Forest Policy and Economics xxx (2013) xxx–xxx

FORPOL-01077; No of Pages 11

Contents lists available at ScienceDirect

Forest Policy and Economics

j ourna l homepage: www.e lsev ie r .com/ locate / fo rpo l

Whole-tree harvesting with stump removal versus stem-only harvesting in peatlandswhen water quality, biodiversity conservation and climate change mitigation matter☆

Jenni Miettinen a,⁎, Markku Ollikainen a, Tiina M. Nieminen b, Liisa Ukonmaanaho b,Ari Laurén c, Jari Hynynen b, Mika Lehtonen b, Lauri Valsta d

a University of Helsinki, Department of Economics and Management, P.O. Box 27, 00014 University of Helsinki, Finlandb Finnish Forest Research Institute, P.O. Box 18, 01301 Vantaa, Finlandc Finnish Forest Research Institute, P.O. Box 68, 80101 Joensuu, Finlandd University of Helsinki, Department of Forest Sciences, P.O. Box 27, 00014 University of Helsinki, Finland

☆ This article belongs to the Special Issue: Forests andNew Policy Options.⁎ Corresponding author. Tel.: +358 40 534 1285; fax: +

E-mail addresses: jenni.miettinen@helsinki.fi (J. Miettimarkku.ollikainen@helsinki.fi (M. Ollikainen), tiina.niemiliisa.ukonmaanaho@metla.fi (L. Ukonmaanaho), ari.laurenjari.hynynen@metla.fi (J. Hynynen), mika.lehtonen@metlalauri.valsta@helsinki.fi (L. Valsta).

1389-9341/$ – see front matter © 2013 Elsevier B.V. All rihttp://dx.doi.org/10.1016/j.forpol.2013.08.005

Please cite this article as: Miettinen, J., et al.,quality, biodiversity conservation and climat

a b s t r a c t

a r t i c l e i n f o

Article history:Received 30 November 2012Received in revised form 27 July 2013Accepted 12 August 2013Available online xxxx

Keywords:Whole-tree harvestingStem-only harvestingClimate policyNutrient loadMercury loadBiodiversity conservation

This article examines alternative forest harvesting regimes when ecosystem services in terms of water quality,biodiversity conservation and climate change mitigation are included in the analysis. The harvesting regimesare whole-tree harvesting with stump removal and conventional stem-only harvesting. The harvesting regimesare evaluated under two alternative climate policy contexts. The first alternative is a carbon neutral bioenergypolicy, which assumes the carbon dioxide (CO2) neutrality of bioenergy and produces substitution benefits, asbioenergy replaces fossil fuels. The second alternative climate policy, a carbon non-neutral bioenergy policy,takes into account the fact that bioenergy causes carbon dioxide emissions, producing substitution costs, andthat harvested woody biomass affects the ability of a forest to act as a carbon sink. We extend the traditionalFaustmann (1849) rotation model to include nutrient load damage, biodiversity benefits, and climate impacts.The empirical analysis is based on Finnish data from a catchment experiment carried out on drainedpeatland for-ests. The empirical results show that under a carbon neutral bioenergy policy, whole-tree harvesting with stumpremoval produces the highest net social benefits. However, if a carbon non-neutral bioenergy policy is assumed,the net social benefits are greater under stem-only harvesting.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Increasing the use of renewable energy is an important tool of climatemitigation policy in both the EU and US. Bioenergy, such as forest resi-dues, is regarded as a potential means for substituting conventional fossilfuels, since the use of biomass for energy production reduces fossil fuelemissions and replaces non-renewable energy sources. Due to the in-creasing economic potential of forest residues, there is a need to reconsid-er alternative harvesting methods and policy regimes. In conventionalstem-only harvesting, only the stems are harvested and the logging resi-dues are left on the site. In contrast, whole-tree harvesting also removesforest residues (tree tops, branches and foliage) from the site. Further-more, in addition to the above-ground tree biomass, the tree stumpscan also be removed from the site and used as an energy source.

Ecosystem Services: Outlines for

358 9 191 58096.nen),nen@metla.fi (T.M. Nieminen),@metla.fi (A. Laurén),.fi (M. Lehtonen),

ghts reserved.

Whole-tree harvesting with se change mitigation matter, F

The two possible harvesting regimes impact differently on the eco-system services provided by forests. In addition to climate change miti-gation, forests provide timber, biodiversity conservation, amenities, andwater quality (Fisher et al., 2009). Forest ecosystems cover provisioningservices, regulating services, cultural services, and supporting services(Millennium Ecosystem Assessment, 2005, page v). The sustainabilityof whole-tree harvesting, in particular, has been challenged due to itspotential negative impacts on several forest ecosystem services.

A primary concern of whole-tree harvesting and stump removal isthe depletion of soil nutrients and its effect on future forest productivity(e.g. Mann et al., 1988; Bengtsson and Wikström, 1993; Olsson et al.,1996; Jacobson et al., 2000; Egnell and Valinger, 2003; Merino et al.,2005; Walmsley et al., 2009). Furthermore, the extraction of both har-vest residues and stumps causes habitat loss for saproxylic species,thereby affecting biodiversity (Jonsell, 2007; Lattimore et al., 2009;Walmsley and Godbold, 2010; Jonsell and Hansson, 2011; Bougetet al., 2012). Whole-tree harvesting also weakens the ability of a forestto act as a carbon sink (Repo et al., 2011).

Several studies have discerned high total mercury (TotHg) andmethylmercury (MeHg) concentrations in water leaching from treeharvesting areas (Munthe and Hultberg, 2004; Porvari et al., 2003;Sørensen et al., 2009). Moreover, logging residues left in a clear-cutarea are a potential source of heavy metals when the dry deposited

tump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005

2 J. Miettinen et al. / Forest Policy and Economics xxx (2013) xxx–xxx

substances that have accumulated on the surface of the logging residuesare leaching away. All forest management measures, includingharvesting, are considered to increase sediment and nutrient export towatercourses (e.g. Grip, 1982; Ahtiainen and Huttunen, 1999; Laurénet al., 2005; Miettinen et al., 2012).

The aim of this article is to consider the optimal forest managementpractice when various ecosystem services are included in the analysis.We extend the traditional Faustmann (1849) rotation model to includenutrient load damage, climate impacts, and biodiversity benefits. Weaddress the selection of the optimal rotation age so as to balance differ-ent environmental externalities caused by harvesting under the two al-ternative harvesting regimes: stem-only harvesting and whole-treeharvesting with stump removal. We assess these harvesting regimesfrom two alternative climate policy perspectives. First, in the spirit ofKyoto Protocol, we postulate a carbon neutral bioenergy policy, whichassumes the CO2 neutrality of bioenergy such as forest fuels and pro-duces substitution benefits as bioenergy substitutes for fossil fuels.Our second perspective is a carbon non-neutral bioenergy policy,which takes into account that using forest biomass in bioenergy produc-tion releases CO2, producing substitution costs, and that harvestedwoody biomass affects the ability of a forest to act as a carbon sink.

Our choice of carbon neutral and non-neutral perspectives reflectscurrent discussion. Forest bioenergy has traditionally been regarded as acarbon neutral (or low-carbon) energy source, because the carbon diox-ide emissions released into the atmosphere when harvested vegetationis converted to energy are taken up again by the growth of the new treegeneration (e.g.Wihersaari, 2005; Stupak et al., 2007). Recently, however,the carbon neutrality of forest bioenergy has been questioned (Repo et al.,2011, 2012; Searchinger et al., 2009) by the argument that neutrality cru-cially depends on the time horizon and on indirect carbon dioxideemissions.

More specifically, to estimate the efficiency of forest bioenergy in re-ducing carbon dioxide emissions and mitigating climate change, in ad-dition to direct emissions from the bioenergy production chain(Palosuo et al., 2001; Mälkki and Virtanen, 2003; Wihersaari, 2005;Lattimore et al., 2009), the temporal dimension and indirect carbon di-oxide emissions into the atmosphere must be considered (e.g.Schlamadinger et al., 1995; Palosuo et al., 2001; Lattimore et al., 2009;Melin et al., 2010; Repo et al., 2011, 2012). In a carbon non-neutralbioenergy policy, we assume that carbon dioxide emissions resultingfrom bioenergy combustion are in the short-term greater than emis-sions from fossil fuels (the default emission factor is 94.6 tCO2/TJ forcoal and 109.6 tCO2/TJ for forest fuelwood (Statistics Finland, 2011)).

Ourmodel combines twodifferent lines of literature.We incorporatewater protection issues in forestry using the approach provided inMiettinen et al. (2012), who included a nutrient load damage functionin the Faustmann model. 1 Modeling of carbon sequestration is basedon van Kooten et al. (1995), who included the carbon sequestered bythe growth of timber and carbon released due to harvesting in theHartman (1976) model.2 We include biodiversity benefits followingmodeling in the conventional Hartman model. In an empirical analysis,we use data including nitrogen, phosphorus and mercury loads from acatchment experiment carried out in eastern Finland. Under the two al-ternative climate policies we examine for a given rotation length whichharvesting regime produces higher net social benefits for drainedpeatland forests.We are unaware of previous empirical analyses similarto this. The closest studies to ours have included, for instance,Kaltschmitt et al. (1997), Miranda and Hale (2001), Gan and Smith

1 Articles focusing on water protection applications in forestry: Miller and Everett(1975), Clinnick (1985), Matero (1996, 2002, 2004), Matero and Saastamoinen (1998),Creedy and Wurzbacher (2001), Sun (2006), Laurén et al. (2007), Matta et al. (2009),and Eriksson et al. (2011). For more economic studies on water protection in forestrysee Miettinen et al. (2012).

2 For reviews of economic carbon sequestration studies in forestry, see Sedjo et al.(1995), van Kooten et al. (2004), Richards and Stokes (2004), and Boyland (2006).

Please cite this article as: Miettinen, J., et al., Whole-tree harvesting withquality, biodiversity conservation and climate change mitigation matter, F

(2006, 2007), Olschewski and Benítez (2010), Valente et al. (2011),and Alavalapati and Lal (in press).

The rest of the paper is organized as follows. In the next section wepresent the theoretical frame of the study. In Section 3, the study areaand methods used in the ecological study are described. The economicdata and estimation of social costs and benefits used in the empirical anal-ysis are provided in Section 4. In Section 5 the empirical analysis and theresults of alternative harvesting regimes are presented Section 6 containsthe sensitivity analysis. Finally, in Section 7 we provide our conclusions.

2. Socially optimal harvesting under environmental costs and benefits

Here, we examine social welfare from a representative forest standunder carbon neutral and non-neutral bioenergy policies. Under bothclimate policies, the economic problem of the social planner is to choosethe harvesting regime and the optimal rotation age so as to maximizesocial welfare from forestry subject to relevant ecosystem services(water quality, biodiversity conservation and climate change mitiga-tion). In the following, we start with the carbon neutral policy.

2.1. The socially optimal rotation age under a carbon neutral bioenergy policy

In the steady state, the social planner starts with bare land. We de-note rotation age under two management alternatives by Ti, i = 1,2(stem-only harvesting is denoted by superscript 1 and whole-treeharvesting with stump removal by superscript 2). The growth functionas a function of rotation age is denoted by Q(Ti) with Q′(Ti) N 0 and Q″(Ti) b 0 over the relevant range of forest age. The wood price, P(Ti), isthe average sum of thewood prices for sawlogs, pulpwood, and loggingresidues and stumps, pl, pp and ph respectively, weighted by the propor-tions of sawlogs, αl(Ti), pulpwood, αp(Ti), and logging residues andstumps, αh(Ti) from the total amount of wood harvested, Q(Ti). Hence,the wood price is defined as follows:

P Ti� �

¼ αl Ti� �

pl þ αp Ti� �

pp þ αh Ti� �

ph: ð1Þ

In stem-only harvesting, the proportion of logging residuesand stumps is zero, αh(Ti) = 0. Let c denote regeneration costs and rthe real interest rate. The Faustmann net present value of harvest reve-nue, Vi, from the harvested sawlogs, pulpwood, and logging residues

and stumps in the absence of environmental aspects is defined as: Vi ¼

P Ti� �

Q Ti� �

e−rTi−ch i

1−e−rTi� �−1

. It is straightforward to demon-

strate that in the Faustmann model with Eq. (1), the optimal rotationage is implicitly determined by the optimality condition,

ViTi ¼ P′ Ti

� �Q Ti� �

þ P Ti� �

Q ′ Ti� �

−rP Ti� �

Q Ti� �

−rVi ¼ 0: ð2Þ

We next introduce ecosystem services in the model and start withnegative externalities. Without the loss of generality, we focus on nutri-ent loads (the empirical analysis also includes mercury loads). Nutrientloading starts one period after the stand is harvested. The load first in-creases, and then with the growth of the stand it decreases to the back-ground level.3 FollowingMiettinen et al. (2012), let ki be the number ofyears that nutrient loads occur under each harvesting regime i. The nu-trient load after harvesting, g(s), is expressed as a function of time, s. Thenutrient load damage D(zi) as a function of the present value of the pe-riodic loads, zi, can then be written as:

D zi� �

¼ dZki0

g sð Þe−rsds; ð3Þ

where d denotes the constant marginal damage.

3 According to Sillanpää et al. (2006), final harvesting and site preparation increases therunoff for 7 to 11 years at the most.

stump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005

4 The default emission factor is 94.6 tCO2/TJ for coal and 109.6 tCO2/TJ for forest fuelwood (Statistics Finland, 2011).

3J. Miettinen et al. / Forest Policy and Economics xxx (2013) xxx–xxx

Biodiversity benefits depend on time, s, and are denoted by the valua-tion function Fi(s). According to the Hartmanmodel, biodiversity benefitscan be assumed to depend positively on the age of the forest, Ui′(Ti) N 0(Hartman, 1976; Koskela et al., 2007; Amacher et al., 2009), so that

Ui Ti� �

¼ZTi

0

Fi sð Þe−rsds: ð4Þ

Note that both nutrient load damage and biodiversity benefits differbetween the harvesting regimes. Based on the catchment experimentpresented in Section 3, whole-tree harvesting with stump removalcauses higher nutrient loads than stem-only harvesting. It also reducesthe habitats available for saproxylic species more than stem-onlyharvesting (e.g. Jonsell and Hansson, 2011).

Including climate impacts in themodel is slightly more complicated.Climate benefits from replacing fossil fuels can be given as:

G Ti� �

¼ ε bc−bwð Þφαh Ti� �

Q Ti� �

; ð5Þ

where the symbols are defined as follows. Themarginal climate damagedue to carbon dioxide emissions is ε. Carbon dioxide emissions from thegiven amount of energy produced by fossil fuels (the default emissionfactor for fossil fuels) are defined by bc, and emissions from the givenamount of energy produced by logging residues and stumps (thedefaultemission factor for energy wood) are defined by bw. When we considerthe carbon neutral bioenergy policy (bw = 0), we have bc − bw N 0.Themultiplierφ indicates howmuch energy is produced by 1 m3 of log-ging residues and stumps.

Forwhole-tree harvestingwith stump removal, themarginal climatebenefits from replacing fossil fuels are given by

GTi ¼ ε bc−bwð Þφ αh Ti� �

Q 0 Ti� �

þ α0h Ti� �

Q Ti� �h i

: ð6Þ

In Eq. (6), the amount of logging residues and stumps as a propor-tion of the forest growth rate, αh(Ti)Q′(Ti), is positive as long as forestgrowth is positive. The second term, themarginal change in the propor-tion of logging residues and stumps from the total amount of woodharvested, αh′(Ti)Q(Ti), tends to decrease as a function of the rotationage (Lehtonen et al., 2004). Here, we assume that αh(Ti)Q′(Ti) N αh

′(Ti)Q(Ti), so that the marginal climate benefits from replacing fossilfuels are positive. Note that for the carbon neutral bioenergy policy,bw = 0 and Eq. (6) reduces to εbcφ[αh(Ti)Q′(Ti) + αh′(Ti)Q(Ti)]. Finally,we let γi denote carbon dioxide emissions from vehicles in harvestingthe stand under both regimes and define the net climate benefits asKi = ε[(bc − bw)φαh(Ti)Q(Ti) − γi].

Under the carbon neutral bioenergy policy, combining theFaustmann net present value of harvest revenue, Vi, with the terms de-scribing the ecosystem services described above yields the following so-cial welfare function:

Wi ¼ Vi þ −e−rTi

D zi� �

þ Ui Ti� �

þ Kie−rTi� �

1−e−rTi� �−1

: ð7Þ

The economic problem for the planner is to choose harvesting re-gime, i = 1,2, thatmaximizes social welfare in Eq. (7) under an optimalrotation age. The problem is solved recursively by first defining the op-timal rotation age under both regimes and then choosing the regimethat leads to the highest social welfare.

To examine how the Faustmann rotation age changes in the presenceof ecosystem service valuation, we express the optimal rotation age withthe help of the Faustmann optimality condition (Eq. (2)) as follows:

ViTi ¼ − rD zi

� �1−e−rTi� �−1

− Fi Ti� �

−rUi Ti� �

1−e−rTi� �−1� �

− KiTi−rKi 1−e−rTi

� �−1� �;

ð8Þ

Please cite this article as: Miettinen, J., et al., Whole-tree harvesting with squality, biodiversity conservation and climate change mitigation matter, F

whereKiTi ¼ GTi, aswas defined in Eq. (6). Thefirst right-hand side term is

negative and represents the reduction in the nutrient load damage due toa longer rotation age. The second negative right-hand side term reflectsbiodiversity benefits and it lengthens the Faustmann rotation age due tothe valuation of biodiversity benefits from older stands. While these fea-tures hold for both harvesting regimes, the third term, representing cli-mate impacts, differs between the harvesting regimes. The third term isambiguous for whole-tree harvesting, but it has an intuitive interpreta-

tion. If forest growth increases strongly with the rotation age (KiTi is

high), or emissions from transportation are high (Ki is low), the lastterm is negative, tending to lengthen the rotation age so as to increasethe climate benefits or to decrease the emissions from harvesting (how-

ever, if KiTi is low or Ki is high, the rotation age tends to shorten). For

stem-only harvesting, the third term reduces to −rεγi 1−e−rTi� �−1

only, because logging residues and stumps are not harvested. Thus, theoptimal rotation age unambiguously lengthens for stem-only harvesting.

We now use our findings to establish the social decision making onharvesting regimes and provide new comparative static results(Appendix A reports the details) in Lemma 1.

Lemma 1. The optimal rotation age under climate change mitigation andother ecosystem services under a carbon neutral bioenergy policy.

When CO2 emissions fromwood stocks are regarded as carbon neu-tral in a bioenergy policy, the optimal rotation age under stem-only andwhole-tree harvesting regimes has the following properties:

a) For stem-only harvesting, the optimal rotation age is always longerthan the Faustmann rotation age when climate impacts, nutrientload damage and biodiversity benefits count. For whole-treeharvesting with stump removal, the optimal rotation age tends tolengthen, but not unambiguously, as climate impacts are ambiguous.

b) For both regimes, the effect of regeneration costs on the optimal ro-tation is positive, but the impacts of timber prices and the real inter-est rate on the optimal rotation age remain ambiguous.

c) Higher nutrient load damage lengthens the optimal rotation ageunder both regimes, whereas higher climate damage has the sameeffect for stem-only harvesting, but its effect is ambiguous underwhole-tree harvesting with stump removal.

Lemma 1 provides new insights into the comprehensive evaluationof ecosystem services for the first time in the literature. That biodiversi-ty benefits and nutrient load damage unambiguously lengthen the rota-tion age has previously been shown (see Amacher et al., 2009;Miettinen et al., 2012). A new result is that for stem-only harvesting, cli-mate damage reinforces the effects of biodiversity benefits and nutrientload damage, because emissions from harvesting and transportationcount, and they are postponed by a longer rotation age. For whole-tree harvesting with stump removal, the impact of the net climate ben-efit remains ambiguous: rotation age may shorten or lengthen. Wecharacterize the choice of harvesting regime later and next ask whetherthe features will prevail if the bioenergy policy takes into account theactual CO2 emissions from logging residues and stumps.

2.2. The socially optimal rotation age under a carbon non-neutral bioenergypolicy

Under a non-neutral bioenergy policy, the climate costs of replacingfossil fuels continue to be as in Eq. (5), but now we have bw N bc.4

Under a non-neutral bioenergy policy, we need to include the role ofgrowing forests as a carbon sink and the role of harvesting as a sourceof carbon release.

tump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005

4 J. Miettinen et al. / Forest Policy and Economics xxx (2013) xxx–xxx

Drawing on van Kooten et al. (1995), we assume carbon benefits tobe a function of the change in biomass and the amount of carbon percubic meter of biomass. The ton of carbon sequestered per cubicmeter of wood biomass is denoted by q. Thus, carbon sequestration isgiven by [αl(Ti) + αp(Ti) + αh(Ti)]qQ′(Ti), where Q′(Ti) is the growthrate of wood biomass. The present value of carbon uptake benefitsL(Ti) over a rotation period Ti is as follows:

L Ti� �

¼ εqZTi

0

αl xð Þ þ αp xð Þ þ αh xð Þh i

Q 0 xð Þe−rxdx: ð9Þ

Integration by parts gives:

L Ti� �

¼ εq HQ Ti� �

e−rTi

þ rZTi

0

h xð ÞQ xð Þe−rxdx

264

375; ð10Þ

where H = [αl(Ti) + αp(Ti) + αh(Ti)] and h(x) = [αl(x) + αp(x) +αh(x)].

Harvesting reduces the carbon stock. Let the fraction of wood thatremains in long-term storage (e.g. in structures and landfills) after

harvesting be β. The discounted cost of the carbon released is: εq 1−βð ÞHQ Ti� �

e−rTi. Using the same approach as van Kooten et al. (1995), we

express the present value of carbon sequestration benefits and thecosts from the carbon released due to harvesting as follows:

Ji ¼ εq HQ Ti� �

e−rTi

− 1−βð ÞHQ Ti� �

e−rTi

þ rZTi

0

h xð ÞQ xð Þe−rxdx

264

375: ð11Þ

Note that Ji is higher for stem-only harvesting than whole-treeharvesting with stump removal, because stumps and residues are lefton the ground in stem-only harvesting.

The social welfare under our two harvesting regimes is given by

Wi∧

¼ Vi þ −e−rTi

D zi� �

þ Ui Ti� �

þ Kie−rTi

þ Ji� �

1−e−rTi� �−1

: ð12Þ

We follow the same decomposition strategy as before and expressthe necessary condition of the optimal rotation age under a carbonnon-neutral bioenergy policy as:

ViTi ¼ −rD zi

� �1−e−rTi� �−1

− Fi Ti� �

−rUi Ti� �

1−e−rTi� �−1� �

− KiTi−rKi 1−e−rTi

� �−1� �− JiTi−r Ji 1−e−rTi

� �−1� � ð13Þ

where JiTi ¼ εq βHTiQ Ti� �

þ βHQ 0 Ti� �

þ r 1−βð ÞHQ Ti� �h i

and HTi ¼αl

0 Ti� �

þ αp0 Ti� �

þ αh0 Ti� �

.

Again, the left-hand side describes the Faustmann optimality condi-tion. The biodiversity benefit and nutrient load damage terms work asbefore and unambiguously lengthen the rotation age for bothharvesting regimes. In the third term (the substitution of fossil fuels),

we have bw N bc. Therefore, K2T2b0 but−rK2 1−e−rT2

� �−1N0. Provided

that the former term dominates, the substitution term tends to shortenthe rotation age (in order to reduce emissions from replacing fossilfuels) under whole-tree harvesting. However, under stem-only

harvesting thewhole term again reduces to−rεγi 1−e−rTi� �−1

and un-

ambiguously lengthens the rotation age. Finally, JiTi , the new term de-scribing benefits from forests as a carbon sink, is positive and tends to

Please cite this article as: Miettinen, J., et al., Whole-tree harvesting withquality, biodiversity conservation and climate change mitigation matter, F

lengthen the optimal rotation age under both regimes provided that it

dominates the last term in brackets, −r Ji 1−e−rTi� �−1

.

The comparative static effects of exogenous variables remain for themost part the same as before, but now the proportion of woodremaining in storage also matters. For these we have:

Lemma 2. The optimal rotation age under climate change mitigation andother ecosystem services under a carbon non-neutral bioenergy policy.

When CO2 emissions from wood stocks count in the bioenergypolicy, the optimal rotation age under stem-only and whole-treeharvesting regimes has the following properties:

a) For stem-only harvesting, the optimal rotation age is always longerthan the Faustmann rotation age when climate impacts, nutrientload damage, and biodiversity benefits count. For whole-treeharvestingwith stump removal, the climate impacts are ambiguous,as the substitution term tends to shorten the rotation age and theterm describing benefits from forests as a carbon sink tends tolengthen it.

b) The impacts of the timber price, regeneration costs, and real interestrate on the optimal rotation age are qualitatively similar to that inLemma 1.

c) The effects of nutrient load and climate damage on the optimal rota-tion age are qualitatively the same as in Lemma 1, except that nowthe impact of climate damage also remains ambiguous for stem-only harvesting. The effect of an increase in the proportion of woodremaining in storage remains ambiguous.

2.3. The socially optimal harvesting regime: stem-only versus whole-treeharvesting with stump removal

Recall, the challenge of a bioenergy policy is ultimately to assesswhether it is wise to favor traditional stem-only harvesting or to followwhole-tree harvesting with stump removal. The latter harvesting re-gime produces logging residues and stumps to be used as bioenergy.Both harvesting regimes impact on biodiversity benefits, nutrient loaddamage, and net climate benefits.

A comparison of the first-order conditions of both regimes under acarbon neutral bioenergy policy (Eq. (8)) suggests that whole-treeharvesting with stump removal may entail higher climate benefits dueto the replacement of fossil fuel use, but lower biodiversity benefitsand higher nutrient load damage than stem-only harvesting. Accordingto Lemma 1, both harvesting regimes depend on exogenous parametersin a similar way, except for climate damage. From Eq. (13), a similarcomparison of rotation ages can bemade under a non-neutral bioenergypolicy. Because of the opposite signs of the several terms, we cannotconclude in general which of the two harvesting regimes produces alonger rotation age under the two alternative policy options. However,the choice of harvesting regime is straightforward and is presented inthe following proposition.

Proposition. The choice of the harvesting regime under carbon neutraland non-neutral bioenergy policies.

Irrespective of how emissions from wood stocks are regarded in thebioenergy policy, thewhole-tree harvesting regime should only be cho-sen if it yields higher maximum social welfare than stem-onlyharvesting (that is, if W1 ⁎[T1 ⁎(P,c,r,d,ε)] b W2 ⁎[T2 ⁎(P,c,r,d,ε)]), andstem-only harvesting if the opposite holds.

This proposition together with Lemmas 1 and 2 shows how the car-bon in forestry under climate policies has an important influence on thechoice of the optimal harvesting regime and respective rotation age.Therefore, for instance, the recent discussion in the EU on the sustainabil-ity criteria of bioenergy production has an important bearing on thechoice of harvesting regime. If the use of wood, for instance stumps, is

stump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005

5J. Miettinen et al. / Forest Policy and Economics xxx (2013) xxx–xxx

given a CO2 emission coefficient, this leads to the merging of our basicpolicy regimes and increases the relative returns to stem-only harvesting.

3. Study area and methods used in the ecological study

The empirical application in this paper is based on a catchment ex-periment carried out in eastern Finland. We evaluated the net socialbenefits of stem-only harvesting and whole-tree harvesting withstump removal in two representative forest stands: a low fertility siteand a high fertility site.

3.1. Description of the study area

The studywas conducted onfive drained peatland forest catchmentslocated in Sotkamo, eastern Finland (Table 1). Tree stands were clear-cut in winter 2009, either as conventional stem-only harvest (SOH) orwhole-tree harvest followed by stump lifting (WTH). Logging residuesand stumpswere exported from the sites andmoundingwas conductedin summer 2010. Two catchmentswere left unharvested (controls). TheSOH site was relatively infertile (S24) with a low-volume tree stand,while the WTH sites (ML07, ML10) had higher production rates due tohigher site fertility.

In 2007, the study plots were established and the tree stands mea-sured. A KPL model was used in tree measurement analyses(Heinonen, 1994) and a stand-level decision support tool, MOTTI(Salminen et al., 2005), to assess the biomass of the stand and carbon di-oxide emissions from vehicles used during harvesting and transport op-erations. In MOTTI, the biomass of an individual tree was based on thefunctions of Repola et al. (2007) and Repola (2009).

3.2. Hydrochemical measurements and modeling

Discharge water samples for total nitrogen (N) and phosphorus (P)analysis were collected biweekly from the outlet of a ditch (upstreamof a gauge installed in 2007) from 2008 until 2010 during snow-free pe-riods, with 2008 representing a calibration year.Mercury (Hg) samplingwas less frequent, ranging from 4 to 12 times per year. Total Nwasmea-sured using a flow-injection analyzer and total P colorimetrically after

Table 1Characteristics of the catchments.

Catchment field code ML07 ML10 S24 KV13 ML09

Harvesting regime WTHa WTHa SOHb Control Control

Relative fertility level High High Low Low High

Catchment size, ha 0.7 0.5 1.6 2.9 1.1Stem number, n/ha 1541 1208 1147 1237 1113Scots pine, % 13 15 57 54 48Norway spruce, % 47 69 22 29 30Hardwood, % 40 16 21 17 22Basal area, m2/ha 23.4 24.2 12.5 12.4 12.1Diameter, cm 12.22 15.21 10.72 10.41 10.47Height, m 11.41 12.86 8.8 9.22 8.8Sawlog and pulpwood,m3/ha

154.9 158.4 59.7 59.4 59.6

Logging residues, m3/ha 74.68 74.0 40.7Average age of the trees,years

70 70 50 60 68

Thickness of peatlayer, cm

90–160 100–160 100–150 100–150 120–200

Site type Mtkgc Mtkgc Ptkgc Ptkgc Mtkgc/Vatkgc

a Whole-tree harvesting with stump removal.b Stem-only harvesting.c Peatland forest types according to Laine and Vasander (2005): Mtkg = Vaccinium

myrtillus drained forest type; Ptkg = Vaccinium vitis-idaea drained forest type; Vatkg =dwarf-shrub drained forest type.

Please cite this article as: Miettinen, J., et al., Whole-tree harvesting with squality, biodiversity conservation and climate change mitigation matter, F

sulfide digestion using spectrophotometry. Total Hg was determinedusing cold vapor atomic fluorescence spectrometry.

The hydrological model FEMMA (Koivusalo et al., 2008) was used toestimate the daily runoff from each catchment. FEMMA calculateswaterand element fluxes along the typical water flow path. The annual ele-ment export was computed using the simulated daily runoff and themeasured P, N, and Hg concentrations. The missing daily concentrationvalues between the sampling dateswere interpolated, and the obtainedconcentration was multiplied by the simulated daily runoff to calculatethe daily export load. The daily export loads were further aggregated toannual export loads.

The treatment effect indicates howmuch the treatment changes theexport load with respect to the background load5 in terms of the differ-ence between the background and observed loads in the treatmentcatchment (Scott, 1999;Watson et al., 2001). The pre-treatment datasetdefines the relationship between thedata series of the control and treat-ment catchments (Grip, 1982; Neal, 2002; Nieminen, 2004). The treat-ment effect is defined as:

Bi ¼ Ei−CiEpCp

ð14Þ

where Βi is the treatment effect in year i, Ei is the export load from thetreatment catchment in year i, and Ep is the export load from the treat-ment catchment in the pre-treatment year, Ci is the export load from thecontrol catchment in year i, and Cp is the export load from the controlcatchment in the pre-treatment year.

In order to obtain treatment effect values for the two harvesting re-gimes at sites of both fertility levels, we assumed that the effect of theharvesting regime would be relatively similar, irrespective of site fertil-ity. Hence, we calculated the treatment effect of whole-tree harvestingwith stump removal at the low-fertility site on the basis of empirical re-sults obtained from the whole-tree harvest (average of ML07 andML10) and control (ML09) sites of high fertility. Respectively, the treat-ment effects of stem-only harvesting at the high-fertility site were cal-culated based on empirical results from the stem-only harvest (S24)and control (KV13) sites of low fertility.6 The treatment effect of stem-only harvesting on the high-fertility site and that of whole-treeharvesting on the low-fertility site is presented in Appendix B.

Element loads for the post-empirical years 2011 to 2018 have beenestimated as follows. For N, the treatment effect was assumed to remainat the same level for three years after harvesting, and thereafter the loadwas assumed to diminish by half each year. In the case of P, the mea-sured load increased only during the first year after harvesting and al-ready decreased in the second year. The P load for the following eightyears was assumed to decrease at the same rate as it had decreased be-tween the second and third years. The treatment effect of Hg indicatedan increase for the measurement two years after harvesting, but we as-sumed that the Hg load would start to decrease by 20% thereafter.

Although no ditch network maintenance was conducted in thesecatchments, we added an estimated effect of ditch network mainte-nance on the export load based on previous studies (Finér et al.,2010). We assumed that ditch network maintenance had no effect onthe N load, but the P treatment effect increased four-fold in the firstyear after ditch networkmaintenance, and in the following years the ef-fect was equal to the harvesting effect. Due to the lack of empirical data,

5 Because hydrological years can be very different, it is not enough to compare exportloads before and after treatment, but we have to estimate how much the export loadwould be from the treatment catchment under the assumption that the treatment hadnot occurred (“background load”).

6 As in Section 4, we estimated the social costs and benefits of stem-only harvesting andwhole-tree harvesting with stump removal in both of these representative catchments.We denote the low-fertility site (S24) as catchment 1 and the average of the high-fertility sites (ML07 and ML10) as catchment 2.

tump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005

Table 2Parameter values used in the empirical analysis.

Descriptions Unit Value Source

BenefitsTimber pricesScots pine FFRI (2010)a

Sawlogs €/m3 43.97Pulpwood €/m3 13.79

Norway spruce FFRI (2010)a

Sawlogs €/m3 44.81Pulpwood €/m3 17.79

Hardwood FFRI (2010)a

Sawlogs €/m3 38.01Pulpwood €/m3 13.30

Logging residues €/m3 2 Huttunen (2012)Stumps €/m3 1 Huttunen (2012)

Bare land value €/ha 106.00 Paananen (2007)Costs

Ditch network maintenance cost €/ha 157 FFRI (2010)a

Mounding cost €/ha 296 FFRI (2010)a

Seeding cost €/ha 200 FFRI (2010)a

Nutrient load damage €/kg N 8.69 Gren (2001)Mercury load damage €/kg Hg 23 077 Hylander and Goodsite

(2006)Marginal damage of carbon dioxideemissions

€/tCO2 21.27 Tol (2005)

a Finnish Forest Research Institute.

6 J. Miettinen et al. / Forest Policy and Economics xxx (2013) xxx–xxx

we excluded Hg from the calculations. The treatment effect ofharvesting was summed up with the treatment effect of ditch networkmaintenance and presented as a nutrient damage load.

8 The default emission factor is 94.6 tCO2/TJ for coal and 109.6 tCO2/TJ for forest fuel-wood (Statistics Finland, 2011).

9 We are not able to optimize the rotation age, as we have no growth data or forestgrowth function from the catchments studied. The estimated amounts of harvested saw-

4. Economic data and estimation of social costs and benefits

The parameter values of costs and benefits used in the determina-tion of social welfare under both regimes are provided in Table 2.7 Thereal interest rate used here was 2.5%.

In Table 2, for both harvesting regimes the value of the harvestedtimber and the bare land value are included as benefits. The value ofharvested sawlogs and pulpwood was calculated using average stump-age prices from 2009 (Finnish Forest Research Institute, 2010). Inwhole-tree harvesting with stump removal, in addition to sawlogs andpulpwood, harvest revenues from logging residues and stumps arealso included in the calculation of net social benefits.

The effect of using logging residues and stumps to replace fossil fuel(coal) in energy production is included in the analysis for whole-treeharvesting with stump removal. Because in the carbon neutralbioenergy policy it is assumed that logging residues and stumps donot produce any carbon dioxide emissions, the use of logging residuesand stumps provides substitution benefits as a substitute for coal. Fur-thermore, the avoided site preparation costs could be considered asbenefits in whole-tree harvesting with stump removal. However, inthis study, mounding was also carried out for whole-tree harvestingareas.

The costs of both harvesting regimes in Table 2 include the nutrientand mercury load damage, regeneration costs and climate costs fromcarbon dioxide emitted by harvesting vehicles. The nutrient load in-cludes total nitrogen and phosphorus. Forest regeneration costs includeditch network maintenance, mounding and seeding. Carbon dioxideemissions from vehicles in harvesting and transportation were estimat-ed using MOTTI (Salminen et al., 2005).

For the carbon non-neutral bioenergy policy, the value of the lostwoody biomass as a carbon sink is included as a cost for society. Basedon Pingoud and Perälä (2000), we assumed that 44% of sawlogs would

7 A detailed description of the data is available from the authors upon request.

Please cite this article as: Miettinen, J., et al., Whole-tree harvesting withquality, biodiversity conservation and climate change mitigation matter, F

be used as end products. The remaining sawlogs (56%) and all of thepulpwood were allocated to lost woody biomass as a carbon sink. Thecarbon content of sawlogs and pulpwood was assumed to be 50% ofdry wood. For whole-tree harvesting with stump removal, the value ofthe lost woody biomass as a carbon sink was estimated in similar wayas for stem-only harvesting, except that the carbon content of loggingresidues and stumps and roots was also included in the estimate ofthe lost woody biomass. Finally, because the carbon dioxide emissionsare higher for energy wood than for coal,8 using logging residues andstumps as a substitute for coal produces costs to society, as we are con-sidering the carbon non-neutral bioenergy policy.

5. Empirical analysis: net social benefits of the harvesting regimes

We are finally in a position to evaluate the net social benefits of thetwo harvesting regimes under the carbon neutral bioenergy policy andunder the carbon non-neutral bioenergy policy. The net social benefitsof the twoharvesting regime are compared for an exogenously given ro-tation length.9 Draining costs and thinning revenues are assumed to besunk costs. We start by reporting the results under a carbon neutralbioenergy policy in Table 3.

Table 3 shows that under a carbon neutral bioenergy policy, whole-tree harvestingwith stump removal provides the highest net social ben-efits in both catchments. Whole-tree harvesting with stump removalproduces significantly high climate benefits, as it is assumed to replacefossil fuels. Note, however, that it also causes higher nutrient load dam-age than stem-only harvesting, but the difference in the climate benefitsofwhole-tree harvestingwith stump removal strongly dominates nutri-ent load damage.

We next report the results under a carbon non-neutral bioenergypolicy in Table 4.

Table 4 indicates that under a carbon non-neutral bioenergy policythe order is reversed: stem-only harvesting now provides highest netsocial benefits in both of the catchments, although they are negativein catchment 1 for stem-only harvesting and in both catchments forwhole-tree harvesting. The benefits and costs of stem-only harvestingare the same as under a carbon neutral bioenergy policy, except thatnow for both harvesting regimes the lost woody biomass as a carbonsink is included in costs. In whole-tree harvesting with stump removal,the benefits are the same as under a carbon neutral bioenergy policy,except that the substitution of fossil fuels is not a benefit but a costitem. The most significant difference between harvesting regimescomes from the lost woody biomass as a carbon sink being higher inwhole-tree harvestingwith stump removal. However, as under a carbonneutral bioenergy policy, here the difference in nutrient load damage isalso significant, being higher in whole-tree harvesting with stumpremoval.

The reason for the negative net social benefits (except for stem-onlyharvesting in catchment 2) under a carbon non-neutral bioenergy poli-cy lies in the high climate costs, especially due to lost woody biomass asa carbon sink. Furthermore, the harvest revenue is quite low, as the siteis peatland,where the site productivity is lower than in upland stands. Ifwe omit the water quality and climate impacts and consider only theharvest revenue, bare land value, and silvicultural and forest improve-ment costs,wefind that all values are positive, although net harvest rev-enues are quite low (in catchment 1, €723/ha for stem-only harvestingand €795/ha for whole-tree harvesting with stump removal, and incatchment 2, €3681/ha for stem-only harvesting and €3808/ha forwhole-tree harvesting with stump removal).

logs and pulpwood, logging residues and stumps are based on tree stand measurementsfrom 2007.

stump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005

Table 3Net social benefits of the harvesting regimes evaluated at the time of harvest under a carbon neutral bioenergy policy, (€/ha).

Stem-only harvesting Whole-tree harvesting with stump removal

Catchment 1a Catchment 2b Catchment 1a Catchment 2b

BenefitsHarvest revenue

Sawlogs and pulpwood 1253.54 4211.55 1253.54 4211.55Logging residues – – 62.30 104.36Stumps – – 9.55 22.20

Bare land value 106.00 106.00 106.00 106.00Climate impacts

Substitution of fossil fuels – – 599.45 1105.36Total 1359.54 4317.55 2030.84 5549.47Costs

Ditch network maintenance cost 157.00 157.00 157.00 157.00Mounding cost 288.78 288.78 288.78 288.78Seeding cost 190.36 190.36 190.36 190.36

Water quality impactsNutrient load damage 175.56 239.95 457.58 629.76Mercury load damage 2.04 2.21 3.71 3.99

Climate impactsCO2 emitted by harvesting vehicles 25.01 46.49 33.64 58.45

Total 838.76 924.78 1131.08 1328.33Net social benefits 520.79 3392.77 899.76 4221.14

a Catchment 1 is the low-fertility site.b Catchment 2 is the high-fertility site.

Table 4Net social benefits of the harvesting regimes evaluated at the time of harvest under a carbon non-neutral bioenergy policy, (€/ha).

Stem-only harvesting Whole-tree harvesting with stump removal

Catchment 1a Catchment 2b Catchment 1a Catchment 2b

BenefitsHarvest revenue

Sawlogs and pulpwood 1253.54 4211.55 1253.54 4211.55Logging residues – – 62.30 104.36Stumps – – 9.55 22.20

Bare land value 106.00 106.00 106.00 106.00Total 1359.54 4317.55 1431.39 4444.11Costs

Ditch network maintenance cost 157.00 157.00 157.00 157.00Mounding cost 288.78 288.78 288.78 288.78Seeding cost 190.36 190.36 190.36 190.36

Water quality impactsNutrient load damage 175.56 239.95 457.58 629.76Mercury load damage 2.04 2.21 3.71 3.99

Climate impactsCO2 emitted by harvesting vehicles 25.01 46.49 33.64 58.45Substitution of fossil fuels – – 95.05 175.27Lost woody biomass as a carbon sink 819.37 1973.75 1865.23 4281.62

Total 1658.13 2898.53 3091.35 5785.23Net social benefits −298.59 1419.03 −1659.96 −1341.12

a Catchment 1 is the low-fertility site.b Catchment 2 is the high-fertility site.

10 We also conducted conventional partial sensitivity analysis for both carbon neutraland carbon non-neutral bioenergy policies by allowing values to vary 25% below or abovethe baseline estimates. The results of partial sensitivity analysis for marginal climate dam-age, the price of logging residues and stumps, and for nutrient and mercury load damageare reported in Appendix C in Table C.1. The partial sensitivity analysiswas also conductedfor the real interest rate, timber prices, seeding cost, mounding cost, and ditch networkmaintenance cost and is available from the authors upon request. However, analysisshowed that there were no changes in the ranking order of the harvesting regimes.

7J. Miettinen et al. / Forest Policy and Economics xxx (2013) xxx–xxx

Our empirical analysis was based on an exogenously given rotationage due to the lack of growth data. Changing the rotation agewould nat-urally impact on the costs and benefits for both harvesting regimes.Under a carbon neutral bioenergy policy, the change in rotation agewould mostly change the harvest revenues for both harvesting regimesand benefits from substitution of fossil fuels for whole-tree harvesting,but have a lower effect on the other costs and benefits. Therefore, weare tempted to argue that the ranking of the harvesting methodswould not change. Under a carbon non-neutral bioenergy policy, thechange in the rotation age would mostly change the harvest revenuesand the costs from lost woody biomass as a carbon sink for bothharvesting regimes and costs from substitution of fossil fuels forwhole-tree harvesting. Given the high value of especially the lostwoody biomass as a carbon sink, it is possible that the ranking of theharvesting methods would change.

Please cite this article as: Miettinen, J., et al., Whole-tree harvesting with squality, biodiversity conservation and climate change mitigation matter, F

6. Sensitivity analysis

Wenext examinehow sensitive our results are to the chosen param-eter values. We applied Monte Carlo simulations using probabilitydistributions of the key parameters in order to determine how the si-multaneous change of parameters impacts on the ranking of harvestingregimes.10 We conducted Monte Carlo sensitivity analyses using the

tump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005

Table 5Net social benefits under carbon neutral bioenergy policy: Monte Carlo simulation.

Minimum Maximum Mean welfare (ranking) Welfare in Table 3 (ranking)

Stem-only harvestingCatchment 1a −142.93 673.50 328.40 (2) 520.79 (2)Catchment 2b 2693.86 3581.91 3195.96 (2) 3392.77 (2)

Whole-tree harvesting with stump removalCatchment 1a −167.86 2047.49 885.90 (1) 899.76 (1)Catchment 2b 2716.54 6215.92 4363.78 (1) 4221.14 (1)

a Catchment 1 site is the low-fertility site.b Catchment 2 is the high-fertility site.

Table 6Net social benefits under carbon non-neutral bioenergy policy: Monte Carlo-simulation.

Minimum Maximum Mean welfare (ranking) Welfare in Table 4 (ranking)

Stem-only harvestingCatchment 1a −2295.23 433.52 −743.31 (1) −298.59 (1)Catchment 2b −2600.63 3103.99 614.36 (1) 1419.03 (1)

Whole-tree harvesting with stump removalCatchment 1a −5716.15 78.82 −2462.13 (2) −1659.96 (2)Catchment 2b −9740.79 2527.04 −2911.48 (2) −1341.12 (2)

a Catchment 1 site is the low-fertility site.b Catchment 2 is the high-fertility site.

8 J. Miettinen et al. / Forest Policy and Economics xxx (2013) xxx–xxx

program @risk 5.7. For simulation, we determined the number of itera-tions as 100,000 and the probability distribution was set as triangle. Forthe probability distributions, we set the lowest and highest possiblevalues for each parameter, mainly based on the previous literature(Table C.2 in Appendix C).

From Tables 5 and 6 we can see that even though the means of theMonte Carlo simulations differ from the results of our empirical analysis(especially under a carbon non-neutral bioenergy policy), the rankingorder of the harvesting regimes does not change, despite the fact thatthe values of marginal climate damage, stand establishment costs,ditch network maintenance cost, nutrient and mercury load damageand real interest rate simultaneously change.

All in all, both our baseline analysis and sensitivity analysis showvery robust results. The net social benefits of stem-only harvesting ver-sus whole-tree harvesting with stump removal predominantly dependon how society considers carbon dioxide emissions from wood.

7. Conclusions

We examined how stem-only harvesting and whole-tree harvestingwith stump removal perform in terms of the net social benefits they pro-vide when various ecosystem services including water quality, biodiver-sity conservation and climate change mitigation are accounted for. Toreflect recent discussion on the use of wood biomass, we postulatedtwo alternatives of social preferences towards the climate: a carbon neu-tral bioenergy policy and a carbon non-neutral bioenergy policy. The for-mer alternative is justified by the fact that the carbon dioxide emissionsreleased into the atmosphere when harvested biomass is converted toenergy will be taken up again by the growth of the new tree generation.However, this carbon neutrality of forest bioenergy depends on the tem-poral dimension and indirect carbon dioxide emissions.

We extended the Faustmann rotationmodel to include nutrient loaddamage, biodiversity benefits and climate impacts and examined howthese ecosystem services impacted jointly and separately on the rota-tion age. The empirical application was based on Finnish data, includingnitrogen, phosphorus and mercury loads from a catchment experimentcarried out on drained peatland. In the empirical analysis we assessedfor a given rotation length which harvesting regime produces highernet social benefits under the alternative climate policy options.

Please cite this article as: Miettinen, J., et al., Whole-tree harvesting withquality, biodiversity conservation and climate change mitigation matter, F

Our empirical results showed that under a carbon neutral bioenergypolicy, whole-tree harvestingwith stump removal has the highest net so-cial benefits. Meanwhile, if a carbon non-neutral bioenergy policy is as-sumed, stem-only harvesting has the highest net social benefits.However, in this latter case the net social benefits were negative (exceptfor one catchment). This provides two important lessons. First, societal at-titudes towards the temporal dimensions of released emissions from theuse of wood biomass do indeed matter. Second, the returns to forestry inpeatland may be low depending on the soil fertility.

The empirical results of this study were dependent on the data avail-able, andhere the empirical analysiswould clearly benefit fromadditionaldata. First, we were unable to endogenously solve the optimal rotationage due to the lack of data on forest growth. In Finland, there is notenough experience of forestry in peatlands concerning the second andfurther generations of trees. Second, measurements of nutrient and mer-cury loads over longer time periods would be needed to obtain more re-liable estimates of nutrient load damage caused by the two alternativeharvesting regimes. Third, ecological data on the effects of the two alter-native harvesting regimes on biodiversity and environmental valuationestimates of biodiversity benefits would make it possible to also includebiodiversity impacts in the empirical analysis of alternative harvestingregimes.

Acknowledgments

This study was funded by two Academy of Finland projects: Non-Point Source Pollution Economics (NOPEC, project no. 115364) andHydro-biogeochemistry of drained peatlands: impacts of bioenergyharvesting on trace metal transport under different hydrogeologicalsettings (HYPE) (project no. 132447). This studywas also part of the re-search project Stem-only vs. whole-tree harvesting — the effects ofharvesting in forested peatlands on the leaching of nutrients andheavymetals (subproject no. 347702), funded by the Finnish Forest Re-search Institute. In addition, this work was supported by the FinnishCultural Foundation, Kyösti Haataja Foundation, Alfred Kordelin Foun-dation, Yrjö Jahnsson Foundation and Niemi Foundation. Finally, wethank two anonymous referees for constructive comments. We wouldalso like to thank Chiara Lombardini for her advice and assistancewith the Monte Carlo simulations.

stump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005

Table C.1Partial sensitivity analysis under carbon neutral bioenergy policy and carbon non-neutralbioenergy policy:marginal climate damage, energywood prices andnutrient andmercuryload damage. Net social benefits and the ranking order for stem-only harvesting andwhole-tree harvesting with stump removal.

Stem-only harvesting Whole-tree harvestingwith stump removal

Catchment1a

Catchment2b

Catchment1a

Catchment2b

Carbon neutral bioenergy policyBaseline (ranking) 520.79 (2) 3392.77 (2) 899.76 (1) 4221.14 (1)Marginal climate damage (€/tCO2)26.59 (+25%) 514.53 (2) 3381.14 (2) 1041.28 (1) 4482.99 (1)15.95 (−25%) 527.04 (2) 3404.40 (2) 758.24 (1) 3959.28 (1)

Price of logging residues (€/m3)2.5 (+25%) 520.79 (2) 3392.77 (2) 915.33 (1) 4247.23 (1)1.5 (−25%) 520.79 (2) 3392.77 (2) 884.18 (1) 4195.05 (1)

Price of stumps (€/m3)1.25 (+25%) 520.79 (2) 3392.77 (2) 902.15 (1) 4226.69 (1)0.75 (−25%) 520.79 (2) 3392.77 (2) 897.37 (1) 4215.59 (1)

Nutrient load damage (€/kg N)10.86 (+25%) 476.95 (2) 3332.85 (2) 785.50 (1) 4063.88 (1)6.52 (−25%) 564.63 (2) 3452.69 (2) 1014.02 (1) 4378.39 (1)

Mercury load damage (€ kg−1 Hg)28 846 (+25%) 520.28 (2) 3392.22 (2) 898.83 (1) 4220.14 (1)17 308 (−25%) 521.30 (2) 3393.32 (2) 900.69 (1) 4222.13 (1)

Carbon non−neutral bioenergy policyBaseline (ranking) −298.59 (1) 1419.03 (1) −1659.96 (2) −1341.12 (2)Marginal climate damage (€/tCO2)26.59 (+25%) −509.78 (1) 913.73 (1) −2158.68 (2) −2470.49 (2)15.95 (−25%) −87.39 (1) 1924.32 (1) −1161.25 (2) −211.75 (2)

Price of logging residues (€/m3)2.5 (+25%) −298.59 (1) 1419.03 (1) −1644.39 (2) −1315.03 (2)1.5 (−25%) −298.59 (1) 1419.03 (1) −1675.54 (2) −1367. 21(2)

Price of stumps (€/m3)1.25 (+25%) −298.59 (1) 1419.03 (1) −1657.58 (2) −1335.57 (2)0.75 (−25%) −298.59 (1) 1419.03 (1) −1662.35 (2) −1346.67 (2)

Nutrient load damage (€/kg N)10.86 (+25%) −342.43 (1) 1359.11 (1) −1774.23 (2) −1498.38 (2)6.52 (−25%) −254.75 (1) 1478.94 (1) −1545.70 (2) −1183.86 (2)

Mercury load damage (€ kg−1 Hg)28 846.25 (+25%) −299.10 (1) 1418.47 (1) −1660.89 (2) −1342.12 (2)17 307.75 (−25%) −298.08 (1) 1419.58 (1) −1659.04 (2) −1340.12 (2)

a Catchment 1 site is the low-fertility site.b Catchment 2 is the high-fertility site.

9J. Miettinen et al. / Forest Policy and Economics xxx (2013) xxx–xxx

Appendix A. Comparative statics of rotation age under a carbonneutral policy (Section 2.1)

The second-order condition is:

WiTiTi ¼ 2P0 Ti

� �Q 0 Ti� �

þ P00 Ti� �

Q Ti� �

þ P Ti� �

Q 00 Ti� �

− r P0 Ti� �

Q Ti� �

þ P Ti� �

Q 0 Ti� �h i

þ Fi 0 Ti� �

þ Ki 0 0−rKi 0� �b 0:

ðA:1Þ

It must be negative at the optimal interior solution.For both harvesting regimes we have:

∂T∂c ¼ − Wi

Tic

WiTiTi

N0; where WiTic ¼ r 1−e−rTi

� �−1N0: ðA:2Þ

∂T∂pl

¼ −Wi

Tipl

WiTiTi

¼ ?;

where WiTipl

¼ α0l Ti� �

Q Ti� �

þ αl Ti� �

Q 0 Ti� �

− rαl Ti� �

Q Ti� �

1−e−rTi� �−1

:

ðA:3Þ

Similarly, for pp and ph.

∂T∂r ¼ − Wi

Tir

WiTiTi

¼ ?; where WiTir ¼ −P Ti

� �Q Ti� �

þ D zi� �

þ rDzzr−Ki−Wi−rWir: ðA:4Þ

∂T∂d ¼ − Wi

Tid

WiTiTi

N 0; where WiTid ¼ r

Zki0

g sð Þe−rsds 1−e−rTi� �−1

N 0:

ðA:5Þ

The effect of climate damage differs under the two regimes.Under stem-only harvesting:

∂T∂ε ¼ − Wi

Tiε

WiTiTi

N0; where WiTiε ¼ rγi 1−e−rTi

� �−1N0: ðA:6Þ

Under whole-tree harvesting with stump removal:

∂T∂ε ¼ − Wi

Tiε

WiTiTi

¼ ?; where

WiTiε ¼ − r bcφαh Ti

� �Q Ti� �

−γ−1h i

1−e−rTi� �−1

þ bcφ α0h Ti� �

Q Ti� �

þ αh Ti� �

Q 0 Ti� �h i

:

ðA:7Þ

Appendix B. The treatment effect of stem-only harvesting onthe high-fertility site and the treatment effect of whole-treeharvesting on the low-fertility site for hydrochemical modelingin Section (3.2)

The proportion of the treatment effect from the hypothetical exportload in the case where no treatment would have been carried out is cal-culated for stem-only harvesting at the low-fertility site. This proportionis used together with the hypothetical export load for no treatment atthe high-fertility site to obtain a value for the treatment effect ofstem-only harvesting at the high-fertility site according to the following

Please cite this article as: Miettinen, J., et al., Whole-tree harvesting with squality, biodiversity conservation and climate change mitigation matter, F

formula (treatment effect of stem-only harvest on the high-fertilitysite):

Bi high−fertility ¼Bi low−fertility

Ci low−fertility � Ep low−fertility

Cp low−fertility

!266664

377775

� Ci high−fertility � Ep high−fertility

Ci high−fertility

!: ðB:1Þ

Theproportion of thewhole-tree harvest treatment effectwas calcu-lated from the hypothetical export load with no treatment at the high-fertility site. This proportion was used together with the hypotheticalexport load with no treatment at the low-fertility site as follows (treat-ment effect of whole-tree harvest on the low-fertility site):

Bi low−fertility ¼Bi high−fertility

Ci high−fertility � Ep high−fertility

Cp high−fertility

!266664

377775

� Ci low−fertility � Ep low−fertility

Ci low−fertility

!" #: ðB:2Þ

Appendix C

tump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005

Table C.2Lowest and highest values used in the probability distributions in the Monte Carlosimulations.

Lowest value Highest value Values used inthe analysis

Marginal climatedamage (€/tCO2)

3.72Tol (2011)

58.58Tol (2011)

21.27Tol (2005)

Stand establishmentcost (€/ha)

400FFRIa (2010)

1063FFRIa (2010)

496FFRIa (2010)

Ditch networkmaintenancecost (€/ha)

120Ahtikoski et al. (2007)

240Ahtikoski et al.(2007)

157FFRIa (2010)

Nutrient load damage(€/kg N)

3.57Ollikainenand Lankoski (2009)

11.7Hyytiäinen andOllikainen (2012)

8.69Gren (2001)

Mercury load damage(kg−1 Hg)

2769Hylander andGoodsite (2006)

1 015 385Hylander andGoodsite (2006)

23 077Hylander andGoodsite (2006)

Real interest rate (%) 1 3 2.5

a Finnish Forest Research Institute.

10 J. Miettinen et al. / Forest Policy and Economics xxx (2013) xxx–xxx

References

Ahtiainen, M., Huttunen, P., 1999. Long-term effects of forestry managements on waterquality and loading in brooks. Boreal Environ. Res. 4, 101–114.

Ahtikoski, et al., 2007. Turvemaiden metsien käsittely– ja hoito. Taustaraportti. (In Finnish).Alavalapati, J., Lal, P., 2013. Economics of wood-based biofuels. In: Kant, S., Alavalapati, J.

(Eds.), Handbook of Forest Resource Economics. Routledge, Taylor & Francis Group.(in press).

Amacher, G.S., Ollikainen, M., Koskela, E., 2009. Economics of Forest Resources. MIT Press,Cambridge, MA.

Bengtsson, J., Wikström, F., 1993. Effects of whole-tree harvesting on the amount of soilcarbon: model results. N. Z. J. For. Sci. 23, 380–389.

Bouget, C., Lassauce, A., Jonsell, M., 2012. Effects of fuelwood harvesting on biodiversity—a review focused on the situation in Europe. Can. J. For. Res. 42, 1421–1432.

Boyland, M., 2006. The economics of using forests to increase carbon storage. Can. J. For.Res. 36, 2223–2234.

Clinnick, P.F., 1985. Buffer strip management in forest operations: a review. Aust. For. 48(1), 34–45.

Creedy, J., Wurzbacher, A.D., 2001. The economic value of a forested catchment with tim-ber, water and carbon sequestration benefits. Ecol. Econ. 38, 71–83.

Egnell, G., Valinger, E., 2003. Survival, growth, and growth allocation of planted Scots pinetrees after different levels of biomass removal in clear-felling. For. Ecol. Manag. 177,65–74.

Eriksson, L.O., Löfgren, S., Öhman, K., 2011. Implications for forest management of the EUWater Framework Directive's stream water quality requirements — a modeling ap-proach. For. Policy Econ. 13, 284–291.

Faustmann, M., 1849. Calculation of the value which forest land and immature standspossesses. Reprinted in J. For. Econ. 1995 (1), 89–114.

Finér, L., Mattsson, T., Joensuu, S., Koivusalo, H., Laurén, A., Makkonen, T., Nieminen, M.,Tattari, S., Ahti, E., Kortelainen, P., Koskiaho, J., Leinonen, A., Nevalainen, R.,Piirainen, S., Saarelainen, J., Sarkkola, S., Vuollekoski, M., 2010. Metsäisten valuma-alueiden vesistökuormituksen laskenta. [A method for calculating nitrogen,phosphorus and sediment load from forest catchments]. Suomen Ympäristö, 10(In Finnish).

Finnish Forest Research Institute (FFRI), 2010. Finnish Statistical Yearbook of Forestry.Vammalan kirjapaino Oy, Sastamala.

Fisher, B., Turner, R.K., Morling, P., 2009. Defining and classifying ecosystem services fordecision making. Ecol. Econ. 68, 643–653.

Gan, J., Smith, C.T., 2006. A comparative analysis of woody biomass and coal for electricitygeneration under various CO2 emission reductions and taxes. Biomass Bioenergy 30,296–303.

Gan, J., Smith, C.T., 2007. Co-benefits of utilizing logging residues for bioenergy production:the case for East Texas, USA. Biomass Bioenergy 31, 623–630.

Gren, I.-M., 2001. International versus national actions against nitrogen pollution of theBaltic Sea. Environ. Resour. Econ. 20 (1), 41–59.

Grip, H., 1982. Water chemistry and runoff in forest streams in Kloten. UNGI Report 58.Division of Hydrology, Department of Physical Geography, Uppsala University.

Hartman, R., 1976. The harvesting decision when a standing forest has value. Econ. Inq.14, 52–55.

Heinonen, J., 1994. Koealojen puu- ja puustotunnusten laskentaohjelma KPL. Käyttöohje.[KPL— a computer programme package for computing stand and single tree character-istics from sample plot measurements]. Metsäntutkimuslaitoksen tiedonantoja, 504 (InFinnish).

Huttunen, J., 2012. Metsänhoitoyhdistys Sotkamo. Personal communication on 9 April 2012.Hylander, L.D., Goodsite, M.E., 2006. Environmental costs of mercury pollution. Sci. Total.

Environ. 368, 352–370.Hyytiäinen, K., Ollikainen, M., 2012. Taloudellinen näkökulma Itämeren suojeluun.

Ympäristöministeriön raportteja 22/2012. http://www.ymparisto.fi/download.asp?contentid=139647&lan=fi (last accessed on 1 May 2013. In Finnish).

Please cite this article as: Miettinen, J., et al., Whole-tree harvesting withquality, biodiversity conservation and climate change mitigation matter, F

Jacobson, S., Kukkola, M., Mälkönen, E., Tveite, B., 2000. Impact of whole-tree harvestingand compensatory fertilization on growth of coniferous thinning stands. For. Ecol.Manag. 129, 41–51.

Jonsell, M., 2007. Effects on biodiversity of forest fuel extraction, governed by processesworking on a large scale. Biomass Bioenergy 31, 726–732.

Jonsell, M., Hansson, J., 2011. Logs and stumps in clearcuts support similar saproxylic bee-tle diversity: implications for bioenergy harvest. Silva Fennica 45 (5), 1053–1064.

Kaltschmitt, M., Reinhardt, G.A., Stelzer, T., 1997. Life cycle analysis of biofuels under dif-ferent environmental aspects. Biomass Bioenergy 12 (2), 121–134.

Koivusalo, H., Ahti, E., Laurén, A., Kokkonen, T., Karvonen, T., Nevalainen, R., Finér, L., 2008.Impacts of ditch cleaning on hydrological processes in a drained peatland forest.Hydrol. Earth Syst. Sci. 12 (5), 1211–1227.

Koskela, E., Ollikainen, M., Pukkala, T., 2007. Biodiversity conservation in commercial bo-real forestry: the optimal rotation age and retention tree volume. For. Sci. 53 (3),443–452.

Laine, J., Vasander, H., 2005. Suotyypit ja niiden tunnistaminen. Metsäkustannus Oy,Hämeenlinna (In Finnish).

Lattimore, B., Smith, C.T., Titus, B.D., Stupak, I., Egnell, G., 2009. Environmental factors inwoodfuel production: opportunities, risks, and criteria and indicators for sustainablepractices. Biomass Bioenergy 33, 1321–1342.

Laurén, A., Finér, L., Koivusalo, H., Kokkonen, T., Karvonen, T., Kellomäki, S., Mannerkoski,H., Ahtiainen, M., 2005. Water and nitrogen processes along a typical water flowpathand streamwater exports from a forested catchment and changes after clear-cutting:a modelling study. Hydrol. Earth Syst. Sci. 9 (6), 657–674.

Laurén, A., Koivusalo, H., Ahtikoski, A., Kokkonen, T., Finér, L., 2007. Water protection andbuffer zones: how much does it cost to reduce nitrogen load in a forest cutting?Scand. J. For. Res. 22, 537–544.

Lehtonen, A., Mäkipää, R., Heikkinen, J., Sievänen, R., Liski, J., 2004. Biomass expansion fac-tors (BEFs) for Scots pine, Norway spruce and birch according to stand age for borealforests. For. Ecol. Manag. 188, 211–224.

Mälkki, H., Virtanen, Y., 2003. Selected emissions and efficiencies of energy systems basedon logging and sawmill residues. Biomass Bioenergy 24, 321–327.

Mann, L.K., Johnson, D.W., West, D.C., Cole, D.W., Hornbeck, J.W., Martin, C.W., Riekerk, H.,Smith, C.T., Swank, W.T., Tritton, L.M., van Lear, D.H., 1988. Effects of whole-tree andstem-only clearcutting on postharvest hydrologic losses, nutrient capital, andregrowth. For. Sci. 34 (2), 412–428.

Matero, J., 1996. Costs of water pollution abatement in forestry. J. For. Econ. 2, 67–90.Matero, J., 2002. Economic analyses of diffuse load abatement in Finnish forestry. D.Sc.

thesis summary Research Notes 138. University of Joensuu, Faculty of Forestry.Matero, J., 2004. Cost-effective measures for diffuse load abatement in forestry. Silva

Fennica 38 (3), 333–345.Matero, J., Saastamoinen, O., 1998. Monetary assessment of the impacts of forestry on

water-based benefits in Finland. Boreal Environ. Res. 3, 87–96.Matta, J.R., Alavalapati, J.R.R., Stainback, G.A., 2009. Effect of conserving habitat for biodi-

versity on optimal management of non-industrial private forests in Florida. J. For.Econ. 15, 223–235.

Melin, Y., Petersson, H., Egnell, G., 2010. Assessing carbon balance trade-offs betweenbioenergy and carbon sequestration of stumps at varying time scales and harvest in-tensities. For. Ecol. Manag. 260, 536–542.

Merino, A., Balboa, M.A., Soalleiro, R.R., Conzalez, J.G.A., 2005. Nutrient exports under dif-ferent harvesting regimes in fast-growing forest plantations in southern Europe. For.Ecol. Manag. 207, 325–339.

Miettinen, J., Ollikainen, M., Finér, L., Koivusalo, H., Laurén, A., Valsta, L., 2012. Diffuse loadabatement with biodiversity co-benefits: the optimal rotation age and buffer zonesize. For. Sci. 58 (4), 342–352.

Millennium Ecosystem Assessment, 2005. Ecosystems and HumanWell-Being: Synthesis.Island Press, Washington, DC.

Miller, W.L., Everett, H.W., 1975. The economic impact of controlling nonpoint pollutionin hardwood forestland. Am. J. Agric. Econ. 57, 576–583.

Miranda, M.L., Hale, B., 2001. Protecting the forest from the trees: the social costs of ener-gy production in Sweden. Energy 26, 869–889.

Munthe, J., Hultberg, H., 2004. Mercury and methylmercury in runoff from a forestedcatchment — concentrations, fluxes and their response to manipulations. Water AirSoil Pollut. 4, 607–618.

Neal, C., 2002. Assessing environmental impacts on stream water quality: the use of cu-mulative flux and cumulative flux difference approaches to deforestation of theHafren Forest, mid-Wales. Hydrol. Earth Syst. Sci. 6 (3), 421–431.

Nieminen, M., 2004. Export of dissolved organic carbon, nitrogen and phosphorus follow-ing clear-cutting of three Norway spruce forests growing on drained peatlands. SilvaFennica 38 (2), 123–132.

Ollikainen, M., Lankoski, J., 2009. Multifunctionality: environment versus rural viability insocial optima. J. Rural. Dev. 32 (2), 31–57.

Olsson, B.A., Bengtsson, J., Lundkvist, H., 1996. Effects of different forest harvest intensitieson the pools of exchangeable cations in coniferous forest soils. For. Ecol. Manag. 84,134–147.

Olschewski, R., Benítez, P., 2010. Optimizing joint production of timber and carbon se-questration of afforestation projects. J. For. Econ. 16, 1–10.

Paananen, R., 2007. Summa-arvomenetelmän aputaulukot.Metsätalouden kehittämiskeskusTapio. (In Finnish).

Palosuo, T., Wihersaari, M., Liski, J., 2001. Net greenhouse gas emissions due to energy useof forest residues — impact of soil carbon balance. In: Pelkonen, P., Hakkila, P.,Karjalainen, T., Schlamadinger, B. (Eds.), Wood Biomass as an Energy Source —

Challenges in Europe. European Forest Institute, Joensuu, pp. 115–122 (EFI Proceed-ings no 39).

Pingoud, K., Perälä, A.-L., 2000. Arvioita puurakentamisen kasvihuonevaikutuksesta. VTTjulkaisuja 840 (In Finnish).

stump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005

11J. Miettinen et al. / Forest Policy and Economics xxx (2013) xxx–xxx

Porvari, P., Verta, M., Munthe, J., Haapanen, M., 2003. Forestry practices increase mercuryand methyl mercury output from boreal forest catchments. Environ. Sci. Technol. 37,2389–2393.

Repo, A., Tuomi, M., Liski, J., 2011. Indirect carbon dioxide emissions from producingbioenergy from forest harvest residues. Glob. Chang. Biol. Bioenergy 3, 107–115.

Repo, A., Känkänen, R., Tuovinen, J.-P., Antikainen, R., Tuomi, M., Vanhala, P., Liski, J., 2012.Bioenergy climate mitigation from forestry can be improved by allocating forest res-idue removal. Glob. Chang. Biol. Bioenergy 4, 202–212.

Repola, J., 2009. Biomass equations for Scots pine and Norway spruce in Finland. SilvaFennica 43 (4), 625–647.

Repola, J., Ojansuu, R., Kukkola, M., 2007. Biomass functions for Scots pine, Norway spruceand birch in Finland. Metlan työraportteja/Working Papers of the Finnish ForestResearch Institute 2007, 53.

Richards, K., Stokes, C., 2004. A review of forest carbon sequestration cost studies: a dozenyears of research. Clim. Chang. 63, 1–48.

Salminen, H., Lehtonen, M., Hynynen, J., 2005. Reusing legacy Fortran in the MOTTIgrowth and yield simulator. Comput. Electron. Agric. 49, 103–113.

Schlamadinger, B., Spitzer, J., Kohlmaier, G.H., Lüdeke, M., 1995. Carbon balance ofbioenergy from logging residues. Biomass Bioenergy 8, 221–234.

Scott, D., 1999. Managing riparian zone vegetation to sustain streamflow: results of pairedcatchment experiment in South Africa. Can. J. For. Res. 29, 1149–1157.

Searchinger, T.D., Hamburg, S.P., Melillo, J., Chameides, W., Havlik, P., Kammen, D.M.,Likens, G.E., Lubowski, R.N., Obersteiner, M., Oppenheimer, M., Robertson, G.P.,Schlesinger, W.H., Tilman, G.D., 2009. Fixing a critical climate accounting error. Sci-ence 326, 527–528.

Sedjo, R.A., Wisniewski, J., Sample, A.V., Kinsman, J.D., 1995. The economics of managingcarbon via forestry: assessment of existing studies. Environ. Resour. Econ. 6,139–165.

Sillanpää, P., Bilaletdin, Ä., Kaipainen, H., Frisk, T., Sallantaus, T., 2006. Metsätaloudenaiheuttaman kuormituksen laskentamenetelmä. The Finnish Environment, 817.

Sørensen, R., Meili, M., Lambertsson, L., von Brömssen, C., Bishop, K., 2009. The effects offorest harvest operations on mercury and methylmercury in two boreal streams: rel-atively small changes in the first two years prior to site preparation. Ambio 38 (7),364–372.

Please cite this article as: Miettinen, J., et al., Whole-tree harvesting with squality, biodiversity conservation and climate change mitigation matter, F

Statistics Finland, 2011. Fuel classification and emission coefficients 2011. http://tilastokeskus.fi/tup/khkinv/khkaasut_polttoaineluokitus.html (last accessed on 10March 2012).

Stupak, I., Asikainen, A., Jonsell, M., Karltun, E., Lunnan, A., Mizaraitė, D., Pasanen, K., Pärn,H., Raulund-Rasmussen, K., Röser, D., Schroeder, M., Varnagirytė, I., Vilkriste, L.,Callesen, I., Clarke, N., Gaitnieks, T., Ingerslev, M., Mandre, M., Ozolincius, R.,Saarsalmi, A., Armolaitis, K., Helmisaari, H.-S., Indriksons, A., Kairiukstis, L.,Katzensteiner, K., Kukkola, M., Ots, K., Ravn, H.P., Tamminen, P., 2007. Sustainableutilisation of forest biomass for energy — possibilities and problems: policy, legisla-tion, certification, and recommendations and guidelines in the Nordic, Baltic, andother European countries. Biomass Bioenergy 31, 666–684.

Sun, C., 2006. Welfare effects of forestry best management practices in the United States.Can. J. For. Res. 36, 1674–1683.

Tol, R., 2005. The marginal damage costs of carbon dioxide emissions: an assessment ofthe uncertainties. Energy policy 33, 2064–2074.

Tol, R., 2011. The social cost of carbon. Ann. Rev. Resour. Econ. 3, 419–443.Valente, C., Spinelli, R., Hillring, B.G., 2011. LCA of environmental and socio-economic im-

pacts related to wood energy production in alpine conditions: Valle di Fiemme(Italy). J. Clean. Prod. 19, 1931–1938.

van Kooten, G.C., Binkley, C.S., Delcourt, G., 1995. Effect of carbon taxes and subsidies onoptimal forest rotation age and supply of carbon services. Am. J. Agric. Econ. 77,365–374.

van Kooten, G.C., Eagle, A.J., Manley, J., Smolak, T., 2004. How costly are carbon offsets? Ameta-analysis of carbon forest sinks. Environ. Sci. Pol. 7, 239–251.

Walmsley, J.D., Godbold, D.L., 2010. Stump harvesting for bioenergy— a review of the en-vironmental impacts. Forestry 83 (1), 17–38.

Walmsley, J.D., Jones, D.L., Reynolds, B., Price, M.H., Healey, J.R., 2009. Whole treeharvesting can reduce second rotation forest productivity. For. Ecol. Manag. 257,1104–1111.

Watson, F., Vertessy, R., McMahon, T., Rhodes, B., Watson, I., 2001. Improved methods toassess water yield changes from paired-catchment studies: application to theMaroondah catchments. For. Ecol. Manag. 143, 189–204.

Wihersaari, M., 2005. Greenhouse gas emissions from final harvest fuel chip production inFinland. Biomass Bioenergy 28, 435–443.

tump removal versus stem-only harvesting in peatlands when wateror. Policy Econ. (2013), http://dx.doi.org/10.1016/j.forpol.2013.08.005