S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 4 ( 2 0 0 8 ) 2 8 4 – 2 8 9

ava i l ab l e a t www.sc i enced i r ec t . com

www.e l sev i e r. com/ loca te / sc i to tenv

The influence of N load and harvest intensity on the risk of Plimitation in Swedish forest soils

Cecilia Akselssona,⁎, Olle Westlinga, Mattias Alvetegb, Gunnar Thelinb,Ann-Mari Franssonc, Sofie Hellstena

aIVL Swedish Environmental Research Institute, P.O. Box 5302, SE-400 14 Göteborg, SwedenbDepartment of Chemical Engineering, Lund University, P.O Box 124, SE-221 00, Lund, SwedencSection of Plant Ecology and Systematics, Ecology Building, Lund University, SE-223 62, Lund, Sweden

A R T I C L E I N F O

⁎ Corresponding author. Tel.: +46 4631 725620E-mail address: cecilia.akselsson@ivl.se (C.

0048-9697/$ – see front matter © 2007 Elsevidoi:10.1016/j.scitotenv.2007.11.017

A B S T R A C T

Available online 26 December 2007

Nitrogen (N) is often considered to be themajor factor limiting tree growth in northern forestecosystems. An increased N availability, however, increases the demand for other nutrientssuch as base cations and phosphorous (P) which in turn may change which nutrient is thelimiting factor. If P or base cations become limiting, N will start to leach which means a riskof increased eutrophication of surface waters. As many studies focus on base cations, thisstudy instead aims at estimating P budgets on a regional scale for different harvestingscenarios relevant for Swedish conditions.P budget calculations were carried out for 14,550 coniferous sites from the Swedish NationalForest Inventory, as weathering+deposition–harvesting–leaching. Three scenarios withdifferent harvest intensities were used: 1) no harvesting, 2) stem harvesting and 3) whole-tree harvesting. The input data were derived from measurements and model results.The P budget estimates indicate that harvesting, especially whole-tree harvesting, result innet losses of P in large parts of Sweden. The highest losses were found in southern Swedendue to high growth rate in this area. In the whole-tree harvesting scenario the lossesexceeded 1 kg ha−1 y−1 on many sites. N budget calculations on the same sites indicate thatN generally accumulates in the whole country and especially in the southern parts.Consequently, the N and P budget calculations indicate that the forests in southern Swedenare in a transition phase from N-to P-limitation to growth. This transition will proceed aslong as the accumulation of N continues. These results are important in a sustainableforestry context, as a basis for assessing the risk of future N leaching, and in designingrecommendations for abatement strategies of N deposition and for application of wood ashrecycling and N fertilization.

© 2007 Elsevier B.V. All rights reserved.

Keywords:PhosphorousNitrogenBudgetForest soilsSwedenHarvestingWeathering

1. Introduction

In northern forest ecosystems nitrogen (N) is often consideredto limit tree growth (Tamm, 1991). The high N depositioncaused by anthropogenic emissions may, however, lead to Nsaturation in the forest ecosystems and thus N leaching (Aberet al., 1989), something which has already been observed in

0; fax: +46 4631 7256290.Akselsson).

er B.V. All rights reserved

central Europe (Gundersen et al., 2006). N saturation meansthat other nutrients become growth limiting, e.g. potassium(K), magnesium (Mg) and phosphorous (P). Whereas basecation availability has been investigated in several studies (e.g.Akselsson et al., 2006b; Thelin et al., 1998) there are not manystudies on P budgets on mineral soils even though P is oftenmentioned as one of the most interesting nutrients from a

.

285S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 4 ( 2 0 0 8 ) 2 8 4 – 2 8 9

nutrient limitation perspective, due to the high demand andlow supply (Abel et al., 2002).

The development fromN to P limitation to growth in ageingecosystems is a natural process as shown by Wardle et al.(2004) in chronosequence studies on nine different locationsworldwide. The transition is explained by the low externalP input to the system, a limited and declining input fromweathering as the soil ages and a slow but continuous accu-mulation of N with biological and atmospheric fixation.Measures that increase the P loss and/or increase the N accu-mulation in the system will speed up the process. In Swedenit is well known that P often limits growth on peat soils butthere are indications of P limitation to forest growth also onmineral soils (Thelin, 2006, Månsson, 2005, Giesler et al., 2002).Furthermore high N deposition levels have been shown tocause P deficiency in different areas (Mohren et al., 1986, Tengand Timmer, 1995, Fenn et al., 2006) and N addition in matureNorway spruce stands has been shown to induce critically lowP concentrations in needles in southern Sweden (Rosengren-Brinck and Nihlgård, 1995) and Denmark (Gundersen, 1998).

The aim of this study was tomake high resolution P budgetcalculations on a regional scale for Swedish forests onmineralsoil and to estimate the risk of future P limitation by com-paring P budget calculationswith N budget calculations for thesame sites. This comparison should be seen as a first attemptto assess the P status in relation to the N status. The complexinteractions between N, P and carbon limitation in the plantsare not considered, but the results can be a valuable input inthe discussions about Progressive N Limitation (PNL) (Johnson,2006). Although the results, partly due to the uncertainties inimportant model inputs, cannot give clear answers aboutwhich nutrient that will limit growth, the estimated accumu-lation or net loss of N and P may give indications as to wherein Sweden there is a risk of future P limitation and thus Nleaching. The results constitute a good basis for more complexanalyses where the pools, the nutrient demand of trees andthe time aspects are included. This will be useful in designingrecommendations for abatement strategies of N depositionand for application of wood ash recycling and N fertilization.

2. Materials and methods

2.1. Budget calculations

The P and N budget calculations in the present study wereperformed by means of a simple mass balance model wherethe inputs to the forest ecosystem were compared with theoutputs. Only the inputs and outputs of the system wereconsidered, not the internal circulation of organic matter. Ifthe inputs exceed the outputs there is a net accumulationwhile larger outputs than inputs indicate net loss. Net changesin either direction occur in natural ecosystems, but normallyat low rates. Accumulation and net losses at high rates mayindicate a risk of adverse environmental effects.

The budget calculations were based on static massbalances with several assumptions:

• The budget terms are assumed to be constant over time.• Only vertical percolation is considered.

• The harvesting is equal to the net growth, i.e. no long-termaccumulation of standing biomass.

• The root zone, where nutrients are taken up, is 40 cm forspruce and 50 cm for pine (organic layer included).

2.2. P budget calculations for different forestry scenarios

Since the intensity of forest management greatly influencesthe P status (Yanai, 1998), the P budgetwas calculated for threedifferent forestry intensity scenarios: 1) no harvesting, 2) stemharvesting and 3) whole-tree harvesting. In the whole-treeharvesting scenario 75% of the branches were assumed to beremoved, in accordance with an intensive harvesting scenario(Swedish Forest Agency, 2000). Further, 75% of the needleswere assumed to accompany the branches, based on a studyof needle loss in slash removal (S. Jacobson, pers. comm).

The P budget calculations were performed for the root zoneon 14,550 sites with coniferous forest from the National ForestInventory in Sweden (Hägglund, 1985). Atmospheric deposi-tion and weathering of soil minerals constitute the input of Pto the forest ecosystem. The outflow consists of harvestedbiomass and leaching. The nutrient budgets for P can be cal-culated as:

D ¼ DepositionþWeathering �Harvesting � Leaching ð1Þ

Whereas current rates, or approximations of current rates,can be used for the deposition,weathering and leaching terms,the harvesting term must be regarded in the perspective of awhole forest rotation. Thus, the results of the calculations givethe annual net change as an average for a forest rotation,provided that the other terms are constant over time. If theinputs are greater than the outputs, P accumulates in the soil.If, however, the outputs are greater than the inputs, thosepools are reduced. The pools include the exchangeable P andthe potentially available P in different forms.

Most of the phosphorous in Swedish soils is found in themineral apatite (Ca10(PO4) 6(OH) 2), and all weathered P wasassumed to come derive the weathering of apatite, accordingto earlier studies (Sverdrup and Warfvinge, 1993). The weath-ering rate in the rooting zone was modelled using the PROFILEmodel (Sverdrup and Warfvinge, 1993). PROFILE is a steadystate model including process-oriented descriptions of solu-tion equilibrium reactions, chemical weathering of mineralsand leaching and accumulation of dissolved chemical com-ponents. The data acquisition is described thoroughly inAkselsson et al. (2006b), and the same root depths was usedas in that study, i.e. 40 cm (including humus layer) for siteswith spruce forest and 50 cm for sites with either pine forest ormixed spruce and pine forest.

Data on wet deposition of P from sites in Sweden and thesurrounding countries were based on available measurementsfrom collectors on open field in Sweden (Knulst; 2001; Westlingand Hultberg, 1990/91). In these studies the wet deposition ofphosphorous varied between 0.05 and 0.27 kg per hectareand year on two sites in southern Sweden, two sites incentral Sweden, one site in northern Sweden, one site inDenmark and 18 sites in Norway. There were no indicationsof a geographical gradient of P deposition in Sweden and thusan average deposition was used. The total deposition (wetand dry) was estimated to be 0.20 kg ha−1 y−1, i. e. 40% higher

Table 1 – Statistics for input parameters (deposition,weathering, harvesting, leaching) and results(accumulation) in the P budget calculations

Average 5percentile

95percentile

Kg ha−1 y−1

Deposition 0.20 0.20 0.20Weathering 0.09 0.01 0.24Stem harvesting 0.26 0.09 0.54Whole-tree harvesting 0.54 0.16 1.27Leaching 0.04 0.04 0.04Accumulation noharvesting

0.25 0.18 0.40

Accumulation stemharvesting

−0.01 −0.29 0.17

Accumulation whole-treeharvesting

−0.29 −1.01 0.08

286 S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 4 ( 2 0 0 8 ) 2 8 4 – 2 8 9

than the wet deposition, based on a known relationshipbetween wet and dry deposition of the non sea-salt part of thebase cations Ca, Mg and K (Akselsson et al., 2006b), assumingthat the properties, e.g. particle size, of the deposited P and basecations are about the same. Losses of P through harvestingwereestimated based on data from the National Forest Inventory inSweden and data on P concentration in stems, branches andneedles. As a first step a probable growth was estimated byreducing the estimated site productivity by 20% for each site.Volume growth was recalculated to mass growth using thedensities 430 kgm−3 for spruce and 490 kgm−3 for pine. Harvestlosses for the stem harvesting scenario were then estimated bymultiplying the growth by P concentration in stems (0.14mg g−1

for spruce and 0.10 mg g−1 for pine). To estimate harvest lossesfor the whole-tree harvesting scenario the amount of branchesand needleswas estimated using standardmethods (Marklund,

Fig. 1 –P budget in Sweden for three different forestry scenarios:harvesting (right).

1988), which are based on empirical correlations between thediameter at breast hight and themeasured amount of branchesand needles on a large number of trees in Sweden. Theestimated amount of branches and needles removed was thenmultiplied by P concentrations in branches (0.61 mg g−1 forspruce and 0.35mg g−1 for pine) and needles (1.3mg g−1 for bothspruce and pine). P concentrations in different tree parts arebased on data compiled by Jacobson and Mattson (1998) andEgnell et al. (1998). Total P leaching was derived from Swedishmeasurements in 25 streams in forest areas in Sweden (Ugglaand Westling, 2003). The leaching varied between 0.02 and0.08 kg ha−1 y−1, except for one sitewith higher leaching, 0.14 kgha−1 y−1. No geographical gradient was found, thus an average,0.04 kg ha−1 y−1, was used.

2.3. N budget calculations

N budget calculations have been performed on a GIS (Geo-graphical Information System) platform in a 5×5 km grid inSweden for different nitrogen deposition and forestry scenar-ios (Akselsson and Westling, 2005). In the present study Nbudget calculations were performed using the same principleson the 14,550 sites, to make it comparable with the P budgetresults. The accumulation in soil (Δ) was calculated as:

D ¼ Depositionþ Fixation�Harvesting � Leaching ð2ÞThe best available regional-scaled data was incorporated

into the input database for each of the four input parametersand mass balance calculations were then made for all the sites(Table 1). Deposition was derived from the Swedish MATCHmodel in the 5×5 km resolution (Langner et al., 1996) andfixation was set to a constant value, 1.5 kg ha−1 y−1, based on astudy innorthern Sweden byDeLuca et al. (2002). Harvest lossesof N were estimated based on growth data from the NationalForest Inventory, in a similar way as for P. The variation for N

no harvesting (left), stem harvesting (centre) and whole-tree

Fig. 2 –N budget in Sweden for three different forestry scenarios: no harvesting (left), stem harvesting (centre) and whole-treeharvesting (right).

287S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 4 ( 2 0 0 8 ) 2 8 4 – 2 8 9

concentration in runoff water from forest soils in Sweden issmall, due to high N retention. This justifies the use of oneaverage value for southern Sweden and one for central andnorthern Sweden. For central and northern Sweden N leachingwas calculated based on an empirical relationship includingrunoff. In southern Sweden N concentrations in surface waterfrom 23 catchments were combined with runoff data. Thecompilation of input data is described in Akselsson andWestling (2005). The same scenarios as for P were run.

3. Results

3.1. P budgets for different forestry scenarios

The harvest intensity is decisive for the P budget (Fig. 1) in thestem harvesting scenario as well as the whole-tree harvestingscenario (Table 1). Harvest intensity becomes increasinglyimportant from the north to the south. Since deposition ishigher than leaching thePbudget ispositive for all sites in thenoharvest scenario, regardless of weathering input. The majorityof the sites accumulated up to 0.4 kg P ha−1 y−1 according to thecalculations. In the stemharvesting scenariomost sites showeda budget between−0.3 and+0.2 kg ha−1 y−1 with mainly ac-cumulation in the north and losses in the south. In thesouthernmost part net losses of up to 0.4 kg ha−1 y−1 werecommon. At whole-tree harvesting only the inner parts ofnorthern Sweden showed accumulation of P. In the southern-most parts net losses ofmore than 1 kg ha−1 y−1 were common.In the northern part of Sweden the budget was around ±0.

3.2. N budgets for different forestry scenarios

The N budget showed a clear gradient in Sweden from thesouthwest to the north (Fig. 2). If no harvest was applied more

than 12 kg ha−1 y−1 accumulated in southwestern Swedenaccording to the calculations. Even in northern Sweden, wherethe deposition is much lower, there was a relatively high Naccumulation, more than 4 kg ha−1 y−1 on most sites. In thestem harvesting scenario the accumulation was still high, butlower than in the no harvesting scenario. In southwesternSweden the accumulation was between 8 and 12 kg ha−1 y−1

and in northern Sweden between 0 and 4 kg ha−1 y−1 on mostof the sites. Whole-tree harvesting reduced the accumulationfurther, but still the accumulation in the southwestern partwas high, up to 12 kg ha−1 y−1. Several sites in northern, centraland southeastern Sweden showed net losses of N.

4. Discussion

There was a clear gradient for P with more net losses in thesouthern parts of Sweden for both harvesting scenarios, due tothe higher forestry intensity in the south with more favour-able climate and thus more growth. Also for N there was aclear geographical gradient, with the highest accumulation ofN in southern Sweden with the greatest N loads. The high Naccumulation indicates a risk of future N saturation, whichmeans that other nutrients become limiting factors and Nstarts to leach. This is important knowledge in the context ofsustainable forestry, when designing recommendations forabatement strategies of N deposition and for application ofwood ash recycling and N fertilization. Since there are netlosses of P in the south it is likely that P will become limiting.High N concentrations may hamper the activity of the fun-gus Hygrophoropsis aurantiaca which stimulates weatheringand increases dissolution of humus, and thus increases theP availability (Fransson et al., 2004). This may reinforce therisk of P limitation. It is difficult to assess the effect of whole-tree harvesting on risk of P limitation, since it reduces the risk

288 S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 4 ( 2 0 0 8 ) 2 8 4 – 2 8 9

of N saturation but at the same time reduces the availabilityof P.

The interpretation of the results is complicated by thecomplex phosphorous activities in soil, e.g. P sorption andP ageing. Some of the phosphate weathered from apatite isreprecipitated into secondary minerals, i.e. Ca-, Fe-and Al-phosphates of different strength, and is thus not availablefor the trees. Accordingly, soils can become P limited althoughthe total amount of P is high (Giesler et al., 2002). Secondaryapatites become more recalcitrant with time. On the otherhand plant species of different kinds are known to utilize dif-ferent parts of the soil P (Trolove et al., 1996, Fransson et al.,2003). It has been proposed that the activity of phosphatase,which decomposes organic P, may increase at low concentra-tions of P, which would decrease the risk of P shortage for thetrees (Vance, 2001). In the stem harvesting scenario the excessN might compensate for the negative P budget through an in-creased enzymatic avtivity (Treseder and Vitousek, 2001), butin the whole-tree harvesting scenario this is less probable, dueto less or no excess N. Thus, the P budget calculations only givethe external conditions for P limitation and cannot be usedalone for assessments of the risk of P limitation.

All input parameters are to some extent associated withuncertainties. The main contribution to the overall uncertain-ties in the P budget calculations probably comes from thedeposition term. There are only a few measurements of P de-position available in Sweden and in surrounding countries,and thus it is not possible to map the geographical variationsof P deposition. However, in spite of the large uncertainties, itcan be concluded that the intensity of the forestry plays amajor role in the P budget and that the budget is substantiallynegative at whole-tree harvesting in large parts of Sweden,even if the P deposition would be twice as high.

The method used in this study is associated with uncer-tainties both in the calculated weathering rate and in theplant-availability of the released P. The uncertainty in weath-ering rate as calculated by the PROFILE model in a regionalapplication is typically more related to the quality of the inputdata than to themodel. Themineralogy, an important input, isbased on a total elemental analysis of the soil and all P deter-mined is considered to be apatite bound. This is, however,not entirely correct since the soil also may contain Al and Feboundphosphates to a varying degree. As these compounds areconsidered less prone to weathering than apatite, the weath-ering rate might be overestimated. Furthermore Göransson(2006) showed that the P accessible in different soil layers isclosely related to the root density since themobility of P is verylow. Thus, in deep soil layers with low root density the P inputfrom weathering should be less than the assumed 100%.

The uncertainties in harvest losses of P aremainly caused byuncertainties in the estimated growth rates and the estimatedamount of branches and needles on the trees, which is furtherdiscussed in Akselsson et al. (2006a), as well as uncertainties inthe P concentrations in different tree parts. For P concentrationsan averagewas used, and since there is no obvious geographicalgradient inP concentrations theerrorsduetovariationsbetweendifferent stands will most likely even out when interpretingthe whole dataset of 14,550 sites. The effect of the uncertaintiesin the harvesting term on the budget results is estimated tobe relatively small in comparison with uncertainties from the

deposition and weathering terms. However, it is important tofind ways to be able to vary the P concentrations in tree partsbased on e.g. the proximity to agricultural land or standcharacteristics. Although the uncertainties in the leaching dataare rather large, the effect on the total uncertainties is small,since the leaching term is small compared to the other terms.

To reduce the uncertainties of the budget calculations, andto be able to drawmore far-reaching conclusions, more P datais required. In 2006 P deposition measurements were intro-duced on about 30 environmental monitoring sites in Sweden,and when these data are available they can be used for im-proving the P deposition input. Further the P pools and trans-formation processes in soil need to be investigated, in order tobe able to estimate the effects of the net losses, and the resultshave to be interpreted together with results from base cationcalculations, in order to be able to rank the risk of limitationof different nutrients. To include the time aspects and makelong-term predictions, dynamicmodels are required as a com-plement, e.g. ForSAFE (Wallman et al., 2005).

5. Conclusions

This study shows that forestry intensity is of great importancefor the P budget. High N accumulation in the south, combinedwith high P losses, indicate that P may become limiting forgrowth. For the forestry to be sustainable, the N depositionhas to be reduced, and wood ash recycling is needed at highharvest intensities. More data on P deposition are required inorder to reduce the uncertainties of the calculations. Furtherstudies, including soil pools of P and ecosystem dynamics canmake it possible to predict the future development.

R E F E R E N C E S

Abel S, Ticconi CA, Delatorre CA. Phosphate sensing in higherplants. Physiol Plant 2002;115:1–8.

Aber JD, Nadelhoffer KJ, Steudler P, Melillo JM. Nitrogen saturationin northern forest ecosystems. Bioscience 1989;39:378–86.

Akselsson C, Sverdrup H, Holmqvist J. Estimating weatheringrates of Swedish forest soils in different scales using thePROFILE model and affiliated databases. J Sustain For2006a;21:119–31.

Akselsson C, Westling O. Regionalized nitrogen budgets in forestsoils for different deposition and forestry scenarios in Sweden.Glob Ecol Biogeogr 2005;14:85–95.

Akselsson C, Westling O, Sverdrup H, Holmqvist J, Thelin G, UgglaE, et al. Impact of harvest intensity on long-term base cationbudgets in Swedish forest soils. Water, Air, & Soil Pollution.Focus 2006b;7:201–10.

DeLuca H, Zackrisson O, Nilsson MC, Sellstedt A. Quantifyingnitrogen-fixation in feather moss carpets of boreal forests.Nature 2002;419:917–20.

Egnell G, Nohrstedt HÖ, Weslien J, Westling O, Örlander G.Miljökonsekvensbeskrivning av skogsbränsleuttag,asktillförsel och övrig näringskompensation. Report 1:1998.Jönköping, Sweden: National Board of Forestry; 1998.(In Swedish).

Fenn ME, Perea-Estrada VM, de Bauer LI, Pérez-Suárez M, ParkerDR, Cetina-Alcalá VM. Nutrient status and plant growth effectsof forest soils in the Basin of Mexico. Environmen Pollut2006;140:187–99.

289S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 4 ( 2 0 0 8 ) 2 8 4 – 2 8 9

Fransson A-M, Valeur I, Wallander H. The wood-decaying fungusHygrophoropsis aurantiaca increases P availability in acid forestsoil, while N addition hampers this effect. Soil Biol Biochem2004;36:1699–705.

Fransson A-M, van Aarle IM, Olsson PA, Tyler G. Differencesbetween Plantago lanceolata L. and Rumex acetosella in utilisationof soil phosphorus. Plant Soil 2003;248(1–2):285–95.

Giesler R, Pettersson T, Högberg P. Phosphorus limitation in borealforests: effects of aluminum and iron accumulation in thehumus layer. Ecosystems 2002;5:300–14.

Göransson H. The vertical distribution of roots, mycorrhizalmycelia and nutrient acquisition in mature forest trees.PhD-thesis, Lund University, 2006.

Gundersen P. Effects of enhanced nitrogen deposition in a spruceforest at Klosterhede, Denmark, examined by moderateNH4NO3 addition. For Ecol Manag 1998;101:251–68.

Gundersen P, Schmidt IK, Raulund-Rasmussen K. Leaching ofnitrogen from temperate forests—effects of air pollution andforest management. Environ Rev 2006;14:1–57.

Hägglund B. En ny svensk riksskogstaxering (A new SwedishNational Forest Survey). Swedish University of AgriculturalSciences, Report 37, Uppsala, Sweden; 1985.(In Swedish with English summary).

Jacobson S, Mattson S. Snurran”—an Excel program for calculatingsite nutrient levels in logging residues. The ForestryResearch Institute of Sweden, Report No.1; 1998.(In Swedish with English summary).

Johnson D. Progressive N Limitation in forests: review andimplications for long-term responses to elevated CO2.Ecology 2006;87:64–75.

Knulst JC. Phosphorous inprecipitation.Results frommeasurementsduring the 1990s. IVL Swedish Environmental Research InstituteLtd., Report B 1442; 2001. (In Swedish with English summary).

Langner J, Persson C, Robertson L, Ullerstig A. Air pollutionassessment study using the MATCH Modelling System.Application to sulphur and nitrogen compounds over Sweden1994. Swedish Meteorological and Hydrological Institute,Report no. 69, Norrköping, Sweden; 1996.

Månsson K. Plant–Bacterial and Plant–Fungal Competition forNitrogen and Phosphorus. PhD-thesis, Lund University, 2005.

Marklund LG. Biomass functions for pine, spruce, and birch inSweden. Department of Forest Taxation, Swedish Universityof Agricultural Sciences, SLU, Report 45; 1988. (In Swedish withEnglish summary).

Mohren GMJ, Van den Burg J, Burger FW. Phosphorous deficiencyinduced by nitrogen input in Douglas fir in the Netherlands.Plant Soil 1986;95:191–200.

Rosengren-Brinck U, Nihlgård B. Nutritional status in needlesof Norway spruce in relation to water and nutrient supply.Ecol Bull 1995;44:168–77.

Sverdrup H, Warfvinge P. Calculating field weathering rates usinga mechanistic geochemical model (PROFILE). J Appl Geochem1993;8:273–83.

Swedish Forest Agency. Skogliga konsekvensanalyser 1999. Report2:2000. Jönköping, Sweden: National Board of Forestry; 2000.(In Swedish).

Tamm CO. Nitrogen in terrestrial ecosystems. Ecological Studies.Berlin, Germany: Springer Verlag; 1991. p. 81.

Teng Y, Timmer VR. Rhizosphere phosphorous depletion inducedby heavy nitrogen fertilization in forest nursery soils. Soil SciSoc Am J 1995;59:227–33.

Thelin G. Ash recycling to spruce and beech stands—effects onnutrition, growth, N dynamics, and C balance. Värmeforsk,report nr 965; 2006. (In Swedish with English summary).

Thelin G, Rosengren-Brinck U, Nihlgård B, Barkman A. Trends inneedle and soil chemistry of Norway spruce and Scots pinestands in South Sweden 1985–1994. Environ Pollut1998;99:149–58.

Treseder K, Vitousek P. Effects of soil nutrient availability oninvestment in acquisition of N and P in Hawaiian rainforests.Ecology 2001;82:946–54.

Trolove S, Hedley M, Caradus J, Macay A. Uptake of phosphorusfrom different sources by Lotus penunclatus and threegenotypes of Trifolium repens. 2. Forms of phosphate utilisedand acidification of the rhizosphere. Aust J Soil Res1996;34:1027–40.

Uggla E, Westling O. Utlakning av fosfor från brukad skogsmark.IVL Swedish Environmental Research Institute Ltd., Report B1549; 2003. (In Swedish).

Vance C. Symbiotic nitrogen fixation and phosphorousacquisition. Plant Nutrition in a world of decliningrenewable resources. Plant Physiology 2001;127:390–7.

Wallman P, Svensson M, Sverdrup H, Belyazid S. ForSAFE—anintegrated process-oriented forest model for long-termsustainability assessments. For Ecol Manag 2005;207:19–36.

Wardle D, Walker L, Bardgett R. Ecosystem properties and forestdecline in contrasting long-term chronosequences. Science2004;305:509–13.

Westling O, Hultberg H. Liming and fertilization of acid forest soil:short-term effects on runoff from small catchments. Water,Air, and Soil Pollution 1990/91;54:391–407.

Yanai R. The effect of whole-tree harvest on phosphorouscycling in a northern hardwood forest. For Ecol Manag1998;104:281–95.