Review of indexing tools for identifying high risk areas of phosphorus loss in Nordic catchments

This article was published in an Elsevier journal. The attached copyis furnished to the author for non-commercial research and

education use, including for instruction at the author’s institution,sharing with colleagues and providing to institution administration.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

http://www.elsevier.com/copyright

Author's personal copy

Review of indexing tools for identifying high riskareas of phosphorus loss in Nordic catchments

G. Heckrath a,*, M. Bechmann b, P. Ekholm c, B. Ulen d, F. Djodjic e,H.E. Andersen f

a University of Aarhus, Department of Agroecology and Environment, Research Centre Foulum, P.O. Box 50, 8830 Tjele,Denmarkb Bioforsk, Norwegian Institute for Agricultural and Environmental Research, Fred. A. Dahls Vei 20, 1432 As, Norwayc Finnish Environment Institute, P.O. Box 140, 00251 Helsinki, Finlandd Swedish University of Agricultural Sciences, Department of Soil Science, P.O. Box 7014, 75007 Uppsala, Swedene Swedish University of Agricultural Sciences, Department of Environmental Assessment, P.O. Box 7050, 75007 Uppsala,Swedenf University of Aarhus, National Environment Research Institute, Department of Freshwater Ecology, 8600 Silkeborg,Denmark

KEYWORDSRegulation;P Index;Nordic countries;Agriculture;Water quality

Summary Compliance with the Water Framework Directive (WFD) will require substantialreductions in agricultural phosphorus (P) losses in the Nordic countries Denmark, Norway,Sweden and Finland. Falling P surpluses in agriculture for more than a decade and voluntaryprogrammes of good agricultural practice have not reduced P losses to surfacewaters, whilegeneral regulatory measures have primarily focused on nitrogen. Without addressing therole of critical source areas for P loss, policy measures to abate diffuse P losses are likelyto be ineffective. This has created a demand by environmental authorities for instrumentsthat assess the risk of P losses fromagricultural land and facilitate the planning ofmitigationmeasures. In Nordic countries index-type risk assessment tools for diffuse P losses are underdevelopment inspired by experiences with P indexing in the USA. A common feature is thatthey are empirical, risk-based, user-friendly decision tools with low data requirements.Phosphorus indices vary between the four Nordic countries in response to different agricul-ture, soil and climate. These differences also result in different recent average annual agri-cultural P load estimates to the sea of 0.3, 0.5, 0.5 and 1.1 kg total P ha�1 in Denmark,Norway, Sweden and Finland, respectively. In initial evaluations, Nordic P indices explaineda large degree of variance in P losses at the field or catchment scale, but comparativedata are still limited. To gain acceptance amongst stakeholders and inform river basin

0022-1694/$ - see front matter ª 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.jhydrol.2007.10.039

* Corresponding author. Tel.: +45 8999 1715; fax: +45 8999 1719.E-mail address: [email protected] (G. Heckrat).

Journal of Hydrology (2008) 349, 68–87

ava i lab le at www.sc iencedi rec t . com

journal homepage: www.elsevier .com/ locate / jhydro l


management planning in Nordic catchments as part of the WFD, it is crucial to more thor-oughly evaluate the performance of these indices’ at the field and catchment scale.

ª 2007 Elsevier B.V. All rights reserved.

Introduction

Diffuse phosphorus (P) losses from agricultural land to waterhave become a major environmental concern in the Nordiccountries (e.g. Selvik et al., 2006; Kronvang et al., 2005;Ekholm and Mitikka, 2006; Boesch et al., 2006). In the lightof the Water Framework Directive (WFD) of the EuropeanUnion, which commits member states to establishing highwater quality standards in all aquatic environments, it isessential to devise targeted measures for the control of Plosses from agricultural land, if the ambitious quality goalsfor surface waters in the Nordic countries (Denmark, Fin-land, Norway, Sweden) are to be achieved. In view of thelarge amounts of P that have already accumulated in agri-cultural soils in the Nordic region (e.g. Andersson et al.,1998; Haraldsen et al., 2001; Saarela, 2002), general regula-tory measures designed solely to reduce the P surplus inagriculture, or to limit the use of P fertilizers, are unlikelyto result in the necessary reductions of P transfer to waters(e.g. Stalnacke et al., 2004; Ekholm et al., 2005; Granlundet al., 2005). Instead, a decline in diffuse P losses has beenachieved in Norwegian (Bechmann and Stalnacke, 2005),and to a lesser degree Swedish (Ulen et al., 2004), catch-ments through targeting of erosion control measures inhigh-risk areas of P loss.

Phosphorus loss is a complex function of climate, topog-raphy, soil type and land management and varies both tem-porally and spatially. High-risk areas contributesubstantially more to P losses than other areas within agri-cultural catchments (e.g. Gburek et al., 2000; Uusitaloand Jansson, 2002). The identification of these so-called‘‘critical source areas’’ (CSA), therefore, is increasinglyseen as a prerequisite for implementing cost-effective mit-igation measures (e.g. Johansson and Randall, 2003; Ekholmet al., 2005). Due to the nature of P loss processes, system-atic mitigation planning across larger areas calls for anobjective risk assessment framework warranting equal stan-dards. Modelling tools differ widely in their approach to pre-dicting diffuse P losses from soil to water, i.e. in theircomplexity, their representation of processes and pathways(empirical, physical) and their resource requirements (data,time). A basic distinction can be made between simplescreening tools for identifying high-risk areas and complex,physically based scenario analysis tools (ICECREAM, GAMES,SWAT; e.g. Schoumans and Silgram, 2003) for exploringmanagement options in abating P losses from agriculturalland (Heathwaite et al., 2005a). Essentially, the specificneeds of end-users and the costs they are willing to allocateto mitigation planning determine the choice of model. Asphysically based models require extensive input data forsite-specific parameterization, they may be too expensiveto use and unsuitable for end-users conducting practical riskassessment and mitigation planning (e.g. Grayson et al.,1992). Phosphorus delivery to water bodies can be reducedby decreasing the potential for P mobilization in soils or at

the soil surface and by employing hydrotechnical and agri-cultural measures that strip P from runoff (Ulen and Jakobs-son, 2005). Hence, many mitigation strategies requiremanagement decisions best taken at the field and farmscale. Together this has created an urgent demand by envi-ronmental authorities and land-users in Nordic countries foruser-friendly decision tools that assess the risk of P lossesfrom agricultural land and facilitate the planning of mitiga-tion measures (Dørge and Windolf, 2003; Rekolainen et al.,2003).

Our aim in this paper is to provide an overview of the cur-rent development of decision tools to aid abatement of Plosses from Nordic agriculture. Such tools for integrated riskassessment of P loss, including sources, transport and deliv-ery of P, are not part of the official framework for watermanagement in any Nordic country. However, respondingto the challenges presented by the WFD, research on thedevelopment and feasibility of different risk assessmenttools has been conducted in Denmark, Finland, Norwayand Sweden in recent years.

Nordic agriculture and phosphorus losses to water

Nutrient transport in watercourses has been intensivelymonitored in the Nordic countries. The contribution of agri-cultural P losses to riverine P transport has typically beendetermined by means of indirect methods such as sourceapportionment (e.g. Kronvang et al., 2001) and in combina-tion with detailed field studies. Throughout the Nordic re-gion P discharges from sewage treatment works andindustry have declined dramatically over the past three dec-ades, leaving agriculture as one of the main anthropogenicsources of P input to the aquatic environment in Denmark(Kronvang et al., 2005), Finland (Vuorenmaa et al., 2002),Norway (Selvik et al., 2006) and Sweden (Ejhed et al., inpress). In Nordic agricultural catchments P losses vary sub-stantially in space and time with mean annual P loads rang-ing from a few hundred grams to several kilograms perhectare (e.g. Kronvang et al., 2007). Catchments vulnerableto erosion tend to lose most P. Phosphorus losses often haveshown no correlation with annual P surpluses in the catch-ments, but are explained by annual variations in runoff(Vagstad et al., 2001). Testing different riverine P transportmodels Andersen et al. (2005) obtained the same results fora large number of Danish catchments. Similarly, P surplus oranimal density had no effect on riverine P concentrations in27 long-term monitored agricultural catchments in Sweden(Kyllmar et al., 2006).

Denmark is characterised by an intensive agricultural pro-duction, which occupies 61% of the land area. About half ofthe agricultural land is tile-drained, and 90% of all watercourses have been widened or canalized to improve drainageof agricultural land (Kronvang et al., 1997). The average Psurplus calculated from a recent farm-gate balance is about13 kg P ha�1 yr�1 (Table 1), down from 29 kg P ha�1 yr�1 in

Review of indexing tools for identifying high risk areas of phosphorus loss in Nordic catchments 69


1980 (Kyllingsbæk and Hansen, 2007). Net input of P duringthe 20th century has accumulated 1400 kg P ha�1 in agricul-tural soils (Dalgaard and Rubæk, 2005). Despite a reductionin P discharge from point sources of 80% since the 1980s(Boutrup et al., 2006), the water quality of Danish lakesand estuaries continues to deteriorate due to diffuse P losses(Kronvang et al., 2002). The Danish National EnvironmentalResearch Institute suggests a maximum in-lake total P con-centration of 0.05 mg P l�1 for shallow lakes and0.025 mg P l�1 for deep lakes to achieve good ecologicalquality in accordance with the WFD (Søndergaard et al.,2005). At present the median in-lake total P concentrationin 692 Danish lakes >1 ha is 0.138 mg P l�1 and 75% of alllakes have a concentration above 0.058 mg P l�1 (Sønderg-aard et al., 2005). The median total P concentration andmedian P load in streams of typical agricultural catchmentswere about 0.1 mg P l�1 and 0.3 kg P ha�1, respectively(Fig. 1a). In Denmark agricultural P losses make up almost40% of the P loading to coastal waters (Table 1). Surface run-off and water erosion are considered less important for P lossthan subsurface runoff (Kronvang and Rubæk, 2005).

In Norway the agricultural area is small compared to thetotal land area (3% for the whole country). However, in thearea most vulnerable to eutrophication as defined by theOSPAR (2003) and the North Sea Declaration (south easternand southern Norway), agriculture is the dominating sourceand contributes somewhat less than 50% of the total anthro-pogenic loading of P (Selvik et al., 2006). Within this prob-lem area, agricultural land constitutes ca. 5% of the totalland area and is dominated by arable crop production (Table1). Lakes in the agricultural areas are generally eutrophicand P is the limiting nutrient (Berge et al., 2001). Mean an-nual P losses from agricultural land in Norway vary greatly inresponse to soils, climate and production systems (Fig. 1b).The smallest losses occur in extensive upland areas

(0.2 kg P ha�1, Fig. 1b). The main cause for P losses fromarable land is erosion (Lundekvam, 2007). The sloping landand the silty and clayey soils in south-eastern Norway areparticularly vulnerable, exemplified by average annual Plosses of 1.2 and 1.6 kg P ha�1 in the Mørdre and the Skute-rud catchment (Fig. 1b). Snowmelt contributes to the higherosion risk in this area. Additionally, land levelling in the1950–1960s has increased the erodibility of soils (Øygarden,2000). In the southern most parts (e.g. Skuterud) wintercereals have increased the intensity of autumn tillage andhence, increased the erosion risk. To lengthen the growingseason most agricultural soils (60%) are intensively tiledrained creating an important transport pathway for P.About 12–60% of the total P losses from agricultural landoccur in drainage (Bechmann et al., 2005). The annual P sur-plus in Norwegian agriculture has recently decreased to13 kg P ha�1 yr�1 (Table 1). Areas with intensive livestockproduction and grassland have substantially higher P sur-pluses and often relatively high P losses (Fig. 1b). Despiteextensive agriculture, areas dominated by organic soils alsoare vulnerable to P losses in Norway (e.g. Naurstad; Fig. 1b).

In Sweden only about 8% of the land area, or 2.7 mil-lion ha, is used for agriculture, 45% of which is tile drained(Table 1). Phosphorus surplus additions in the 20th centuryhave resulted in the accumulation of 600–700 kg P ha�1 intopsoils (Andersson et al., 2000), corresponding to half theenrichment reached in Denmark. Recently the annual P sur-plus has fallen from 5.2 kg P ha�1 in 1995 to a comparativelylow level of 2.1 kg P ha�1 in 2003 (Statistics Sweden, 2005).Considerable differences exist between regions due to dif-ferent livestock densities. The gross P load from agriculturalland has been estimated to contribute 40% of the total Pload to the Baltic Sea (Brandt and Ejhed, 2003). AgriculturalP losses have also enhanced eutrophication in many smallinland lakes (SEPA, 2003). The long-term annual average

Table 1 Sources of calculated annual P loads to the sea and relevant country statistics in Nordic countries

Denmark Norway OSPAR areaa Sweden Finland

Area (km2) 43,098 100,000 450,295 338,000Agricultural landb (%) 60 5 8 7Forested land (%) 11 38 49 78P surplusc (kg ha�1 yr�1) 13 (2004) 13d (2002–2004) 2.1 (2003) 10 (2002)

P source (tons yr�1)Background 400 163e 652 2700Agricultural 810 247 1590 2600Forestry n.d. n.d. 1240 320Scattered dwellings 210 n.d. 260 355Point sources (industry) 750 279 990 546Total P load to the sea 2170 695 4950 6521

Agricultural P loads (kg ha�1 yr�1) 0.3 0.5 0.5 1.1Year 2004 2005 2005 2005Reference Bøgestrand (2005) Selvik et al. (2006) Ejhed et al. (in press) www.ymparisto.fi

n.d.: not determined.a OSPAR area from the Swedish border to Lindesnes in southern Norway.b Including grassland.c Year shown in parentheses.d OECD (in press).e Including forestry.

70 G. Heckrath et al.


total P losses from 22 small Swedish agricultural catchments(1.8–35 km2) varied between 0.09 kg ha�1 and 0.87 kg ha�1

(Fig. 1c; Kyllmar and Grill, 2007). Similar results were ob-tained in field studies from different parts of the country(Fig. 1c). The proportion of dissolved reactive P rangedwidely between 22% and 86% of the total P and reflected dif-ferent transport processes (Johansson and Gustafson, 2005).Phosphorus losses in drainage dominate on the tile drainedclay soils in the region around big lakes in south and centralSweden and on the east coast. Surface runoff and erosionare more important on distinctly silty soils in the river val-leys of central and northern Sweden (Bergstrom et al.,2007). Losses of P from sandy soils in southern Sweden areusually low unless the soil’s sorption capacity is low (Ulen,2006; Djodjic and Bergstrom, 2005).

Only 7.4% of the land area of Finland is agricultural land.Yet, crop production and animal husbandry are responsiblefor about 60% of the anthropogenic load of phosphorus tosurface waters (Table 1). Total P concentrations in streamstypically are related to the intensity of agricultural produc-tion and soil texture (Vuorenmaa et al., 2002; Fig. 1d).Accordingly, in agricultural catchments many lakes, riversand coastal waters are eutrophic or hypertrophic (Meeuwiget al., 2000; Kauppila et al., 2003; Raike et al., 2003;Ekholm and Mitikka, 2006). A total of 58% of all agriculturalland is tile drained and another 27% has open ditches. In2004 the average P surplus at field level was10 kg P ha�1 yr�1 – down from 30 kg P ha �1 yr�1 in 1990(Ministry of Agriculture and Forestry, 2004). Net input of Pduring the 20th century has resulted in P accumulation of

Catchment type

I II III

TP

con

cent

ratio

n (m

g P

l-1)

0.0

0.1

0.2

0.3

0.4

0.5

I II III

TP

load

(kg

P h

a-1)

0

1

2

3

Skuterud Kolstad Volbu

TP

load

(kg

P h

a-1)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Catchment

Mørdre Skas-Heigre Naurstad

Monitoring facility

I II

TP

con

cent

ratio

n (m

g P

l-1)

0.0

0.1

0.2

0.3

0.4

0.5

I II

TP

load

(kg

P h

a-1

)

0

1

2

3

Catchment

I II III

TP

con

cent

ratio

n (m

g P

l-1)

0.0

0.1

0.2

0.3

0.4

0.5

I II III

TP

load

(kg

P h

a-1)

0

1

2

3

a b

c d

Figure 1 Examples of P losses from Nordic countries. Median, lower and upper quartile and the 10th and 90th percentiles areshown. (a) Discharge weighted TP concentration and TP loads from three types of monitored catchments in Denmark in 2004. I:nature areas, catchments with <10% agricultural land and no point sources (n = 10); II: agricultural catchments without point sources(n = 62); III: agricultural catchments with point sources (n = 64). (b) Annual TP loads from continuously monitored catchments withflow proportional water quality sampling in Norway, 1998–2006. Skuterud, Mørdre and Kolstad in south-eastern Norway aredominated by cereal production; Skag-Heigre in the south-west is dominated by dairy and grassland; Volbu, upland catchment incentral Norway, and Naustad in the north represent extensive livestock production. (c) Discharge weighted annual TP concentrationsand loads (I) from 13 cultivated monitoring fields in Sweden mainly for the period 1982–2005 (Johansson and Gustafson, 2007) and(II) from 23 small agricultural catchments with few point sources from scattered dwellings monitored between 7 and 21 years(Kyllmar and Grill, 2007). (d) Total P concentrations and annual TP loads in streams of three intensively monitored Finnishcatchments with increasing intensity of agricultural land use in the order Ylaneenjoki (I), Savijoki (II), Loytaneenoja (III), 1990–2004. Percentiles of TP concentrations are based on individual samples collected both manually and automatically; (I) n = 604, (II)n = 398, (III) n = 372. Based on Vuorenmaa et al. (2002) and the database of the Finnish Environment Institute.



about 1000 kg P ha�1 in agricultural soils (Saarela, 2002).The mean long-term P loss from agricultural land was1.1 kg P ha�1 yr�1 (Vuorenmaa et al., 2002), but the varia-tion from plot to plot was large. In three experimental fieldson fine soils in southern Finland, the 4-year average annual Ploss ranged from 0.32 to 2.49 kg ha�1 (Uusitalo, 2004). Theproportion of dissolved reactive P (DRP) in total P varied be-tween 14% and 23%. In general, the total P losses are a func-tion of soil erosion and soil texture, whereas DRP losses tendto increase with soil P status (Uusitalo, 2004). The contribu-tion of surface runoff to total runoff from fields rangeswidely between 10% and 90%, increasing with runoff vol-ume, slope, vegetation cover and ineffective drainage,and decreasing with macropore density and clay content(Turtola and Paajanen, 1995; Aijo and Tattari, 2000).

Regulatory measures addressing phosphorus loss inNordic countries

The regulatory measures implemented through national ‘Ac-tion Plans’ to reduce nutrient losses to the aquatic environ-ment in Denmark (Poulsen and Rubæk, 2005) have mostlikely been the main factor driving the continuing decreaseof the P surplus in Danish agriculture. This regulation wasinitiated in 1986 focusing on nitrogen (N) use efficiencyand has been tightened several times since (Table 2). It in-cludes restrictions on livestock densities, mandatory N bud-gets for crops and manure application techniques, as well asa requirement for nine months’ manure storage capacity(shifting manure application from autumn towards spring).In recent years new feeding strategies based on reducedmineral P supplements and the use of phytase for increasingP digestibility in feeds have become common in pig andpoultry production (Poulsen and Rubæk, 2005). Trading slur-ry amongst farms is a standard procedure. Animal slurry sep-aration is becoming more widespread and has provided theopportunity for export of a P-rich solid fraction from thefarm. Altogether this has considerably improved the utilisa-tion of P in Danish agriculture. However, the phosphorussurplus problem and P losses were addressed by legislationonly recently, in 2004 and 2007, and then to a limited de-gree. It is expected that measures already introduced bythe national Action Plans and related legislation will eventu-ally stabilize net P accumulation in agricultural soils at avery low level.

Since the early 1980s there has been focus on P transferfrom agricultural areas in Norway. Legislation relates tomanure management, storage capacity, time of applicationand livestock density, however, P losses are primarily beingaddressed through voluntary conservation schemes in closecollaboration with the advisory service (Table 2). This issupported by a financial incentives programme focusing onerosion and offering farmers graded support for installinga variety of conservation measures. Thus, higher conserva-tion subsidies are awarded for reduced tillage in areas ofgreater erosion risk (e.g. Lundekvam et al., 2002). Subsidiesare also given to establish vegetated buffer zones alongopen water and constructed wetlands in small streams. Asa new strategy, the agri-environmental programme has be-come regionalised since 2004, with locally adapted subsidiesand incentives (Table 2).

Sweden pursues an ambitious national strategy to eradi-cate eutrophication (Swedish Environmental ProtectionAgency, 2004). By 2010 anthropogenic emissions of P tolakes and streams in Sweden will have to be reduced by20% compared to the 1995 level. There is a two-tier ap-proach of regulation and voluntary programmes to achievethis aim. Legislation has restricted livestock density, de-fined a minimum storage capacity for animal manures andrestricted the application of manures and fertilizers on fro-zen, snow-covered or waterlogged soils (Table 2). Regula-tion does not differentiate between actual soilcharacteristics. Livestock density limits are based on P con-centrations in manures and more severe than in other Nor-dic countries. Generally, regulation is stricter in southerncoastal areas and thus takes account of the vulnerabilityof this part of the Baltic Sea. A voluntary farm advisory pro-gramme and information campaign ‘‘Focus on Nutrients’’since 2001 has strongly recommended balanced P fertiliza-tion, i.e. P applications no larger than P offtake in crops(Focus on Nutrients, 2006). Additionally, it was campaignedfor a reduction of P contents in pig and poultry feeds. Somevoluntary measures are also backed by financial incentivesschemes. Since 1995 EU subsidies have been used to furtherthe establishment of grassed buffer strips along water-courses and wetlands on former agricultural land as partof the effort to reduce P losses.

The first comprehensive efforts to reduce nutrient lossesfrom agriculture were started in Finland in 1995, when thecountry joined the European Union and implemented the Ni-trates Directive (EEC, 1991) and the Agri-environmental Pro-gramme (EEC, 1992). Prior to 1995, environmental schemesinvolving agriculture focused on voluntary actions by farm-ers backed by minor financial incentives (Valpasvuo-Jaati-nen et al., 1997). To be eligible for agri-environmentalsupport, farmers today have to implement all the basicmeasures plus one additional measure. The basic measuresinclude farm-scale environmental planning and monitoring,balanced fertilisation, good agricultural practices on live-stock farms and buffer strips along waterways (Table 2).The additional measures on arable farms comprise account-ing for soil test P in fertilization, plant cover in winter or re-duced tillage, or biodiversity options. Reducing ammoniaand gaseous emissions, treating dairy wastewater or improv-ing animal welfare are the additional options for livestockfarms. Voluntary actions are encouraged, such as imple-menting riparian zones, constructing wetlands or practicingorganic farming (Ministry of Agriculture and Forestry, 2004).The implementation of the Nitrates Directive contains pro-visions on good agricultural practices, storage of manureand methods and rates of fertilizer application (Mitikkaet al., 2005). Recently, the Finnish Government has issueda decision-in-principle ‘National Water Protection PolicyOutlines to 2015’, which aims at decreasing the eutrophyingloading to surface waters by 30% by 2015. A longer-termgoal is to cut loading by 50% from the level of 2001–2005.

Practical indexing tools for assessing phosphorusloss

The planning and implementation of efficient mitigationmeasures requires end-users (e.g. farmers, advisory services



Table 2 Current regulatory measures affecting P applications to agricultural land and/or the P loss potential in the Nordic countries

Denmark Norway Sweden Finland

General targets By 1993, 80% reduction in P dischargefrom point sources (achieved).Halving of P surplus in agriculture by2015 by mandatory reduction of Psupplements in feeds; establishmentof 50,000 ha new 10 m wide bufferzones along water courses

By 2015, 50% reduction inanthropogenic P losses for the areafrom the Swedish boarder toLindesnes

By 2010, 20% reduction ofanthropogenic P losses towaters compared to 1995level

National Water ProtectionPolicy Outlines to 2015: todecrease eutrophying loadingto surface waters.Agricultural nutrient loadingto be reduced by 30% until2015. Longer-term goal is 50%reduction from the level of2001–2005. Decision-in-principle by the FinnishGovernment

Mineral P fertilizerrestrictions

None None Balanced fertilisation ofarable crops, when signed upto incentives scheme

Animal manurerestrictions

Indirect effect of regulation directedat limiting N applications; Livestockdensity limits based on manure N (1LU equals 100 kg manure N abstorage): 1.4 LU pigs, 1.7 LU cattle,1.4 LU others corresponding to ca.19, 19 and 25 kg P ha�1 forrespectively finishing pigs, dairy cowsand broilers

Livestock density limits based onmanure P contents corresponding to35 kg P ha�1 yr�1; manure storagefacilities for 8 months required

Livestock density limits basedon manure P contents;22 kg P ha�1 permittedmaximum annual applicationrate in manures (Ulen et al.,2004)

Storage and application ofmanure is mainly regulated bythe Nitrates directive. Itcontains provisions on goodagricultural practices, storageof manure, spreading andallowable quantities offertilizers and silage liquor

Manure application,timing and method

No liquid manure applicationbetween harvest, or 1 November onestablished grassland or oil seedrape, and 1 February. No restrictionon timing of farmyard manureapplication

No manure application between 1November and 15 February.Incorporation of manure from 1September. No manure on frozen soilpermitted. Manure incorporation onbare soil within 18 h

No manure applicationbetween 1 August and 30November, except ongrassland and before sowingwinter crops, and 1 Januaryand 15 February in designatedsensitivea areas. Manureinjection into frozen soilpermitted. Manureincorporation within 4 h ofapplication on bare soil inparts of the country

Storage and application ofmanure is mainly regulated bythe Nitrates Directive

Ban on broadcast liquid manurespreading. Manure incorporationwithin 6 h of spreading on bare soil

(continued on next page)

Revie

wofindexin

gtools

foridentifyin

ghigh

riskare

asofphosphorusloss

inNordic

catchments

73


Table 2 (continued)


Good agriculturalpractice

No manure spreading underconditions (weather, topography,soil) that present an obvious risk fornutrient transport to water courses

Nutrient application according toplant needs and soil P status

No fertilizer applications onwater-saturated, snow-covered or deeply frozen soil

Farm-scale environmental planningand monitoring. Treatment of dairywastewater. Storage and applicationof manure according to the NitratesDirective

Erosion control 2 m wide buffers along all naturalstreams and streams with a highobjective. Additional establishmentof 50,000 ha new 10 m wide bufferzones along water courses

Voluntary options, reduced tillagedepending on erosion risk class;grassed water ways

Grassed buffers zones alongwater courses

Plant cover in winter or reducedtillage

Riparian buffer zones 2 m wide buffers along all naturalstreams and streams with a highobjective. Additional establishmentof 50,000 ha new 10 m wide bufferzones along water courses

Voluntary options, riparian buffers,constructed wetlands

5 m wide along blue-markedwatercourses

1–3 m along waterways, wider zonesvoluntary

Incentive schemes EU-backed subsidies under an agri-environmental programme

National Agri-environmentalprogramme, subsidies for reducedtillage, riparian buffer zones andsedimentation ponds

EU-backed subsidies forriparian buffer zones andsedimentation ponds

EU-backed Agri-environmentalProgramme; combination ofcompulsory, optional, and voluntarymeasures

Competent authority Regional and municipalenvironmental authorities

Regional authorities Swedish EnvironmentalProtection Agency

Regional Environment Centres;Municipal environmental authorities

Policy Actions,Ministerial Orders

Action Plan for the AquaticEnvironment III 2005–2015 (Kronvanget al., in press)

The National Environmentalprogramme 2004 (NorwegianAgricultural Authority, 2004)

Swedens EnvironmentalObjectives (SwedishEnvironmental ProtectionAgency, 2004)

Nitrates Directiveb; Agri-environmental Programme (Ministryof Agriculture and Forestry, 2004)

a Along the coast and Lake Malaren coinciding mainly with land affected by the EU Nitrate Directive.b In Finland, the Nitrates Directive is transposed into national legislation through the Environmental Protection Act and Government Decree on reducing the release of nitrates from

agricultural sources into water bodies. The provisions of the Government Decree apply to the whole national territory of Finland without regional or local differentiation.

74

G.Heckrath

etal.


and government agencies) to be equipped with simple, user-friendly management tools that estimate the risk of P lossbased on limited data. For detailed recent reviews of suchrisk assessment tools the reader is referred to Heathwaiteet al. (2005a) or Buczko and Kuchenbuch (2007). One ofthe first examples of such a tool was the original P Indexdeveloped in the United States by Lemunyon and Gilbert(1993) (Table 3). The authors suggested a simple tool toidentify high-risk areas and recommended managementalternatives to reduce the risk of P loss. The original P Indexadded up the scores of all risk factors, including soil erosion,irrigation erosion, runoff class, soil P test, P fertilizer andmanure application rate and method to yield the final indexscore (Lemunyon and Gilbert, 1993). A variety of moresophisticated P Indices has evolved from the original ap-proach (Sharpley et al., 2003). Selected P Indices are pre-sented in Table 3. One of the most significant changesfrom the original approach was to assess source and trans-port factors separately before combining the integratedsource and transport factors multiplicatively to better rep-resent landscape processes (Gburek et al., 2000). Therebythe P transport factor provides a scaling of the P source fac-tor and hence the amount of potentially mobile P in fields.Thus P indices take account of critical source areas, i.e.that most P is lost to water during a few large storm eventsfrom only a small area within the catchment. This concept isalso applicable to artificially drained land when it is consid-ered as contributing area (e.g. Heathwaite et al., 2003).

Most of the indices comprehensively account for bothsource (soil test P, fertilizer and manure management)and transport factors (erosion, runoff, leaching, connectiv-ity to receiving waters) for ranking fields according to theirvulnerability to potential P loss (Gburek et al., 2000). Phos-phorus indices use readily available information on sourceand transport factors and are specifically aimed at land-users and advisors. The Pennsylvania P Index (Table 3) isone of the most advanced approaches and much researchhas gone into refining its representation of both the sourceand the transport factors (e.g. Sharpley et al., 2001; McDo-well et al., 2001; Weld et al., 2001). Some indices, e.g. theIowa P Index (Table 3), estimate P losses instead of rankingfields after their relative vulnerability (e.g. Mallarinoet al., 2001). Others, e.g. the New York P Index (Table3), have separate indices for particulate and dissolved Pfor dealing with effects of land management on the trans-port of different P forms (Czymmek et al., 2001). Gener-ally, P Indices provide site-specific management guidanceto minimize P loss based on identified critical sources ortransport factors, or both. In response to the USDA-EPArequiring animal feeding operations to set up comprehen-sive P management plans most states of the US have nowadopted P Indices for their local conditions (Sharpleyet al., 2003). This makes it the only officially recom-mended empirical risk assessment tool for P loss. Phospho-rus Indices usually identify only relatively few fields withinthe catchment that require improved P management,hence limiting the economic impact on producers whilecapping the largest P losses. Combined with the flexibilityto choose site-specific measures this makes P index-basedmitigation cost-effective and acceptable to land users(Johansson and Randall, 2003).

Experiences with phosphorus index tools in theNordic countries

In Denmark, Andersen and Kronvang (2006) based their workof developing a P loss assessment tool suitable for Danishconditions on the Pennsylvania P Index (Sharpley et al.,2003). However, that P Index was originally developed forareas where soil erosion and surface runoff were major Ploss pathways. Due to the Danish topography, low intensityrainfall regime and soils generally only moderately erodible,soil erosion and surface runoff are not considered major Ploss pathways in Denmark (Kronvang and Rubæk, 2005).Andersen et al. (2005), on the other hand, demonstratedthat tile drains are an important pathway for P losses fromfields to surface waters in Denmark. A Danish P Index shouldalso consider the risk of P leaching from sandy soils with rel-atively low P-binding capacity, since decades of large net Pinputs to soils have increased the risk of P leaching (Kronv-ang et al., 2002). Consequently, a new soil type dependentfactor, the leaching potential, has been introduced in theDanish P Index (Table 4), as also suggested by Coale et al.(2002). This factor expresses the risk of P being transportedfrom the root zone to tile drains or to a shallow groundwatertable. Loamy soils have been attributed a higher weight, i.e.a higher risk of losing P, in the Danish P Index than sandysoils due to the risk of macropore flow and rapid transportof both particulate and dissolved P to drains. The need toinclude preferential flow losses of P in a P index was alsosuggested by Djodjic et al. (2002). The P-binding capacityof soils is almost exclusively associated with the mineralfraction. Some organic soils in Denmark have a low P-bind-ing capacity, others are characterized by a large pool of re-dox-sensitive P. Therefore, typically tile-drained, organicsoils tend to be particularly vulnerable to P losses (e.g.Poulsen and Rubæk, 2005) and have been assigned the high-est risk of P leaching in the Danish P Index (Table 4). Theimportance of tile drains for the overall risk of P beingtransported from soils to surface waters has been strength-ened in the Danish P index compared to the Pennsylvania Pindex. In the Danish version all artificially drained fieldshave been included in the P index calculation regardlessof their distance to receiving waters. The weightings ofthe factors in the Danish P Index were generally inheritedfrom the Pennsylvania P Index. The weighting of the leach-ing potential has been kept in the same order of magnitudeas for the transport factors erosion and surface runoff.Based on Danish field research, modified connectivity hasbeen modified to account for both the buffer strip widthand the soil erosion risk of the evaluated field.

To test the performance of the Danish P index Andersenand Kronvang (2006) examined the ability of the index tocorrectly rank measured annual diffuse P losses from 12 sub-catchments within the 1000-km2 Odense Fjord catchment.Vulnerability to P loss in runoff as calculated by the DanishP index on a subcatchment basis was closely related to ac-tual measured losses (Fig. 2a). This suggested that the Dan-ish P index was able to correctly describe the mostimportant factors determining P losses to surface watersin Danish catchments. It is, however, important to stressthat even though the index worked well at the catchmentscale, it is not yet clear whether the index can correctly



Table 3 Different P indexing approaches in the USA

Version Originala Pennsylvaniab New Yorkc Iowad

Multiplicative MultiplicativeSpecial remarks Additive Multiplicative Considers both particulate and

dissolved PEstimates loads

Considers enrichment ratio

Source factors

Soil P test Method not specified Mehlich 3 Morgan Bray, Mehlich 3, Olsen

P application rate Annual rate of added P Annual rate of added P,availability of manure P

Annual rate of added P Annual rate of added P

P appl. time and method Incorporation andtiming of added P

Incorporation andtiming of added P

Incorporation and timing of added P Incorporation and timing of added P

Transport factors

Soil erosion USLEe + irrigation erosion RUSLEe RUSLE RUSLE + gully

Surface runoff/flooding Slope and hydraulicconductivity

Soil permeabilityclass and slope

Flooding frequency Curve number

Concentrated flow Precipitation

Subsurface drainage/flooding – None/random/patternedartificial drainage or rapidpermeability soil near a stream

Poorly to welldrained

Presence of tiles

Flooding frequency Water flow to tile drains

Watershed

Contributing distance/Modified connectivity

Distance to stream

Riparian buffer/grassedwaterway/direct connection

Distance to stream Distance to stream

Riparian buffera Original P index: Lemunyon and Gilbert (1993).b Pennsylvania P index: http://nmsp.css.cornell.edu/publications/pindex.asp.c New York P index: http://nmsp.css.cornell.edu/publications/pindex.asp.d Iowa P index: http://www.ia.nrcs.usda.gov/Technical/Phosphorus/phosphorusstandard.html.e USLE Universal Soil Loss Equation (Wischmeier and Smith, 1978); RUSLE Revised USLE (Renard et al., 1997).

76

G.Heckrath

etal.


rank the vulnerability of individual fields to P losses. This isimportant because the chief purpose of applying a P Index isto identify those areas (fields) that contribute the most tototal P losses and thus target cost-effective mitigationoptions.

In Norway, Bechmann et al. (2007) also based their P in-dex on the Pennsylvania P Index (Sharpley et al., 2003).

Modifications were made to reflect the conditions importantfor the risk of P losses in Norway mainly related to the Nor-wegian climate. During winter P may be released from fro-zen plant material from grass subjected to consecutivefreeze–thaw events under laboratory conditions (Bechmannet al., 2005; Bechmann et al., 2007). These studies docu-mented the potential release of all the P contained in

Table 4 The Danish P Index (after Andersen and Kronvang, 2006)

Evaluation category

Part A – Screening toolSoil test P >200 mg P kg�1 If yes to either factor then proceed to Part BContributing distance <45 mContributing distance >45 m and field artificially drained

Part B – Source factorsSoil test P Soil test P (mg P kg�1) (Olsen-P translated to Mehlich-III-P)

Soil test P rating = 0.20 · soil test P (mg P kg�1)

Fertilizer P rate Fertilizer P (kg ha�1)

Manure P rate Manure P (kg ha�1)

P source application method 0.2 0.4 0.6 0.8 1.0Placed or injected5 cm or more deep

Incorporated<1 week

Incorporated >1week or notincorporatedApril–October

Incorporated >1 week ornot incorporatedNovember–March

Surface appliedto frozen orsnow coveredsoil

Fertilizer rating =rate · method

Manure P availability 0.5 0.8 1.0Treated manure/biosolids Dairy Poultry/Pigs

Manure rating = rate · method · availabilitySource factor = Soil test P rating + fertilizer rating + manure rating

Part C – Transport factorErosion Soil loss (tonnes ha�1)

Runoff potential 0 2 4 6 8Very low Low Medium High Very high

Leaching potential 2 4 6Sandy soil Loamy soil Organic soil

Subsurface drainage 0 1 2No artificial drains Few ditches or tile drains with wide spacing Field is on a tile

drainage system

Contributing distance 8 0<45 m >45 m

Modified connectivity 0.03 0.24 0.65Riparianbuffer = 2 m

Riparian buffer = 2 m Riparian buffer = 2 m

Erosion negligible Erosion medium Erosion high0.02 0.20 0.59Riparian buffer>2 m

Riparian buffer >2 m Riparian buffer >2 m

Erosion negligible Erosion medium Erosion high

Transport factor = [(erosion + runoff potential + contributing distance) · modified connectivity + (sub-surface drainage + leachingpotential)]/22Phosphorus index value = 2 · source factor · transport factor



aboveground biomass of annual grasses (Fig. 3), and about50% of the P in perennial grasses was released during ex-treme freeze–thaw events. This effect has also been con-firmed in field studies with pasture (Eltun et al., 2002),and a factor was included in the Norwegian P Index. Com-pared to the Pennsylvania P Index the representation of Ptransport from the soil surface to tile drains was improved.These changes were similar to the changes made in the Dan-ish P index (Andersen and Kronvang, 2006), focusing on theincreased risk of macropore flow in loamy, structured soilscompared to sandy soils. For Norwegian soils, Sveistrupet al. (2005) documented a significant translocation of claymineral particles through soil macropores on clay and siltyclay soils lending the clay soils a higher risk of P loss thanthe sandy soils. Organic soils often have an even higher riskof P loss due to several factors such as low contents of P-binding constituents, reductive release of iron-bound P ormineralization of organic P (e.g. Rupp et al., 2004). This ef-fect has also been shown in the Norwegian monitoring pro-gramme (Bechmann et al., 2005), where P losses werehighest from organic soils compared to other soil types.

Snow and snowmelt also are very important for theamount of P transferred in Norway. Phosphorus losses duringsnowmelt contribute on average 30% of the total annual Pload in small agricultural catchments in south-eastern Nor-

way (Bechmann et al., 2005). Based on maps of slope andsoil type for agricultural land in Norway the risk of erosionhas been calculated by the USLE (Universal Soil Loss Equa-tion; Wischmeier and Smith, 1978) for a standard slopelength of 100 meter (Norwegian Soil Maping, 2007). Theseerosion estimates were calibrated for south-eastern

Sub-catchment P Index score

0 5 10 15 20

TP

load

(kg

P h

a-1 y

r-1)

0.00

0.25

0.50

0.75

1.00

P Index score

0 20 40 60 80 100

Rel

ativ

e an

nual

P c

once

ntra

tion

(%)

0

100

200

300

Sub-catchment Field

P Index score0 25 50 75 100 125 150

TP

load

(kg

P h

a-1 y

r-1)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

F1F2F3F4F7F11F12

a b

c

Figure 2 Relationships between P index scores and P losses from trials in Nordic countries. (a) the Danish P Index at the catchmentscale. (b) the Norwegian P index at both the field and subcatchment scale. Relative P concentrations are mean TP concentrationsrelative to catchment outlet concentrations for 15–20 grab samples per field/subcatchment. (c) the Swedish PI set up for sevenfields.

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8No of freese-thaw cycles

% T

P re

leas

e

Annual Ryegrass

Perennial Ryegrass

Festuca

Figure 3 Water extractable P in different grass species afterexposure to a number of freeze–thaw cycles in the laboratory(adapted from Bechmann et al., 2005).



Norway. Regionally adapted soil management (C) factors(Lundekvam, 2002) were included in calculating the erosionrisk. Other changes in the P index were made to reflect Nor-wegian legislation regarding manure incorporation (<18 hafter application). In Norway, flooding has been includedas a factor that increases the P index to account for the in-creased risk of erosion and P release from soils during flood-ing. A net retention may be obtained when areas are left instubble during winter, but on autumn-ploughed areas, therisk of P losses increases with flooding (Eggestad et al.,2007). The factor has been included to promote no-till onareas at high risk of flooding. The P removal factor accountsfor the capacity of different plants to remove P from thesoil. The weighting of factors in the Norwegian P indexwas largely based on the Pennsylvania P Index. However,in Norway soil erosion is considered comparatively moreimportant for P transport to waters. The definition of thethreshold between medium and high erosion risk in Norway(2 tons ha�1 yr�1; Norwegian Soil Mapping, 2007) is onetenth of the corresponding threshold in Pennsylvania (Gbu-rek et al., 2000). Hence, the weighting of the erosion riskin the Norwegian P Index was increased accordingly.

Calibration of the first version of the Norwegian P indexwas carried out at both the catchment and the field scale.Results indicated that the index ranked fields according tothe actual P losses (Fig. 2b), suggesting that the P indexcould be a valuable field-scale decision support tool fornutrient management planning in Norway. The testing, how-ever, was carried out on areas with limited variability in

soils, climate and production system. For example, the P re-moval was lower than applied P for nearly all fields or catch-ments. Additionally, the testing did not include areaswithout tile drainage and only a few grasslands. Accord-ingly, until further testing has been carried out, the use ofthe Norwegian P index should be limited to tile-drainedareas in arable production with a positive P balance.

The Swedish P index resembles the north American Pindices to a lesser degree than the one adopted in Norwayand Denmark, though it is still an empirical, risk-basedassessment tool essentially structured into source and trans-port factors and aimed at farmers and agricultural advisorsas an educational and P management tool. The developmentof a Swedish index was governed by several conditions spe-cific to Swedish agriculture. Firstly, the P surplus in Swedenis rather low, i.e. fertilizer and manure P inputs are roughlyin balance with crop P offtake. Secondly, most of the arableland is located on level areas where the permeability of thesoils is generally higher than rainfall intensity and surfacerunoff rarely occurs. Erosion intensity, therefore, is usuallylow, but still sufficient to cause relevant P losses. Watererosion generally occurs during snowmelt (Alstrom andBergman, 1990). Freezing may also cause P loss from cropresidues. Thirdly, a large part of the agricultural land istile-drained, including the majority of cultivated clay soilsincreasing the risk of P losses through preferential flow (Djo-djic, 2001). Finally, substantial P surplus additions duringthe 1960s and 1970s have led to an increase in soil P testin the topsoil. However, the degree of P saturation (DPS)

Table 5 Factors included in the Swedish conditional P index

Factor Descriptiona

Transport factorsSurface runoff (SR) Function of soil permeability class and field slope, transports RP and UPPreferential flow (PF) Function of saturated hydraulic conductivity and soil structure, transports RP and UPMatric flow (MF) Function of saturated hydraulic conductivity and soil structure, transports only RP

Source factorsErosion RUSLEEnrichment ratio (ER) Determines UP content on the eroded soil, function of soil texture and P-ALb

Application rate Based on crop yield and soil P content (P-AL) in top- and subsoil, kg/haApplication timing Month of applicationApplication method Time to incorporationSoil P content P-AL, in topsoil (affects SR and PF) and subsoil (affects MF)Soil P sorption PSIc, governs distribution of added P into RP and UPDegree of P saturation DPS = (P-AL)/PSI, calculated for topsoil (affects SR and PF) and subsoil (affects MF)

P formsReactive phosphorus (RP) Soil P as a function of DPS + added P as a function of PSIUnreactive phosphorus (UP) Soil P as a function of RUSLE and ER + added P as a function of PSI

ConnectivityContributing distance Edge of field distanceExistence of tile drains Influences PF losses and field connectivityBuffer strips Influences UP losses in surface runoffInlets for surface runoff Influences connectivity of SRPonding conditions Enhances PF and losses of RP

a See Djodjic and Bergstrom (2005) for explanation.b Ammonium lactate/acetic acid.c Single point sorption index.



is usually low due to the high sorption capacity of Swedishsoils. For instance, a survey of 230 non-calcareous topsoilsfrom southern Sweden showed that typical DPS values basedon the commonly used ammonium lactate method (Egneret al., 1960) ranged from 10% to 20% (Ulen, 2006).

The political decision to reduce P losses from agriculturalland furthered the demand for a simple, user-friendly toolto assess the risk of P loss. Inspired by Swedish experienceswith reducing N losses, where collaboration between farm-ers, extension services and policy makers and knowledgedissemination were key issues for improvements, the educa-tional value of such a tool was also given attention (Focus onNutrients, 2006).

The main factors included in the Swedish PI are listed inTable 5. Table 6 shows the influence of selected factors on Pforms in different runoff types. Precipitation is divided intosurface runoff and infiltration based on field slope and soilpermeability. A further division of infiltrating water intomatric and preferential flow is based on soil structure andsaturated hydraulic conductivity. Matric flow dominates insandy soils, while preferential flow is important in struc-tured heavy soils with high hydraulic conductivity. Essen-tially dissolved reactive P (RP) and sediment-boundunreactive P (UP) are considered separately in differentrunoff types. Erosion is predicted with the RUSLE (Renardet al., 1997) but is treated as a source of UP rather thanas a transport process in the index. The P enrichment ratioin sediment was used to calculate the amount of UP oneroded soil material. Repeated freeze/thaw cycles influ-ence both erosion by reducing soil permeability and P re-lease from crop residues (Zuzel and Pikul, 1987;Bechmann et al., 2005). Based on the P content in crop res-idues and the number of freeze/thaw cycles, a fraction of Pin crop residues is transferred to the soil. The potential forlosses of RP is calculated based on the DPS concept, whereDPS is calculated as the ratio between P-AL soil test and a

single-point P sorption index (PSI, Borling et al., 2004).The DPS in topsoils is used to estimate the potential RPlosses with surface runoff and preferential flow, whereasthe DPS in subsoils is used to derive potential P losses withmatric flow. The apportionment of P additions into RP andUP is based on the relationship between PSI and CaCl2-extractable RP (Djodjic and Bergstrom, 2005). There areno explicit weighting factors in the Swedish PI. By calculat-ing P loads the apportionments of both transport mecha-nisms (surface runoff, matric and preferential flow) and Pforms (UP and RP, in soil and in P additions) determinethe relative importance of different transport processes.The presence of tile drains and inlets for surface runoff in-crease field connectivity and thereby P loss potential. Theintroduction of several new calculation steps, conditionalrules (for instance by-passing the subsoil sorption capacitywith preferential flow), and distinction between P forms(reactive and unreactive P) have made computations morecomplex. Therefore, a Visual Basic application was built tofacilitate calculations and offer users field-specific educa-tional information and management recommendations. Amore detailed description of Swedish PI is given in Djodjicand Bergstrom (2005).

The performance of the Swedish P index was initiallytested by comparing the calculated index values with mea-sured P transport from seven field sites included in thewater quality monitoring programme (Fig. 2c). The SwedishP index has also been used to identify fields with a high po-tential for P losses in a small agricultural catchment (Nat-terlund, 2004). Due to the uniformity of the tile-drained,heavy clay soils within the catchment there was little vari-ation in the P Index values regarding transport factors.Therefore, P source factors had a decisive role for field clas-sification. Fields where Salix spp. was grown had the highestrisk of P loss due to large inputs of P in sewage sludge coin-ciding with preferential flow on the clay soils. According to

Table 6 Factors in the Swedish conditional P Index affecting reactive (RP) and unreactive P (UP) forms in surface and subsurfacerunoff as functions of either field slope and soil permeability or saturated hydraulic conductivity and soil structure

Factor Representation in PI Field slope, soil permeability

Surface runoff Subsurface runoff

Saturated hydraulic conductivity, soilstructure

Preferential flow Matrix flow

P fertilization Rate, method, timing RPa, UP RP, UP RPDPSb P-AL/PSIc RP RP RPErosionrate RUSLEd UPPEe Soil texture, P-AL UP UP

Connectivity Drains, ponding conditions RP, UPBuffer strips UPContributing distance UP RP

a <0.45 lm molybdate reactive P; unreactive, sediment-bound P.b Degree of phosphorus saturation.c Soil P content measured with ammonium lactate/acetic acid; phosphorus sorption index.d Revised universal soil loss equation.e P enrichment in transported sediment.



current Swedish legislation it is allowed to apply up to245 kg P ha�1 in sewage sludge.

In Finland, the VIHMA (Viljelyalueiden valumavesienhallintamalli) assessment tool is being developed to esti-mate the P load from agricultural land and to target miti-gation measures according to cost effectiveness. Sincedifferent winter climate conditions have been found tohave a significant effect on P losses (Puustinen et al.,2005, 2006), the VIHMA assessment tool can also be usedto assess the effect of annual variations in climatic vari-ables and runoff on mitigation planning (Puustinen and

Tattari, 2005). The VIHMA tool (Fig. 4) is based on field-scale data, but can in future be scaled up to account forentire catchments. The risk of P loss is determined as afunction of field slope, soil type, cultivation practice andP status – all generally accessible data in Finland. Theinfluence of each factor on the estimated P load and theeffect of mitigation measures were derived from fieldexperiments. The relationship between P loss and abovevariables was separately examined for mild’winters’and’normal’ winters (differentiation based on climaticand hydrological variables). Runoff pathways are notexplicitly taken into account, but assumed to be specificfor each soil type group. Costs were determined for thedifferent mitigation measures for several crops and envi-ronmental practices (Table 7).

Using the VIHMA assessment tool it is possible to esti-mate the present load from an individual field and to com-pare the effectiveness of various P reduction measures indecreasing the absolute P loading (kg P ha�1 yr�1) and theircost efficiency (€ kg�1 P yr�1). In addition, correspondinginformation can be obtained for an entire catchment in or-der to find the most efficient set of measures to achieve thedesired load reduction. The mitigation measures include se-ven different cultivation methods (e.g. autumn ploughing,stubble tillage, direct drilling), riparian buffer zones andconstructed wetlands. The basic idea behind the VIHMA ap-proach is that the effect of individual measures crucially de-pends on local conditions, such as soil type and slope(Puustinen et al., 2005). For example, a riparian buffer zoneon a flat field may have a negligible effect on the P losses.On the other hand, some measures may be ineffective whenjudged by the percentage of P reduction, but may still re-move a large absolute amount of P in a catchment with ahigh risk for P losses, and thus may yield the desired effecton recipient loads. Therefore, the choice of mitigation mea-sures must be flexible and based on local conditions (Puus-tinen and Tattari, 2005). The variety of farming conditionsin combination with different mitigation options availableconfounds mitigation planning. The VIHMA tool makes themitigation process more transparent and at the same timeenables assessment of the allocation of environmental sub-sidies. It may help in better targeting of the agri-environ-ment measures. The intended users include particularlyenvironmental authorities implementing WFD, and exten-sion services.

Present load• Losses to recipient• P, (N), sediment

Target load

Field characteristics• Slope class• Soil type group• Soil P status• Surface/subsurface runoff risk• Cultivation practice

Mitigation options• Cultivation methods• Buffer strips• Constructed wetlands

Mitigation costs• Private• Public

Choice of mitigation• Effectiveness (kg ha–1 a–1)• Cost efficiency ( kg–1 a–1)• Number of measures• Allocation of measures

Required loadreduction

Achieved loadreduction

Figure 4 Concept and the components of the Finnish VIHMAassessment tool. Annual costs are presented in € kg�1 ha�1

reduction in total P load. Effectiveness stands for the annualreduction in P losses in kilograms per hectare, cost efficiencythe average cost in euros per one kilogram of reduced P loss perhectare and year.

Table 7 Examples of costs of different mitigation options for reducing P losses incurred by landowners or the public

Mitigation measure Landowner Publica

Wetland constructed by excavatingb 85–290 25–48Wetland constructed by dammingb 25–75 25–113Buffer stripsc 0–50 30–65From ploughing to stubbled 0–40 50–160From ploughing to direct sowingd 10–65 70–150From ploughing to grassd 0–100 65–135

Based on the Finnish VIHMA assessment tool. Annual costs are presented in € kg�1 ha�1 reduction in total P load.a According to the current Finnish Agri-environmental programme.b Assuming that the area of wetland is 1 ha, i.e. 1% of the catchment having 30–100% fieldland.c Assuming that the area of field is 2.2 ha, its mean slope is >3%, it is ploughed in autumn and is on cereals or sugar beet.d Assuming that the field has a mean slope of >3%.



Discussion

Nordic countries have different traditions in addressing themitigation of P losses to surface waters. Denmark, and to alesser extent Sweden, have mainly opted for general regula-

tory measures, while Finland and Norway have put greateremphasis on voluntary actions backed by financial incen-tives (Table 2). For all countries, regulatory measures direc-ted specifically at P losses have only recently beenintroduced (Table 2). Previously efforts to control N

Table 8 Phosphorus indexing approaches in the Nordic countries


US P index-type Yes Yes Partly No

Source factors Olsen P P-ALa P-AL, DPSb, PSIc P-AAd

P application rate P application rate P application ratetiming and method ofP application

Tillage

Timing andincorporation of Papplication

Timing andincorporation of Papplication

Plant cover

P release by freezingof crop residues

P release by freezingof crop residues

P offtake

Erosion risk USLE USLE-approach basedon national erosionrisk maps

RUSLE including Penrichment ratio insediment

C-factors in relationto soil management

Surface transport Soil permeability andslope

Soil hydraulicconductivity and slope

Soil permeability andslope

Soil type

Subsurface transport Artificial drainage Artificial drainage Artificial drainagePreferential flow Matrix & preferential

flow

Connectivity Buffer zone width Distance from field tostream

Existence of tiledrains

Buffer zone

Buffer strips Buffer strips WetlandsInlets for surfacerunoff

Inlets for surfacerunoff

P index score Product of source andtransport factor sums

Product of source andtransport factor sums

Product of source andtransport factor sums

Function of slopeclass, soil type group,soil P status, surface/sub-surface runoffrisk, cultivationpractice

Output variables Index score Index score Index score Loss of P, N andsedimentCosts

End-user Local/regionalenvironmentalauthority

Extension service Extension service Local environmentalauthority

Development stage Prototype forenvironmentalplanning

Experimental Educational Experimental

a Ammonium lactate/acetic acid.b Degree of P saturation.c P sorption index (see Djodjic and Bergstrom, 2005 for explanation).d Ammonium acetate.



emissions in compliance with the EU Nitrates Directive haveindirectly been a driver for reducing P surpluses in agricul-tures, though not P losses (Kronvang et al., 2005). In con-trast, erosion control programmes in Norway have also ledto somewhat lower P losses.

Since 1995 the Finnish Agri-Environmental Programme, afinancial incentives scheme, has been widely subscribed toand costly. As more than 90% of the agricultural land iscovered by the programme, it ought to have effectively re-duced P emissions. It was estimated that the Agri-environ-mental Programme would decrease both soil erosion and Plosses by about 20–40% (Valpasvuo-Jaatinen et al., 1997).However, the programme has not noticeably reduced Plosses from agricultural land, partly because it failed to tar-get the areas most vulnerable to P losses (Granlund et al.,2005). Further telling examples are the Baltic countrieswhere fertilizer and manure P inputs have decreased sub-stantially since the late 1980s without generally resultingin reduced P losses (Stalnacke et al., 2004). In other words,without addressing the role of CSAs, policy measures toabate diffuse P losses are unlikely to be effective. Also, akey to devising effective mitigation strategies for P lossesis to direct measures at transport pathways, and not solelyat soil P accumulation (Heathwaite et al., 2005b). Site vul-nerability, or P indexing, tools lately have been shown toserve this purpose well, since they make good use of expertknowledge and spatial data (e.g. Harmel et al., 2005;Schendel et al., 2004).

It is vital to recognize the limitations of these simpleempirical models. In general it remains a challenge to up-scale fine-scale P transport processes to field and catch-ment scale patterns of diffuse P loss. The importance of aspecific P index architecture and linkage of different factorsfor the predictive ability of P indices is not well researched(Heathwaite et al., 2005a). Therefore, a drawback of manyP indices is the lack of direct calibration and the uncertaintyrelated to delineation of risk classes. However, an extensiverunoff plot study (Sharpley et al., 2001) and a recent com-parison of measured and simulated P losses with indexedsite vulnerability (Veith et al., 2005) both indicated thatthe P index provides a reliable assessment of where P lossesoccur within a catchment. Testing different US P indicesHarmel et al. (2005) found satisfactory agreement betweenmeasured dissolved reactive P loads in surface runoff and Pindex scores in ten small Texan monitoring catchments. To alesser degree the indices captured variations in total P loadsdue to the inability to accurately predict erosion. In con-trast, the relationship between Norwegian P Index scoresand P losses from erosion-vulnerable sites was satisfactorilyclose (Fig. 2b).

Another practical limitation of the P index from a user’sperspective is the cost of obtaining and evaluating the geo-physical data used in estimating erosion, runoff and leach-ing potential. This problem may be alleviated by linkinggeographic databases directly to P index assessments (Bee-gle et al., 2000). Finally, and most importantly, P indicescannot quantify the ecological impact of P loss in waters be-cause they are not quantitative and dynamic. That is, theyonly crudely, if at all, distinguish loss of different P formsand, crucially, only predict potential P loss.

Only recently the usefulness of P indexing tools has beenappreciated in Nordic countries (Heckrath et al., 2005).

Therefore, the development of adapted tools for Nordiccountries still is at an early stage and has been much in-spired by the US P indices. Table 8 summarizes and com-pares the key features of the Nordic P indexingapproaches. While the Danish and the Norwegian P Indicesare modifications of the Pennsylvania P Index, the SwedishPI is more sophisticated in its representation of subsurfaceP transport and P retention in subsoils and potentially mo-bile P forms. The Finnish VIHMA tool is most remote fromthe P Index concept. Though it also resembles an empiricalpredictor of the risk of P loss and uses much the same inputfactors, it notably integrates an assessment of the cost ofcertain mitigation methods (Table 8, Fig. 4). The differentapproaches account for regional pedologic, topographic,climatic, hydrologic and management conditions. For exam-ple, freezing-out of P from crop residues is part of the Nor-wegian and the Swedish P Index. The lack of appropriateweighting of individual factors according to their relativecontribution to a combined source or transport factor re-mains a weak point of the simple Danish and Norwegian PIndices. First, the relative importance of different transportprocesses under field conditions and their representation inP indices are not sufficiently understood. Additionally, sen-sitivity analyses ought to be employed to elucidate the influ-ence of the chosen factor weightings on P index scores andto facilitate their adjustment (e.g. Buczko and Kuchenbuch,2007). Currently the purposes of different Nordic P indicesrange from experimental, scientific tools in Norway and Fin-land, educational tool in Sweden to environmental planningtool in Denmark. To date the field and catchment scale eval-uations of the Nordic P indexing approaches are still verylimited. To gain acceptance with potential stakeholders, abroader database against which the indices’ performancecan be tested is urgently required.

In Norway, policy instruments have since 1993 beenlinked to risk of erosion, giving higher subsidies for reducedtillage in areas with high erosion risk (Lundekvam et al.,2002). This targeted tillage payment has had a great influ-ence on farmers behaviour and reduced soil erosion in Nor-way. With P enrichment in soils still rising and in recognitionof subsurface transport of P, the adoption of P indexingbroadens the risk assessment approach for P loss. The VIH-MA model is still being adjusted and tested, and will poten-tially be a major tool in targeted mitigation planning andabatement of P losses in Finland. VIHMA describes wellthe effect of different cultivation practices on P lossesand approximates their costs. However, it is less suitablefor identifying vulnerable fields on livestock operationsdue to an insufficient representation of manure applica-tions. This is a clear drawback, since animal husbandry is ex-pected to have a major impact on P losses in the comingyears.

While both environmental managers and agriculturaladvisors in Nordic countries request simple and objectiverisk assessment tools ensuring equal standards within na-tional agricultures (e.g. Bidstrup, 2005), policy makers havebeen somewhat reluctant to adopt P indexing approaches asa regulatory option. Instead uniform reduction policies haveremained attractive, since they can be linked to measurablefluxes of feed and fertilizer P and, hence, are easily imple-mented, monitored and common sense. Denmark to ourknowledge is the only country in Europe, where the mapping



of high risk areas for P loss as integral part of environmentalplanning has been written into law (Annonymous, 2004).Therefore, the development of a P index better adaptedto conditions in Denmark is currently actively pursued andthe release of a prototype P index-based planning tool is ex-pected for 2008 (Andersen et al., 2007).

An important perspective for the further development ofindexing tools is their use as a pollution indicator to whichpolicy instruments and mitigation evaluations may belinked. Given the wide range of options available to policymakers, models linking mitigation of P losses to the costof achieving mitigation are essential for informed decisions.However, economic analyses of P mitigation are very lim-ited in Nordic countries and ought to be prioritized in thefuture.

Conclusions

Interdisciplinary approaches with strong end-user participa-tion are essential for cost-effective targeting of mitigationactions to reduce P losses in critical source areas in catch-ments. Farmers and environmental managers experiencingsevere cost constraints are best served by user-friendly,qualitative decision tools for hands-on mitigation planning,which must be compatible with market-based incentives.The P index is one such decision tool that has been widelydeveloped to help identify what actions need to be takenand where they should be targeted. In a Nordic context,important lessons can be learned from experiences with Pindexing in the US regarding simplification of process knowl-edge and meeting end-user needs. Hence the P index ap-proach has heavily influenced the risk assessment toolsrecently tested under Nordic conditions. Their performanceappears to be promising and it is expected that they willplay an important role in the WFD process in the Nordiccountries.

Acknowledgements

This paper is a product of the Nordic interdisciplinary work-shop ‘‘Tools for Assessing Phosphorus Loss from Nordic Agri-culture’’ in Foulum Denmark, February 2004. The authorswish to thank the Nordic Council of Ministers for providingfunding for the workshop.

References

Aijo, H., Tattari, S., 2000. Management system for runoff fromarable lands. The Finnish Environment 442, 68 pp. (in Finnishwith an English abstract).

Alstrom, K., Bergman, A., 1990. Water erosion on arable land insouthern Sweden. In: Boardman, J., Foster, I.D.L., Dearing, J.A.(Eds.), Soil Erosion on Agricultural Land. John Wiley & Sons Ltd.,Chichester, pp. 107–117.

Andersen, H.E., Kronvang, B., 2006. Modifying and evaluating a Pindex for Denmark. Water, Air and Soil Pollution 174, 341–353.

Andersen, H.E., Kronvang, B., Larsen, S.E., 2005. Development,validation and application of Danish empirical phosphorusmodels. Journal of Hydrology 304, 355–365.

Andersen, H.E., Heckrath, G., Hoffmann, C.C., Iversen, B.V.,Jacobsen, O.H., Kjaergaard, C., Kronvang, B., Lægdsmand, M.,

Rubæk,G.H., 2007.Aphosphorus indexapproach forDenmark. In:Heckrath, G., Rubæk, G.H., Kronvang, B. (Eds.), Diffuse Phos-phorus Loss. Risk Assessment, Mitigation Options, and EcologicalEffects in River Basins. DJF Plant Science No. 130, pp. 255–258.

Andersson, A., Andersson, R., Eriksson, J., Mattsson, L., 1998.Phosphorus accumulation in Swedish agricultural soils. SwedishEnvironmental Protection Agency, Stockholm, Report 4919,ISSN: 0282-7298, 33 pp.

Andersson, A., Eriksson, J., Mattsson, L., 2000. Phosphorus accu-mulation in Swedish agricultural soils. Swedish EnvironmentalProtection Agency (SEPA) Report 5110. Stockholm.

Annonymous, 2004. Agreement on the Action Plan for the AquaticEnvironment III 2005–2015 between the Danish Government, theDanish People’s Party and the Christian Demokrats. http://www.vmp3.dk/Files/Filer/English%20version/engelsk_oversa-ettelse.pdf (Verified 20.09.07).

Bechmann, M., Stalnacke, P., 2005. Effect of policy-inducedmeasures on suspended sediments and total phosphorus con-centrations from three Norwegian agricultural catchments.Science of the Total Environment 344, 129–142.

Bechmann, M., Krogstad, T., Sharpley, A.N., 2005. A phosphorusindex for Norway. Acta Agriculturae Scandinavica, Section Soiland Plant Science 55 (3), 205–213.

Bechmann, M.E., Stalnacke, P., Kværnø, S.K., 2007. Testing theNorwegian phosphorus index at the field scale. Agriculture,Ecosystems and Environment 120, 107–128.

Beegle, D.B., Sharpley, A.N., Gburek, W.J., Weld, J., 2000.Integrating the Phosphorus Index and Existing Data bases forUse in Nutrient Management planning. Department of Agronomy,Penn State University, PA.

Berge, D., Vandsemb, S.M., Bechmann, M. 2001. Monitoring ofagricultural dominated lakes. Measures, water quality andtrends. Norwegian Institute for Water Research, NIVA Report,ISBN: 82-577-3947-2, 90 p (in Norwegian).

Bergstrom, L., Djodjic, F., Kirchmann, H., Nilsson, I., Ulen, B.,2007. Phosphorus from farmland to water. Report Food 21, no.4/2007. Swedish University of Agricultural Sciences, Uppsala,Sweden.

Bidstrup, J., 2005. Using phosphorus loss assessment tools inpractise. The river basin manager’s perspective. In: Heckrath,G., Bechmann, M., Ekholm, P., Djodjic, F., Ulen, B., Andersen,H.E., Olsen, P., (Eds.), Tools for Assessing Phosphorus Loss fromNordic Agriculture. TemaNord 2005:583, pp. 111–114.

Boesch, D., Hecky, R., O’Melia, C., Schindler, D., Seitzinger, S.,2006. Eutrophication of the Swedish Seas. Swedish Environmen-tal Protection Agency Report 5509, 68 pp.

Bøgestrand, J. (Ed.), 2005. Vandløb 2004. NOVANA. DanmarksMiljøundersøgelser. Faglig rapport fra DMU nr. 554, 82 pp.

Borling, K., Otabbong, E., Barberis, E., 2004. Soil variables forpredicting potential phosphorus release in Swedish noncalcare-ous soils. Journal of Environmental Quality 33, 99–106.

Boutrup, S., Bijl, L. van der, Svendsen, L.M., Grant, R., Ellermann,T., Bruus, M., 2006. Vandmiljø og Natur 2005. Tilstand ogudvikling - faglig sammenfatning. Danmarks Miljøundersøgelser.Faglig rapport fra DMU nr. 597, 50 pp.

Brandt, M., Ejhed, H., 2003. Transport, retention och kallfordeling– Belastning pa haven. SNV Rapport 5247, Stockholm. ISSN 0282-7298.

Buczko, U., Kuchenbuch, R.O., 2007. Phosphorus indices as risk-assessment tools in the USA and Europe – a review. Journal ofPlant Nutrition and Soil Science 170, 445–460.

Coale, F.J., Sims, J.T., Leytem, A.B., 2002. Accelerated deploy-ment of an agricultural nutrient management tool: the MarylandPhosphorus Site Index. Journal of Environmental Quality 31,1471–1476.

Czymmek, K., Ketterings, Q., Geohring, L., 2001. Phosphorus andagriculture VII: the New Phosphorus Index for New York State.What’s Cropping Up? 11 (4), 1–3.



Dalgaard, T., Rubæk, G.H., 2005. Fosfortilførsel, ophobning og -mætningsgrad i danske landbrugsjorde. In: Poulsen, H.D.,Rubæk, G.H. (Eds.), Fosfor i dansk landbrug. Omsætning, tabog virkemidler (phosphorus in Danish agriculture. Turnover,losses and mitigation measures). DJF rapport Husdyrbrug no. 68,pp. 93–101.

Djodjic, F., 2001. Displacement of phosphorus in structured soils.PhD thesis, Acta Universitatis Agriculturae Sueciae, Agraria 283,Swedish University of Agricultural Sciences, Uppsala, Sweden.

Djodjic, F., Bergstrom, L., 2005. Conditional phosphorus index as aneducational tool for risk assessment and phosphorus manage-ment. Ambio 34 (3–4), 296–300.

Djodjic, F., Montas, H., Shirmohammadi, A., Bergstrom, L., Ulen,B., 2002. A decision support system for phosphorus managementat a watershed scale. Journal of Environmental Quality 31 (3),937–945.

Dørge, J., Windolf, J., 2003. Implementation of the Water Frame-work Directive – can we use models as a tool in integrated riverbasin management? International Journal of River Basin Man-agement 1, 165–171.

EEC, 1991. Council Directive 91/676/EEC of 12 December 1991concerning the protection of waters against pollution caused bynitrates from agricultural sources. Official Journal L 375, 1–8.

EEC, 1992. Council Regulation (EEC) No 2078/92 of 30 June onagricultural production methods compatible with the require-ment of the protection of the environment and the maintenanceof the countryside, Official Journal L 215, 85–90.

Eggestad H.O., Sveistrup, T., Øgaard, A.F., Kitterød, N.O., 2007:Effekter av flom/oversvømmelser pa vannkvalitet, med fokus pasediment og fosfor - en litteraturgjennomgang. Bioforsk report48/07. ISBN 978-82-17-00020-4, 18 pp. (in Norwegian).

Egner, H., Riehm, H., Domingo, W.R., 1960. Untersuchungen uberdie chemische Bodenanalyse als Grundlage fur die Beurteilungde Nahrstoffzustandes der Boden. Kungl LantbrukshogskolansAnnaler 26, 199–215.

Ejhed, H., Brandt, M., Djodjic, F., Olshammar, M., Ryegard, A.,Johnsson, H., Larsson, M., Nisell, J., Rapp, L., Branvall, G., inpress. Miljomalsuppfoljning. Ingen overgodning 1995 och 2005.Swedish Methodology for Environmental Data Report.

Ekholm, P., Mitikka, S., 2006. Agricultural lakes in Finland: currentwater quality and trends. Environmental Monitoring & Assess-ment 116, 111–135.

Ekholm, P., Turtola, E., Gronroos, J., Seuri, P., Ylivainio, K., 2005.Phosphorus loss from different farming systems estimated fromsoil surface phosphorus balance. Agriculture, Ecosystems andEnvironment 110, 266–278.

Eltun, R., Korsaeth, A., Nordheim, O., 2002. A comparison ofenvironmental, soil fertility, yield, and economical effects in sixcropping systems based on an 8-year experiment in Norway.Agriculture, Ecosystems and Environment 90, 155–168.

Focus on Nutrients, 2006. Available at: http://www.greppa.nu [22November 2006]. The Swedish Board of Agriculture, The CountyAdministration Boards, and the Federation of Swedish Farmers.

Gburek, W.J., Sharpley, A.N., Heathwaite, A.L., Folmar, G., 2000.Phosphorus management at the watershed scale: a modificationof the P Index. Journal of Environmental Quality 29, 130–144.

Granlund, K., Raike, A., Ekholm, P., Rankinen, K., Rekolainen, S.,2005. Assessment of water protection targets for agriculturalnutrient loading in Finland. Journal of Hydrology 304, 251–260.

Grayson, R.B., Moore, I.D., McMahon, T., 1992. Physically basedhydrologic modelling. 2. Is the concept realistic? WaterResources Research 26 (10), 2659–2666.

Haraldsen, T., Loigu, E., Iital, A., Jansons, V., Vagstad, N., 2001.Plant nutrients in soils and cereals in Norway and the Balticcountries. Jordforsk report no. 105/01, Norwegian Centre forSoil and Environmental Research, 22 pp.

Harmel, R.D., Torbert, H.A., DeLaune, P.B., Haggard, B.E., Haney,R.L., 2005. Field evaluation of three phosphorus indices on new

application sites in Texas. Journal of Soil and Water Conserva-tion 60 (1), 29–42.

Heathwaite, A.L., Sharpley, A.N., Bechmann, M., 2003. Theconceptual basis for a decision support framework to assessthe risk of phosphorus loss at the field scale across Europe.Journal of Plant Nutrition and Soil Science 166, 447–458.

Heathwaite, A.L., Sharpley, A.N., Bechmann, M., Rekolainen, S.,2005a. Assessing the risk and magnitude of agricultural nonpointsource phosphorus pollution. In: Sims, T., Sharpley, A.N. (Eds.),Phosphorus: Agriculture and the Environment. American Societyof Agronomy Madison, pp. 981–1021.

Heathwaite, A.L., Quinn, P.F., Hewett, C.J.M., 2005b. Modellingand managing critical source areas of diffuse pollution fromagricultural land using flow connectivity simulation. Journal ofHydrology 304, 446–461.

Heckrath, G., Bechmann, M., Ekholm, P., Djodjic, F., Ulen, B.,Andersen, H.E., Olsen, P., (Eds.), 2005. Tools for AssessingPhosphorus Loss from Nordic Agriculture. TemaNord 2005:583,139 pp.

Johansson, G., Gustafson, A., 2005. Observationsfalten pa aker-mark. Teknisk rapport 107, Avd. Vattenvardslara SLU, 38 pp.

Johansson, G., Gustafson, A., 2007. Observation fields on arableland. Discharge and nutrient losses for the agro-hydrologicalyear 2005/06 and a long-term review. Technical Report 115 (inSwedish), 38 pp.

Johansson, R.C., Randall, J.R., 2003. Incorporating economics intothe phosphorus index: an application to US watersheds. Journalof Soil and Water Conservation 58 (5), 224–231.

Kauppila, P., Meeuwig, J.J., Pitkanen, H., 2003. Predicting oxygenin small estuaries of the Baltic Sea: a comparative approach.Estuarine, Coastal and Shelf Science 57, 1115–1126.

Kronvang, B., Rubæk, G.H., 2005. Kvantificering af dyrkningsbi-draget af fosfor til vandløb og søer. In: Poulsen, H.D., Rubæk,G.H. (Eds.), Fosfor i dansk landbrug. Omsætning, tab ogvirkemidler (Phosphorus in Danish agriculture. Turnover, lossesand mitigation measures). DJF rapport Husdyrbrug no. 68, pp.132–146.

Kronvang, B., Svendsen, L.M., Friberg, N., Dørge, J., 1997. Streamrestoration in Denmark: assessment of the ecological benefits.Zentralblatt fur Geologie und Palaontologie Teil I 10, 969–981.

Kronvang, B., Iversen, H.L., Jørgensen, J.O., Paulsen, I., Jensen,J.P., Conley, D., Ellermann, T., Laursen, K.D., Wiggers, L.,Flindt Jørgensen, L., Stockmarr, J., 2001. Fosfor i jord og vand.Udvikling, status og perspektiver. National EnvironmentResearch Institute, NERI Report no. 380, 88 pp.

Kronvang, B., Grant, R., Laubel, A.R., Pedersen, M.L., 2002.Quantifying sediment and nutrient pathways within Danishagricultural catchments. In: Haygarth, P.M., Jarvis, S.C. (Eds.),Agriculture, Hydrology and Water Quality. CABI Publishing, pp.282–301.

Kronvang, B., Jeppesen, E., Conley, D.J., Søndergaard, M., Larsen,S.E., Ovesen, N.B., Carstensen, J., 2005. Nutrient pressures andecological responses to nutrient loading reductions in Danishstreams, lakes and coastal waters. Journal of Hydrology 304,274–288.

Kronvang, B., Vagstad, N., Behrendt, H., Bøgestrand, J., Larsen,S.E., 2007. Phosphorus losses at the catchment scale withinEurope: an overview. Soil Use and Management 23 (Suppl. 1),104–116.

Kronvang, B., Andersen, H.E., Børgesen, C.D., Dalgaard, T., Larsen,S.E., Bøgestrand, J., Blicher-Mathiasen, G., in press. Effects ofpolicy measures implemented in Denmark on nitrogen pollutionof the aquatic environment. Environmental Science and Policy.

Kyllingsbæk, A., Hansen, J.F., 2007. Development in nutrientbalances in Danish agriculture 1980–2004. Nutrient Cycling inAgriculture 79, 267–280.

Kyllmar, K., Grill, K., 2007. Nutrient losses in small agriculturalcatchments 2005/2006. Division of Water Quality Management,



Department of Soil Sciences, Swedish University of AgriculturalSciences. Ekohydrologi 98, 32 pp. (in Swedish).

Kyllmar, K., Carlsson, C., Gustafson, A., Ulen, B., Johnsson, H.,2006. Nutrient discharge from small agricultural catchments inSweden. Characterisation and trends. Agriculture, Ecosystemsand Environment 115, 15–26.

Lemunyon, J.L., Gilbert, R.G., 1993. The concept and need for aphosphorus assessment tool. Journal of Production Agriculture6, 483–496.

Lundekvam, H., 2002. ERONOR/USLENO – Empirical erosion modelsfor Norwegian conditions. Report no. 6/2002. ISBN: 82-483-0022-6, 40 pp.

Lundekvam, H., 2007. Plot studies and modelling of hydrology anderosion in southeast Norway. Catena 71 (2), 200–209.

Lundekvam, H., Romstad, E., Øygaarden, L., 2002. Agriculturalpolicies in Norway and effects on soil erosion. EnvironmentalScience Policy 6, 57–67.

Mallarino, A.P., Stewart, B.M., Baker, J.L., Downing, J.A., Sawyer,J., 2001. Background and basic concepts of the Iowa phosphorusindex. A support document to the NRCS field office technicalnote 25. Iowa State University.

McDowell, R.W., Sharpley, A.N., Beegle, D., Weld, J.L., 2001.Comparing phosphorus management strategies at a watershedscale. Journal of Soil and Water Conservation 56, 306–315.

Meeuwig, J.J., Kauppila, P., Pitkanen, H., 2000. Predicting coastaleutrophication in the Baltic: a limnological approach. CanadianJournal of Fisheries and Aquatic Sciences 57, 844–855.

Ministry of Agriculture and Forestry, 2004. Horisontaalisen maaseu-dun kehittamisohjelman valiarviointi. Mid-term evaluation ofthe Horizontal Rural Development Programme, Manner-Suomi.MMM:n julkaisuja 1/2004. 273 pp.

Mitikka, S., Britschgi, R., Granlund, J., Gronroos, J., Kauppila, P.,Makinen, R., Niemi, J., Pyykkonen, S., Raateland, A., Silvo, K.,2005. Report on the implementation of the Nitrates Directive inFinland 2004. The Finnish Environment 741, 92 pp.

Natterlund, H., 2004. Risk assessment of P losses – calculationswith Swedish Phosphorus Index. M.Sc. Thesis Swedish Universityof Agricultural Sciences (in Swedish).

Norwegian Agricultural Authority, 2004. The National Environmen-tal Programme, 2004. 26 pp.

Norwegian Soil Maping, 2007. Norwegian Forest and LandscapeInstitute. http://www.skogoglandskap.no/english/index.html(verified 24.09.07).

OECD, in press. Environmental Indicator for Agriculture, Volume 4.http://www.oecd.org/department/0,3355,en_2649_33795_1_1_1_1_1,00.html.

OSPAR, 2003. Nutrients in the Convention area. Implementation ofPARCOM Recommendation 88/2 and 89/4. Eutrophication andNutrient Series, OSPAR Commision 2003. ISBN 1-904426-16-6.

Øygarden, L., 2000. Soil erosion in small agricultural catchments,south-eastern Norway. PhD Thesis 2000:8, Department of Soiland Water Sciences, Agricultural University of Norway.

Poulsen, H.D., Rubæk, G.H. (Eds.), 2005. Phosphorus in DanishAgriculture – Cycling, Loss and Mitigation. DIAS report, AnimalProduction 68, 211 pp.

Puustinen, M., Tattari, S., 2005. The VIHMA model – assessingmeasures to reduce erosion and P loss from agricultural land inFinland. In: Heckrath, G., Bechmann, M., Ekholm, P., Djodjic,F., Ulen, B., Andersen, H.E., Olsen, P., (Eds.) Tools for AssessingPhosphorus Loss from Nordic Agriculture. TemaNord 2005:583,pp 75–81.

Puustinen, M., Koskiaho, J., Peltonen, K., 2005. Influence ofcultivation methods on suspended solids and phosphorusconcentrations in surface runoff on clayey sloped fields inboreal climate. Agriculture Ecosystems and Environment 105,565–579.

Puustinen, M., Tattari, S., Koskiaho, J., Linjama, J., 2006. Influenceof seasonal and annual hydrological variations on erosion and

phosphorus transport from arable areas in Finland. Soil andTillage Research 93, 44–55.

Raike, A., Pietilainen, O.-P., Rekolainen, S., Kauppila, P., Pitkanen,H., Niemi, J., Raateland, A., 2003. Trends of phosphorus,nitrogen and chlorophyll a concentrations in Finnish rivers andlakes in 1975–2000. Science of the Total Environment 310, 47–59.

Rekolainen, S., Kamari, J., Hiltunen, M., 2003. A conceptualframework for identifying the need and role of models in theimplementation of the Water Framework Directive. Interna-tional Journal of River Basin Management 1, 165–171.

Renard, K.G., Foster, R.G., Weesies, G.A., McCool, D.K., Yoder,D.C., 1997. Predicting soil erosion by water: a guide toconservation planning with the revised universal soil lossequation (RUSLE). USDA-ARS Agriculture Handbook 703. USGovernment Print Office, Washington, DC. 384 pp.

Rupp, H., Meissner, R., Leinweber, P., 2004. Effects of extensiveland use and re-wetting on diffuse phosphorus pollution in fenareas – results from a case study in the Dromling catchment,Germany. J. Plant Nutr. Soil Sci. 167, 408–416.

Saarela, I., 2002. Phosphorus in Finnish soils in the 1900s withparticular reference to the acid ammonium acetate soil test.Agricultural and Food Science in Finland 11, 257–271.

Schendel, E.K., Schreier, H., Lavkulich, L.M., 2004. Linkagesbetween phosphorus index estimates and environmental qualityindicators. Journal of Soil and Water Conservation 59 (6), 243–251.

Schoumans, O.F., Silgram, M. (Eds.), 2003. Review and LiteratureEvaluation of Nutrient Quantification Tools. Application andevaluation of phosphorus and nitrogen loss quantification tools.EuroHarp WP 3 and 4. European Commission FrameworkProgramme 5, 122 pp.

Selvik, J.R., Tjomsland, T., Borgvang, S.A., Eggestad, H.O., 2006.Tilførsler av næringsstoffer til Norges kystomrader i 2005,beregnet med tilførselsmodellen TEOTIL2. NIVA-report 5330-2007 (in Norwegian).

SEPA (Swedish Environmental Protection Agency), 2003. Miljokval-itetsnormer for fosfor i sjoar. Swedish Environmental ProtectionAgency Report 5288.

Sharpley, A.N., McDowell, R.W., Weld, J.L., Kleinman, P., 2001.Assessing site vulnerability to phosphorus loss in an agriculturalwatershed. Journal of Environmental Quality 30, 2026–2036.

Sharpley, A.N., Weld, J.L., Beegle, D.B., Kleinman, P.J.A., Gburek,W.J., Moore, P.A., Mullins, G., 2003. Development of phospho-rus indices for nutrient management planning strategies in theUnited States. Journal of Soil and Water Conservation 58 (3),137–152.

Søndergaard, M., Jeppesen, E., Jensen, J.P., 2005. Water Frame-work Directive: ecological classification of Danish lakes. Journalof Applied Ecology 42, 616–629.

Stalnacke, P., Vandsemb, S.M., Vassiljev, A., Grimvall, A., Jolankai,G., 2004. Changes in nutrient levels in some Eastern Europeanrivers in response to large-scale changes in agriculture. WaterScience and Technology 49 (3), 29–36.

Statistics Sweden, 2005. Nitrogen and phosphorus balances inagricultural land and in agricultural sector in 2003. MI 40SM0501.

Sveistrup, T.E., Haraldsen, T.K., Langohr, R., Marcelino, V.,Kværner, J., 2005. Impact of land use and seasonal freezingon morphological and physical properties of silty Norwegiansoils. Soil and Tillage Research 81, 39–56.

Swedish Environmental Protection Agency, 2004. Sweden’s envi-ronmental objectives, de facto 2004 (http://www.miljomal.nu).

Turtola, E., Paajanen, A., 1995. Influence of improved subsurfacedrainage on phosphorus losses and nitrogen leaching from aheavy clay soil. Agricultural Water Management 28, 295–310.

Ulen, B., 2006. A simplified risk assessment for losses of dissolvedreactive phosphorus through drainage pipes from agricultural



soils. Acta Agriculturae Scandinavica Section B – Soil and PlantScience 56, 307–314.

Ulen, B., Jakobsson, C., 2005. Critical evaluations of measures tomitigate phosphorus losses from agricultural land to surfacewaters in Sweden. Science of Total Environment 344, 37–50.

Ulen, B., Calsson, C., Lidberg, B., 2004. Recent trends and patternsof nutrient concentrations in small agricultural streams inSweden. Environmental Monitoring and Assessment 98, 307–322.

Uusitalo, R., 2004. Potential bioavailability of particulate phospho-rus in runoff from arable clayey soils. Agrifood Research Reports53, 99 pp.

Uusitalo, R., Jansson, H., 2002. Dissolved reactive phosphorus inrunoff assessed by soil extraction with an acetate buffer.Agricultural and Food Science in Finland 11, 343–354.

Vagstad, N., Stalnacke, P., Andersen, H.E., Deelstra, J., Gustav-son, A., Iital, A., Jansons, V., Kyllmar, K., Loigu, E.,Rekolainen, S., Vouorenmaa, J., 2001. Nutrient losses fromAgriculture in the Nordic and Baltic Regiona. Tema NordReport 2001:591.

Valpasvuo-Jaatinen, P., Rekolainen, S., Latostenmaa, H., 1997. Finnishagriculture and its sustainability: environmental impacts. Ambio 7,448–455.

Veith, T.L., Sharpley, A.N., Weld, J.L., Gburek, W.J., 2005.Comparison of measured and simulated phosphorus losses withindexed site vulnerability. Transactions of the ASAE 48, 557–565.

Vuorenmaa, J., Rekolainen, S., Lepisto, A., Kenttamies, K.,Kauppila, P., 2002. Losses of nitrogen and phosphorus fromagricultural and forest areas in Finland during the 1980s and1990s. Environmental Monitoring and Assessment 76, 213–248.

Weld, J.L., Sharpley, A.N., Beegle, D.B., Gburek, W.J., 2001. Identifyingcritical sources of phosphorus export from agricultural watersheds.Nutrient Cycling in Agroecosystems 59, 29–38.

Wischmeier, W.H., Smith, D.D., 1978. Predicting rainfall erosionlosses – a guide to conservation planning. Agriculture handbookno 537, US Department of Agriculture.

Zuzel, J.F., Pikul Jr., J.L., 1987. Infiltration into a seasonally frozenagricultural soil. Journal of Soil and Water Conservation 42,447–450.


Review of indexing tools for identifying high risk areas of phosphorus loss in Nordic catchments

Documents