these fields together with a similar experimental field have been studied for at least 10 years and most of them for much longer. Transport of total phosphorus (TP) from four of the fields together accounted for 74% of the total transport. Five other fields together accounted for another 19%, while transport from the other five fields was more or less negligible. Based on factor combinations, a simple regression model for calculating TP concentration was used for prediction of the losses. The parameters included were livestock density (LD), HCl extractable P (P–HCl) in the topsoil, duration of high water flow (DHF) and soil specific areas (SSAs). The model was found to satisfactorily predict the TP transport in the streams of 32 small agriculture-dominated catchments in different parts of the country, with a P loss from arable land that was calculated as varying from 0.03 to 0.50 kg ha−1 yr−1. However, for a very few catchments, the model worked only for transport by drainage water and not for P transport measured in the stream water. Eleven of the observation fields had been investigated since 1977 or earlier. During this period, a surplus of P added to the soil during the initial years was transformed to a slight negative balance of P. Transport from most of the fields was found to be relatively constant, but the two fields with the highest P transport showed an increased trend during the 1970s and 1980s.

Drainage discharge and loss of dissolved (PO4-P) and particulate (part-P) P from four artificially drained plots with clay soils were studied over a four-year period. The loads of total P observed during this time ranged from 421 g P ha−1 to 805 g P ha−1. Annual losses and losses during high-flow events were measured. Leaching of P occurred episodically with the largest losses during high-flow events. Concentrations of PO4-P during high-flow events were positively related to discharge rates and the curves obtained were grouped into two classes based on similar slopes within the class. Concentrations of part-P were related to concentrations of suspended soil material, and increased with increasing discharge rate, but the slope of the curves obtained varied over a wide range. High P concentrations were observed in high-flow episodes which occurred after P fertilizer applications.

Winter cover crops are used as a method of reducing nitrogen (N) losses from arable land in several countries, but their effect on phosphorus (P) losses is poorly documented. Run-off and losses of nutrients and soil were measured from a clay loam with autumn-ploughed and spring-ploughed plots and from plots with winter wheat during three winter seasons (1993–1996) in Holland County in south western Sweden. The run-off water was collected in troughs dug into the soil at the end of collecting slopes placed in the experimental plots. As a result of the weather, there was only one winter in which surface run-off occurred to any great extent. On average, 75% of P was in particulate form (Ppart). Neither winter wheat (Triticum aestivum L.) nor catch crops of English ryegrass (Lolium perenne L.) reduced losses of Ppart when compared with losses from autumn-ploughed soil; and losses from spring-ploughed soil containing stubble and weeds were no lower than those from autumn-ploughed soil. Losses of Ppart from all treatments were moderate considering its low bio-availability. Concentrations of phosphate phosphorus (PO4-P) were low, with a mean 0.04 mg 1−1. Despite a significant increase in losses of PO4-P from spring-ploughed soil covered with stubble and catch crops or weeds compared with that in autumn-ploughed soil, the extra input from this P source was at most 2 g ha−1 yr−1. This mass loss was equal to 0.5 g kg−1 of the total mass of P in the vegetation. Thus, only very small extra P surface losses were found with winter cover crops compared with those with bare soils. N losses in run-off were low in all treatments.