512

AbstractTransport of phosphorus (P) from agricultural fields to water bodies deteriorates water quality and causes eutrophication. To reduce P losses and optimize P use efficiency by crops, better knowledge is needed of P turnover in soil and the efficiency of best management practices (BMPs). In this review, we examined these issues using results from 10 Swedish long-term soil fertility trials and various studies on subsurface losses of P. The fertility trials are more than 50 years old and consist of two cropping systems with farmyard manure and mineral fertilizer. One major finding was that replacement of P removed by crops with fertilizer P was not sufficient to maintain soil P concentrations, determined with acid ammonium lactate extraction. The BMPs for reducing P leaching losses reviewed here included catch crops, constructed wetlands, structure liming of clay soils, and various manure application strategies. None of the eight catch crops tested reduced P leaching significantly, whereas total P loads were reduced by 36% by wetland installation, by 39 to 55% by structure liming (tested at two sites), and by 50% by incorporation of pig slurry into a clay soil instead of surface application. Trend analysis of P monitoring data since the 1980s for a number of small Swedish catchments in which various BMPs have been implemented showed no clear pattern, and both upward and downward trends were observed. However, other factors, such as weather conditions and soil type, have profound effects on P losses, which can mask the effects of BMPs.

Turnover and Losses of Phosphorus in Swedish Agricultural Soils: Long-Term Changes, Leaching Trends, and Mitigation Measures

Lars Bergström,* Holger Kirchmann, Faruk Djodjic, Katarina Kyllmar, Barbro Ulén, Jian Liu, Helena Andersson, Helena Aronsson, Gunnar Börjesson, Pia Kynkäänniemi, Annika Svanbäck, and Ana Villa

Eutrophication of inland and coastal waters is one of the largest environmental problems in many countries around the world. In Sweden, eutrophication of waters

has been recognized for a long time. For rivers, lakes, and other freshwater systems, eutrophication is mainly associated with high phosphorus (P) levels (Schindler, 1977), whereas for coastal waters, nitrogen (N) has been considered to be the main limit-ing nutrient (Boesch et al., 2008). However, for the Baltic Sea in northern Europe, where the salinity is predominantly below 1% (Feistel et al., 2010) due to mixing of seawater from the North Sea and freshwater from rivers draining into the Baltic Sea, P has been identified as being the limiting nutrient for eutrophication (Boesch et al., 2006). The total P (TP) load to the Baltic Sea amounts to about 37,000 t yr-1 (Helsinki Commission, 2013a), to which anthropogenic sources are an important contributor. For Sweden, it has been estimated that agriculture contributes around 40% of the total anthropogenic load to the Baltic Sea (Brandt et al., 2009). Therefore, one of the most important actions in reducing P loads to the Baltic Sea is to minimize non-point source losses of P from agricultural land. To do that, better knowledge is needed of P turnover in soil and of the efficiency of best management practices (BMPs) in reducing P losses.

The rate at which P is transferred among the different components of the plant/soil system is critical when assessing P turnover in agricultural soils (McLaughlin et al., 1988). This, together with application of P fertilizers, determines the amount of P available to plants and the fraction that could potentially be lost from the system and cause environmental problems. The most important factors determining the amount of P available to plants are the concentration of dissolved reactive P (DRP) in the soil solution and the soil’s ability to replenish P taken up by plants (Syers et al., 2008). Syers et al. (2008) introduced the concept of a “critical value” for a given soil, which is the soil P content where no further addition of P is justifiable from either

Abbreviations: BMPs, best management practices; DPS, degree of phosphorus saturation; DRP, dissolved reactive phosphorus; FTCs, freezing-thawing cycles; P-AL, plant-available P; PP, particulate phosphorus; TP, total phosphorus.

L. Bergström, H. Kirchmann, K. Kyllmar, B. Ulén, J. Liu, H. Andersson, H. Aronsson, G. Börjesson, P. Kynkäänniemi, and A. Svanbäck, Dep. of Soil and Environment, Swedish Univ. of Agricultural Sciences, P.O. Box 7014, SE-75007 Uppsala, Sweden; F. Djodjic and A. Villa, Dep. of Aquatic Sciences and Assessment, Swedish Univ. of Agricultural Sciences, P.O. Box 7050, SE-75007 Uppsala, Sweden. Assigned to Associate Editor Peter Kleinman.

Copyright © American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America. 5585 Guilford Rd., Madison, WI 53711 USA. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. J. Environ. Qual. 44:512–523 (2015) doi:10.2134/jeq2014.04.0165Freely available online through the author-supported open-access option. Received 15 Apr. 2014. Accepted 1 Dec. 2014. *Corresponding author (lars.bergstrom@slu.se).

Journal of Environmental QualityPHOSPHORUS FATE, MANAGEMENT, AND MODELING IN ARTIFICIALLY DRAINED SYSTEMS

SPECIAL SECTION

Published March 11, 2015

www.agronomy.org • www.crops.org • www.soils.org 513

an agronomic or environmental standpoint. The question is how long it takes to reach this critical value in a given soil with small P reserves or in a soil with a P surplus.

To reduce environmental pollution in the Baltic Sea, the Baltic Sea Action Plan was jointly adopted by the environment ministers of the Baltic Sea countries and the European Commission in 2007, aiming at achieving good environmental status by 2021 (SEPA, 2009). The Country Allocated Reduction Targets in the Baltic Sea Action Plan were revised in 2013 (Helsinki Commission, 2013b). To meet these targets and to achieve the P load reductions required, substantial mitigation efforts are needed. In addition to national regulations and various international conventions (e.g., the EU Water Framework Directive), there are several BMPs that can be adopted at farm and field level to reduce diffuse losses of P. However, to achieve successful results, certain conditions need to be known. For example, the transport mechanism by which P is moving at various scales—through surface runoff or accompanying water moving through soil to tile drains—is critical information when selecting appropriate BMPs. In Sweden, almost 50% of arable soils are systematically tile drained (Statistics Sweden, 2014), and most of these soils have a relatively high clay content. A drainage system can either decrease P leaching by allowing improved and more uniform infiltration, without water-saturated topsoil, or increase P leaching, providing a more direct route to recipient waters. In situations with high P concentrations (either dissolved or particle-bound) in drainage water, a tile drainage system can therefore carry large amounts of P to surface waters (Gelbrecht et al., 2005). If P is bound to colloidal clay particles, the distance this P is transported can also be great (Ulén, 2004).

To determine the response to various BMPs and environmental policies introduced, it is essential to have well-designed monitoring of watercourses and watersheds (Bechmann et al., 2008; Kyllmar et al., 2014) and single fields of concern (Ulén et al., 2012). Information on P concentration trends in watercourses and drainage water, together with information on the agricultural practices performed, can enable politicians and authorities to set realistic water quality goals and targets. In an evaluation of trends in nutrient concentrations in 12 Swedish rivers in agricultural areas from 1993 to 2004, Ulén and Fölster (2007) found that 5 of the 12 rivers had decreasing (2–4% yr-1) residual-P concentrations. In a similar analysis, Grimvall and Nordgaard (2004) found a statistically significant downward trend in P and N concentrations in rivers in the extreme south of Sweden, with strong indications that P had decreased more and that the N:P ratio had increased. However, trend analyses have to be combined with detailed experiments and knowledge of trends for other sources (e.g., sewage water) to corroborate the effects of BMPs on nutrient levels in recipient water bodies.

In this review paper, we summarize the main findings of a number of Swedish studies examining turnover of P in soil, BMPs to reduce diffuse losses of P from agricultural soils, and monitoring programs of long-term P trends in water recipients. Based on this, the main objective of this study was to propose some recommendations that can be taken forward to reduce the environmental impact of P use in Sweden. The focus is on Nordic climate conditions, where leaching plays a major role in the transport of P from fields to surface waters.

P Turnover and Changes in SoilTo exemplify long-term changes in P turnover through release

and binding of P in soil, to draw up P budgets, and to determine the influence of different P fertilizer on P forms in soil, we selected 10 Swedish soil fertility sites included in the Swedish long-term fertilization trials (Carlgren and Mattsson, 2001) that have received different P inputs since the 1950s. Each site has two cropping systems, with farmyard manure addition and mineral fertilization only. For this review, data for the highest N fertilization rate (120–150 kg N ha-1 yr-1) were analyzed in combination with data for the four P fertilization treatments applied at the sites (no P addition, replacement of P removed by crops, and double and triple replacement of P removed by crops). The most important form of P in soil for crop production is referred to as plant-available P (P-AL), which at the fertility sites is determined by extraction with ammonium acetate lactate solution (Egnér et al., 1960). More information about the soil fertility sites can be found in Carlgren and Mattsson (2001).

National and regional P budgets for Sweden based on data originating from official Swedish statistics (Statistics Sweden 2001, 2003a, 2005, 2007, 2009, 2011, 2013) were analyzed. The P budget for agricultural land is calculated according to the soil surface method, in which P inputs consist of mineral fertilizer, soil amendments, barn and grazing manure, seed, and sewage sludge. Crop yields and harvested plant residues represent nutrient outputs. The difference between P inputs and outputs results in a balance that is either positive or negative.

Trends in Plant-Available Soil P in Long-Term Field Experiments

Concentrations of P-AL appeared similar at all 10 sites analyzed, increasing or decreasing over time according to the level of fertilizer P applied (Fig. 1; Table 1). Plant-available P was depleted in plots when no fertilizer P was applied and enriched when P addition exceeded P removal by crops. An important finding was a decline in P-AL in a rotation with farmyard manure, in which P removed by crops was partially replaced as manure. Furthermore, replacement of the P removed by crops with fertilizer P resulted in a decline in P-AL concentration in the soil (Fig. 1). This is somewhat different from the concept proposed by Syers et al. (2008), in which the amount of P required to maintain an optimal P level in soil is often similar to the amount removed by the crop. In the present analysis, replacement of removed P by fertilizer P was not sufficient to counteract depletion of P-AL in soil over time. Nevertheless, crop yield was satisfactory in most years.

In the two treatments where additional P was applied, exceeding P replacement, an increase in P-AL was observed (Fig. 1). These P-AL changes were linear over time. Variations between years within treatments were found for high fertilizer P addition, whereas almost no variations were found in treatments with low or no P addition. Variations in treatments with high P addition sometimes amounted to 75 to 80% between years. One possible explanation is that dissolution of P fertilizer granules does not lead to a uniform P distribution in soil. Binding of dissolved fertilizer P on soil particles and movement of phosphate anions may be limited to a couple of millimeters around the fertilizer granule (Havlin et al., 2005), leading to unequal allocation.

514 Journal of Environmental Quality

Fig. 1. Depletion/enrichment of plant-available soil P (P-AL) caused by different P fertilization regimes at five sites. (I) Crop rotation combining animal manure and mineral fertilizer. (II) rotation with mineral fertilizer only. Letters indicate different P fertilization rates: A = 0 kg P, B = replacement of P removed with harvested crop, C = replacement + 15 kg, and D = replacement + 30 kg P; “3” indicates high N fertilization level (120–150 kg N ha-1).

Table 1. Mean change in plant-available P in soil in the 10 Swedish soil fertility experiments. Rates derived from slopes in Fig. 1.

Type of P addition Total P applied Rate of P-AL† changeP-depleting treatments kg P ha-1 yr-1 mg P 100 g-1 soil yr-1

No P 0 -0.062 Manure P only (5 t wet weight yr-1) 7.5 -0.049 Replacement of P removed by crops 15 -0.032P-enriching treatments Replacement + 15 kg fertilizer P 30 0.055 Replacement + 20 kg fertilizer P 35 0.086 Replacement + 30 kg fertilizer P 45 0.157

† Plant-available P.

www.agronomy.org • www.crops.org • www.soils.org 515

Furthermore, regular soil tillage seems unable to counteract this effect within 1 or 2 yr.

The P-AL value ranged from 7 to 32 mg P-AL 100 g-1 soil at the different sites and treatments analyzed (Fig. 1), with a linear change ranging from -0.06 to 0.15 mg P-AL 100 g-1 soil and year (Table 1). Mean slope for the same treatment at all sites was used for further evaluations. Plotting rate of change in P-AL against application of fertilizer P gave a highly significant relationship (R2 = 0.97; significant at the 0.01 probability level), according to which about 17 kg fertilizer P ha-1 yr-1 are needed to maintain P-AL levels in soil (Fig. 2). Although this value cannot be generally applied because yield level influences removal rate, it may be a useful approximation for high-yielding systems in countries in Northern Europe. According to the same relationship, it will take about 17 yr for a decrease of 1 mg P-AL per 100 g soil when no P is applied. At the other extreme, an increase of 1 mg P-AL per 100 g soil can be achieved within 7 yr when 45 kg P ha-1 is applied annually. Consequently, building up high soil P status is easier than lowering soil P status.

Long-Term P Budget in Agricultural SoilsAt one of the soil fertility sites (Fjärdingslöv), a complete

P budget was calculated for an experimental period of 52 yr in which recorded inputs of P and harvested P (yield × P content) were summed (Table 2). This revealed that removal of P by crops was not fully compensated for by fertilizer P, resulting in a deficit of about 1 kg P ha-1 yr-1. In the non–P-fertilized treatments, large amounts of P were still removed with the harvested crops, resulting in depletion of about 8 kg P ha-1 yr-1. Applications of fertilizer P exceeding crop demand (double and triple replacement) resulted in a positive P balance and soil enrichment. Between 670 and 1460 kg P ha-1 were added in excess of crop demand over the 52 yr in those treatments. These figures can be compared with reported accumulation of fertilizer P in Swedish arable soils of on average 700 kg P ha-1 over a 50-yr period, with 590 to 620 kg P ha-1 accumulated in the topsoil (0–20 cm) and the remainder in the subsoil (Andersson et al., 1998). The P balance was similar for treatments with and without animal manure because inorganic P was the main P source.

The P use efficiency by crops was calculated according to the balance method (Ladha et al., 2005), which includes recovery from both recent and previous P fertilizer applications and thus represents the long-term recovery of fertilizer P. These calculations revealed that fertilizer P was completely utilized when applied at the replacement rate (Table 2). At application rates in excess

of crop demand, the P use efficiency decreased to 40 to 70%. However, these figures show that P fertilizer is efficiently utilized by crops, confirming findings by Syers et al. (2008), and that P forms other than the fraction extracted with ammonium lactate can be utilized by crops. Fertilizer P in soil that is not P-AL soluble thus remains available to plants and is not “fixed,” as previously assumed. Sorption of inorganic fertilizer P takes place on the surface of iron (Fe) and aluminum (Al) oxides, increasing the labile inorganic P fractions (Negassa and Leinweber, 2009).

National and Regional P Budgets in SwedenApproximately 700 kg P ha-1 have accumulated in Swedish

arable soils since the 1950s (Andersson et al., 1998). However, eutrophication problems have prompted targeted reduction of P inputs since the 1990s. These efforts have resulted in considerable improvements regarding the balance between P additions and P removal with harvested crops. A plot of the P balance in agricultural land from 1995 to 2011 for each Swedish production region and for Sweden as a whole shows that there has been a clear decreasing trend in the P balance regionally and nationally (Fig. 3). For example, there was a surplus of 6 kg P ha-1 yr-1 at the national level in 1995, but there was a balance between P inputs and outputs in 2011. However, regional differences are quite large. The P surplus has been reduced or eliminated by

Fig. 2. Correlation between rate of change in plant-available soil P (P-AL) and application of fertilizer P.

Table 2. Phosphorus budget and crop utilization of fertilizer P at the Fjärdingslöv site.

Treatment Fertilizer P Manure P Total Harvested P P budget P use efficiency

——————————————————— kg ha-1 ——————————————————— %Rotation with fertilizer and manure 7.5 kg P 0 326 326 740 -414 – 15 kg P 898 326 1224 1272 -48 104 30 kg P 1803 326 2129 1462 667 69 45 kg P 2633 326 2959 1510 1449 41Rotation with fertilizer 0 kg P 0 0 0 438 -438 – 15 kg P 816 0 816 868 -52 106 30 kg P 1672 0 1672 993 679 59 45 kg P 2500 0 2500 1037 1463 41

516 Journal of Environmental Quality

decreasing the use of mineral P fertilizer (from 6 kg P ha-1 yr-1 in 1995 to 3 kg P ha-1 yr-1 in 2011 at the national level) and the use of manure (from 7 kg P ha-1 yr-1 in 1995 to 5 kg P ha-1 yr-1 in 2011 at the national level) while maintaining or increasing crop yields (Statistics Sweden, 2013).

The regional patterns largely reflect the animal density distribution in Sweden (Statistics Sweden, 2003b). The higher animal densities in, for example, central districts (PO2) and forest districts in Götaland (PO5) result in a P surplus, whereas the low animal densities in plain districts of northern Götaland and in Svealand (PO3 and PO4) lead to a substantial P deficit (Fig. 3). This also means that P is exported with harvested crops from Swedish regions dominated by crop production (PO3 and PO4) and imported with animal feeds to regions with significant animal husbandry (PO2 and PO5). The uneven distribution of P addition is also evident at the local level. For example, Djodjic and Kyllmar (2011) showed that approximately 5% of arable soils in Sweden receive >60 kg P ha-1 annually, whereas 43% of arable land is not fertilized at all. Overall, there has been a positive trend in reducing the P surplus, and there is now a balance between P inputs and outputs at the national level. However, smoothing the substantial differences at regional and local scales will be an important task in further improving P use. Because most of the P surplus is in the form of manure, transport costs, which are high, also need to be considered.

Impact of Different P Fertilization Treatments on P Forms in Soil

The P budget at the Fjärdingslöv site discussed above reveals that different P treatments enriched or depleted soil P reserves.

Some key questions we examined here were: (i) How were the main P fractions in soil affected by P fertilization? (ii) Did organic and inorganic fertilizers cause differences? and (iii) Were sandy soils affected differently than clay soils?

Plotting data on net P enrichment/depletion in soil against P forms in soil (Fig. 4) revealed that organic P in soil was not affected by the addition of fertilizer P and that even manure addition had little or no impact on the amount of organic P in soil. Instead, enrichment/depletion of soil P was only detectable as changes in inorganic P, and no differences were found between sandy and clay soils (data not shown). The inorganic P pool in soil increased or decreased and the organic pool remained constant regardless of the type of P fertilizer added, crop rotation, and soil texture (Fig. 4). These findings are in agreement with results reviewed by Negassa and Leinweber (2009) showing that P fertilization increases the fraction of inorganic P, but not organic P, in soil. This has also been shown by Ahlgren et al. (2013), who

Fig. 3. Left: Phosphorus balances calculated for different Swedish production regions (1999–2011) and for Sweden as a whole (1995–2011). Right: Map of Swedish production regions.

Fig. 4. Changes in organic and inorganic P in soil after 53 yr of enrichment/depletion of P.

www.agronomy.org • www.crops.org • www.soils.org 517

analyzed monoesters and two other organic P compounds with nuclear magnetic resonance spectroscopy.

Inorganic soil P is thus the main storage pool for P added with different fertilizers. The question is whether this inorganic P is mainly bound in easily exchangeable or strongly bound form (e.g., Syers et al., 2008). Van der Salm et al. (2009) showed that the largest reduction in soil P can be achieved for weakly bound P pools, whereas reductions in more strongly bound P forms are relatively difficult. We examined which fraction (labile or strongly bound P) was mainly affected by fertilization for the Fjärdingslöv soil, assuming that a single extraction with P-AL solution represents the labile pool in soil that can be taken up by crops. Changes in P-AL were expressed as amounts using the bulk density and actual topsoil depth and were compared with P budget data (Fig. 5). Subtraction of P-AL from the budget showed that most of the P enriched/depleted in that soil was not derived from the plant-available pool. This indicates that strongly bound and mineral P were greatly involved in the turnover of P during enrichment and depletion. This type of P binding has previously been assumed to more or less “fix” P (see review by Syers et al. [2008]), but our review shows that this fraction is reversible to easily plant-available forms. The more pronounced depletion of the strongly bound P pool compared with the P-AL pool (Fig. 5) suggests that the general interpretation of this fraction needs to be revised. In fact, early work by Walker and Syers (1976) on soil P during pedogenesis showed that mineral and strongly bound P changed to a larger extent than P-AL. Our results are in line with those findings and show that the process can be reversed when P fertilizer is added to soil. This highlights the need to fertilize only in line with crop requirements so as not to elevate the soil P status.

Mitigation of P Losses from Agricultural SoilsEven when the addition of P to agricultural soils is in balance

with its removal, large P losses may occur, mainly due to the episodic nature of diffuse losses of P from agricultural fields (Ulén 1995; Haygarth and Sharpley, 2000). Phosphorus losses may also show great spatial variation over a field or watershed, requiring identification of critical source areas where BMPs should be placed (Gburek et al., 2000). Additionally, even though there may be great spatial variation in P losses within fields, variations in terms of soil P content may be small (Svanbäck et al., 2014). A sound, functional package of BMPs should have a positive effect on the environment and should be economically feasible for farmers applying the measures in the field.

In this review, we examined results from studies in which P leaching in response to applications of animal manure was determined in clay and sandy soils and in soils with differing P-AL status. In these studies, leaching was measured in topsoil lysimeters exposed to irrigation in the laboratory and in tile-drained field plots. Structure liming, a method to reduce P leaching losses from clay soils, was tested in two field studies performed in tile-drained plots. The efficiency of catch crops to reduce P losses was tested in irrigated topsoil lysimeters. To determine the importance of subsoil properties on P leaching, some results of a study performed in lysimeters filled with 1-m-long intact soil columns with topsoil and subsoil and with only subsoil of four soils were analyzed. The freely drained topsoil lysimeters used in some of the studies would represent conditions with a high groundwater level. The 1-m-long, gravity-drained soil columns would better resemble the

conditions in tile-drained plots, where the tiles are placed at 1 m depth. In both cases, the measured P concentrations in drainage water would be typical of those in water leaving the root zone of most crops. To further reduce P losses beyond the field boundary, lime filters and wetlands can be used. We reviewed some results from a constructed wetland receiving tile-drain effluent from horse paddocks.

Reducing P Losses in the FieldManure Management

Appropriate manure management is necessary for mitigating P losses from agricultural systems because manure is an unavoidable by-product of animal production. Approximately 25,000 t P yr-1 are excreted by farm animals in Sweden, constituting a major fraction (72%) of the P input to agricultural soils (SCB, 2012). To minimize potential P losses, Sweden has established rather strict legislation and regulations regarding manure management. For instance, farmers are not allowed to apply more than 110 kg P ha-1 with animal manure during a 5-yr period (average, 22 kg P ha-1 yr-1) on an entire farm basis to ensure good water quality (SBA, 2010). However, in practice, application of manure at a rate exceeding 22 kg P ha-1 is common in a single year, especially in southern Sweden, with its intensive animal production. Such applications of manure may create a great risk of P losses and thus need to be reduced.

An important mitigation measure to improve the distribution of P with manure is adaptation of application rates to reduce the long-term risks of P leaching related to build-up of soil P. Svanbäck et al. (2013) demonstrated clear relationships between P-AL and P leaching from topsoil columns of five Swedish soils ranging from sand to clay that had received different P inputs with manure for several decades (Fig. 6). However, the relationship between P-AL and P leachate concentrations varied between soils. The P-AL values in the soil columns ranged from 15 to 236 mg kg-1, representing poor to excessive P-AL values. The four different P-AL levels for each of those soils were the result of different rates of long-term P application with manure in different field plots of the Swedish long-term fertilization trials (Carlgren and Mattsson, 2001). The results confirm the controlling role of P-AL in P leaching from topsoils

Fig. 5. Soil P budget data (Table 2) combined with data on plant-available P (Fig. 1), showing that strongly bound P mainly participates in P turnover during enrichment or depletion of P in soil.

518 Journal of Environmental Quality

and are in agreement with Börling et al. (2004), who observed clear relationships between P-AL and CaCl2–extractable P (a surrogate for leachate P) for topsoils (0–20 cm) from the same field experiments. Therefore, to reduce P leaching in the long-term, application of manure at a P rate slightly over or balancing crop removal is an important mitigation strategy. This is also true for soils in which the current risk of P leaching is low (e.g., sandy soils with high P sorption capacity). In a long-term field leaching study on a tile-drained sandy soil in southern Sweden, annual P leaching was only 0.2 kg ha-1, despite the fact that high rates of slurry P had been applied since 1983 (Liu et al., 2012b). This moderate P leaching load was attributed to sorption of P by Fe and/or Al oxides and hydroxides in the soil, especially in the subsoil, which had a low degree of P saturation (DPS). However, it was observed that the plow layer of this specific soil had high DPS (>40%), due to the long-term slurry application at high rates (Liu et al., 2012c). Continuing manure application at high rates on such soil will eventually result in high P leaching because the subsoil will become saturated with P.

For incidental leaching losses of P, the timing and application technique for manure are important. This has been demonstrated

in a field leaching experiment on a clay soil (Aronsson et al., 2014), where topsoil lysimeters were also collected for laboratory irrigation studies (Liu et al., 2012a). The lysimeter study demonstrated that incorporation of manure into the soil was an effective mitigation strategy for a well-structured clay soil with a pore system allowing preferential flow, whereas for a sandy soil high P sorption capacity caused a relatively small or no risk of P leaching regardless of whether the manure was surface applied or incorporated. Incorporation of manure allows sorption of labile manure P to the soil matrix and breaks the continuity of macropores through which various forms of P can be transported. Incorporation of pig slurry into the clay soil in question instead of applying it to the surface reduced potential TP leaching from the topsoil by 50% and DRP leaching by 64% (Fig. 7). Application of slurry to the soil surface under wet field conditions (i.e., in late October) also increased P leaching through tile drains, irrespective of whether it was applied after a cereal crop or in a growing grass/clover crop (Aronsson et al., 2014). Thus, manure application during wet conditions on soils with preferential flow paths should be avoided if possible or followed by incorporation.

Structure LimingStructure liming has received considerable attention in

Sweden during recent years as a method to reduce P leaching from agricultural fields. When structure lime, either as quicklime (calcium oxide, CaO) or hydrated (slaked) lime [Ca(OH)2], is mixed with a clay soil, several reactions take place with soil aggregates, and improvements in structure, stability, porosity, and aggregate strength have been reported (Choquette et al., 1987). The reactions include cation exchange, flocculation, and agglomeration, together with a slower cementing pozzolanic reaction (Kavak and Baykal, 2012). This is applicable to most clay soils, especially those of moderate to high plasticity. In addition, chemical precipitation of Ca-P precipitates or Ca-P complexes may occur, as indicated for an illite-clay soil in southwestern Sweden with a history of pig manure addition (Ulén and Snäll, 2007).

In a study to investigate the effects of liming, two field sites (Wiad and Bornsjön) in eastern Sweden with tile-drained

Fig. 6. Increase in leaching of dissolved reactive P (DRP) at varying plant-available soil P (P-AL) for topsoil (0–20 cm) lysimeters from five soils (adapted from Svanbäck et al. [2013]). Results refer to a cropping system, including forage crops and moderate manure applications in the field, after manure application to soil columns in the laboratory.

Fig. 7. Cumulative P leaching from clay loam and loamy sand topsoils before and after slurry application and from a control without slurry application. An 80-mm rain event was simulated before and after the slurry treatments. DRP, dissolved reactive P. Other-P = total P – DRP. Different lowercase letters indicate significant difference at p < 0.05 (adapted from Liu et al. [2012a]).

www.agronomy.org • www.crops.org • www.soils.org 519

experimental plots and differing clay content (25 and 60% clay, respectively) were amended with structure lime (Ulén and Etana, 2014). The P-AL content in the topsoil was approximately fourfold higher in the Wiad soil than in the Bornsjön soil. Due to this fact and to a high content of aluminum in the Bornsjön soil (Andersson et al., 2013), the degree of P saturation (DPS-AL), determined according to Ulén (2006), was low at Bornsjön (6%) but was relatively high at Wiad (30%). At both sites, structure lime was cultivated into the topsoil in early autumn, and subsequent P leaching was monitored (Ulén and Etana, 2014). Due to the large spatial variation (visualized as SE in Fig. 8), a spatial power covariance structure was adapted to the general mixed model used for statistical evaluation (Littell et al., 2006).

Only P leaching in particulate P (PP) form was significantly reduced after structure liming at Bornsjön, whereas only DRP leaching was significantly reduced at Wiad. Total P leaching at Wiad was moderate (mean, 0.20 kg ha-1 yr-1) but was significantly higher from the Bornsjön plots (mean, 0.86 kg ha-1 yr-1). Both sites demonstrated TP leaching losses that were significantly (pr > F < 0.002) lower from plots with structure liming than from the control plots without lime addition. These findings indicate that structure liming is an effective mitigation measure to reduce leaching of particulate-bound P on soils with a high clay content or high soil P status.

Catch CropsCatch crops (i.e., crops grown to capture mineral N and reduce

leaching between two main crops when the soil is otherwise bare) (Thorup-Kristensen et al., 2003) have been proposed as another potential option for reduction of P leaching (Lemunyon, 2006). Catch crops are frequently grown in southern Sweden and Denmark, with government subsidies available to promote the practice (Ulén et al., 2007). However, their effect in reducing P leaching under Nordic conditions has been poorly investigated. In a Swedish long-term field experiment, perennial ryegrass (Lolium perenne L.) as a catch crop had no influence on P concentrations in tile-drainage water (Liu et al., 2012b). Likewise, in a leaching study in which eight catch crop species (broad-leaved plants and grasses) were grown on topsoil lysimeters with clay soils, none of the catch crops reduced P leaching compared with the control without a catch crop (Liu et al., 2015). This is despite the fact that, in the field where the lysimeters were collected, the catch crops took up considerable amounts of P, which were stored in above- and below-ground plant parts. Under the best growing conditions, the catch crops took up as much as 6 to 15 kg P ha-1, with 20 to 80% of the P in roots depending on species.

A concern under Nordic conditions with cold winters is that catch crops can become sources of P losses after they are exposed to freezing-thawing events. In winter, plant cells may burst due to formation of ice crystals and frost damage ( Jones, 1992), which can lead to the release of inter-/intracellular P from catch crops. Several studies with cut plant material of catch crops grown in greenhouses have demonstrated that all the P in the crops is potentially released and may end up in water percolating through soil after a few freezing-thawing cycles (FTCs) (Bechmann et al., 2005; Liu et al., 2013). The use of such materials on clay soils may thereby cause high P leaching (Riddle and Bergström, 2013). In the lysimeter study described above, the influence of catch crops on P leaching from the lysimeters after exposure to FTCs was also studied. Among the species tested, perennial ryegrass and oilseed radish (Raphanus sativus L.) significantly increased TP concentrations in leachate from the lysimeters after exposure to FTCs compared with before exposure (Liu et al., 2015) (Fig. 9). Such losses of P from catch crops must be considered when using catch crops in the context of climate change because increasing numbers of soil FTCs of greater intensity are predicted to occur in Scandinavia during the next 50 to 100 yr (Mellander et al., 2007). However, according to the studies reviewed here, it seems that catch crops have limited value as an option to reduce P leaching from agricultural soils under cold climate conditions and that P leaching is regulated by other factors, such as P levels in soil.

The Importance of Subsoil Properties on P LeachingThe leaching studies in lysimeters referred to above were all

limited to the topsoil (0–20 cm), which is the portion of the agricultural soil profile that often contains the most P. Some studies have also shown that P leaching is strongly correlated to extractable P in topsoil (e.g., Maguire and Sims, 2002). However, other studies have found that subsoil properties may confound the influence of topsoil properties on P leaching (Sinaj et al., 2002; Djodjic et al., 2004). The large impact of subsoil properties on P leaching was demonstrated in a study performed by Andersson et al. (2015). Of the two sandy soils included in that study, one with a considerably higher P content in the topsoil (Mellby; >80 mg Olsen-P kg-1 soil) had much smaller P leaching loads than the other (Nåntuna; <30 mg Olsen-P kg-1 soil), with average loads of 0.25 and 4.39 kg TP ha-1 yr-1, respectively (Fig. 10). This was mainly attributed to the much higher sorption capacity in the Mellby subsoil, which contained higher concentrations of Fe oxides than the Nåntuna subsoil (Andersson et al., 2013). In the context of finding BMPs to reduce P leaching, this example shows that knowledge about subsoil properties is essential, especially if the subsoil contains large amounts of P and has low

Fig. 8. Seasonal (3-mo) leaching losses of total P (TP) in plots at Bornsjön with or without structure liming. The bars represent the seasons spring (Mar.–May), summer (June–Aug.), autumn (Sept.–Nov.), and winter (Dec.–Feb.). Error bars show SD (adapted from Svanbäck et al. [2014]).

520 Journal of Environmental Quality

sorption capacity. Phosphorus in drainage water from such soils could be reduced by using, for example, constructed wetlands (to reduce PP) or gypsum filters installed beyond the field boundary or added to tile-drain backfill (to reduce DRP) (Bryant et al., 2012). Increasing the sorption capacity of the subsoil would also be an option, if this was possible.

Reducing P Losses beyond the Field BoundaryFree water surface wetlands can mitigate nutrient losses

from agricultural areas. However, there are few quantitative measurements showing the efficiency of wetlands as P traps under Nordic climate conditions. In a compilation of data from 17 wetlands located in Scandinavia, Switzerland, and Illinois, it was found that factors such as wetland area in relation to catchment area and wetland age were of great importance for P retention, which varied between 1 and 88% for TP (Braskerud et al., 2005). The variation was even greater for DRP (-19 to 89%).

Since the 1990s, farmers in Sweden have been able to obtain environmental subsidies to construct wetlands that increase biodiversity or remove N but not P. In 2010, a new subsidy for small wetlands (0.1–0.5% of catchment area) focusing on P retention was implemented based on good efficiency in small Norwegian wetlands, in which P retention is reported to be 21 to 44% (Braskerud, 2002). The P wetlands comprise a deep pond followed by a shallow vegetated area. The first small P wetland (0.3% of the catchment area) with this design in Sweden was

constructed in 2009, next to the drainage culvert for a small catchment (26 ha) consisting of an agricultural field and horse paddocks (Kynkäänniemi et al., 2013). All drainage water from the area now enters the wetland. Flow-proportional composite water samples were collected at the wetland inlet and outlet and analyzed for TP, DRP, and PP, and retention was estimated for the first 2 yr (2010–2011) after construction. Outflow P concentrations were generally lower than inflow concentrations. On an annual basis, the P wetland acted as a net P sink, with mean specific retention of 69 kg TP ha-1 yr-1 (Kynkäänniemi et al., 2013). The relative TP retention was 36%, which is similar to that reported for the Norwegian wetlands mentioned above. Further studies on this and other wetlands have shown similar relative retention (Kynkäänniemi et al., unpublished data). The P trapped in the wetland was mostly PP, and only 9% of DRP was retained (Kynkäänniemi et al., 2013). The TP reduction was 0.22 kg ha-1 yr-1 of the load from the agricultural catchment, which indicates that small constructed wetlands have the ability to reduce P losses in drainage water from agricultural clay soils in regions with a cold climate. Over the long-term, this requires proper management of the wetlands, with, for example, removal of sediments that have accumulated on the bottom of the deep pond of a wetland with the design described here. Neglecting this would cause the wetland to become a source of P rather than a sink (Uusi-Kämppä et al., 2000).

Reflections on the Use of Best Management Practices to Reduce P Leaching in the Nordic Climate

The BMPs evaluated in this review were selected based on the fact that they have all been tested in Sweden in recent years for their efficiency in reducing P leaching and are eligible for governmental subsidies aimed at promoting their use. However, several other measures that have the potential to reduce agricultural P losses, such as those reducing surface runoff losses of P (e.g., buffer strips), were not included in this review because the main focus was on subsurface losses of P.

In terms of BMPs to reduce leaching of P from agricultural soils, it is our opinion that, in principle, measures are most effective when performed as early as possible, preventing P mobilization from the field, with measures outside the field boundary (e.g., constructed wetlands) as a complement. In terms

Fig. 9. Net change in total P concentration in the leachate from soil lysimeters after freezing-thawing cycles compared with before in all experiments (n = 8–20 for each catch crop species) (adapted from Liu et al. [2015]).

Fig. 10. Average annual load of total P from the Mellby sand and the Nåntuna sand with and without topsoil during the period 1 Sept. 2010 to 31 Aug. 2013. Error bars show SD (adapted from Andersson et al. [2015]).

www.agronomy.org • www.crops.org • www.soils.org 521

of the Baltic Sea, it is also important to consider the internal load of P from anoxic bottom sediments when estimating the eutrophication risk. In 2005, this internal load was estimated to be 2.3 g P m-2 yr-1 (i.e., about 92,000 t P yr-1 from sediments in the Baltic Sea Proper) (Stigebrandt et al., 2014). This can be compared with the load originating from agricultural nonpoint sources, which is considerably smaller. In other words, the effects of using BMPs to reduce P load to the Baltic Sea could, in many cases, be overshadowed by leakage of P from sediments. The average P load from tile-drained agricultural fields in Sweden is about 0.4 kg P ha-1 yr-1 (Bergström et al., 2007), which is rather small from an international perspective. For example, in Finland the average annual agricultural P load to the sea is estimated to be 1.1 kg TP ha-1 (Heckrath et al., 2008). The question is how to reduce such small losses further to meet the Swedish environmental quality goals triggered by the European Union Water Framework Directive. The BMPs presented here are all intended for use in achieving such a reduction.

On farms with animals, appropriate manure management is a necessity. It must include sufficient storage capacity to ensure that the manure is spread when the risk of P losses is small and to enable manure management related to application rates, which are strictly regulated in Sweden. Incorporation of manure immediately after spreading on clay soils, as exemplified in this review, was shown to be an important measure to reduce P leaching. This has also been shown in many other studies (e.g., Glæsner et al., 2011) and in studies focusing on surface runoff (e.g., Kleinman et al., 2002). In addition to manure-related BMPs performed on agricultural fields, a range of other measures can be adopted to reduce agricultural P losses, including additives to animal feed (phytase) and manure (e.g., aluminum sulfate). The latter contributes to P binding and thus decreases losses when the manure is applied on fields (Sims and Luka-McCafferty, 2002). However, such solutions to temporarily bind P, thereby rendering it less accessible to dissemination in nature, have never achieved any great breakthrough in Sweden, where solutions based on avoiding excess balances of P and improved use efficiency have considerably greater acceptance. Of the BMPs included in this overview, structure liming is most likely to have the potential for extensive use to reduce P losses in Sweden. At present, structure liming is officially recommended by the Swedish Board of Agriculture as a BMP to improve soil structure in clay soils and thereby reduce P leaching (SBA, 2013). It has also received great acceptance by the farming community. There are several reasons for this: besides reducing P losses, there is usually a yield benefit associated with liming (Coventry et al., 1987), which increases the cost-effectiveness of the method. However, there are still unanswered questions related to structure liming that require more field studies (e.g., to clarify the effect of structure liming on P leaching as a function of available soil P content alone and in combination with different clay contents in soil). Such studies should be of a long-term nature because lime distribution into the soil and soil aggregate formation by biological activities take time. It is also important to stress that structure liming is only applicable to clay soils. For other soil types, BMPs other than those included in this review are preferable.

Monitoring Programs to Examine Long-Term P Trends in Water Recipients

It is important to have good general knowledge of the mobilization and transport of P in soil but also of the time taken to obtain a response to any BMPs and environmental regulations implemented. For this purpose, losses of P from agriculture to surrounding waters are monitored at several scales in Sweden. In monitoring programs covering large rivers, losses of P from agriculture are seldom the single major source because the contribution from forested areas and point sources may be considerable. For further determination of the impact of agriculture on water quality in recipient water bodies, monitoring in small agricultural catchments was started by local authorities in Sweden in the early 1990s. Catchment streams were monitored continuously for water flow rates and biweekly for water quality, and crop management data were collected occasionally by interviewing the farmers within the catchments. In 2002, eight catchments were transferred to a national program run by the Swedish Environmental Protection Agency to secure the long-term status of the program and to intensify monitoring by introducing, for example, flow-proportional water sampling, annual crop management surveys, and groundwater measurements (Kyllmar et al., 2014). The national program is complemented by a regional program consisting of 10 catchments with a less intensive monitoring approach and three pilot catchments where the strict monitoring approach is abandoned. In these pilot catchments, farmers, agricultural advisors, researchers, and authorities are working together to find effective BMPs against P losses.

In the available data, the area-specific losses of P in stream outlets of the catchments studied, calculated as long-term annual averages, varied between 0.1 and 1.9 kg ha-1. In general, the losses were larger in catchments with clay soils than in those with more coarse-textured soils due to erosion in soil profiles and drainage systems, on the soil surface, and in streams.

In time series of P losses for the small agricultural catchments, no clear pattern was observed, and the tendency was for both upward and downward trends (Kyllmar et al., 2006; Fölster et al., 2012; Pengerud et al., 2015). In comparison, for N losses, significant decreasing trends were detected for several catchments, and the tendency as a whole was for decreasing trends. The lack of trends in P losses relates to the fact that, to date, measures against P losses have been less widespread than those against N losses. Another factor is that P losses vary widely within and between years, especially in clay soil catchments (Kyllmar et al., 2014), making detection of trends more difficult. A decreasing trend in one of the national catchments can be explained by a larger proportion of winter crops and forage crops on arable land and by lower livestock density and consequently less spreading of manure, especially in autumn. One of the pilot catchments also had decreasing trends. Measures such as structure liming, reduced soil tillage, and buffer strips have been widely implemented in that catchment during recent years. Regarding significant trends in some regional catchments, no local information on possible agricultural changes was available. However, it can be assumed that they follow the regional trends, with establishment of buffer strips along water courses, more autumn-sown crops and forage crops, less spreading of manure, decreased P balances, and

522 Journal of Environmental Quality

decreasing area of arable land. A decreasing contribution from point sources (scattered households and manure storage) may also be an explanation. Trends in N losses in small catchments are correlated to the downward trends reported for several large rivers in agricultural areas of southwestern Sweden (Fölster et al., 2012). For P losses, the diverse picture apparent in the trends for the small catchments was also seen for large rivers. In small agricultural monitoring catchments in other Nordic-Baltic countries (Denmark, Norway, Finland, Estonia, Latvia, and Lithuania), downward trends in P losses have been reported for around 20% of these catchments (Pengerud et al., 2015).

ConclusionsBased on the results presented in this review, it is important to

achieve a P level in soil that promotes efficient P use by crops and reduces the risk of P transfers to rivers and lakes and, in northern Europe, to the Baltic Sea. Analysis of long-term data showed that limiting P application to soil to “replacement of P removed by crops” resulted in slow depletion of P-AL over time. The data also showed that, when large amounts of P were applied, an increase in P-AL occurred over time, thereby increasing the risk of P leaching. Therefore, P supplied with fertilizer or manure should be applied at a rate slightly over or balancing crop removal to reduce P leaching in the long term. Examples of BMPs that have the ability to considerably reduce drainage losses of P from clay soils in cold climate regions according to the leaching studies reviewed here include liming, incorporation of manure into soil, and small constructed wetlands. However, catch crops appear to have limited value as an option to reduce P leaching under such conditions. To determine the time taken to obtain an effect of BMPs on P levels in recipient water bodies, well-designed monitoring programs running for long periods are needed. These can capture the often slow processes involved in P turnover in agricultural soils.

AcknowledgmentsThe authors are grateful for financial support from The Swedish Farmers’ Foundation for Agricultural Research and the Swedish University of Agricultural Sciences.

ReferencesAhlgren, J., F. Djodjic, G. Börjesson, and L. Mattsson. 2013. Identification and

quantification of organic phosphorus forms in soils from fertility experiments. Soil Use Manage. 29:24–35. doi:10.1111/sum.12014

Andersson, H., L. Bergström, F. Djodjic, B. Ulén, and H. Kirchmann. 2013. Topsoil and subsoil properties influence on phosphorus leaching from four agricultural soils. J. Environ. Qual. 42:455–463. doi:10.2134/jeq2012.0224

Andersson, H., L. Bergström, B. Ulén, F. Djodjic, and H. Kirchmann. 2015. The role of subsoil properties as a source or sink for phosphorus leaching. J. Environ. Qual. doi:10.2134/jeq2014.04.0186

Andersson, A., J. Eriksson, L. Mattsson, and R. Andersson. 1998. Fosforupplagring i svensk jordbruksmark. Report 4919. (In Swedish.) Swedish Environmental Protection Agency, Stockholm.

Aronsson, H., J. Liu, E. Ekre, G. Torstensson, and E. Salomon. 2014. Effects of pig and dairy slurry application on N and P leaching from crop rotations with spring cereals and forage leys. Nutr. Cycling Agroecosyst. 98:281–293. doi:10.1007/s10705-014-9611-3

Bechmann, M.E., P.J.A. Kleinman, A.N. Sharpley, and L.S. Saporito. 2005. Freeze-thaw effects on phosphorus loss in run-off from manures and catch-cropped soils. J. Environ. Qual. 34:2301–2309. doi:10.2134/jeq2004.0415

Bechmann, M., J. Deelstra, P. Stålnacke, H.O. Eggestad, L. Øygarden, and A. Pengerud. 2008. Monitoring catchment scale agricultural pollution in Norway: Policy instruments, implementation of mitigation methods and trends in nutrient and sediment losses. Environ. Sci. Policy 11:102–114. doi:10.1016/j.envsci.2007.10.005

Bergström, L., F. Djodjic, H. Kirchmann, I. Nilsson, and B. Ulén. 2007. Phosphorus from farmland to water– status, flows and preventive measures in a Nordic perspective. Report Food 21 no. 4/2007. Swedish Univ. of Agric. Sci., Uppsala.

Boesch, D., R. Hechy, C. O’Melia, D. Schindler, and S. Seitzinger. 2006. Eutrophication of the Swedish seas. Rep. 5509. Swedish Environ. Protect. Agency, Stockholm.

Boesch, D., J. Carstensen, H.W. Paerl, H.R. Skjodal, and H. Voss. 2008. Eutrophication of the seas along Sweden’s West Coast. Rep. 5898. Swedish Environ. Protect. Agency, Stockholm.

Börling, K., E. Otabbong, and E. Barberis. 2004. Soil variables for predicting potential phosphorus release in Swedish noncalcareous soils. J. Environ. Qual. 33:99–106. doi:10.2134/jeq2004.9900

Brandt, M., H. Ejhed, and L. Rapp. 2009. Nutrient loads to the Swedish marine environment in 2006. Report 5995. Swedish Environ. Protect. Agency, Stockholm.

Braskerud, B.C. 2002. Factors affecting phosphorus retention in small constructed wetlands treating agricultural non-point source pollution. Ecol. Eng. 19:41–61. doi:10.1016/S0925-8574(02)00014-9

Braskerud, B.C., K.S. Tonderski, B. Wedding, R. Bakke, A.-G.-B. Blankenberg, B. Ulén, and J. Koskiaho. 2005. Can constructed wetlands reduce the diffuse phosphorus loads to eutrophic water in cold temperature regions? J. Environ. Qual. 34:2145–2155. doi:10.2134/jeq2004.0466

Bryant, R.B., A.R. Buda, P.J.A. Kleinman, C.D. Church, L.S. Saporito, G.L. Folmar, S. Bose, and A.L. Allen. 2012. Using flue gas desulfurization gypsum to remove dissolved phosphorus from agricultural drainage water. J. Environ. Qual. 41:664–671. doi:10.2134/jeq2011.0294

Carlgren, K., and L. Mattsson. 2001. Swedish soil fertility experiments. Acta Agric. Scand. B 51:49–78.

Choquette, M., M.-A. Bérube, and J. Locat. 1987. Minerological and microtextural changes associated with lime stabilization of marine clays from eastern Canada. Appl. Clay Sci. 2:215–232. doi:10.1016/0169-1317(87)90032-9

Coventry, D.R., T.G. Reeves, H.D. Brookes, A. Ellington, and W.J. Staffery. 1987. Increasing wheat yields in north eastern Victoria by liming and deep ripping. Aust. J. Exp. Agric. 27:679–685. doi:10.1071/EA9870679

Djodjic, F., K. Börling, and L. Bergström. 2004. Phosphorus leaching in relation to soil type and soil phosphorus content. J. Environ. Qual. 33:678–684. doi:10.2134/jeq2004.6780

Djodjic, F., and K. Kyllmar. 2011. Spridning av gödselmedel på åkermark. (In Swedish.) Swedish University of Agricultural Sciences, Department of Water and Environment, Uppsala.

Egnér, H., H. Riehm, and W.R. Domingo. 1960. Untersuchungen über die chemische Bodenanalyse als Grundlage für die Beurteiling des Nährstoffzustandes der Böden. II Chemische Extrationsmethoden zur Phosphor- och Kaliumbestimmung. (In German.) Kungliga Lantbrukshögskolans Annaler 26:199–215.

Feistel, R., S. Weinreben, H. Wolf, S. Seitz, P. Spitzer, B. Adel, G. Nausch, B. Schneider, and D.G. Wright. 2010. Density and absolute salinity of the Baltic Sea 2006-2009. Ocean Sci. 6:3–24. doi:10.5194/os-6-3-2010

Fölster, J., K. Kyllmar, M. Wallin, and S. Hellgren. 2012. Kväve- och fosfortrender i jordbruksvattendrag: Har åtgärderna gett effekt? [In Swedish, with English summary.) Report 2012:1 Swedish University of Agricultural Sciences, Department of Aquatic Sciences and Assessment, Uppsala.

Gburek, W.J., A.N. Sharpley, Á.L. Heathwaite, and G.J. Folman. 2000. Phosphorus management at the watershed scale: A modification of the Phosphorus Index. J. Environ. Qual. 29:130–144. doi:10.2134/jeq2000.00472425002900010017x

Gelbrecht, J., H. Lemgsfeld, R. Pötzig, and P. Opitz. 2005. Temporal and spatial variation of phosphorus input, retention and loss in a small catchment of NE Germany. J. Hydrol. 304:151–165. doi:10.1016/j.jhydrol.2004.07.028

Glæsner, N., C. Kjaergaard, G.H. Rudbæk, and J. Magid. 2011. Interaction between soil texture and placement of dairy slurry application: II. Leaching of phosphorus forms. J. Environ. Qual. 40:344–351. doi:10.2134/jeq2010.0318

Grimvall, A., and A. Nordgaard, 2004. Sjöar och vattendrag i Skåne: går utvecklingen åt rätt håll? Statistisk utvärdering av vattenkvalitet och provtagningsprogram i Skåne län. (In Swedish.) County Board of the County Scania, the Environmental Div., Scania, Sweden.

Havlin, J.L., J.D. Beaton, S.L. Tisdale, and W.L. Nelson. 2005. Soil fertility and fertilizers. 7th ed. Prentice Hall, Upper Saddle River, NJ.

Haygarth, P.M., and A.N. Sharpley. 2000. Terminology of phosphorus transfer. J. Environ. Qual. 29:10–15. doi:10.2134/jeq2000.00472425002900010002x

Heckrath, G., M. Bechmann, P. Ekholm, B. Ulén, F. Djodjic, and H.E. Andersen. 2008. Review of indexing tools for identifying high risk areas of phosphorus loss in Nordic catchments. J. Hydrol. 349:68–87. doi:10.1016/j.jhydrol.2007.10.039

Helsinki Commission. 2013a. Review of the Fifth Baltic Sea Pollution Load Compilation for the 2013 HELCOM Ministry of Meeting. Baltic Sea Environment Proceedings no. 141. Baltic Marine Environ. Protect. Commiss., Copenhagen, Denmark.

Helsinki Commission. 2013b. 2013 HELCOM Copenhagen ministerial declaration 2013. Helsinki Commission, Copenhagen, Denmark.

www.agronomy.org • www.crops.org • www.soils.org 523

Jones, H.G. 1992. Plants and microclimate. 2nd ed. Cambridge Univ. Press, Cambridge, UK.

Kavak, A., and G. Baykal. 2012. Long-term behaviour of lime-stabilized kaolinite clay. Environ. Earth Sci. 66:1943–1955. doi:10.1007/s12665-011-1419-8

Kleinman, P.J.A., A.N. Sharpley, B.G. Moyer, and G.F. Elwinger. 2002. Effect of mineral and manure phosphorus sources on runoff phosphorus. J. Environ. Qual. 31:2026–2033. doi:10.2134/jeq2002.2026

Kyllmar, K., C. Carlsson, A. Gustafson, B. Ulén, and H. Johnsson. 2006. Nutrient discharge from small agricultural catchments in Sweden. Agric. Ecosyst. Environ. 115:15–26. doi:10.1016/j.agee.2005.12.004

Kyllmar, K., L. Stjernman Forsberg, S. Andersson, and K. Mårtensson. 2014. Small agricultural monitoring catchments in Sweden representing environmental impact. Agric. Ecosyst. Environ. 198:25–35.

Kynkäänniemi, P., B. Ulén, G. Torstensson, and K.S. Tonderski. 2013. Phosphorus retention in a newly constructed wetland receiving agricultural tile drainage water. J. Environ. Qual. 42:596–605. doi:10.2134/jeq2012.0266

Ladha, J.K., H. Pathak, T.J. Krupnik, J. Six, and C. van Kessel. 2005. Efficiency of fertilizer nitrogen in cereal production: Retrospects and prospects. Adv. Agron. 87:85–156. doi:10.1016/S0065-2113(05)87003-8

Lemunyon, J. 2006. Cover crops. SERA-17: Minimizing phosphorus losses from agriculture. http://sera17.ext.vt.edu/ (accessed Feb. 2007).

Littell, R., G. Milliken, W. Stroup, R. Wolfinger, and O. Schabenberger. 2006. SAS for mixed models. 2nd ed. SAS Inst., Cary, NC.

Liu, J., H. Aronsson, L. Bergström, and A.N. Sharpley. 2012a. Phosphorus leaching from loamy sand and clay loam topsoils after application of pig slurry. SpringerPlus 1: Paper 53. http://www.springerplus.com/content/1/1/53 (accessed Dec. 2013).

Liu, J., H. Aronsson, K. Blombäck, K. Persson, and L. Bergström. 2012b. Long-term measurements and model simulations of phosphorus leaching from a manured sandy soil. J. Soil Water Conserv. 67:101–110. doi:10.2489/jswc.67.2.101

Liu, J., H. Aronsson, B. Ulén, and L. Bergström. 2012c. Potential phosphorus leaching from sandy topsoils with different fertiliser histories before and after application of pig slurry. Soil Use Manage. 28:457–467. doi:10.1111/j.1475-2743.2012.00442.x

Liu, J., R. Khalaf, B. Ulén, and G. Bergkvist. 2013. Potential phosphorus release from catch crop shoots and roots after freezing-thawing. Plant Soil 371:543–557. doi:10.1007/s11104-013-1716-y

Liu, J., B. Ulén, G. Bergkvist, and H. Aronsson. 2015. Freezing-thawing effects on phosphorus leaching from potential catch crops. Nutr. Cycling Agroecosyst. (in press).

Maguire, R.O., and J.T. Sims. 2002. Soil testing to predict phosphorous leaching. J. Environ. Qual. 31:1601–1610. doi:10.2134/jeq2002.1601

McLaughlin, M.J., A.M. Alston, and J.K. Martin. 1988. Phosphorus cycling in wheat-pasture rotations: III. Organic phosphorus turnover and cycling. Aust. J. Soil Res. 26:343–353. doi:10.1071/SR9880343

Mellander, P., M.O. Löfvenius, and H. Laudon. 2007. Climate change impact on snow and soil temperature in boreal Scots pine stands. Clim. Change 85:179–193. doi:10.1007/s10584-007-9254-3

Negassa, W., and P. Leinweber. 2009. How does the Hedley sequential phosphorus fractionation reflect impacts of land use and management on soil phosphorus: A review. J. Plant Nutr. Soil Sci. 172:305–325. doi:10.1002/jpln.200800223

Pengerud, A., S. Stålnacke, M. Bechmann, G. Blicher-Mathiesen, A. Iital, J. Koskiaho, K. Kyllmar, A. Lagzdins, and A. Povilaitis. 2015. Temporal trends in phosphorus concentrations and losses from agricultural catchments in the Nordic and Baltic countries. Acta Agric. Scand. (in press).

Riddle, M.U., and L. Bergström. 2013. Phosphorus leaching from two soils with catch crops exposed to freeze-thaw cycles. Agron. J. 105:803–811. doi:10.2134/agronj2012.0052

SBA. 2010. Föreskrifter om ändring i Statens jordbruksverks föreskrifter och allmänna råd (SJVFS 2004:62) om miljöhänsyn i jordbruket vad avser växtnäring. SJVFS 2010:55. (In Swedish.) Swedish Board of Agriculture, Jönköping.

SBA. 2013. Swedish Board of Agriculture guidelines for fertilization and liming 2014. Report Jo 13:11, 2013-11-18. (In Swedish.) Swedish Board of Agriculture, Jönköping.

SCB. 2012. Jordbruksstatistisk årsbok [yearbook of agricultural statistics]. (In Swedish, with English summary.) Statistiska Centralbyrån, Stockholm.

Schindler, D.W. 1977. Evolution of phosphorus limitation in lakes. Science 195:260–262. doi:10.1126/science.195.4275.260

SEPA. 2009. Sweden’s commitment under the Baltic Sea Action Plan. Rep. 5989. Swedish Environ. Protect. Agency, Stockholm.

Sims, J.T., and N.J. Luka-McCafferty. 2002. On-farm evaluation of aluminum sulfate (Alum) as a poultry litter amendment: Effects on litter properties. J. Environ. Qual. 31:2066–2073. doi:10.2134/jeq2002.2066

Sinaj, S., C. Stamm, G.S. Toor, L.M. Condron, T. Hendry, H.J. Di, K.C. Cameron, and E. Froussard. 2002. Phosphorus exchangeability and leaching losses from two grassland soils. J. Environ. Qual. 31:319–330. doi:10.2134/jeq2002.3190

Statistics Sweden. 2001. Nitrogen and phosphorus balances for agricultural land and the agricultural sector in 1999. (In Swedish, with English summary.) Statistiska meddelanden MI 40 SM 0101.

Statistics Sweden. 2003a. Nitrogen and phosphorus balances for agricultural land and the agricultural sector in 2001. (In Swedish, with English summary.) Statistiska meddelanden MI 40 SM 0301.

Statistics Sweden. 2003b. Statistical data for drainage areas 2000. (In Swedish, with English summary.) Statistiska meddelanden MI 11 SM 0301.

Statistics Sweden. 2005. Nitrogen and phosphorus balances for agricultural land and the agricultural sector in 2003. (In Swedish, with English summary.) Statistiska meddelanden MI 40 SM 0501.

Statistics Sweden. 2007. Nitrogen and phosphorus balances for agricultural land and the agricultural sector in 2005. (In Swedish, with English summary.) Statistiska meddelanden MI 40 SM 0701.

Statistics Sweden. 2009. Nitrogen and phosphorus balances for agricultural land and the agricultural sector in 2007. (In Swedish, with English summary.) Statistiska meddelanden MI 40 SM 1101.

Statistics Sweden. 2011. Nitrogen and phosphorus balances for agricultural land and the agricultural sector in 2009. (In Swedish, with English summary.) Statistiska meddelanden MI 40 SM 1102.

Statistics Sweden. 2013. Nitrogen and phosphorus balances for agricultural land and the agricultural sector in 2011. (In Swedish, with English summary.) Statistiska meddelanden MI 40 SM 1301.

Statistics Sweden. 2014. Drainage of agricultural land, preliminary statistics. Serie JO- Jordbruk, skogsbruk och fiske JO 10 SM 1401.

Stigebrandt, A., L. Rahm, L. Viktorsson, M. Ödalen, P.O.J. Hall, and B. Liljebladh. 2014. A new phosphorus paradigm for the Baltic Proper. Ambio 43:634–643. doi:10.1007/s13280-013-0441-3

Svanbäck, A., B. Ulén, and A. Etana. 2014. Mitigation of phosphorus leaching losses via subsurface drains from a cracking marine soil. Agric. Ecosyst. Environ. 184:124–134. doi:10.1016/j.agee.2013.11.017

Svanbäck, A., B. Ulén, A. Etana, L. Bergström, P.J.A. Kleinman, and L. Mattsson. 2013. Influence of soil phosphorus and manure on phosphorus leaching in Swedish topsoils. Nutr. Cycling Agroecosyst. 96:133–147. doi:10.1007/s10705-013-9582-9

Syers, J.K., A.E. Johnston, and D. Curtin. 2008. Efficiency of soil and fertilizer phosphorus use. FAO Fertilizer and Plant Nutrition Bull. 18. FAO, Rome.

Thorup-Kristensen, K., J. Magid, and L.S. Jensen. 2003. Catch crops and green manures as biological tools in nitrogen management in temperate zones. Adv. Agron. 79:227–302. doi:10.1016/S0065-2113(02)79005-6

Ulén, B. 1995. Episodic precipitation and discharge events and their influence on losses of phosphorus and nitrogen from tile drained arable fields. Swed. J. Agric. Res. 25:25–31.

Ulén, B. 2004. Size and settling velocities of phosphorus-containing particles in water from agricultural drains. Water Air Soil Pollut. 157:331–343.

Ulén, B. 2006. A simplified risk assessment for losses of dissolved reactive phosphorus through drainage pipes from agricultural soils. Acta Agric. Scand. B (Soil and Plant Sci.) 56:307–314.

Ulén, B., and J. Fölster. 2007. Recent trends in nutrient concentrations in Swedish agricultural rivers. Sci. Total Environ. 373:473–487. doi:10.1016/j.scitotenv.2006.11.032

Ulén, B., M. Bechmann, J. Fölster, H.P. Jarvie, and H. Tunney. 2007. Agriculture as a phosphorus source for eutrophication in the north-west European countries, Norway, Sweden, United Kingdom and Ireland: A review. Soil Use Manage. 23:5–15. doi:10.1111/j.1475-2743.2007.00115.x

Ulén, B., and A. Etana. 2014. Phosphorus leaching from clay soils can be counteracted by structure liming. Acta Agric. Scand. Sec. B (Soil and Plant Sci.) 64:425–433.

Ulén, B., and S. Snäll. 2007. Forms and reduction of phosphorus in an illite-clay soil profile with a history of fertilization with pig manure and mineral fertilizers. Geoderma 137:455–465. doi:10.1016/j.geoderma.2006.10.003

Ulén, B., C. von Brömsen, G. Johansson, G. Torstensson, and L. Stjernman-Forsberg. 2012. Trends in nutrient concentrations in drainage water from single fields under ordinary cultivation. Agric. Ecosyst. Environ. 151:61–69. doi:10.1016/j.agee.2012.02.005

Uusi-Kämppä, J., B. Braskerud, H. Jansson, N. Syversen, and R. Uusitalo. 2000. Buffer zones and constructed wetlands as filters for agricultural phosphorus. J. Environ. Qual. 29:151–158. doi:10.2134/jeq2000.00472425002900010019x

Van der Salm, C., W.J. Chardon, G.F. Koopmans, J.C. van Middelkoop, and P.A.I. Ehlert. 2009. Phytoextraction of phosphorus-enriched grassland soils. J. Environ. Qual. 38:751–761. doi:10.2134/jeq2008.0068

Walker, T.W., and J.K. Syers. 1976. The fate of phosphorus during pedogenesis. Geoderma 15:1–19. doi:10.1016/0016-7061(76)90066-5