Biogeosciences, 14, 1947–1967, 2017
www.biogeosciences.ne /14/1947/2017/
doi:10.5194/bg-14-1947-2017
© Au ho (s) 2017. CC A ibu ion 3.0 License.
Me hane exchange a he pea land o es loo – au oma ic chambe
sys em exposes he dynamics o small luxes
Mika Ko kiakoski1, Juha-Pekka Tuo inen1, Mika Au ela1, Ma kku Koskinen2, Ka i Minkkinen2, Paa o Ojanen3,
Timo Pen ilä3, Juuso Rainne1, Tuomas Lau ila1, and Annalea Lohila1
1Finnish Me eo ological Ins i u e, A mosphe ic Composi ion Resea ch, P.O. Box 503, 00101 Helsinki, Finland
2Uni e si y o Helsinki, Depa men o Fo es Sciences, P.O. Box 27, 00014 Uni e si y o Helsinki, Finland
3Na u al Resou ces Ins i u e Finland, Viikinkaa i 4, 00790 Helsinki, Finland
Co espondence o: Mika Ko kiakoski ([email p o ec ed])
Recei ed: 2 June 2016 – Discussion s a ed: 26 July 2016
Re ised: 18 Janua y 2017 – Accep ed: 13 Ma ch 2017 – Published: 10 Ap il 2017
Abs ac . We measu ed me hane (CH4)exchange a es wi h
au oma ic chambe s a he o es loo o a nu ien - ich
d ained pea land in 2011–2013. The en, loca ed in sou h-
e n Finland, was d ained o o es y in 1969 and he ee
s and is now a mix u e o Sco s pine, No way sp uce, and
pubescen bi ch. Ou measu emen sys em consis ed o six
anspa en chambe s and s ainless s eel ames, posi ioned
on a numbe o di e en ield and moss laye composi-
ions. Gas concen a ions we e measu ed wi h an online ca -
i y ing-down spec oscopy gas analyze . Fluxes we e cal-
cula ed wi h bo h linea and exponen ial eg ession. The
use o linea eg ession esul ed in sys ema ically smalle
CH4 luxes by 10–45 % as compa ed o exponen ial eg es-
sion. Howe e , he use o exponen ial eg ession wi h small
luxes (<2.5 µg CH4m−2h−1) ypically esul ed in anoma-
lously la ge absolu e luxes and high hou - o-hou de ia ions.
The e o e, we ecommend ha luxes a e ini ially calcula ed
wi h linea eg ession o de e mine he h eshold o “low”
luxes and ha highe luxes a e hen ecalcula ed using ex-
ponen ial eg ession. The exponen ial lux was clea ly a -
ec ed by he leng h o he i ing pe iod when his pe iod
was <190 s, bu s abilized wi h longe pe iods. Thus, we
also ecommend he use o a i ing pe iod o se e al minu es
o s abilize he esul s and dec ease he lux de ec ion limi .
The e we e clea seasonal dynamics in he CH4 lux: he o -
es loo ac ed as a CH4sink pa icula ly om ea ly summe
un il he end o he yea , while in la e win e he lux was e y
small and luc ua ed a ound ze o. Howe e , he magni ude
o luxes was ela i ely small h oughou he yea , anging
mainly om −130 o +100 µg CH4m−2h−1. CH4emission
peaks we e obse ed occasionally, mos ly in summe du -
ing hea y ain all e en s. Diu nal a ia ion, showing a lowe
CH4up ake a e du ing he day ime, was obse ed in all o
he chambe s, mainly in he summe and la e sp ing, pa ic-
ula ly in d y condi ions. I was a ibu ed mo e o changes in
wind speed han ai o soil empe a u e, which sugges ha
physical a he han biological phenomena a e esponsible
o he obse ed a ia ion. The annual ne CH4exchange a -
ied om −104 ±30 o −505 ±39 mg CH4m−2y −1among
he six chambe s, wi h an a e age o −219 mg CH4m−2y −1
o e he 2-yea measu emen pe iod.
1 In oduc ion
Me hane (CH4)is one o he mos impo an a mosphe ic
g eenhouse gases due o i s capabili y o abso b he mal
adia ion and wa m he clima e (IPCC, 2014). One o he
main sou ces o CH4globally is pea land (e.g., Denman e
al., 2007), whe e CH4is p oduced by he decomposi ion
o o ganic ma e in anae obic condi ions. A ound 3 % (ca.
4 000 000 km2)o he Ea h’s land su ace is co e ed by pea -
lands (Cla ke and Rieley, 2010) and he majo i y o hese a e
loca ed in he bo eal egion (Fischlin e al., 2007). Abou one
hi d (104 000 km2)o Eu opean mi e and pea esou ces a e
loca ed in Finland (Mon ana ella e al., 2006) and mo e han
hal (55 000 km2)o his a ea has been d ained o o es y
(Päi änen and Hånell, 2012).
Me hane can be bo h p oduced and consumed in soil so
ha he ne CH4 lux depends on he a e o CH4p oduc-
Published by Cope nicus Publica ions on behal o he Eu opean Geosciences Union.
1948 M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo
ion in anoxic soil laye s and on he a e o CH4oxida ion in
he oxic soil laye s. In pea lands, he hickness and dep h o
CH4p oducing and oxidizing laye s a e la gely de e mined
by he wa e able (Bubie and Moo e, 1994), which con ols
he e ical dis ibu ion o oxygen in he soil p o ile. CH4is
p oduced unde anae obic condi ions by mic obes known as
me hanogens. The p oduc ion a e is dependen on he a ail-
abili y o o ganic subs a es a low edox po en ial (Eh) al-
ues and is con olled by soil empe a u e and pH (Dun ield e
al., 1993; Wang e al., 1993; Sege s, 1998; Ko syu benko e
al., 2004). In con as , oxida ion o CH4occu s in he oxic
soil laye close o he su ace and po en ially also in he
moss laye (La mola e al., 2010). Like he p oduc ion a e,
he oxida ion a e is con olled by soil empe a u e and pH
(Dun ield e al., 1993; Scheu z and Kjeldsen, 2004; Boeckx
and Van Cleempu , 1996), bu many o he ac o s also a -
ec oxida ion p ocesses, such as soil wa e con en , soil ex-
u e, nu ien s, and CH4and oxygen concen a ion (Boeckx
and Van Cleempu , 1996; Ridgwell e al., 1999; Scheu z and
Kjeldsen, 2004). In addi ion o he di ec con ol o p oduc-
ion and oxida ion a es, he e a e o he phenomena which
may a ec he obse ed ne CH4 lux abo e he soil su ace,
including la e al CH4 anspo in he soil (Ch is ophe sen
and Kjeldsen, 2001) and subsu ace s o age (Hu chinson e
al., 2000).
In en i onmen s wi h low soil CH4p oduc ion, such as
upland o es soils, g asslands, and und a, up ake o a -
mosphe ic CH4by he me hano ophic mic obes domina es
(Du au and Ve cho , 2007). This is also wha commonly
happens a e he d ainage o pea lands, which esul s in
wa e le el d awdown and inc eased oxic laye hickness.
The eby, CH4p oduc ion is dec eased and he ac ion o
oxidized CH4inc eased (e.g., Moo e and Knowles, 1989;
Roule e al., 1992). Consequen ly, he CH4oxida ion a e in
he ae a ed su ace soil and mosses ypically exceeds CH4
p oduc ion ha occu s deepe in he soil, hus u ning in
pa icula well-d ained pea lands in o ne CH4sinks (Ma -
ikainen e al., 1995; Minkkinen e al., 2007; Ojanen e al.,
2010; Lohila e al., 2011). Howe e , poo ly d ained si es may
emain o ac as CH4sou ces (Ojanen e al., 2010). In addi-
ion, he d ainage di ches e en a well-d ained si es ypically
con inue o emi CH4a a es simila o p is ine bo eal pea -
lands (Minkkinen e al., 1997; Minkkinen and Laine, 2006;
Luan and Wu, 2015).
Closed chambe s a e commonly used in he measu emen
o g eenhouse gas exchange be ween he o es loo and
he a mosphe e (e.g., Denmead, 2008; Fo b ich e al., 2010;
Koskinen e al., 2014). Unlike he eddy co a iance (EC)
me hod, which is mo e sui able o measu ing luxes a he
ecosys em le el, he chambe me hod pe mi s he in es iga-
ion o small-scale p ocesses, such as he gas exchange o
di e en mic o opog aphic su aces, and enables he quan-
i ica ion o spa ial a ia ion (Kelle e al., 1990; Singh e
al., 1997). Howe e , he e a e a ious de ails ela ed o he
chambe design and he deploymen o his measu emen
echnique in p ac ice ha may ha e a signi ican impac on
he lux ha is es ima ed om he obse ed concen a ion
change in he chambe headspace. Fo example, he lux es-
ima e seems o depend on he dimensions o he chambe
(Pihla ie e al., 2013). In addi ion, chambe s should include
a an o e enly dis ibu e he ai in he chambe headspace
(Pumpanen e al., 2004; Ch is iansen e al., 2011) al hough
he o a ional speed o he an should be kep low o a oid ex-
cessi e u bulence (Koskinen e al., 2014). A majo sou ce o
unce ain y is he impac o he chambe i sel on he gas con-
cen a ion g adien in he soil (Healy e al., 1996; Hu chinson
e al., 2000; Conen and Smi h, 2000; Da idson e al., 2002;
Li ings on e al., 2005) and in he bounda y laye jus abo e
i . The concen a ion g adien is c i ical as i d i es he soil–
a mosphe e gas exchange and hus any ae odynamic dis u -
bance may impac he obse ed lux.
The g adien be ween he soil and he ai inside he cham-
be changes when he gas concen a ion inside he cham-
be changes du ing he measu emen . This changes he lux,
which makes he concen a ion change nonlinea in ime.
Howe e , nonlinea i y o he concen a ion du ing he cham-
be closu e may also esul om chambe leaking. Fo exam-
ple, Pi k e al. (2016) demons a ed ha he deg ee o con ex
cu a u e in he inc easing me hane concen a ion co ela ed
posi i ely wi h wind speed (WS) ou side he chambe . Fu -
he mo e, in he case o soil ac ing as a me hane sink, he
me hane consump ion by soil me hano ophs obeys he i s -
o de eac ion kine ics, which should lead o cu ilinea con-
cen a ion dynamics in he chambe (e.g., Sab eko e al.,
2016). Howe e , he di e en p ocesses esponsible o he
cu a u e in he concen a ion ime se ies may be di icul o
sepa a e om each o he (Ku zbach e al., 2007).
The e a e many s udies ha ha e ecognized ha he use
o linea eg ession in lux calcula ion can cause signi ican
unde es ima ion o he lux (e.g., Healy e al., 1996; Hu chin-
son e al., 2000; Nakano e al., 2004; Li ings on e al., 2005,
2006; Ku zbach e al., 2007; K oon e al., 2008; Pede sen e
al., 2010; Pihla ie e al., 2013). Howe e , many s udies ha e
used linea eg ession (e.g., Laine e al., 2006; Alm e al.,
2007; Jones e al., 2011; Be gie e al., 2013; Fassbinde e
al., 2013), because unde ield condi ions i is mo e obus o
andom measu emen e o s han nonlinea me hods. Mo e-
o e , he use o linea eg ession is p e e ed when compa -
ing measu emen si es as i is no as sensi i e as nonlinea
models o small di e ences in soil p ope ies (Ven e ea e
al., 2009). The selec ion o he op imal i ing me hod is im-
po an as i can be a la ge sou ce o unce ain y in lux cal-
cula ions (Le y e al., 2011; Ven e ea, 2013).
Al hough se e al s udies ha e examined he di e en i -
ing me hods o calcula ing luxes om chambe da a, he e
exis only a ew pape s explo ing he dynamics o CH4 lux
da a ha mainly consis o small up ake luxes and a e mea-
su ed wi h au oma ic chambe s coupled o a high- esolu ion
gas analyze (e.g., Sa age e al., 2014; Ueyama e al., 2015).
In his s udy, we measu ed he CH4 lux be ween a o es
Biogeosciences, 14, 1947–1967, 2017 www.biogeosciences.ne /14/1947/2017/
M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo 1949
Table 1. CN a io, bulk densi y, and ash con en (±SD) o he pea
a Le osuo (n=4).
CN a io Bulk densi y (g cm−3) Ash con en (%)
Humus 29.2 ±1.8 0.005 ±0.003 3.1 ±0.4
0–10 cm 23.9 ±1.0 0.107 ±0.014 6.5 ±1.5
10–20 cm 24.3 ±0.7 0.170 ±0.011 3.4 ±0.4
loo and he a mosphe e con inuously h oughou 2 yea s
a a bo eal nu ien - ich o es y-d ained pea land si e wi h
ypically small CH4exchange a es. We used six au oma ic
soil chambe s and a ca i y ing-down spec oscopy analyze ,
which allowed us o measu e wi h a ela i ely high sampling
a e du ing each chambe closu e and o pe o m mul iple
daily measu emen s wi h each chambe . Ou pa icula aims
wi h his se up we e o answe he ollowing ques ions:
1. Wha is he op imal i ing me hod o calcula ing he
CH4 lux?
2. How la ge a e he diu nal, seasonal, and in e annual
a ia ions in he CH4 lux?
3. Wha is he annual CH4balance o he s udy si e?
2 Ma e ials and me hods
2.1 Si e desc ip ion
The measu emen s we e made in sou he n Finland a Le -
osuo (60◦380N, 23◦570E) (Fig. S1a in he Supplemen ),
which is a nu ien - ich pea land o es ha was d ained in
1969 and e ilized wi h phospho us and po assium soon
a e . The open di ches, loca ed in app oxima ely 45 m in-
e als (Fig. S1b), we e o iginally abou 1 m deep bu
ha e since been pa ly illed wi h new ege a ion. Be o e
d ainage, he ee s and was domina ed by Sco s pine (Pinus
syl es is) wi h some pubescen bi ch (Be ula pubescens).
A e d ainage, he s and has de eloped o a mix u e o Sco s
pine and pubescen bi ch in he dominan canopy laye , wi h
an unde s o ey o No way sp uce (Picea abies) wi h some
sca e ed small-sized pubescen bi ch. The s em olumes a
he ime o his s udy equaled 174, 46, and 28 m3ha−1 o
Sco s pine, pubescen bi ch, and No way sp uce espec i ely.
The ee s and is qui e dense, which esul s in i egula shad-
ing and, consequen ly, pa chy and a iable g ound ege a ion
laye . Fo example, he bs such as D yop e is ca husiana and
T ien alis eu opaea and dwa sh ubs such as Vaccinium my -
illus a e common in he g ound ege a ion (Bhuiyan e al.,
2017). In addi ion, he moss laye is pa chy and is domina ed
by Pleu ozium sch ebe i and Dic anum polyse um wi h some
Sphagna (Sphagnum gi gensohnii, Sphagnum angus i olium,
and Sphagnum ussowii) appea ing in mois pa ches.
Table 2. G ound ege a ion, all-sided maximum ascula g een a ea
(VGAmax; ascula g een su ace, m2, di ided by o es loo , m2)
and co e age (%) o o es mosses (CFM)and Sphagnum mosses
(CSP).
Chambe Vege a ion VGAmax CFM CSP
1Pleu ozium sch ebe i 2.04 56 0
Dic anum polyse um
Vaccinium my illus
2Pleu ozium sch ebe i 0.85 60 0
Dic anum polyse um
Vaccinium i is-idaea
3Maian hemum bi olium 0.01 3 0
Pleu ozium sch ebe i
Dic anum polyse um
4D yop e is ca husiana 2.34 26 0
Vaccinium my illus
Vaccinium i is-idaea
Pleu ozium sch ebe i
Dic anum polyse um
5Pleu ozium sch ebe i 0.11 30 0
Dic anum polyse um
6Sphagnum gi gensohnii – 0 90
CN a io o he su ace pea , sampled a ou poin s loca ed
a a 20–40 m dis ance om he chambe plo s, a e aged 24
o he 0–20 cm laye (Table 1). The ela i ely low CN a io is
ypical o e ile pea land o es s and e lec s he en his o y
o he si e. The bulk densi y o hese samples was 0.11 and
0.17 g cm−3 o he 0–10 and 10–20 cm laye s, espec i ely,
while he a e age bulk densi y o he 0–20 cm laye below
each chambe a ied om 0.03 o 0.13 g cm−3(Koskinen e
al., 2014). The ash con en o he pea a ied om 3.4 o
6.5 %.
The ascula g een a ea (VGA) was es ima ed o each
chambe and ascula plan species e e y 2 weeks du ing he
g owing seasons 2011 and 2012 (Ojanen, unpublished da a).
This was done by es ima ing he numbe and dimensions o
lea es wi hin each chambe and calcula ing g een a ea by
species-speci ic eg ession models be ween lea dimensions
and g een a ea. Fo Vaccinium my illus, he su ace a ea o
he g een s ems was also included in o VGA. The co e age
o he mosses was es ima ed isually. The maximum VGA
and he co e ages o each chambe a e shown in Table 2.
2.2 Flux measu emen sys em and ancilla y
measu emen s
The au oma ic chambe measu emen sys em is he same as
used o CO2exchange by Koskinen e al. (2014). The CO2
lux measu emen s s a ed in au umn 2010, and he CH4ana-
lyze was added o he sys em in Ma ch 2011. He e we epo
he CH4 luxes measu ed om hen un il Ap il 2013. Fo es
loo gas exchange, including he ee oo s, was moni o ed
using six anspa en soil chambe s connec ed o an ins u-
men cabin. The cabin was loca ed a a dis ance o abou 30 m
www.biogeosciences.ne /14/1947/2017/ Biogeosciences, 14, 1947–1967, 2017
1950 M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo
om he 25.5 m all EC mas (Koskinen e al., 2014) om
which he WS abo e he canopy was measu ed. The loca-
ions o he chambe s we e selec ed o maximize he numbe
o di e en g ound ege a ion composi ions (Table 2) wi hin
a ci cle o ca. 15 m adius a ound he cabin.
The de ails o he chambe sys em can be ound in Kosk-
inen e al. (2014), and hus he e we only desc ibe he main
ea u es o he sys em. The size o he chambe boxes we e
57 cm ×57 cm ×30 cm (leng h ×wid h ×heigh ). We used a
pe manen ly ins alled s eel colla (heigh 5 cm, inse ed a a
dep h o 2 cm) below each chambe o minimize he dis u -
bance o he soil and o enhance he sealing be ween he soil
and he chambe . The e was a U p o ile a he bo om o he
chambe edges, insula ed wi h a oam ape, o u he im-
p o e he sealing. In win e , he whole chambe ame was
aised abo e he snow le el by placing one o wo ex ension
colla s (heigh 16 cm) be ween he ame and soil.
A 24 V an (Magle KDE2408PTV1, Sunon L d, Kaohsi-
ung, Taiwan) (size 8 cm ×8 cm) was used o mix he ai in-
side he chambe headspace. The ol age o he an was egu-
la ed o keep he mixing s eady, bu as low as possible (Kosk-
inen e al., 2014). Sample gas was d awn om he chambe
ypically once an hou (wi h some excep ions explained be-
low) a a low a e o abou 1 L min−1and e u ned back o
he chambe om he gas analyze s. CH4, CO2, and wa e
apo concen a ions we e measu ed app oxima ely e e y 4 s
wi h a Pica o G1130 ca i y ing-down spec oscopy gas an-
alyze (Pica o Inc., San a Cla a, CA, USA). The inle and
ou le gas ubes (Fes o Oy, Van aa, Finland) we e made o
polyu e hane and we e 15 m in leng h and had an inne and
ou e diame e o 4 and 6 mm espec i ely.
The ubes we e lushed wi h ambien ai jus be o e he
chambe was closed. When all he chambe s we e open, am-
bien ai was sampled. The delay in he analyze esponse
caused by he long ubing was aken in o accoun using a
lagging sys em in he compu e p og am collec ing he da a,
which labeled each da a poin wi h he espec i e chambe
numbe using a 20 s lag. Howe e , as he low a e a ied
sligh ly in ime, some poin s we e emo ed om he da a be-
o e he lux calcula ion (Sec . 2.4).
Ai and soil empe a u e da a we e collec ed e e y 10 s us-
ing P 100 p obes (PT4T, Noke al Oy, Nokia, Finland) and
Noke al 680 logge s (Noke al Oy, Nokia, Finland). One
p obe was loca ed inside each chambe a a heigh o 30 cm
and posi ioned nex o he an unde a me al hea shield o
p e en di ec sola adia ion om a ec ing he measu e-
men s. Fu he mo e, soil su ace empe a u e was moni o ed
inside each chambe jus below he su ace o he moss o li -
e laye . In addi ion, soil empe a u e p obes we e placed
a dep hs o 2, 5, 10, 20, and 30 cm a one loca ion nea
he chambe s. Wa e able le el (WTL) was moni o ed e -
e y hou a ou di e en poin s a he si e (T uT ack WT-
HR da a logge s, In ech Ins umen s L d, Auckland, New
Zealand). The ai p essu e, p ecipi a ion, and snow dep h
da a we e acqui ed om he nea by Finnish Me eo ological
Ins i u e obse a o y a Jokioinen (∼35 km no hwes o Le -
osuo).
2.3 Me eo ological condi ions
The clima e a he si e has bo h con inen al and ma i ime
in luences. The annual mean empe a u e and p ecipi a ion
a he nea by wea he s a ion in 1981–2010 we e 4.6 ◦C
and 627 mm espec i ely (Pi inen e al., 2012). Du ing his
s udy, he i s measu emen yea (Ap il 2011–Ma ch 2012)
was signi ican ly wa me (annual mean empe a u e 5.8 ◦C)
han he second measu emen yea (Ap il 2012–Ma ch 2013)
(1.4 ◦C) (Fig. 1). The i s yea was sligh ly wa me and he
second yea was signi ican ly colde han he long- e m mean
eco ded a he nea by wea he s a ion (4.6 ◦C). Bo h he
summe (JJA) (17.6 ◦C) and win e (DJF) (2.7 ◦C) empe -
a u es in 2011 we e wa me han hose o 2012 (12.1 and
−2.3 ◦C). In pa icula , he beginning o summe 2012 was
much colde han he same pe iod in 2011.
Annual p ecipi a ion du ing he i s (976 mm) and sec-
ond (780 mm) measu emen yea s was highe han he long-
e m mean (627 mm). Summe ime p ecipi a ion was 9 %
highe in he i s (309 mm) yea as compa ed o he second
(284 mm) yea , while in win e he di e ence was 18 % (577
and 490 mm in he i s and second win e s espec i ely). The
i s snow appea ed on 5 Decembe in 2011 and 25 Oc obe
in 2012, and he i s pe manen snow was eco ded on 7 Jan-
ua y in 2012 and 28 No embe in 2012. In sp ing (MAM)
2011, he snow had mel ed by 13 Ap il. Fo sp ing 2012,
we do no know he exac day o snowmel due o missing
da a, al hough he snow had mel ed a la es by 4 Ap il. F om
he empe a u e da a we es ima e ha he snow co e disap-
pea ed some ime in mid-Ma ch.
WTL a ied om −8 o −59 cm om he soil su ace
(nega i e sign deno es WTL below he su ace) and was
highes in he sp ing and la e au umn (SON). The lowes (i.e.,
deepes ) alues we e eco ded a he end o summe . The a -
e age WTL in summe 2011 was −47.2 ±7.4 cm (±SD) and
−49.1 ±7.1 in summe 2012. Occasional sudden inc eases
in WTL we e obse ed a e ain all e en s and i usually
ook 1–2 weeks o each he WTL obse ed p io o he e en .
2.4 Flux calcula ion
Du ing he s udy pe iod, he chambe s we e ope a ed wi h
a ying closu e imes anging om 2 o 16 min. In 2011,
2 min closu es we e used wi h he excep ion o 6 min mea-
su emen s made ou imes pe day. A e mid-Ma ch 2012,
he minimum closu e ime was 6 min. Thus, each chambe
was ypically sampled once an hou , wi h he excep ion o
summe 2012 (JJA) when a longe closu e ime o 16 min
was es ed and each chambe was sampled e e y 2 h. Fo he
analysis o CH4exchange dynamics (Sec . 3.3–3.4), we used
he luxes calcula ed wi h a 6 min closu e ime (as jus i ied in
Sec . 3.2). In addi ion o emo ing 20 s om he s a o he
Biogeosciences, 14, 1947–1967, 2017 www.biogeosciences.ne /14/1947/2017/
M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo 1951
1 Jun 2011 1 Sep 2011 1 Dec 2011 1 Ma 2012 1 Jun 2012 1 Sep 2012 1 Dec 2012 1 Ma 2013
30
20
10
0
10
20
30
Tempe a u e [ C]
Ai empe a u e
2 cm soil empe a u e
0
20
40
60
Snow dep h [cm]
1 Jun 2011 1 Sep 2011 1 Dec 2011 1 Ma 2012 1 Jun 2012 1 Sep 2012 1 Dec 2012 1 Ma 2013
60
50
40
30
20
10
0
WTL [cm]
0
20
40
60
P ecipi a ion [mm]
(a)
(b)
Figu e 1. (a) The daily mean o ai empe a u e ( ed) and soil empe a u e a 2 cm dep h (blue) a Le osuo du ing he measu emen pe iod
(1 Ap il 2011 o 31 Ma ch 2013) and he daily snow dep h (ba s) measu ed a he nea by Jokioinen obse a o y. (b) The daily mean wa e
able (WTL) (line) and i s s anda d de ia ion (shading) om ou di e en poin s a Le osuo and he daily p ecipi a ion (ba s) measu ed a
Jokioinen.
measu emen due o lag caused by long ubing (Sec . 2.2),
18 s was disca ded om he s a o a measu emen o ensu e
ha he ai inside he chambe was p ope ly mixed. Dilu-
ion and spec al co ec ed CH4concen a ions epo ed by
Pica o G1130 we e used o calcula e he luxes.
Two di e en eg ession ypes we e i ed o he da a:
linea and exponen ial. The linea unc ion desc ibing he
change in he concen a ion, C, as a unc ion o ime was
C( )=alin +blin , (1)
whe e alin and blin a e pa ame e s and is he ime om he
s a o he closu e. In his model, he slope blin equals he
concen a ion change in ime.
The exponen ial unc ion we i ed was
C( )=aexp +bexp exp(cexp ), (2)
whe e aexp,bexp, and cexp a e pa ame e s. When di e en ia -
ing Eq. (2) wi h espec o ime and inspec ing he momen
when he chambe closes ( =0), i ollows ha he concen-
a ion change wi h ime is he p oduc o pa ame e s bexp
and cexp. I is gene ally conside ed ha his ini ial a e o
concen a ion change bes ep esen s he lux a ha ime.
Howe e , when i ing he exponen ial unc ion o he da a
using he leas -squa es app oach, he i ing equen ly ails
due o local minima. To o e come his and o a oid o e -
pa ame e iza ion, a Taylo powe se ies expansion (Ku zbach
e al., 2007) was i ed o he da a o de e mine ini ial es i-
ma es o he pa ame e s o he exponen ial eg ession.
The exponen ial eg ession should cap u e he lux be e
han he linea eg ession as i akes in o accoun he change
in he g adien be ween soil and chambe headspace du ing
chambe closu e, which is e iden when di usion lux is de-
c easing he concen a ion di e ence. Howe e , exponen ial
eg ession is e y sensi i e o possible dis u bances o he
da a a he beginning o chambe closu e. In ou s udy, we a -
emp ed o minimize hese dis u bances by closing he cham-
be slowly and smoo hly, which seemed o p e en p essu e
luc ua ions ela ed o chambe closing. Fo he analysis o
CH4exchange dynamics (Sec . 3.3–3.4), we used lux da a
ha a e based on a combina ion o linea and exponen ial i s:
i s all luxes we e calcula ed using he linea eg ession, and
below and abo e a limi o 2.5 µg CH4m−2h−1 he luxes
we e calcula ed wi h he linea and exponen ial me hod e-
spec i ely ( o jus i ica ion see Sec . 3.1).
The CH4 lux (F, µg CH4m−2h−1)was calcula ed ac-
co ding o Eq. (3), which is based on he ideal gas law:
F=dC( )
d  =0
MP V
RT A 3600 s
h,(3)
whe e dC( )
d  =0is he ime de i a i e (ppm s−1)o a lin-
ea (blin)o exponen ial (bexp ×cexp) eg ession a he be-
ginning o he closu e, Mis he molecula mass o CH4
(16.042 g mol−1),Pis ai p essu e (Pa), Ris he uni e sal
gas cons an (8.31446 J mol−1K−1),Tis he mean cham-
be headspace empe a u e du ing closu e (K), and Vand
Aa e he olume (m3)and he base a ea (m2)o he cham-
be headspace espec i ely. He e, a mic ome eo ological sign
con en ion is used: a posi i e lux indica es a lux om he
ecosys em o he a mosphe e (emission) and a nega i e lux
indica es a lux om he a mosphe e in o he ecosys em (up-
ake).
www.biogeosciences.ne /14/1947/2017/ Biogeosciences, 14, 1947–1967, 2017

1952 M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo
When es ima ing he olume o he chambe headspace,
he heigh o he moss and snow su aces was assumed o ep-
esen he in e ace be ween he soil and ai . In o he wo ds,
he po e space in he soil and snow was igno ed om he
headspace olume. The e o caused by his in lux calcula-
ions was es ima ed o be only a ew pe cen (Koskinen e
al., 2014). To c ea e a con inuous da a se o snow dep h, he
manual measu emen s ca ied ou i egula ly a he si e we e
combined wi h hose measu ed daily a he Jokioinen obse -
a o y. In addi ion o snow dep h, he heigh o he chambe
headspace was measu ed a he s a and end o he g owing
season om 16 poin s inside each colla by gen ly placing he
end o a ape measu e on op o he su ace mosses (Koskinen
e al., 2014). The heigh o he chambe headspace be ween
hese manual measu emen s was de e mined wi h linea in-
e pola ion.
All he calcula ions and analyses we e made wi h he
Py hon p og amming language (Py hon So wa e Founda-
ion, e sion 2.7, h ps://www.py hon.o g) using he ol-
lowing lib a ies: NumPy (h p://www.numpy.o g/), SciPy
(h p://www.scipy.o g/), Pandas (h p://pandas.pyda a.o g/),
and ma plo lib (h p://www.ma plo lib.o g). All he Py hon
sc ip s we e de eloped speci ically o his s udy. Fo he
i s, he leas -squa es me hod was used h ough he “poly i ”
unc ion o NumPy lib a y o he linea eg ession and he
“cu e_ i ” unc ion o SciPy lib a y o he nonlinea i s.
2.5 Fil e ing o he lux da a
A e he luxes we e calcula ed, se e al il e s we e applied
o emo e cases whe e he measu emen sys em did no wo k
adequa ely. The mos common eason o disca ding da a
was due o he p oblems wi h he chambe ope a ion, o
example, o he imp ope unc ioning o a linea ac ua o ,
which caused he chambe s o emain s uck ei he open o
closed. These cases we e de ec ed by moni o ing he simul a-
neously measu ed CO2concen a ion da a du ing he closu e.
The goodness o i was checked by calcula ing he no mal-
ized oo mean squa e e o (NRMSE) (e.g., Ch is iansen e
al., 2011; Pihla ie e al., 2013) o each i :
NRMSE =s1
n
n
P
i=1
(C i ,i −Ci)2
Cmax −Cmin
,(4)
whe e nis he numbe o measu emen poin s, C i ,i is he
CO2concen a ion calcula ed om he i , Ciis he mea-
su ed CO2concen a ion, and Cmax and Cmin a e he high-
es and lowes concen a ions measu ed du ing closu e. I he
NRMSE was la ge han 0.05, he CH4da a om ha closu e
we e disca ded. I should be no ed ha he applica ion o his
c i e ion emo es closu es wi h no change in CO2concen-
a ion, which may esul when pho osyn hesis a e equals
espi a ion a e. He e we ound <20 o such cases, mean-
ing ha his c i e ion could be applied wi hou emo ing a
signi ican amoun o po en ially sui able da a.
Figu e 2. Concen a ion da a du ing one chambe closu e o
a case wi h a highe (alinea and exponen ial: −90 and
104 µg CH4m−2h−1 espec i ely) and lowe (blin −3.5, exp
−4.3 µg CH4m−2h−1) lux.
In addi ion o NRMSE il e ing, he unning mean o CH4
lux (FCH4)wi h a ime window o 14 days (shi ing one day
a a ime) and he co esponding s anda d de ia ion (σ) we e
calcula ed o emo e andom spiking in he da a. The da a
poin s ha ailed o all wi hin FCH4±10σwe e emo ed i -
e a i ely. In o al, 71229 closu es we e eco ded om which
14 % (n=9987) we e disca ded due o la ge NRMSE alues
(p oblems wi h he chambe s) and <0.001 % (n=40) we e
emo ed wi h he σ il e as ou lie s.
2.6 De ec ion limi
The minimum de ec able lux (MDF) was es ima ed by using
he me ic o iginally de eloped by Ch is iansen e al. (2015),
which was modi ied by Nicke son (2016) o make i mo e
sui able o high- equency measu emen s:
Biogeosciences, 14, 1947–1967, 2017 www.biogeosciences.ne /14/1947/2017/
M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo 1953
80 60 40 20 0 20
Flux, linea i [ gCH4
m2h]
300
200
100
0
100
200
Flux, exponen ial i [ gCH4
m2h]
76543210
10
5
0
5
Figu e 3. Bin a e ages (n=500) o he linea and exponen ial luxes o he whole da a se (6 min closu es only). In he small zoomed igu e
he ed e ical line deno es he selec ed lux limi o 2.5 µg CH4m−2h−1. Ve ical and ho izon al e o ba s show he s anda d de ia ion o
lux de e mined wi h he exponen ial and linea i espec i ely.
MDF =

PI
cq c
ps

V P
ART M, (5)
whe e PIis he analy ical p ecision o he ins umen (ppm),
cis he closu e ime o he chambe (h), and psis he sam-
pling pe iodici y (h). The PI o he Pica o G1130 analyze ,
es ed and epo ed by he manu ac u e o he speci ic in-
s umen used in his s udy, was 0.256 ppb and pswas 5 s. On
a ypical summe day (T=20 ◦C), he MDF o he sys em
was abou 0.06 µg CH4m−2h−1. Howe e , du ing win e he
MDF was highe due o lowe empe a u es and he use o he
ex ension colla s, which oge he abou double he headspace
olume (wi hou snow) and he e o e also he MDF.
2.7 The annual balance and i s unce ain y
The annual balance o CH4was es ima ed o each chambe
by i s calcula ing he daily lux sums om he hou ly luxes
and hen summing hese o e a yea . The gaps in he da a
we e illed by using linea in e pola ion be ween he exis ing
hou ly and daily luxes. As mos o he luxes in 2011 and in
he i s qua e o 2012 we e measu ed wi h a 2 min closu e
ime, which was conside ed oo sho o he exponen ial e-
g ession (Sec . 3.2), we co ec ed he luxes calcula ed wi h
linea eg ession om he 2 min closu es o co espond o
hose measu ed using a 6 min closu e a ailable ou imes a
day du ing his pe iod as a e e ence. This co ec ion was im-
plemen ed by calcula ing he daily median a ios be ween he
luxes om 6 and 2 min closu e imes, which we e smoo hed
by a unning median wi h a mo ing window o 14 days. Fi-
nally, he 2 min da a om 2011 o Ma ch 2012 we e mul i-
plied by his a io (Fig. S2).
The unce ain y o he CH4balance es ima e de i ed om
he measu emen s was e alua ed by iden i ying h ee key e -
o sou ces: (1) he andom e o o eg ession, (2) he e o
caused by gap illing, and (3) he e o caused by he co -
ec ion o he luxes measu ed using he 2 min closu e ime
du ing he i s measu emen yea . Fi s , because he annual
balance o each chambe was calcula ed om he mean daily
luxes, we es ima ed he daily andom e o as he squa ed
sum o he unce ain ies o he hou ly lux da a o each day.
Assuming ha he goodness o i e lec s all he unce ain ies
ela ed o a single lux measu emen , he s anda d de ia ion
o he slope es ima e ob ained dC( )
d  =0;Eq.3p o ides
a measu e o his unce ain y.
Nex , he e o caused by he gap- illing p ocedu e was es-
ima ed by emo ing 1 mon h o lux da a om di e en pa s
o he whole da a se and inspec ing how his a ec ed he an-
nual balance o he di e en chambe s. The a e age alue o
he e ec o hese mon hly gaps was calcula ed and down-
scaled o ep esen he e ec o one missing day. Mul iply-
ing his alue by he numbe o missing days du ing he yea
gi es an es ima e o he gap- illing e o . I mus be no ed ha
he leng h o he emo ed pe iod was simila o he longes
gap obse ed in ou da a.
Las , he e o es ima e ela ed o he a io used o ans-
o m he luxes calcula ed om 2 min closu e o ep esen he
6 min closu e was es ima ed om he median absolu e de i-
a ion assuming no mally dis ibu ed medians. Finally, hese
h ee e o es ima es we e added oge he by using he s an-
da d accumula ion p inciple o independen e o s.
www.biogeosciences.ne /14/1947/2017/ Biogeosciences, 14, 1947–1967, 2017
1954 M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo
Figu e 4. The linea (a) and he exponen ial lux (b) as a unc ion
o i ing pe iod. The luxes a e scaled by he lux calcula ed wi h
he longes i ing pe iod (900 s). The e o ba s show he 95 % con-
idence in e als. The da a a e om summe 2012.
As ou measu emen s s a ed in Ap il 2011 and ended in
Ma ch 2013, a e exac ly 2 yea s, om now on in his pa-
pe he exp essions “ i s yea ” and “second yea ” deno e
he ime pe iods o Ap il 2011–Ma ch 2012 and Ap il 2012–
Ma ch 2013 espec i ely.
3 Resul s
3.1 Flux calcula ion me hod
Examples o ypical concen a ion de elopmen inside a
chambe du ing one measu emen a e shown in Fig. 2, o
bo h “high lux” case in summe (Fig. 2a) and a “low lux”
case in win e (Fig. 2b). In summe and au umn, when he
0 100 200 300 400 500
Fi s a ing poin [s]
29
28
27
26
25
24
23
22
21
CH4 lux [ gCH4
m2h]
Linea
Exponen ial
Flux limi me hod
Figu e 5. Fluxes calcula ed wi h he linea , exponen ial, and “ lux
limi ” me hods using a 6 min i ing pe iod wi h di e en s a ing
poin s o he i s. The e o ba s show he 95 % con idence in e als.
The da a a e om summe 2012.
luxes we e he highes , he concen a ion de elopmen in-
side a chambe usually was no adequa ely app oxima ed by
a linea unc ion and hus he slope calcula ed wi h he lin-
ea eg ession (Eq. 1) did no p ope ly ep esen he ini ial
“undis u bed” slope om which he lux should be calcu-
la ed. As a esul , linea eg ession esul ed in lowe lux es-
ima es o hese cases han exponen ial eg ession (Eq. 2).
Howe e , du ing he pe iods when he lux app oached he
de ec ion limi and he concen a ion da a became noisie ,
he use o exponen ial eg ession esul ed in noisie lux da a.
O en, exponen ial eg ession c ea ed a sha p slope a he be-
ginning o he i in he concen a ion ime se ies ha esul ed
in unphysically high luxes. To be able o eliably es ima e
he CH4exchange o he whole ange o luxes, we de e -
mined he lux limi below which he exponen ial eg ession
esul ed in un eliable lux es ima es and he linea i should
be p e e ed. This limi was es ima ed by compa ing bin
(n=500) a e ages o linea and exponen ial luxes o he
whole da a se (using a 6 min closu e ime) (Fig. 3). When he
linea ly calcula ed luxes ell below ca. 2.5 µg CH4m−2h−1,
he noise in he lux calcula ed using he exponen ial e-
g ession inc eased s eeply and he shape o he ela ionship
changed (Fig. 3). The e o e, we decided o i s calcula e he
lux wi h he linea eg ession and o ecalcula e all he luxes
exceeding he limi o 2.5 µg CH4m−2h−1wi h he exponen-
ial eg ession. Hence o h, all he da a shown in his pape
ha e been calcula ed in his way unless s a ed o he wise.
The whole 2-yea da a se showed ha he CH4 luxes cal-
cula ed wi h linea eg ession (Eq. 1) we e sys ema ically
and signi ican ly lowe han hose calcula ed wi h exponen-
ial eg ession (Eq. 2) (Table 3). The seasonal a e age lux
Biogeosciences, 14, 1947–1967, 2017 www.biogeosciences.ne /14/1947/2017/
M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo 1955
Jun 2011
Sep 2011
Dec 2011
Ma 2012
Jun 2012
Sep 2012
Dec 2012
Ma 2013
150
100
50
0
50
100
150
200
Flux [ gCH4
m2h]
Chambe 1
Jun 2011
Sep 2011
Dec 2011
Ma 2012
Jun 2012
Sep 2012
Dec 2012
Ma 2013
150
100
50
0
50
100
150
200 Chambe 2
Jun 2011
Sep 2011
Dec 2011
Ma 2012
Jun 2012
Sep 2012
Dec 2012
Ma 2013
150
100
50
0
50
100
150
200 Chambe 3
Jun 2011
Sep 2011
Dec 2011
Ma 2012
Jun 2012
Sep 2012
Dec 2012
Ma 2013
150
100
50
0
50
100
150
200
Flux [ gCH4
m2h]
Chambe 4
Jun 2011
Sep 2011
Dec 2011
Ma 2012
Jun 2012
Sep 2012
Dec 2012
Ma 2013
150
100
50
0
50
100
150
200 Chambe 5
Jun 2011
Sep 2011
Dec 2011
Ma 2012
Jun 2012
Sep 2012
Dec 2012
Ma 2013
150
100
50
0
50
100
150
200 Chambe 6
Figu e 6. Hou ly CH4 luxes om Ap il 2011 o Ma ch 2013 measu ed in each chambe . Nega i e alues indica e up ake by he soil, and
posi i e alues indica e emission o he a mosphe e. Fluxes ha e been calcula ed using he exponen ial i unless he alue o he lux ob ained
om he linea i was below 2.5 µg CH4m−2h−1.
di e ence be ween he linea and exponen ial eg essions
a ied wi hin 10.9–44.4 % (a e age o e 2 yea 27.5 ±0.3 %,
±95 % con idence in e al). The mean ela i e di e ence
was dependen on he ime o he yea : i was la ges du ing
he win e and sp ing (24.9–44.4 %) when he soil CH4sink
was a i s lowes and smalles in summe and au umn (10.9–
14.4 %) when he sink was a i s highes . When compa -
ing indi idual measu emen s, he a e age ela i e di e ence
be ween he linea and exponen ial eg ession was sligh ly
smalle in 2012 compa ed o 2011. Also, he unce ain ies
associa ed wi h he luxes we e sligh ly la ge in 2011 due
p obably o he ewe measu emen s a ailable wi h 6 min clo-
su e ime.
3.2 E ec o closu e ime on luxes
The e ec o he di e en i ing ime windows was es ed
by bo h inc easing he i ing pe iod om he beginning o
he closu e wi h 10 s s eps and by keeping he i ing win-
dow cons an bu mo ing i s s a ing poin . Fo hese es s,
we used he da a om summe 2012, when he measu emen s
we e made wi h a 16 min closu e ime. The lux om he ex-
ponen ial i was clea ly a ec ed by he leng h o he i when
he i ing pe iod was <190 s (Fig. 4b). A e ha , he mean
di e ence was mos ly s a is ically insigni ican (p > 0.05),
as compa ed o he lux calcula ed wi h he 900 s pe iod.
Howe e , he es ima ed linea lux s ayed abou he same o
he i s 140 s esul ing in 16.2 ±0.6 % highe luxes han ob-
ained wi h he 900 s i ing window (Fig. 4a). Howe e , u -
he inc ease o he i ing pe iod sys ema ically dec eased
he es ima ed lux by abou 1.3 % pe 60 s. A dec ease o
17.3 ±3.0 % was also obse ed when he s a ing poin o
he i was delayed by 530 s, bu he i ing pe iod was kep
cons an a 6 min (Fig. 5).
E en hough he esul s abo e migh suppo he selec ion
o a i ing pe iod o 190 s, a 6 min i ing pe iod was applied
in u he analysis. This was selec ed based on h ee a gu-
men s: (1) i made he exponen ial eg ession esul s mo e
s able; (2) we wan ed o use he same i ing pe iod in bo h
linea and exponen ial eg essions; and (3) a longe i ing
pe iod dec eases he de ec ion limi (Eq. 5). The las poin
was mainly ela ed o win e measu emen s when he de ec-
ion limi was inc eased by lowe empe a u es and he use
o ex ension colla s (inc easing he e ec i e olume be o e
he colla s we e illed wi h snow). Howe e , in 2011 and in
he i s qua e o 2012, a 2 min closu e ime was mos ly
used, which p o ed o be oo sho o accu a e es ima es
wi h he exponen ial i . As a esul , he esul s om hese
sho e measu emen s we e co ec ed o co espond o hose
ob ained wi h he 6 min closu e ime (Sec . 2.7).
www.biogeosciences.ne /14/1947/2017/ Biogeosciences, 14, 1947–1967, 2017
1962 M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo
many o he ac o s, such as he ege a ion composi ion and
small-scale soil p ope ies. The smalles sink was obse ed
in chambe s 1, 3, and 4, which we e cha ac e ized wi h he
lowes (3) and highes (1, 4) ascula g een a ea (VGAmax)
alues. Thus i seems ha i was no he amoun o g ound
ege a ion which a ec ed he sink, bu a mo e ele an ac o
could ha e been he co e age o mosses s. ha o ascula
plan s wi hin he colla , especially ha o he o es mosses
Pleu ozium sch ebe i and Dic anum polyse um, which we e
pa icula ly abundan in he highly oxidizing chambe s 2
and 5. Due o he small numbe o chambe s, howe e , he
ela ionship be ween he o es loo ege a ion and he CH4
exchange may be coinciden al and can only be specula ed.
4.2.4 Diu nal a ia ion
All he chambe s eco ded diu nal a ia ion in CH4 lux a
some ime du ing he s udy pe iod wi h mos o he a ia-
ion obse ed du ing la e sp ing and ea ly summe . Typically,
CH4up ake was a i s highes du ing he nigh and dec eased
owa ds he a e noon. The diu nal a ia ion was mo e com-
mon and occu ed mo e o en in all chambe s in 2012, while
in 2011 i occu ed mos ly in chambe 2. This a ia ion usu-
ally ceased o was a leas g ea ly diminished du ing and a e
ain all e en s, bu usually i appea ed again a e a couple o
days. WTL as such, howe e , did no ha e an impac on he
diu nal cycle, which sugges s ha he condi ions in he soil
su ace we e much mo e impo an o his phenomenon.
Al hough he diu nal a ia ion seemingly ollowed he pa -
e ns in he ai and soil su ace empe a u es, i was bes
explained by he WS measu ed abo e he canopy (below
canopy WS is no a ailable). To u he s udy his ela ion-
ship, we es ed he co ela ion be ween he pa ame e cexp
(Eq. 2) and WS. cexp ep esen s he cu a u e in he exponen-
ial i , being nega i e whene e he concen a ion inc ease
du ing a chambe closu e shows a slowing shape. As we only
selec ed nega i e, i.e., up ake, luxes he e, i ollows ha a
mo e nega i e cexp indica es a highe cu a u e in he con-
cen a ion e olu ion. Should leaking be esponsible o he
smalle CH4up ake du ing day ime, as he obse ed ela-
ionship be ween WS and CH4exchange implied, i would
be logical o ind highe cu a u es wi h highe WS. Such a
ela ionship was ecen ly ound by Pi k e al. (2016) o CH4
emission chambe lux da a om p is ine pea lands. How-
e e , we did no obse e such a ela ionship in ou CH4up-
ake da a. Fo example, in chambe 2, in which he diu nal
cycle was mos explici , an inc easing cexp was de e mined
o mos o he pe iods s udied. The e we e only a ew cham-
be s and pe iods when cexp dec eased wi h inc easing WS.
Thus we mus conclude ha he diu nal a ia ion in ou da a
is ela ed o he echnical ope a ion o he chambe a he
han en i onmen al condi ions. Ne e heless, as he empe -
a u e and WS co ela ed s ongly, i is possible ha some o
he obse ed pa e n was due o some mic obial o en i on-
men al ac o .
We hypo hesize ha , a he han chambe leaking, he
main unde lying ac o o he clea nega i e co ela ion be-
ween WS and CH4up ake is ela ed o changes in he soil
s o age and hus he changes imposed by chambe closu e o
he concen a ion g adien wi hin he op soil and he adja-
cen ai laye . P io o he closu e, his g adien is con olled
by a mosphe ic mixing and hence s ongly a ec ed by he
ambien WS. Du ing a calm nigh wi h a cool soil su ace,
u bulen mixing is s ongly supp essed and molecula di u-
sion gains impo ance, while windy and sunny condi ions e-
sul in much smalle e ical g adien s due o igo ous u bu-
lence ha is also able o pe u b he op-soil po e space. A -
e he chambe is closed, he concen a ion g adien adjus s
o he cons an mixing gene a ed by a an. Thus, he change
in concen a ion g adien depends on he mixing condi ions
ha p e ail abo e he a ge su ace jus be o e he chambe
is in oduced and how hese ela e o he mixing a e o he
chambe headspace ai . In he noc u nal case ou lined abo e,
mixing is enhanced a e he chambe closu e, esul ing in a
highe CH4up ake in he chambe .
The absence o he diu nal cycle in win e , and du ing
and a e he ain in summe , can be explained by he in-
c eased soil mois u e con en , which dec eases he ai - illed
po e space in soil, hus hampe ing he wind-induced mixing
e ec a he soil–a mosphe e in e ace and by slowing down
he di usion a e (Pi k e al., 2016).
I should be no ed ha he si ua ion is di e en when CH4
exchange is measu ed abo e a o es canopy wi h he EC
me hod. In ha case, he measu emen does no signi ican ly
dis u b a mosphe ic mixing and inc eased mechanical u bu-
lence po en ially enhances e ical gas exchange. Such pos-
i i e co ela ion be ween he downwa d CH4 lux and WS,
wi h highe sinks du ing he day ime, has been epo ed by
Wang e al. (2013). This is consis en wi h he esul s o ou
an-speed es , desc ibed in Koskinen e al. (2014) who mea-
su ed CO2 espi a ion by he same chambe sys em. The CH4
lux da a om he same es showed a highe CH4up ake
wi h highe an speed (da a no shown).
A wind-induced diu nal cycle sugges s ha he cu en
chambe se up po en ially leads o an o e - o unde es i-
ma e o he ac ual up ake a e du ing lowe o highe an-
induced mixing, espec i ely, as compa ed o ambien mix-
ing by wind. The chambe cons uc ion could be imp o ed by
making he an speed a y as a unc ion o he ambien wind
speed, so as o mimic he a ia ions in a mosphe ic mixing.
Howe e , we can expec ha he sys ema ic bias esul ing
om he wind esponse is minimized when employing au-
oma ed sampling ha acili a es con inuous measu emen s.
Ou esul s imply ha spo adic sampling wi h manual cham-
be s, which is ypically limi ed o he day ime, would ha e
esul ed in lowe up ake es ima es o his si e han he ex-
ensi e da a collec ed wi h ou au oma ic sys em.
Biogeosciences, 14, 1947–1967, 2017 www.biogeosciences.ne /14/1947/2017/

M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo 1963
4.3 D i ing ac o s o he CH4e lux
The seasonal CH4 luxes co ela ed bes wi h soil empe -
a u es a he dep hs o 20 cm and 30 cm, bu signi ican
co ela ions occu ed also wi h soil empe a u es a o he
dep hs o mos o he s udy pe iod. The co ela ions wi h
ai and soil su ace empe a u es we e lowe . The co ela-
ions we e always nega i e, indica ing ha highe empe a-
u es p omo ed he soil CH4up ake. This obse a ion could
be a ibu ed o inc eased consump ion o CH4by me han-
o ophs in highe empe a u es ha enhance me hano ophic
ac i i y (e.g., Mohan y e al., 2007). Howe e , i is likely ha
in addi ion o – o e en ins ead o – he inc eased me han-
o ophic ac i i y, he e a e o he easons behind his ela-
ionship. The co a ia ion o empe a u e wi h o he a iables,
such as g ound wa e le el and phenology, all ypically peak-
ing in July–Augus , may lead o spu ious co ela ion be ween
empe a u e and CH4 lux. Indeed, he lux was also co e-
la ed wi h WTL, he co ela ion being signi ican ly posi i e
(highe up ake wi h lowe WTL) in sp ing, summe , and au-
umn, bu nega i e in win e . A ou si e, he soil laye s mos
a o able o me hane p oduc ion and oxida ion a e loca ed
a clea ly di e en dep hs in he soil, he i s being ound
below he wa e able and he la e much close o he soil
su ace (A. Pu kinen, unpublished da a). Bo h o hese ha e
dis inc empe a u e and mois u e esponses, which a e p ac-
ically impossible o disen angle by examining he ne CH4
lux obse ed a he su ace.
In addi ion o he co ela ions ound in he hou ly da a, we
ound e idence ha lowe ing WTL inc eases he daily CH4
up ake in he la e pa o summe , when WTL <−40 cm.
In he beginning o he summe , he daily luxes we e be e
explained by he soil empe a u e, while a e mid-July he
WTL o e shadowed he empe a u e as a con ol o he daily
luxes.
In p is ine pea lands, empe a u e has been shown o co -
ela e posi i ely wi h he CH4emission a e (e.g., Mikkelä e
al., 1995; Bellisa io e al., 1999; Mas epano e al., 2013).
In d ained pea land o es s, signi ican co ela ions be ween
CH4 lux and empe a u e ha e been ound in CH4-emi ing
ecosys ems (Nykänen e al., 1998; Minkkinen and Laine,
2006), al hough he di ec ion o he co ela ion has been
ound o di e be ween ens and bogs. In con as , no signi -
ican co ela ions wi h empe a u e we e ound in pea land
o es s ha mainly showed CH4up ake (Ojanen e al., 2010;
Wang e al., 2013). Bo h CH4emission and up ake ha e been
ound o co ela e wi h WTL (e.g., Bellisa io e al., 1999;
Ojanen e al., 2010).
In addi ion o empe a u e, WTL, and WS, CH4 lux has
been ound o co ela e wi h ee s and olume (Ojanen e
al., 2010; Minkkinen e al., 2007), which is, in con as , an
indi ec measu e o he WTL. Also, PAR has been obse ed
o co ela e posi i ely wi h CH4emissions in a Sphagnum-
domina ed mixed mi e (Mikkelä e al., 1995). In his s udy,
howe e , he co ela ions wi h PAR we e low o absen .
5 Conclusions
In his pape , we ha e p esen ed a 2-yea da a se o CH4
exchange measu ed a he o es loo o a bo eal o es y-
d ained en. These esul s show ha au oma ed chambe
measu emen s wi h an accu a e online gas analyze make i
possible o obse e small CH4 luxes e en du ing he win e
wi h snow co e . Al hough he luxes a ou si e we e ela-
i ely low h oughou he yea , we succeeded in ca ching he
annual cycle in he CH4up ake. Ou esul s indica e ha he
o es loo o his pea land si e ac ed as a small annual CH4
sink (mean balance −219 mg CH4m−2y −1), al hough com-
ple ing he balance wi h he emissions om di ches indica es
ha he si e is likely o be a small CH4sou ce.
In spi e o he low lux de ec ion limi o he measu emen
sys em, ou esul s indica e ha i is necessa y o pay a en-
ion o he lux calcula ion me hods, and ins ead o choosing
be ween linea and exponen ial i s we decided on a combi-
na ion o bo h. E en hough he i based on linea eg ession
was obse ed o ypically esul in a smalle lux han an ex-
ponen ial i , i s use was jus i ied o low luxes by showing
ha i p oduced mo e obus es ima es when he concen a-
ion change du ing chambe closu e was small and hus mo e
a ec ed by measu emen noise. In addi ion, we demons a ed
ha bo h he leng h o he i ing pe iod and he s a ing ime
o his window had a signi ican e ec on he lux es ima es
and hus canno be selec ed a bi a ily.
The CH4up ake, measu ed wi h closed chambe s, was ob-
se ed o co ela e wi h wind speed, which caused a co e-
sponding diu nal cycle. Howe e , his was pa ly a ibu ed
o ae odynamic e ec s due o chambe closu e, which a e
dependen on a mosphe ic mixing p io o he closu e. Thus,
he chambe cons uc ion could be po en ially imp o ed by
adjus ing he chambe an speed acco ding o he ambien
wind speed. As his a ia ion is pa ly ela ed o changes in
he soil CH4s o age, he e o in oduced in he annual bal-
ance es ima ed om sho - e m luxes can be diminished by
con inuous measu emen s ully co e ing he diu nal cycle.
Con inuous long- e m measu emen s also acili a e he anal-
ysis o he en i onmen al ac o s ha con ol CH4exchange.
Howe e , in o de o unde s and he biological p ocesses in-
ol ed in CH4p oduc ion and oxida ion, i.e., he p ocesses
behind he ne CH4 lux obse ed, addi ional measu emen s
a e necessa y, ocusing on he p oduc ion and oxida ion po-
en ials and he wi hin-soil gas g adien s.
Since he conside a ions o he measu emen sys em pe -
o mance a e si e and sys em speci ic, we ecommend ha
any u u e s udy should add ess he p ocedu es in ol ed in
lux calcula ion, including he i ing me hod and he leng h
and delay o he i ing pe iod, based on he analysis p e-
sen ed abo e. In pa icula , we ecommend using he lux
limi me hod applied in his s udy, i.e., using linea eg ession
o low luxes and exponen ial eg ession o luxes abo e a
h eshold o be de e mined.
www.biogeosciences.ne /14/1947/2017/ Biogeosciences, 14, 1947–1967, 2017
1964 M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo
Da a a ailabili y. The calcula ed luxes and measu ed headspace
empe a u es a e included in he Supplemen . The es o he me e-
o ological da a can be eques ed om he co esponding au ho .
The Supplemen ela ed o his a icle is a ailable online
a doi:10.5194/bg-14-1947-2017-supplemen .
Compe ing in e es s. The au ho s decla e ha hey ha e no con lic
o in e es .
Acknowledgemen s. We a e g a e ul o he inancial suppo om
he Maj and To Nessling ounda ion and om he Minis y o
T anspo and Communica ions h ough he In eg a ed Ca bon
Obse ing Sys em (ICOS) esea ch. We would also like o hank
Pen i A man and Te o Hi onen o hei help in da a ea men
and o measu emen s a he si e.
Edi ed by: U. Seib
Re iewed by: C.-M. Gö es and M. Mas epano
Re e ences
Alm, J., Saa nio, S., Nykänen, H., Sil ola, J., and Ma ikainen,
P. J.: Win e CO2, CH4and N2O luxes on some na u al
and d ained bo eal pea lands, Biogeochemis y, 44, 163–186,
doi:10.1023/A:1006074606204, 1999.
Alm, J., Shu pali, N., Tui ila, E.-S., Lau ila, T., Maljanen, M.,
Saa nio, S., and Minkkinen, K.: Me hods o de e mining emis-
sion ac o s o he use o pea and pea lands – lux measu emen s
and modelling, Bo eal En i on. Res., 12, 85–100, 2007.
An hony, W. H., Hu chinson, G. L., and Li ings on, G. P.: Cham-
be measu emen s o soil-a mosphe e gas exchange: linea s.
di usion-based lux models, Soil Sci. Soc. Am. J., 59, 1308–
1310, doi:10.2136/sssaj1995.03615995005900050015x, 1995.
Bellisa io, L. M., Bubie , J. L., Moo e, T. R., and Chan on, J. P.:
Con ols on CH4emissions om a no he n pea land, Global
Biogeochem. Cy., 13, 81–91, doi:10.1029/1998GB900021,
1999.
Be gie , I., Rech, R., Mon ei o, H., and So iano, E.: Use e ec s
on chambe ni ous oxide emissions om oxisol soils in no-
illage maize e i iga ed wi h anae obically diges ed swine ma-
nu e, En i onmen and Na u al Resou ces Resea ch, 3, 1–11,
doi:10.5539/en . 3n4p1, 2013.
Bhuiyan, R., Minkkinen, K., Helmisaa i, H.-S., Ojanen, P., Pen ilä,
T., and Laiho, R.: Es ima ing ine- oo p oduc ion by ee species
and unde s o ey unc ional g oups in wo con as ing pea land
o es s, Plan Soil, 412, 299–316, doi:10.1007/s11104-016-3070-
3, 2017.
Boeckx, P. and Van Cleempu , O.: Me hane oxida ion in a neu-
al land ill co e soil: In luence o mois u e con en , empe -
a u e and ni ogen- u no e , J. En i on. Qual., 25, 178–183,
doi:10.2134/jeq1996.00472425002500010023x, 1996.
Bubie , J. L. and Moo e, T. R.: An ecological pe spec i e on
me hane emissions om no he n we lands, T ends Ecol. E ol.,
9, 460–464, doi:10.1016/0169-5347(94)90309-3, 1994.
Ch is iansen, J. R., Ko honen, J. F. J., Juszczak, R., Giebels, M., and
Pihla ie, M.: Assessing he e ec s o chambe placemen , manual
sampling and headspace mixing on CH4 luxes in a labo a o y
expe imen , Plan Soil, 343, 171–185, doi:10.1007/s11104-010-
0701-y, 2011.
Ch is iansen, J. R., Ou hwai e, J., and Smukle , S. M.: Compa -
ison o CO2, CH4, and N2O soil-a mosphe e exchange mea-
su ed in s a ic chambe s wi h ca i y ing-down spec oscopy
and gas ch oma og aphy, Ag . Fo es Me e ol., 211, 48–57,
doi:10.1016/j.ag o me .2015.06.004, 2015.
Ch is ophe sen, M. and Kjeldsen, P.: La e al gas ans-
po in soil adjacen o an old land ill: ac o s go e n-
ing gas mig a ion, Was e Manage. Res., 19, 579–594,
doi:10.1177/0734242X0101900615, 2001.
Cla ke, D. and Rieley, J.: S a egy o esponsible pea land manage-
men , In e na ional Pea Socie y, Jy äskylä, Finland, 11, 2010.
Conen, F. and Smi h, K. A.: An explana ion o linea inc eases in
gas concen a ion unde closed chambe s used o measu e gas
exchange be ween soil and he a mosphe e, Eu . J. Soil. Sci., 51,
111–117, doi:10.1046/j.1365-2389.2000.00292.x, 2000.
Da idson, E. A., Sa age, K., Ve cho , L. V., and Na a o, R.:
Minimizing a i ac s and biases in chambe -based measu e-
men s o soil espi a ion, Ag . Fo es Me eo ol., 113, 21–37,
doi:10.1016/S0168-1923(02)00100-4, 2002.
Denman, K. L., B asseu , G., Chid haisong, A., Ciais, P., Cox, P.
M., Dickinson, R.E., Hauglus aine, D., Heinze, C., Holland, E.,
Jacob, D., Lohmann, U., Ramachand an, S., da Sil a Dias, P. L.,
Wo sy, S. C., and Zhang, X.: Couplings Be ween Changes in he
Clima e Sys em and Biogeochemis y, in: Clima e Change 2007:
The Physical Science Basis. Con ibu ion o Wo king G oup I o
he Fou h Assessmen Repo o he In e go e nmen al Panel on
Clima e Change, edi ed by: Solomon, S., Qin, D., Manning, M.,
Chen, Z., Ma quis, M., A e y , K. B., Tigno , M., and Mille , H.
L., Camb idge Uni e si y P ess, Camb idge, UK and New Yo k,
NY, USA, 2007.
Denmead, O.: App oaches o measu ing luxes o me hane and ni-
ous oxide be ween landscapes and he a mosphe e, Plan Soil,
309, 5–24, doi:10.1007/s11104-008-9599-z, 2008.
Dun ield, P., Knowles, R., Dumon , R., and Moo e, T. R.: Me hane
p oduc ion and consump ion in empe a e and suba c ic pea
soils: Response o empe a u e and pH, Soil Biol. Biochem., 25,
321–326, doi:10.1016/0038-0717(93)90130-4, 1993.
Du au , L. and Ve cho , V.: A global in en o y o he
soil CH4sink, Global Biogeochem. Cy., 21, GB4013,
doi:10.1029/2006GB002734, 2007.
Fassbinde , J. J., Schul z, N. M., Bake , J. M., and G i is,
T. J.: Au oma ed, low-powe chambe sys em o measu -
ing ni ous oxide emissions, J. En i on. Qual., 42, 606–614,
doi:10.2134/jeq2012.0283, 2013.
Fischlin, A., Midgley, G. F., P ice, J. T., Leemans, R., Gopal, B.,
Tu ley, C., Rounse ell, M. D. A., Dube, O. P., Ta azona, J., and
Velichko, A. A.: Ecosys ems, hei p ope ies, goods, and se -
ices, in: Clima e Change 2007: Impac s, Adap a ion and Vul-
ne abili y. Con ibu ion o wo king g oup II o he Fou h Assess-
men Repo o he In e go e nmen al Panel on Clima e Change,
edi ed by: Pa y, M. L., Canziani, O. F., Palu iko , J. P., an de
Linden, P. J., and Hanson, C. E., Camb idge Uni e si y P ess,
Camb idge, 211–272, 2007.
Biogeosciences, 14, 1947–1967, 2017 www.biogeosciences.ne /14/1947/2017/
M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo 1965
Fo b ich, I., Ku zbach, L., Ho mann, A., and Wilmking, M.: A
compa ison o linea and exponen ial eg ession o es ima -
ing di usi e CH4 luxes by closed-chambe s in pea lands, Soil
Biol. Biochem., 42, 507–515, doi:10.1016/j.soilbio.2009.12.004,
2010.
Healy, R. W., S iegl, R. G., Russell, T. F., Hu chinson, G. L.,
and Li ings on, G. P.: Nume ical e alua ion o s a ic-chambe
measu emen s o soil-a mosphe e gas exchange: iden i ica ion
o physical p ocesses, Soil Sci. Soc. Am. J., 60, 740–747,
doi:10.2136/sssaj1996.03615995006000030009x, 1996.
Hu chinson, G. L., Li ings on, G. P., Healy, R. W., and S iegl, R.
G.: Chambe measu emen o su ace-a mosphe e ace gas ex-
change: nume ical e alua ion o dependence on soil, in e acial
laye , and sou ce/sink p ope ies, J. Geophys. Res.-A mos., 105,
8865–8875, doi:10.1029/1999JD901204, 2000.
IPCC: Clima e Change 2014: Impac s, Adap a ion, and Vulne abil-
i y. Pa A: Global and Sec o al Aspec s. Con ibu ion o Wo k-
ing G oup II o he Fi h Assessmen Repo o he In e go e n-
men al Panel on Clima e Change, edi ed by: Field, C. B., Ba os,
V. R., Dokken, D. J., Mach, K. J., Mas and ea, M. D., Bili , T.
E., Cha e jee, M., Ebim, K. L., Es ada, Y. O., Geno a, R. C.,
Gi ma, B., Kissel, E. S., Le y, A. N., MacC acken, S., Mas an-
d ea, P. R., and Whi e, L. L., Camb idge Uni e si y P ess, Cam-
b idge, UK and New Yo k, NY, USA, 2014.
Jones, S. K., Famula i, D., Di Ma co, C. F., Nemi z, E., Skiba,
U. M., Rees, R. M., and Su on, M. A.: Ni ous oxide emis-
sions om managed g assland: a compa ison o eddy co a iance
and s a ic chambe measu emen s, A mos. Meas. Tech., 4, 2179–
2194, doi:10.5194/am -4-2179-2011, 2011.
Kelle , M., Mi e, M. E., and S alla d, R. F.: Consump ion o
a mosphe ic me hane in soils o cen al Panama: E ec s o
ag icul u al de elopmen , Global Biogeochem. Cy., 4, 21–27,
doi:10.1029/GB004i001p00021, 1990.
Koskinen, M., Minkkinen, K., Ojanen, P., Kämä äinen, M., Lau-
ila, T., and Lohila, A.: Measu emen s o CO2exchange wi h
an au oma ed chambe sys em h oughou he yea : challenges
in measu ing nigh - ime espi a ion on po ous pea soil, Biogeo-
sciences, 11, 347–363, doi:10.5194/bg-11-347-2014, 2014.
Ko syu benko, O. R., Chin, K.-J., Glagole , M. V., S ubne , S.,
Simanko a, M. V., Nozhe niko a, A. N., and Con ad, R.:
Ace oclas ic and hyd ogeno ophic me hane p oduc ion and
me hanogenic popula ions in an acidic Wes -Sibe ian pea
bog, En i on. Mic obiol., 6, 1159–1173, doi:10.1111/j.1462-
2920.2004.00634.x, 2004.
K oon, P. S., Hensen, A., Bulk, W. C. M., Jongejan, P. A. C.,
and Ve meulen, A. T.: The impo ance o educing he sys-
ema ic e o due o non-linea i y in N2O lux measu emen s
by s a ic chambe s, Nu . Cycl. Ag oecosys., 82, 175–186,
doi:10.1007/s10705-008-9179-x, 2008.
Ku zbach, L., Schneide , J., Sachs, T., Giebels, M., Nykänen, H.,
Shu pali, N. J., Ma ikainen, P. J., Alm, J., and Wilmking,
M.: CO2 lux de e mina ion by closed-chambe me hods can
be se iously biased by inapp op ia e applica ion o linea e-
g ession, Biogeosciences, 4, 1005-1025, doi:10.5194/bg-4-1005-
2007, 2007.
Laine, A., So oco nola, M., Kiely, G., By ne, K. A., Wilson, D., and
Tui ila, E.-S.: Es ima ing ne ecosys em exchange in a pa e ned
ecosys em: Example om blanke bog, Ag . Fo es Me eo ol.,
18, 231–243, doi:10.1016/j.ag o me .2006.05.005, 2006.
La mola, T., Tui ila, E.-S., Tii ola, M., Nykänen, H., Ma ikainen,
P. J., Y jälä, K., Tuomi i a, T., and F i ze, H.: The ole o Sphag-
num mosses in he me hane cycling o a bo eal mi e, Ecology, 91,
2356–2365, doi:10.1890/09-1343.1, 2010.
Le y, P. E., G ay, A., Leeson, S. R., Gaiawyn, J., Kelly, M. P. C.,
Coope , M. D. A., Dinsmo e, K. J., Jones, S. K., and Sheppa d,
L. J.: Quan i ica ion o unce ain y in ace gas luxes measu ed
by he s a ic chambe me hod, Eu . J. Soil Sci., 62, 811–821,
doi:10.1111/j.1365-2389.2011.01403.x, 2011.
Li ings on, G. P., Hu chinson, G. L., and Spa alian, K.:
Di usion heo y imp o es chambe -based measu emen s o
ace gas emissions, Geophys. Res. Le ., 32, L24817,
doi:10.1029/2005GL024744, 2005.
Li ings on, G. P., Hu chinson, G. L., and Spa alian, K.: T ace gas
emission in chambe s: a non-s eady-s a e di usion model, Soil
Sci. Soc. Am. J., 70, 1459–1469, doi:10.2136/sssaj2005.0322,
2006.
Lohila, A., Minkkinen, K., Au ela, M., Tuo inen, J.-P., Pen ilä, T.,
Ojanen, P., and Lau ila, T.: G eenhouse gas lux measu emen s
in a o es y-d ained pea land indica e a la ge ca bon sink, Bio-
geosciences, 8, 3203–3218, doi:10.5194/bg-8-3203-2011, 2011.
Lohila, A., Aal o, T., Au ela, M., Ha akka, J., Tuo inen, J.-P.,
Kilkki, J., Pen ilä, T., Vuo enmaa, J., Hänninen, P., Su inen, R.,
Viisanen, R., and Lau ila, T.: La ge con ibu ion o bo eal upland
o es soils o a ca chmen -scale CH4balance in a we yea , Geo-
phys. Res. Le ., 43, 2946–2953, doi:10.1002/2016GL067718,
2016.
Luan, J. and Wu, J.: Long- e m ag icul u al d ainage s imula es
CH4emissions om di ches h ough inc eased subs a e a ail-
abili y in a bo eal pea land, Ag . Ecosys . En i on., 214, 68–77,
doi:10.1016/j.agee.2015.08.020, 2015.
Ma ikainen, P. J., Nykänen, H., Alm, J., and Sil ola, J.: Change
in luxes o ca bon dioxide, me hane and ni ous oxide due o
o es d ainage o mi e si es o di e en ophy, Plan Soil, 168,
571–577, doi:10.1007/BF00029370, 1995.
Mas epano , M., Sigsgaa d, C., Tagesson, T., S öm, L., Tams o ,
M. P., Lund, M., and Ch is ensen, T. R.: Re isi ing ac o s con-
olling me hane emissions om high-A c ic und a, Biogeo-
sciences, 10, 5139–5158, doi:10.5194/bg-10-5139-2013, 2013.
Mikkelä, C., Sundh, I., S ensson, B. H., and Nilsson, M.: Diu -
nal a ia ion in me hane emission in ela ion o he wa e able,
soil empe a u e, clima e and ege a ion co e in a Swedish acid
mi e, Biogeochemis y, 28, 93–114, doi:10.1007/BF02180679,
1995.
Minkkinen, K. and Laine, J.: Vege a ion he e ogenei y and
di ches c ea e spa ial a iabili y in me hane luxes om
pea lands d ained o o es y, Plan Soil, 285, 289–304,
doi:10.1007/s11104-006-9016-4, 2006.
Minkkinen, K., Laine, J., Nykänen, H., and Ma ikainen, P. J.:
Impo ance o d ainage di ches in emissions o me hane om
mi es d ained o o es y, Can. J. Fo es Res., 27, 949–952,
doi:10.1139/cj -27-6-949, 1997.
Minkkinen, K., Pen ilä, T., and Laine, J.: T ee s and olume as a
scala o me hane luxes in o es y-d ained pea lands in Fin-
land, Bo eal En i on. Res., 12, 127–132, 2007.
Mohan y, S. R., Bodelie , P. L. E., and Con ad, R.: E ec o em-
pe a u e on composi ion o he me hano ophic communi y in
ice ield and o es soil, FEMS Mic obiol. Ecol., 62, 24–31,
doi:10.1111/j.1574-6941.2007.00370.x, 2007.
www.biogeosciences.ne /14/1947/2017/ Biogeosciences, 14, 1947–1967, 2017
1966 M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo
Mon ana ella, L., Jones, R. J. A., and Hiede e , R.: The dis ibu ion
o pea land in Eu ope, Mi es and Pea , 1, 1–10, 2006.
Moo e, T. R. and Knowles, R.: The in luence o wa e able le els
on me hane and ca bon dioxide emissions om pea land soils,
Can. J. Soil Sci., 69, 33–38, doi:10.4141/cjss89-004, 1989.
Nakano, T., Sawamo o, T., Mo ishi a, T., Inoue, G., and
Ha ano, R.: A compa ison o eg ession me hods o es-
ima ing soil-a mosphe e di usion gas luxes by a closed
chambe echnique, Soil Biol. Biochem., 36, 107–113,
doi:10.1016/j.soilbio.2003.07.005, 2004.
Nicke son, N.: E alua ing gas emission measu emen s using Min-
imum De ec able Flux (MDF), Eosense Inc., Da mou h, No a
Sco ia, Canada, 2016.
Nykänen, H., Alm, J., Lång, K., Sil ola, J., and Ma ikainen, P.
J.: Emissions o CH4, N2O and CO2 om a i gin en and a
en d ained o g assland in Finland, J. Biogeog ., 22, 351–357,
doi:10.2307/2845930, 1995.
Nykänen, H., Alm, J., Sil ola, J., Tolonen, K., and Ma ikainen,
P. J.: Me hane luxes on bo eal pea lands o di e en e ili y
and he e ec o long- e m expe imen al lowe ing o he wa-
e able on lux a es, Global Biogeochem. Cy., 12, 53–69,
doi:10.1029/97GB02732, 1998.
Ojanen, P., Minkkinen, K., Alm, J., and Pen ilä, T.: Soil–
a mosphe e CO2, CH4and N2O luxes in bo eal o es y-
d ained pea lands, Fo es Ecol. Manage., 260, 411–421,
doi:10.1016/j. o eco.2010.04.036, 2010.
Päi änen, J. and Hånell, B.: Pea land Ecology and Fo es y – a
Sound App oach, Uni e si y o Helsinki Depa men o Fo es
Sciences Publica ions 3, 21 pp., 2012.
Pede sen, A. R., Pe e sen, S. O., and Schelde, K.: A comp ehensi e
app oach o soil-a mosphe e ace-gas lux es ima ion wi h s a ic
chambe s, Eu . J. Soil Sci., 61, 888–902, doi:10.1111/j.1365-
2389.2010.01291.x, 2010.
Pihla ie, M., Ch is iansen, J. R., Aal onen, H., Ko honen, J. F.,
No dbo, A., Rasilo, T., Benan i, G., Giebels, M., Helmy, M.,
Sheehy, J., Jones, S., Juszczak, R., Kle o h, R., Lobo-do Vale,
R., Rosa, A. P., Sch eibe , P., Se ça, D., Vicca, S., Wol , B.,
and Pumpanen, J.: Compa ison o s a ic chambe s o measu e
CH4emissions om soils, Ag . Fo es Me eo ol., 171, 124–136,
doi:10.1016/j.ag o me .2012.11.008, 2013.
Pi inen, P., Simola, H., Aal o, J., Kauko an a, J.-P., Ka lsson, P.,
and Ruuhela, R.: Clima ological s a is ics o Finland 1981–2010,
Repo s, 2012:1, Finnish Me eo ological Ins i u e, Helsinki, Fin-
land, 2012.
Pi k, N., Mas epano , M., Pa men ie , F.-J. W., Lund, M., C ill,
P., and Ch is ensen, T. R.: Calcula ions o au oma ic cham-
be lux measu emen s o me hane and ca bon dioxide using
sho ime se ies o concen a ions, Biogeosciences, 13, 903–
912, doi:10.5194/bg-13-903-2016, 2016.
Pumpanen, J., Kola i, P., Il esniemi, H., Minkkinen, K., Vesala, T.,
Niinis ö, S., Lohila, A., La mola, T., Mo e o, M., Pihla ie, M.,
Janssens, I. A., Yus e, J. C., G ünzweig, J. M., Re h, S., Subke,
J.-A., Sa age, K., Ku sch, W., Øs eng, G., Ziegle , W., An honi,
P., Lind o h, A., and Ha i, P.: Compa ison o di e en chambe
echniques o measu ing soil CO2e lux, Ag . Fo es Me eo ol.,
123, 159–176, doi:10.1016/j.ag o me .2003.12.001, 2004.
Ridgwell, A. J., Ma shall, S. J., and G egson, K.: Consump ion o
a mosphe ic me hane by soils: a p ocess-based model, Global
Biogeochem. Cy., 13, 59–70, doi:10.1029/1998GB900004,
1999.
Roule , N., Moo e, T. R., Bubie , J., and La leu , P.: No he n ens:
me hane lux and clima ic change, Tellus B, 44, 100–105, 1992.
Sab eko , A. F., Glagole , M. V., Alekseychik, P. K., Smolen -
se , B. A., Te en ie a, I. E., K i enok, L. A., and Maksyu o ,
S. S.: A p ocess-based model o me hane consump ion by up-
land soils, En i on. Res. Le ., 11, 075001, doi:10.1088/1748-
9326/11/7/075001, 2016.
Sa age, K., Moo e, T. R., and C ill, P. M.: Me hane and ca -
bon dioxide exchanges be ween he a mosphe e and no h-
e n bo eal o es soils, J. Geophys. Res., 102, 29279–29288,
doi:10.1029/97JD02233, 1997.
Sa age, K., Phillips, R., and Da idson, E.: High empo al equency
measu emen s o g eenhouse gas emissions om soils, Biogeo-
sciences, 11, 2709–2720, doi:10.5194/bg-11-2709-2014, 2014.
Scheu z, C. and Kjeldsen, P.: En i onmen al ac o s in lu-
encing a enua ion o me hane and hyd ochlo o luo oca -
bons in land ill co e soils, J. En i on. Qual., 33, 72–79,
doi:10.2134/jeq2004.0072, 2004.
Sege s, R.: Me hane p oduc ion and me hane consump ion: a e iew
o p ocesses unde lying we land me hane luxes, Biogeochem-
is y, 41, 23–51, doi:10.1023/A:1005929032764, 1998.
Singh, J. S., Singh, S., Raghubanshi, A. S., Singh, S., Kashyap, A.
K., and Reddy, V. S.: E ec o soil ni ogen, ca bon and mois u e
on me hane up ake by d y opical o es soils, Plan Soil, 196,
115–121, doi:10.1023/A:1004233208325, 1997.
Sundq is , E., Mölde , M., C ill, P., Kljun, N., and Lind o h,
A.: Me hane exchange in a bo eal o es es ima ed by g adi-
en me hod, Tellus B, 67, 26688, doi:10.3402/ ellusb. 67.26688,
2015.
Smi h, K. A., Dobbie, K. E., Ball, B. C., Bakken, L. R., Si aula,
B. K., Hansen, S., B umme, R., Bo ken, W., Ch is ensen, S.,
P iemé, A., Fowle , D., Macdonald, J. A., Skiba, U., Klemed s-
son, L., Kasimi -Klemed sson, A., Degó ska, A., and O lan-
ski, P.: Oxida ion o a mosphe ic me hane in No he n Eu o-
pean soils, compa ison wi h o he ecosys ems, and unce ain ies
in he global e es ial sink, Glob. Change Biol., 6, 791–803,
doi:10.1046/j.1365-2486.2000.00356.x, 2000.
Ueyama, M., Takeuchi, R., Takahashi, Y., Ide, R., A aka, M.,
Kosugi, Y., Takahashi, K., and Saigusa, N.: Me hane up-
ake in a empe a e o es soil using con inuous closed-
chambe measu emen s, Ag . Fo es Me eo ol., 213, 1–9,
doi:10.1016/j.ag o me .2015.05.004, 2015.
Ven e ea, R. T.: Theo e ical compa ison o ad anced me h-
ods o calcula ing ni ous oxide luxes using non-s eady
s a e chambe s, Soil Sci. Soc. Am. J., 77, 709–720,
doi:10.2136/sssaj2013.01.0010, 2013.
Ven e ea, R. T., Spokas, K. A., and Bake , J. M.: Accu-
acy and p ecision analysis o chambe -based ni ous oxide
gas lux es ima es, Soil Sci. Soc. Am. J., 73, 1087–1093,
doi:10.2136/sssaj2008.0307, 2009.
Biogeosciences, 14, 1947–1967, 2017 www.biogeosciences.ne /14/1947/2017/
M. Ko kiakoski e al.: Me hane exchange a he pea land o es loo 1967
Wang, J. M., Mu phy, J. G., Geddes, J. A., Winsbo ough, C. L.,
Basiliko, N., and Thomas, S. C.: Me hane luxes measu ed by
eddy co a iance and s a ic chambe echniques a a empe a e
o es in cen al On a io, Canada, Biogeosciences, 10, 4371–
4382, doi:10.5194/bg-10-4371-2013, 2013.
Wang, Z. P., DeLaune, R. D., Masscheleyn, P. H., and Pa ick,
W. H.: Soil edox and pH e ec s on me hane p oduc ion
in a looded ice soil, Soil Sci. Soc. Am., 57, 382–385,
doi:10.2136/sssaj1993.03615995005700020016x, 1993.
www.biogeosciences.ne /14/1947/2017/ Biogeosciences, 14, 1947–1967, 2017