Copy igh © 2013 by he au ho (s). Published he e unde license by he Resilience Alliance.
Tol anen, A., A. Juu inen, and R. S en o. 2012. P e e ences o local people o he use o pea lands: he
case o he iches pea land egion in Finland. Ecology and Socie y 18(2): 19. h p://dx.doi.o g/10.5751/
ES-05496-180219
Resea ch, pa o a Special Fea u e on Ecological Res o a ion in No he n Regions
P e e ences o Local People o he Use o Pea lands: he Case o he
Riches Pea land Region in Finland
Anne Tol anen 1,2, A i Juu inen 1,2,3 and Rauli S en o 3
ABSTRACT. We analyze he po en ial o socioeconomically sus ainable pea land use by in es iga ing con lic ing in e es s,
e ealing ade-o s ha people a e willing o accep , and s udying whe he opinions a e dependen on socioeconomic and
demog aphic ac o s. Opinions owa d i e o ms o pea land use and se en pea land ecosys em se ices we e su eyed in
No he n Os obo hnia in no he n Finland in 2011. Choice expe imen (CE) was used o e eal ade-o s in land use p e e ences,
and g oups o esponden s we e iden i ied using he la en class model (LCM). We iden i ied h ee classes o esponden s in
which en i onmen alis s showed a high p e e ence owa d he cessa ion o pea p oduc ion and inc ease o pea land es o a ion,
he p oduc ion-o ien ed class p e e ed an inc ease in imbe and pea p oduc ion a eas, and he cu en use suppo e s ag eed
on he p esen land use policy. Howe e , all esponden classes ag eed on he inc ease o na u e p o ec ion and he p esen le el
o imbe p oduc ion and disag eed on he cessa ion o es o a ion. The CE e ealed ha en i onmen ally minded people who
a e likely o conside he indi ec use alues and exis ence alues impo an a e less willing o make ade-o s be ween ecosys em
se ices han hose who emphasize di ec use alues. Because pea land es o a ion occu s in comme cially unp oduc i e
pea lands, i imp o es bo h he di ec use and exis ence alues wi hou educing p o isioning se ices o pea lands. The e o e,
es o a ion is commonly accep ed by he public, in con as o managemen op ions ha in ol e clea ade-o s be ween
ecosys em se ices. We conclude ha he unde s anding o p e e ences and ade-o s can enhance sus ainable land use planning.
I may be un ealis ic, howe e , o expec a solu ion ha all in e es g oups would comple ely accep .
Key Wo ds: choice expe imen ; con lic managemen ; ecosys em se ices; land use managemen
INTRODUCTION
The ecosys em se ice concep and he alua ion e o s o
ecosys em se ices a e changing he discussion conce ning
land use policy and na u al esou ce managemen (Cos anza
e al. 1997, de G oo e al. 2010). Inc easing knowledge on he
impo ance o heal hy ecosys ems o human well-being is
leading o si ua ions in which mul i unc ional and ecologically
sus ainable land use is seen as an op ion o simul aneously
gene a e ecological, social, and economic bene i s (e.g.,
Balm o d e al. 2002, de G oo 2006). Valua ion o ecosys em
se ices has been done using se e al me hods (e.g., Boye and
Polasky 2004, B ande e al. 2006), whe eby he mos ecen ly
de eloped echnique is he Choice Expe imen (CE) me hod.
Besides placing a di ec mone a y alue on ecosys em
se ices, CE can be u ilized o in es iga e ac o s ha
de e mine he demand o ecosys em se ices because CE
e eals ade-o s ha people a e willing o accep . T ade-o s
can be, o example, hypo he ical land use op ions, in which
people simul aneously win and lose some bene i s such as
en i onmen al heal h, ec ea ion oppo uni ies, o employmen .
The iden i ica ion o ade-o s may help in inding
possibili ies o build consensus and sol e con lic s.
Pea lands a e widely u ilized pea - o ming we land
ecosys ems h oughou he wo ld. They p o ide a a ie y o
ecosys em se ices such as imbe , pea , ca bon seques a ion,
lood isk educ ion, wa e quali y imp o emen , biodi e si y,
and ec ea ional bene i s (e.g., Zedle and Ke che 2005). The
comme cial use o pea lands is p incipally seen o be in con lic
wi h ecological, en i onmen al, and ec ea ional alues
because la ge-scale d ainage o , e.g., ag icul u e, o es y,
and cons uc ion pu poses has esul ed in de e io a ion o
pea land habi a s, he loss o biodi e si y, la ge-scale
hyd ological p oblems, and inc eased emissions o
g eenhouse gases (Chapman e al. 2003). In Eu ope, pea has
ceased o accumula e wi hin o e 50% o he o me mi e
a eas, and almos 20% o he o iginal pea land a eas no longe
exis as pea land (Joos en and Cla ke 2002). In Finland, he
pea land- iches coun y in he wo ld, abou 50% o he
pea land a ea has been d ained o o es y use, 76,000 hec a es
(less han 1%) a e being used o pea p oduc ion, and 13% is
p o ec ed (Finnish Fo es Resea ch Ins i u e 2011). The
g owing public in e es in sa egua ding pea land biodi e si y
and inc eased knowledge on he capaci y o pea lands o e ain
a mosphe ic ca bon has c ea ed a need o pea land es o a ion.
In Finland, es o a ion is p ac iced in comme cially
unp oduc i e o es y-d ained pea lands loca ed in o nea
p o ec ed a eas. The p incipal me hods a e he emo al o
excess ee s and g own a e d ainage and he blocking o
d ainage di ches, which aim a accele a ing he pea land
succession. Hyd ology and pea mine aliza ion and
decomposi ion a es ha e been obse ed o egene a e quickly
o he le el o und ained pea lands, whe eas he ege a ion
succession is slowe (Laine e al. 2011, Ta ainen e al. 2012).
1Finnish Fo es Resea ch Ins i u e, Finland, 2Thule Ins i u e, Uni e si y o Oulu, Finland, 3Depa men o Economics, Uni e si y o Oulu, Finland
Ecology and Socie y 18(2): 19
h p://www.ecologyandsocie y.o g/ ol18/iss2/a 19/
Inc easing public awa eness calls o he conside a ion o
people’s opinions in pea land managemen , al hough
ela i ely ew scien i ic s udies exis on his opic. In I eland,
sociological me hods gi e p omising esul s on he po en ial
o public pa icipa ion o educe con lic s conce ning a e -use
and ehabili a ion o pea lands a e pea p oduc ion (Collie
and Sco 2008, 2010, Collie 2011). In Finland, a su ey sen
o landowne s and pea indus y p o essionals emphasized an
inc easing need o discussions be ween he s akeholde s
(Selin 1999). To ou knowledge, CE has no been used o alue
he bene i s o pea lands, whe eas in we lands, whe e
comme cial imbe and pea p oduc ion a e no ele an land
use op ions, CE s udies ha e ocused solely on managemen
op ions o main ain he quali y and ex en o we lands
(Mo ison e al. 1999, Ca lsson e al. 2003, Bi ol e al. 2006,
Bi ol and Cox 2007, Wes e be g e al. 2010).
We p esen esul s om a su ey ha was pa o a egional
pea land p og am in No he n Finland. The p og am aims a
i ing pea land managemen op ions in o he egional plan and
i ollows he p inciples p esen ed in he Finnish Go e nmen
P og am o he Sus ainable Use o Mi es and Pea lands
launched in 2012. We in es iga ed esiden s’ opinions owa d
i e pea land use op ions: imbe p oduc ion, pea p oduc ion,
p o ec ion, es o a ion, and ec ea ion. Ou aim was o analyze
he po en ial o socioeconomically sus ainable pea land use
by in es iga ing con lic ing in e es s, by e ealing ade-o s
ha people a e willing o accep , and by s udying whe he
opinions a e dependen on socioeconomic and demog aphic
ac o s. CE was used o e eal ade-o s in land use
p e e ences, and he g oups o esponden s we e iden i ied
using a la en class model (LCM). We expec ed ha he
g ouping o esponden s in o di e en in e es g oups based
on he LCM would e lec he p esen le el o socioeconomical
sus ainabili y in pea land use because in CE he esponden s
had o choose be ween p esen and po en ial u u e scena ios
o pea land use. Based on he esul s, we discuss ac o s
inc easing he socioeconomical sus ainabili y o pea land use.
METHODS
Su ey design
Opinions o people conce ning u u e scena ios o pea land
use we e asked using a ques ionnai e. We ini ially included
i e pea land use op ions: imbe p oduc ion, pea p oduc ion,
p o ec ion, es o a ion, and ec ea ion. The CE was used o
assess esponden s’ p e e ences o hese i e op ions
(Lou ie e e al. 2000). The CE cap u es bo h use and nonuse
alues o ecosys em se ices. Use alues can be di ec such
as p o isioning (ene gy, employmen ) o cul u al se ices
( ec ea ion); o indi ec , such as egula ing se ices
(hyd ology, lood p e en ion, ca bon seques a ion). Nonuse
alues include exis ence alues, o example suppo ing
ecosys em se ices such as biodi e si y (classi ica ion based
on Smi h e al. 2006). In he CE, esponden s we e p esen ed
wi h a choice ask ha included se e al al e na i es, i.e.,
scena ios in his s udy, and asked o selec he al e na i e ha
p o ides hem he highes u ili y in a hypo he ical se ing. The
al e na i es we e desc ibed using a ibu es, i.e., he
cha ac e is ics o he en i onmen al p oblem in ques ion. The
a ibu es, i.e., pea land use op ions in his s udy, we e
add essed a se e al le els o ind he e ec o inc ease o
dec ease in he cha ac e is ics on people’s alua ions.
Typically, he ques ionnai e included se e al choice asks o
ge mul iple obse a ions o each esponden .
The ques ionnai e was de eloped by an expe g oup
ep esen ing knowledge on al e na i e pea land use op ions.
The ques ionnai e consis ed o ou pa s. The i s pa
con ained ques ions ela ed o esponden s’ ou doo ac i i ies
and hei a i udes owa d ecosys em se ices p o ided by
pea lands. The second pa con ained he CE and included
desc ip ions o he a ibu es, i.e., pea land use op ions, in
e ms o imbe p oduc ion, pea p oduc ion, conse a ion,
es o a ion, and ec ea ion acili ies (Table 1). The a ibu es
and hei quan i a i e le els we e iden i ied wi h in o ma ion
om he expe g oup. The hi d pa asked esponden s’
opinions owa d s a emen s p esen ed abou he signi icance
o na u e and pea lands. The ou h pa included ques ions on
esponden s’ socioeconomic backg ound.
To de elop he choice asks p esen ed o esponden s in he
ques ionnai e, we applied an o hogonal ac ional ac o ial
(main e ec ) design, which is equen ly used in empi ical
s udies (Lou ie e e al. 2000). This p ocedu e esul ed in 12
choice asks, which we e blocked in o wo g oups.[1] Thus, we
p esen ed six choice asks o each esponden and in each ask
hey had o selec be ween h ee al e na i es p esen ing h ee
pea land use scena ios. Each choice ask included wo u u e
scena ios and a s a us quo al e na i e, in which he le els o
a ibu es e e o he cu en si ua ion (Table 1). An illus a ion
o a choice ask is p esen ed in Fig. 1.
Su ey da a
In Feb ua y 2011, we sen 1250 ques ionnai es o andomly
selec ed, o e 18-yea s-old inhabi an s li ing in ou
municipali ies in no he n Os obo hnia. We sen 500
ques ionnai es o Oulu, he la ges ci y o he egion, and 250
ques ionnai es o h ee u al municipali ies in he sou he n
(Siikala a), no heas e n (Pudasjä i), and eas e n (U ajä i)
pa o he egion. By add essing hese mail-based su eys o
di e en municipali ies we aimed a co e ing people li ing a
a ious socioeconomic si ua ions. Only 192 (15.4%) mail-
based su eys we e e u ned. No eminde s we e sen ,
howe e , because he same su ey was simul aneously open
o e e yone as a web-based su ey on he In e ne . The link
o he web-based su ey was ad e ised h ough local
newspape s, email lis s, and in he social media h ough
Facebook and Twi e . The web-based su ey ecei ed 725
answe s.
Ecology and Socie y 18(2): 19
h p://www.ecologyandsocie y.o g/ ol18/iss2/a 19/
Table 1. A ibu es (pea land use op ions) and hei le els in he choice asks wi h a iable names used in he analysis.
A ibu e Le el Va iable name‡
Timbe p oduc ion a ea (ha) 1. Smalle han p esen s a e: 700,000 ha Timbe –
2. P esen s a e: 800,000 ha†
3. La ge han p esen s a e: 900,000 ha
Timbe +
Pea p oduc ion a ea (ha) 1. Pea p oduc ion ceased: 0 ha, he educ ion akes place g adually
as he exis ing p oduc ion ields a e exhaus ed Pea -
2. P esen s a e: 18,000 ha†
3. La ge han p esen s a e: 22,000 ha
Pea +
P o ec ion a ea (% o pea land a ea) 1. P esen s a e: 9%†P o ec ion
2. La ge han p esen s a e: 10% P o ec ion
3. Conside ably la ge han p esen s a e: 12%
P o ec ion
Res o a ion a ea (ha/yea ) 1. Res o a ion ceased: 0 ha/yea Res o -
2. P esen s a e: 150 ha/yea †
3. La ge han p esen s a e: 300 ha/yea
Res o +
Rec ea ion acili ies ( ou es, es ing
places, in o ma ion boa ds) 1. Less han p esen s a e Rec ea -
2. P esen s a e†
3. Mo e han p esen s a e Rec ea +
†The a ibu e le el desc ibes he basic al e na i e.
‡We used dummy coding o he a ibu e le els in he analysis excep ha P o ec ion was ea ed as a con inuous a iable.
Fig. 1. An example o a choice ask including h ee
al e na i e pea land use scena ios.
Al oge he 205 esponden s we e emo ed om analyses
because hey did no answe he CE ques ion (14 esponden s)
o hey always selec ed he basic al e na i e (s a us quo) in he
CE (191 esponden s), e en hough hey did no uly conside
ha al e na i e as he bes op ion. Ins ead, hey e ealed some
o he easons o hei choices, such as “ he al e na i es we e
no c edible” o “ he choice se s we e oo complex.” A CE
including se e al a ibu es wi h mul iple le els places a la ge
cogni i e bu den on he esponden s, and he e o e, i is no
su p ising ha his ype o p o es answe eme ges. A logi
model analysis e ealed ha he p obabili y o a p o es answe
inc eased wi h he age o esponden o i he esponden li ed
in a u al a ea. The p obabili y dec eased i he esponden had
a uni e si y deg ee o he esponden was an en ep eneu o
a lowe le el employee. The inal numbe o esponden s was
712.
Model
The CE da a was analyzed using la en class model (LCM;
Boxall and Adamowicz 2002). LCM can be de i ed om a
andom u ili y model, in which he u ili y unc ion o each
esponden is he sum o a de e minis ic e m ha can be
desc ibed as a unc ion o ac o s ha in luence esponden s’
u ili y, and a andom e m ha is unobse able and s ochas ic
o esea che s. Assuming he exis ence o S classes and ha
indi idual n belongs o class s (s=1, ..., S) he unobse able
o e all u ili y U o al e na i e i o a esponden n is
ep esen ed by:
(1)
whe e βs is a class-speci ic ec o o u ili y coe icien s o
obse ed a iable xni o esponden n ep esen ing his/he
p e e ences.
Ecology and Socie y 18(2): 19
h p://www.ecologyandsocie y.o g/ ol18/iss2/a 19/
Assuming a ype I ex eme alue dis ibu ion o εni, he logi
p obabili y o choosing al e na i e i condi ional on class
membe ship is de ined as:
(2)
whe e C deno es he esponden ’s choice se and µs is a class-
speci ic scale pa ame e . Nex we can de ine he join
p obabili y Pn(i) ha indi idual n belongs o class s and
chooses al e na i e i as ollows:
(3)
whe e snk is he membe ship p obabili y. In he LCM,
p e e ence he e ogenei y is accoun ed o by simul aneously
assigning indi iduals in o beha io al classes and es ima ing
he choice model. Thus, p e e ences a e assumed o be
homogenous wi hin each class, bu a y be ween he classes.
Indi idual-speci ic a iables can be used o es ima e he
membe ship p obabili y.
A e in ensi e es ing o di e en model speci ica ions, we
decided o exclude he ec ea ion a ibu e om he inal
model. The coe icien s o he a ibu e le els o ec ea ion
acili ies we e o en no s a is ically signi ican and in some
cases hey had unexpec ed signs. No e ha lea ing ou his
a ibu e does no cause p oblems o he es ima ion because
we used o hogonal design. In addi ion, we ea ed he
p o ec ion a ibu e as a con inuous a iable in he es ima ion,
because i included wo inc easing a ibu e le els in con as
o he o he a ibu es, which included bo h inc easing and
dec easing le els, and because he esponden s had qui e
homogenous p e e ences o pea land p o ec ion. We chose a
3-class model, based on a balanced assessmen o he adjus ed
2, AIC and BIC s a is ics (Bi ol e al. 2006). The 3-classes
solu ion p o ided he bes i o he da a: AIC and BIC s a is ics
dec eased and adjus ed 2 inc eased as mo e classes we e added
o he model, bu he changes we e much smalle om wo o
h ee han om h ee o ou class models.
RESULTS
Responden cha ac e is ics
The esponden s we e on a e age 44 yea s old, and 63% o
hem we e men. The p edominan occupa ional le el was
employee, composing 58% o he esponden s. The majo
household mon hly income le el was 3000-5000 eu os, being
a ained by 31% o he esponden s. This indica es a sligh ly
highe sala y le el han on a e age in Finland, whe e 23% o
inhabi an s belong o his le el, and he p edominan mon hly
income le el is 1000-3000 eu os (O icial S a is ics o Finland
2009). Fi y- h ee pe cen o he esponden s li ed in u al
communi ies o small owns wi h less han 20,000 inhabi an s,
and 46% li ed in owns wi h o e 20,000 inhabi an s o ci ies
wi h o e 100,000 inhabi an s. These alues co espond o he
a e age pe cen ages om he egion o No he n Os obo hnia
(55% and 45%, espec i ely, O icial S a is ics o Finland
2011a). The esponden s had a highe educa ion le el han
a e age inhabi an s in he egion: 69% had a seconda y le el
educa ion (uni e si y o poly echnic), whe eas he same
educa ion le el is eached only by 26% o he inhabi an s in
he egion (O icial S a is ics o Finland 2010). The pe cen age
o landowne s was on a e age 56% in he su eys, whe eas
he amoun o landowne s is conside ably lowe , 10%, in he
egion as a whole (O icial S a is ics o Finland 2011b). The
esponden s we e ac i e pea land use s: 92% isi ed pea lands
mo e han once in a yea , and 47% a leas once a mon h.
The majo i y o esponden s disag eed o s ongly disag eed
wi h he op ions ha sugges ed a la ge imbe p oduc ion a ea
(66%), la ge pea p oduc ion a ea (69%), o cessa ion o
es o a ion (74%) (Fig. 2). On he o he hand, he majo i y
ag eed o s ongly ag eed wi h he op ions sugges ing
cessa ion o pea p oduc ion (61%) o la ge p o ec ion (64%)
and es o a ion (70%) a eas. To iden i y he ela i e
impo ance o he a ibu es we analyzed wha kind o ade-
o s be ween he a ibu es he esponden s we e willing o
accep and how hei p e e ences di e ed be ween he g oups
iden i ied by he LCM.
Fig. 2. Dis ibu ions o opinions owa d a ibu e le els
desc ibing land-use op ions, used in he la en class model
(LCM). Numbe o esponden s = 712.
G ouping o esponden s
Based on he p e e ences owa d pea land use op ions in he
CE, he h ee classes in he LCM we e labeled as
“en i onmen alis ,” “p oduc ion-o ien ed,” and “cu en use
Ecology and Socie y 18(2): 19
h p://www.ecologyandsocie y.o g/ ol18/iss2/a 19/
Table 2. Resul s om he h ee-class la en class model (LCM) including socioeconomic explana o y a iables o class
membe ship. The class membe ship coe icien s o Class 3 we e ixed o iden i y he emaining coe icien s. Numbe o
esponden s = 712.
Class 1:
En i onmen alis Class 2:
P oduc ion-o ien ed Class 3:
Cu en use suppo e
Coe . S and. e o Coe . S and. e o Coe . S and. e o
Timbe +*** -1.387 0.203 *** 0.482 0.155 0.353 0.216
Timbe --0.346 0.478 *** -1.879 0.319 *** -2.820 0.549
Pea +*** -2.050 0.163 *** 0.985 0.102 -0.074 0.159
Pea -*** 2.811 0.244 *** -1.229 0.153 * 0.397 0.227
P o ec ion * 0.379 0.206 *** 0.828 0.156 *** 0.756 0.215
Res o +*** 1.629 0.182 0.071 0.130 -0.150 0.156
Res o -0.125 0.266 *** -0.882 0.169 **-0.490 0.225
In e cep
0.223
0.253
0.242
0.178
*** 2.222
0.238
Class membe ship
pa ame e s
In e cep 0.126 0.213 *** -1.386 0.415
Residen ial a ea *** 0.314 0.065 0.000 0.001
Highe educa ion le el *** 1.059 0.265 -0.013 0.353
Ag ic. en ep eneu ** -1.297 0.652 -0.183 0.647
O he en ep eneu 0.226 0.499 *1.022 0.561
Lowe -le e employee * 0.458 0.245 -0.087 0.334
Landowne *** -0.310 0.066 -0.005 0.013
Income 0.000 0.000 *** 0.408 0.097
Class p obabili ies 0.629 0.180 0.191
RsqAdj 0.421
Log-likelihood -2657.570
AIC 1.285
BIC 1.345
*** signi ican a 0.01 le el, ** signi ican a 0.05 le el, * signi ican a 0.1 le el
suppo e ” classes (Table 2). The en i onmen alis was he
la ges class indica ed by he class p obabili y 0.629. The
esponden s in his class had a s ong p e e ence owa d he
cessa ion o pea p oduc ion. They also esponded nega i ely
o he inc ease in imbe p oduc ion a ea, bu hey we e likely
o accep he cu en imbe p oduc ion le el, because he
dec ease in imbe p oduc ion was no s a is ically signi ican .
The en i onmen alis s a ached posi i e alues o la ge
pea land p o ec ion and es o a ion a eas.
Abou one i h o he esponden s (class p obabili y 0.180)
we e assigned o he p oduc ion-o ien ed class (Table 2). In
pa icula , he p oduc ion-o ien ed esponden s p e e ed an
inc ease in imbe and pea p oduc ion a eas in con as o he
o he wo classes. Howe e , hey also ega ded en i onmen al
issues as impo an because hey ag eed on an inc ease in
pea land p o ec ion a ea and disag eed on he cessa ion o
pea land es o a ion. The cu en use suppo e class was o a
simila size o he p oduc ion-o ien ed class (class p obabili y
0.191). Membe s o he cu en use suppo e class p e e ed
he cu en le el o pea land use, indica ed by he s a is ically
signi ican posi i e in e cep in he LCM (Table 2). They also
disag eed on he educ ion o imbe p oduc ion a ea and
cessa ion o es o a ion. They p e e ed he cessa ion o pea
p oduc ion, bu his coe icien was only weakly signi ican .
Like he o he wo classes, cu en use suppo e s p e e ed a
la ge pea land p o ec ion a ea.
We used esponden s’ socioeconomic cha ac e is ics o
explain class membe ships (Table 2). O he es ed a iables,
esiden ial a ea, highes educa ion le el (uni e si y deg ee),
ag icul u al en ep eneu , o he en ep eneu , lowe -le el
employee, land owne , and household mon hly income be o e
axes we e signi ican and we e he e o e used in he model.[2]
The en i onmen alis class was cha ac e ized by esponden s
li ing in a ci y, ha ing he highes educa ion le el, and being
a lowe -le el employee. They we e no likely o be ag icul u al
o o es y en ep eneu s and ewe o hem owned land
compa ed wi h he wo o he classes. The membe s o he
p oduc ion-o ien ed class we e o en en ep eneu s (no
ag icul u al/ o es y) and had a highe mon hly income le el
han he esponden s in he wo o he classes. The mos ypical
esiden ial a ea o he p oduc ion-o ien ed class and cu en
use suppo e s was u al communi y. Table 3 shows he

Ecology and Socie y 18(2): 19
h p://www.ecologyandsocie y.o g/ ol18/iss2/a 19/
b eakdown o he socioeconomic cha ac e is ics by he h ee
classes. The esponden s we e assigned o one o he h ee
classes on he basis o hei la ges p obabili y sco e. The
a e age age o esponden s was lowe in he en i onmen alis
class (42.2 yea s) han in he p oduc ion-o ien ed and cu en
use suppo e classes (48.0 and 46.7 yea s, espec i ely). The
p opo ion o men was highe in he p oduc ion-o ien ed class
(77.2%) han in he en i onmen alis and cu en use suppo e
classes (57.0% and 63.3%, espec i ely).
Table 3. Pe cen ages o esponden s wi h speci ic
socioeconomic cha ac e is ics in Classes 1, 2, and 3, and
a e age household incomes (eu os/mon h) o esponden s.
Numbe o esponden s = 712.
Class 1:
En i onmen alis
Class 2:
P oduc ion-
o ien ed
Class 3:
Cu en use
suppo e
Mos ypical esiden ial
a ea (%) Ci y, 41.0 Ru al, 20.5 Ru al, 26.6
Highes educa ion le el
(%) 50.4 28.3 20.1
Ag ic. En ep eneu
(%) 1.3 5.5 7.2
O he en ep eneu (%) 5.6 14.2 7.2
Lowe -le e employee
(%) 39.5 29.1 34.5
Landowne (%) 47.3 76.4 60.4
A e age income 3650 4532 3853
Opinions owa d pea lands as p o ide s o ecosys em
se ices
In he ques ionnai e we asked esponden s o ank he
impo ance o ecosys em se ices in pea lands by using a
Like scale om 1 = no impo an o 5 = e y impo an . O
he se en su eyed ecosys em se ices, i e we e signi ican
in he LCM in explaining he class membe ship o esponden s
(Table 4): egional employmen and income e ec s
(employmen and income), hyd ological p oblems due o
o es y d aining and pea p oduc ion (hyd ological
p oblems), pea lands as sou ces o ene gy (ene gy sou ce),
pea lands as habi a s o na i e pea land species (species and
habi a s), and pea lands as ec ea ion si es ( ec ea ion). Flood
p e en ion and ca bon seques a ion se ices we e conside ed
impo an o e y impo an by 77% and 75% o all
esponden s, espec i ely, bu because hese ecosys em
se ices we e no s a is ically signi ican in he LCM, hey a e
no shown in Table 4.
Membe s o he en i onmen alis class conside ed egional
employmen and income and pea lands as sou ces o ene gy
o be less impo an ac o s han did membe s in he cu en
use suppo e class, which was ea ed as he e e ence class
wi h ixed class membe ship coe icien s in his LCM (Table
4).[3] In addi ion, he en i onmen alis class conside ed
hyd ological p oblems, pea lands as habi a s o species, and
pea lands as a place o ec ea ion as mo e impo an han he
o he wo classes. The p oduc ion-o ien ed class was
cha ac e ized by a highe emphasis on pea lands as sou ces o
ene gy.
The equency dis ibu ions show ha hyd ological p oblems,
species and habi a s, and ec ea ion we e conside ed impo an
o e y impo an ecosys em se ices by he majo i y
(95-98%) o en i onmen alis s (Fig. 3). On he o he hand, all
i e ecosys em se ices we e conside ed impo an o e y
impo an by he majo i y (54-81%) o p oduc ion-o ien ed
class and cu en use suppo e s, and he leas impo an
ecosys em se ice o hem was ec ea ion.
Fig. 3. Opinions conside ing he ecosys em se ices o
pea lands in he h ee classes o esponden s, signi ican in
he la en class model (LCM). Numbe o esponden s =
712.
Responden s’ a i udes owa d na u e and pea lands we e also
inco po a ed in o he LCM (Table 5). The model e ealed ha
he esponden s ag eeing wi h he s a emen s ha na u e
p o ides peace ul and good eelings, pea lands ha e a igh o
exis wi hou u iliza ion, and people should ha e mo e espec
o pea lands belonged likely o he en i onmen alis class.
Fu he mo e, membe s o his class disag eed wi h he
s a emen s ha pea lands should se e mul iple pu poses and
legisla ion limi s he pea land use oo much. In con as , he
la e s a emen was suppo ed by he esponden s belonging
o he p oduc ion-o ien ed class.
The equency dis ibu ions show he g ea es de ia ion in he
opinions conce ning he mul iple uses o pea lands and he
legisla ion limi ing pea land use. The majo i y (64% and 85%,
espec i ely) o he en i onmen alis s disag eed wi h hese
s a emen s, whe eas he majo i y (81% and 64%, espec i ely)
o he p oduc ion-o ien ed class ag eed wi h hem (Fig. 4).
Mul iple use o pea lands was also ag eed by he majo i y
(76%) o cu en use suppo e s, whose opinions usually lie
be ween hose o he en i onmen alis s and he p oduc ion-
Ecology and Socie y 18(2): 19
h p://www.ecologyandsocie y.o g/ ol18/iss2/a 19/
Table 4. Resul s o he 3-class la en class model (LCM) including opinions on he ecosys em se ices o pea lands as explana o y
a iables o class membe ship. The class membe ship coe icien s o Class 3 we e ixed in o de o iden i y he emaining
coe icien s. Numbe o esponden s = 712.
Class 1:
En i onmen alis Class 2:
P oduc ion-o ien ed Class 3:
Cu en use suppo e
Coe . S and. e o Coe . S and. e o Coe . S and. e o
Timbe +*** -1.181 0.197 ** 0.347 0.155 -0.153 0.259
Timbe --0.458 0.459 *** -1.494 0.300 *** -3.064 0.647
Pea +*** -2.130 0.163 *** 1.013 0.097 -0.199 0.198
Pea -*** 2.656 0.231 *** -1.338 0.158 0.430 0.274
P o ec ion ** 0.496 0.198 *** 0.539 0.154 *** 0.748 0.251
Res o +*** 1.478 0.173 -0.067 0.128 -0.097 0.177
Res o --0.173 0.258 *** -0.638 0.171 -0.204 0.269
In e cep
0.179
0.243
0.212
0.176
*** 2.318
0.286
Class membe ship
pa ame e s
In e cep *** -0.956 0.317 -0.254 0.200
Employmen and income *** -0.878 0.199 0.074 0.242
Hyd ological p oblems ** 0.388 0.171 -0.237 0.144
Ene gy sou ce *** -0.928 0.149 ** 0.506 0.224
Species and habi a s *** 1.108 0.209 -0.128 0.168
Rec ea ion * 0.318 0.171 -0.215 0.143
Class p obabili ies 0.636 0.197 0.167
RsqAdj 0.476
Log-likelihood -2407.935
AIC 1.164
BIC 1.218
*** signi ican a 0.01 le el, ** signi ican a 0.05 le el, * signi ican a 0.1 le el
o ien ed class. The p oduc ion-o ien ed class showed mos
disag eemen (42% o esponden s in he class), owa d he
s a emen p oposing ha pea lands ha e a igh o exis wi hou
u iliza ion.
Pea lands and pe sonal li e
The esponden s we e asked how pea lands ela e o hei
pe sonal li e. The en i onmen alis and cu en use suppo e
classes we e cha ac e ized by he ac ha hei childhood
home was no ypically loca ed close o pea lands compa ed
wi h he p oduc ion-o ien ed class, which was he e e ence
class in his analysis (Table 6). The wo k o esponden s in
he en i onmen alis and cu en use suppo e classes was less
ypically ela ed o pea lands, whe eas hey we e mo e ac i e
in picking be ies in pea lands. Responden s belonging o he
en i onmen alis class we e mo e likely membe s o a
conse a ion league and hey we e mo e likely o use pea lands
o ec ea ion and well-being han esponden s in he wo o he
classes.
DISCUSSION AND CONCLUSIONS
Ou esul s indica e ha he esponden s had he e ogeneous
p e e ences ega ding pea land use. Using he LCM we
iden i ied h ee classes o esponden s, in which
Fig. 4. Opinions on he signi icance o na u e and pea lands
in he h ee classes o esponden s, signi ican in he la en
class model (LCM). Numbe o esponden s = 712.
en i onmen alis s emphasized ecological and en i onmen al
alues o pea lands, p oduc ion-o ien ed class p e e ed an
inc ease in he comme cial use o pea lands, and he cu en
use suppo e s we e an in e media e class showing a
p e e ence o he p esen land use policy. The esul e lec s
Ecology and Socie y 18(2): 19
h p://www.ecologyandsocie y.o g/ ol18/iss2/a 19/
Table 5. Resul s o he 3-class la en class model (LCM) including opinions on he signi icance o na u e and pea lands as
explana o y a iables o class membe ship. The class membe ship coe icien s o Class 3 we e ixed in o de o iden i y he
emaining coe icien s. Numbe o esponden s = 712.
Class 1:
En i onmen alis Class 2:
P oduc ion-o ien ed Class 3:
Cu en use suppo e
Coe . S and. e o Coe . S and. e o Coe . S and. e o
Timbe +*** -1.182 0.229 *** 0.437 0.147 -0.147 0.210
Timbe --0.735 0.503 *** -1.564 0.297 *** -2.498 0.502
Pea +*** -2.07 0.166 *** 1.038 0.097 ** -0.380 0.168
Pea -*** 3.037 0.263 *** -1.153 0.149 0.166 0.216
P o ec ion *** 0.659 0.231 *** 0.622 0.145 *** 0.625 0.198
Res o +*** 1.584 0.188 -0.077 0.127 0.182 0.150
Res o --0.126 0.280 *** -0.715 0.163 -0.241 0.219
In e cep
0.445
0.279
* 0.288
0.167
*** 1.900
0.227
Class membe ship
pa ame e s
In e cep ** -4.227 2.127 -2.104 1.450
Good eeling * 0.704 0.400 -0.046 0.270
Mul iple uses *** -1.033 0.168 -0.027 0.203
Legisla ion *** -0.871 0.188 *** 0.851 0.172
Righ o exis *** 0.847 0.228 -0.167 0.142
Respec *** 0.768 0.234 -0.006 0.022
Class p obabili ies 0.622 0.194 0.184
RsqAdj 0.478
Log-likelihood -2399.442
AIC 1.160
BIC 1.214
*** signi ican a 0.01 le el, ** signi ican a 0.05 le el, * signi ican a 0.1 le el
he p esen le el o socioeconomic sus ainabili y in he
pea land use among he esponden s. The e o e, oughly one
hi d is es ima ed o be sa is ied wi h he p esen si ua ion o
pea land use.
I is impo an o no ice ha he esponse a e was qui e low
in he mail-based su ey, and he web-based su ey was
openly a ailable, which makes selec ion bias a conce n in his
s udy. Rega ding he ype o esiden ial a ea o esponden s,
ou sample was qui e ep esen a i e o he a e age popula ion
in no he n Finland. In con as , esponden s’ educa ion and
sala y le els we e highe in his s udy han in he s udy egion
in gene al. This is no a su p ise because highly educa ed
people end o answe ques ionnai es mo e o en han hose
wi h less educa ion (Bi ol and Cox 2007, Wes e be g e al.
2010). Also he p opo ion o landowne s was highe in his
s udy han on a e age in he egion. Gi en he socioeconomic
and a i udinal p o iles o he h ee classes, we sugges ha he
en i onmen alis and p oduc ion-o ien ed classes p esen
mo e ex eme opinions owa d pea land use, whe eas he
gene al public, he majo i y o which did no answe he
ques ionnai e, appa en ly has no s ong opinions. The closes
class o he gene al public in ou da a migh be he cu en use
suppo e s, bu because he land owne ship and he p opo ion
o ag icul u al and o es y en ep eneu s we e highe in his
class han in he a e age popula ion, his g oup was also
sligh ly biased. Sending eminde s o people ecei ing
ques ionnai es migh ha e dec eased he selec ion bias
somewha .
Inc easing in o ma ion on he de e io a ion o pea land
habi a s and public deba es on pea land use ha e aised a need
o conside all elemen s o sus ainabili y in pea land planning
(Joos en and Cla ke 2002). Con lic s o in e es , obse ed
h ough he g ouping o esponden s in ou s udy, place a
ema kable challenge o i ing oge he mul iple needs
subjec ed o he same ecosys em. Each pa y may concen a e
on maximizing one land-use op ion wi hou conside ing
possibili ies o o he op ions (de G oo 2006). A i s glance,
ou esul s suppo he con en ional hough o u banized
eli is s e sus lowe educa ed people li ing in he coun y. The
en i onmen alis class, cha ac e ized as highly educa ed
people li ing mos ypically in ci ies, p e e ed he comple e
cessa ion o pea p oduc ion, whe eas he people in he
p oduc ion-o ien ed class, whose pe sonal li es we e
connec ed o pea land h ough wo k, and who mos ypically
li ed in u al communi ies, p e e ed he inc ease in pea
p oduc ion a ea. Howe e , mos esponden s isi ed pea lands
Ecology and Socie y 18(2): 19
h p://www.ecologyandsocie y.o g/ ol18/iss2/a 19/
Table 6. Resul s o he 3-class la en class model (LCM) including ac o s on he ela ionships o pea lands o esponden s’
pe sonal li es as explana o y a iables o class membe ship. The class membe ship coe icien s o Class 2 we e ixed in o de
o iden i y he emaining coe icien s. Numbe o esponden s = 712.
Class 1:
En i onmen alis Class 2:
P oduc ion-o ien ed Class 3:
Cu en use suppo e
Coe . S and. e o Coe . S and. e o Coe . S and. e o
Timbe +*** -1.314 0.205 ** 0.389 0.153 0.024 0.258
Timbe --0.437 0.479 *** -1.599 0.310 *** -3.075 0.611
Pea +*** -2.091 0.164 *** 1.022 0.099 -0.244 0.180
Pea -*** 2.840 0.240 *** -1.157 0.151 0.036 0.266
P o ec ion ** 0.475 0.207 *** 0.608 0.152 *** 0.710 0.250
Res o +*** 1.542 0.179 -0.032 0.124 -0.129 0.171
Res o --0.031 0.266 *** -0.632 0.162 -0.407 0.271
In e cep
0.271
0.253
* 0.208
0.175
*** 2.028
0.278
Class membe ship
pa ame e s
In e cep -0.175 0.275 -0.264 0.328
Childhood home *** -0.788 0.273 * -0.562 0.331
Wo k *** -1.400 0.264 *** -0.956 0.314
Well-being *** 1.780 0.274 0.457 0.316
Conse a ion league *** 3.024 0.431 -0.007 0.032
Picking up be ies *** 0.780 0.295 ** 0.793 0.354
Class p obabili ies 0.632 0.191 0.177
RsqAdj 0.443
Log-likelihood -2557.078
AIC 1.235
BIC 1.289
*** signi ican a 0.01 le el, ** signi ican a 0.05 le el, * signi ican a 0.1 le el
ac i ely, which e lec s a close connec ion o pea lands
i espec i e o he class and esiden ial a ea.
Ou esul s also showed sha ed iews be ween s akeholde
g oups, and hese sha ed iews a e c ucial o building
consensus and sol ing con lic s (Williams e al. 1998). All
h ee classes showed a p e e ence owa d he inc ease o
pea land p o ec ion a ea, while hey simul aneously disag eed
on he educ ion o he pea land es o a ion a ea. Hence, he e
was a common ag eemen on he ecological alues o pea lands
and common suppo owa d new managemen me hods such
as es o a ion. Acco ding o he e iew by Young e al. (2005),
people do no usually objec o conse ing biodi e si y,
p o iding i does no con lic wi h pe sonal o ins i u ional
goals. The ac ha he sha ed iews we e ound in he opinions
owa d he p o ec ion and es o a ion, bu no he mul iple uses
o pea lands (disag eed on by en i onmen alis s), ac ually
e lec s s ic e , mo e pu is a i udes owa d pea land use
among he en i onmen alis s compa ed wi h he o he wo
classes. S ill, e en he en i onmen alis s we e likely o ag ee
on he cu en imbe p oduc ion le el, indica ing ha he
di ision be ween en i onmen alis s and o he g oups was no
s aigh o wa d.
The concep o ecosys em se ices p o ides a use ul
amewo k o land use planning, al hough he e a e s ill
ques ions in he de ini ion, quan i ica ion, and aluing o
ecosys em se ices (de G oo e al. 2010). O he se en
ecosys em se ices conside ed, i e we e signi ican o explain
he class membe ships in he LCM: p o isioning (employmen
and income, ene gy sou ce), egula ing (hyd ological
p oblems), suppo ing (species and habi a s), and cul u al
( ec ea ion) se ices. These ecosys em se ices had
appa en ly di e en impac s on esponden s’ well-being. Fo
example, despi e he ac ha hyd ological p oblems we e
conside ed impo an by o e hal o he esponden s in all
classes and all esponden s we e ac i e use s o pea lands,
p oduc ion-o ien ed and cu en use suppo e classes
accep ed he ade-o be ween he p o isioning se ices and
he o he ecosys em se ices. Pea ene gy p oduc ion occu s
in emo e egions wi h high unemploymen a es. Pe haps o
his eason, esponden s in he p oduc ion-o ien ed and cu en
use suppo e classes, among whom he mos common
esiden ial a ea was u al communi ies, conside ed
employmen and income he mos impo an ecosys em
se ice. Two egula ing ecosys em se ices, i.e., lood isk
educ ion and ca bon seques a ion, we e gene ally highly