RESEARCH ARTICLE
The Snow Mus Go On: G ound Ice
Encasemen , Snow Compac ion and Absence
o Snow Di e en ly Cause Soil Hypoxia, CO
2
Accumula ion and T ee Seedling Damage in
Bo eal Fo es
F ançoise Ma z
1
*, Jaana Vuosku
1
, Anu O askainen
1
, Sa i S a k
1,2
, Pasi Rau io
1
1Na u al Resou ces Ins i u e Finland (Luke), Ro aniemi, Finland, 2A c ic Cen e, Uni e si y o Lapland,
Ro aniemi, Finland
*[email p o ec ed]
Abs ac
A high la i udes, he clima e has wa med a wice he a e o he global a e age wi h mos
changes obse ed in au umn, win e and sp ing. Inc easing win e empe a u es and wide
empe a u e luc ua ions a e leading o mo e equen ain-on-snow e en s and eeze- haw
cycles causing snow compac ion and o ma ion o ice laye s in he snowpack, hus c ea ing
ice encasemen (IE). By dec easing he snowpack insula ion capaci y and es ic ing soil-
a mosphe e gas exchange, modi ica ion o he snow p ope ies may lead o colde soil bu
also o hypoxia and accumula ion o ace gases in he subni ean en i onmen . To es he
e ec s o hese o e win e ing condi ions changes on plan win e su i al and g ow h, we
es ablished a snow manipula ion expe imen in a coni e ous o es in No he n Finland wi h
No way sp uce and Sco s pine seedlings. In addi ion o ambien condi ions and p e en ion
o IE, we applied h ee snow manipula ion le els: IE c ea ed by a i icial ain-on-snow
e en s, snow compac ion and comple e snow emo al. Snow emo al led o deepe soil
os du ing win e , bu no clea e ec o IE o snow compac ion done in ea ly win e was
obse ed on soil empe a u e. Hypoxia and accumula ion o CO
2
we e highes in he IE
plo s bu , mo e impo an ly, he du a ion o CO
2
concen a ion abo e 5% was 17 days in IE
plo s compa ed o 0 days in ambien plo s. IE was he mos damaging win e condi ion o
bo h species, dec easing he p opo ion o heal hy seedlings by 47% o sp uce and 76% o
pine compa ed o ambien condi ions. Seedlings in all h ee ea men s ended o g ow less
han seedlings in ambien condi ions bu only IE had a signi ican e ec on sp uce g ow h.
Ou esul s demons a e a nega i e impac o win e clima e change on bo eal o es egen-
e a ion and p oduc i i y. Changing snow condi ions may hus pa ially mi iga e he posi i e
e ec o inc easing g owing season empe a u es on bo eal o es p oduc i i y.
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 1/18
a11111
OPEN ACCESS
Ci a ion: Ma z F, Vuosku J, O askainen A, S a k S,
Rau io P (2016) The Snow Mus Go On: G ound Ice
Encasemen , Snow Compac ion and Absence o
Snow Di e en ly Cause Soil Hypoxia, CO
2
Accumula ion and T ee Seedling Damage in Bo eal
Fo es . PLoS ONE 11(6): e0156620. doi:10.1371/
jou nal.pone.0156620
Edi o : Ch is ian Rixen, WSL Ins i u e o Snow and
A alanche Resea ch SLF, SWITZERLAND
Recei ed: Janua y 22, 2016
Accep ed: May 17, 2016
Published: June 2, 2016
Copy igh : © 2016 Ma z e al. This is an open
access a icle dis ibu ed unde he e ms o he
C ea i e Commons A ibu ion License, which pe mi s
un es ic ed use, dis ibu ion, and ep oduc ion in any
medium, p o ided he o iginal au ho and sou ce a e
c edi ed.
Da a A ailabili y S a emen : All ele an da a a e
wi hin he pape and i s Suppo ing In o ma ion iles.
Funding: This wo k was suppo ed by Suomen
aka emia (www.aka. i, decision 267092, o PR):
comple e p ojec unding ( ield expe imen
implemen a ion, ma e ial collec ion, da a analysis,
p epa a ion o manusc ip ) and Suomen
Kul uu i ahas o, Lapin Rahas o (www.sk . i, g an
40121486, 2010-2011, o FM): unning a pilo ield
expe imen , ma e ial collec ion and da a analysis.
In oduc ion
A high la i udes, he clima e has wa med a wice he a e o he global a e age, and as a conse-
quence, no he n o es plan s a e expe iencing en i onmen al condi ions ha a e highly a i-
able compa ed o in he pas . The inc ease in empe a u e does no conce n only he g owing
season. In ac , empe a u es du ing he win e mon hs in Finland ha e inc eased e en mo e
han hose du ing he g owing season [1]. Inc easing win e empe a u es ha e se e al impac s
on he p ope ies o snowpack, especially due o wa m spells and ain-on-snow e en s. Da a
has shown ha he equency o ain-on-snow e en s has inc eased in A c ic egions du ing
he las 30 yea s [2] and empe a u es a e p ojec ed o inc ease e en u he in he u u e [3]. In
Finland, he du a ion and hickness o snow co e a e expec ed o dec ease and models p edic
lowe snow wa e equi alen and highe ice con en in he snowpack by he end o he 21s cen-
u y [4]. Al hough bo eal win e condi ions a e changing apidly, su p isingly ew expe imen-
al in es iga ions exis on he e ec s o al e ed snow condi ions on bo eal o es plan s o hei
win e su i al.
Snowpack p ope ies a e o special impo ance o ee seedlings and unde s o y o es
plan s because hey a e co e ed by snow o he whole win e . Nume ous cellula p ocesses a e
supp essed in o e win e ing plan s upon en y in o do mancy in he au umn, howe e main e-
nance o some me abolism is necessa y o sus ain de ense and escue p ocesses in do man is-
sues [5]. Simila o abo eg ound pa s, main enance espi a ion in oo s may con inue
h oughou win e e en a nea - eezing empe a u es, e en hough oo g ow h is conside ed
o be limi ed below abou 4°C [6,7]. Changes in he snowpack al e he en i onmen al o e -
win e ing condi ions o coni e seedlings in wo impo an ways. Fi s ly, a hinne o a dense
snowpack has a lowe insula ion capaci y, which may conside ably dec ease subni ean empe -
a u es du ing he win e . The insula ing snowpack decouples soil empe a u es om ai em-
pe a u es, allowing soil p ocesses gene ally o con inue h oughou he win e and p o iding
os p o ec ion o soil mic oo ganisms, oo s and auna. Secondly, inc easing equency o
ain-on-snow e en s and highe win e empe a u e luc ua ions lead o epea ed eeze- haw
e en s ha can c ea e g ound ice encasemen (IE). The consequences o IE ha e so a been
in es iga ed mos ly in no he n ag icul u al landscapes, whe e i is known o cause inju ies and
kill o e win e ing ce eals [8] and pe ennial g asses [9,10].
Ice laye s in he snow o on he g ound su ace es ic soil-a mosphe e gas exchange, which
leads o he de elopmen o hypoxia and accumula ion o ace gases in he subni ean en i on-
men [11,12]. Re u n o no moxia a e long exposu e o hypoxia also ep esen s a majo oxi-
da i e s ess ha can be a al o plan s [11]. Al hough he e ec s o he abo e changes in he
snowpack on soil empe a u es ha e been qui e well s udied, less is known abou hei in luence
on he amoun o oxygen in soil, and he po en ial consequences hese migh ha e on o e win-
e ing coni e seedlings. Gi en ha he snow-co e ed pe iod can las mo e han hal a yea in
bo eal ecosys ems, his c ea es a majo gap in he unde s anding o how clima e wa ming will
a ec bo eal ecosys em unc ioning.
We s udied he e ec s o changing snow condi ions on soil mic oclima e and soil p ope ies
as well as on he su i al and g ow h o coni e seedlings in a no he n bo eal coni e ous o es .
The s udy ocused on No way sp uce and Sco s pine as hey a e he wo mos common and
economically impo an ee species in Fennoscandian bo eal o es s. We subjec ed expe imen-
al plo s o se e al di e en snow manipula ion ea men s, which each ep esen ed di e en
scena ios o how win e clima e change may a ec snow condi ions and snowpack p ope ies.
By dec easing soil empe a u es o c ea ing hypoxia du ing he win e , we hypo hesized ha
snow co e emo al, snow compac ion and IE will inc ease win e mo ali y and dec ease he
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 2/18
Compe ing In e es s: The au ho s ha e decla ed
ha no compe ing in e es s exis .
eg ow h o young coni e seedlings du ing he ollowing summe , hus inducing nega i e
e ec s on bo h he p oduc i i y and he egene a ion o o es ees.
Ma e ial and Me hods
S udy si e and snow manipula ion ea men s
We es ablished a snow manipula ion expe imen ollowing a andomized block design wi h 10
blocks in a Sco s pine o es (xe ic- sub-xe ic si e) nea he A c ic Ci cle close o he ci y o
Ro aniemi (Ta i aa a, 66°25 '35"N 25°41'42"E). The land used o he ield expe imen
desc ibed below is owned by he Ro aniemi pa ish who allowed us o un he expe imen . The
ield s udy did no in ol e endange ed o p o ec ed species. The blocks we e posi ioned on an
a ea o app oxima ely h ee hec a es and o ganized in such a way ha wi hin each block he
unde s o ey ege a ion and ee laye we e as homogenous as possible. The size o he blocks
was ypically a ound 100–200 m
2
o allow ea men plo s (see below) wi hin each block o be
posi ioned a enough om each o he . Each block was enced o p e en big he bi o es o
en e he ea men plo s. The o es soil a he expe imen al si e is acidic wi h a soil pH o
3.76 ± 0.03 (mean ± SE o h ee sample collec ions o e he g owing season, n = 30). The con-
cen a ion o NH
4
-N, NO
3
-N and ex ac able o ganic ni ogen (N) in he humus laye was
5.7 ± 0.9, 1.2 ± 0.1 and 74.8 ± 7.3 mg kg DW
-1
(n = 30), espec i ely, in ambien (AMB) plo s
du ing he 2014 g owing season. G ound laye ege a ion is domina ed by ed-s emmed
ea he moss (Pleu ozium sch ebe i) and eindee lichens (Cladonia spp.) and ield laye ege-
a ion is domina ed by e e g een and deciduous dwa sh ubs (Vaccinium i is-idaea,V.my -
illus,V.uliginosum and Empe um nig um). Local empe a u e, p ecipi a ion and snow dep h
alues we e eco ded a he pe manen wea he s a ion o he Finnish Me eo ological Ins i u e
a Ro aniemi, Apukka, loca ed 15km no h o he ield si e (Wo ld Me eo ological O ganisa-
ion S a ion 02813, 66°34’N 26°00’E 106 m a.s.l.). The long- e m (1981–2010) mean annual
empe a u e in Apukka is 0.4°C wi h a mean p ecipi a ion o 556 mm [13](S1A Fig). Decem-
be , Janua y and Feb ua y a e he coldes mon hs o he yea . The i s pe manen snow ypi-
cally appea s in mid-No embe , he snowpack is he highes in Ap il wi h 60 cm (in open
a eas) and he a e age ending da e o pe manen snow co e is in ea ly May. Fo es s ypically
ha e less snow co e han open a eas [14]. To es ima e he local e ec o ee canopies and
s and s uc u e on he snow co e a he expe imen al si e, we measu ed he dep h o in ac
snow co e a ou ield si e in win e s 2015 and 2016 (19 Ma ch and 14 Ap il 2015, 19 Feb u-
a y, 31 Ma ch and 4 Ap il 2016). Compa ison wi h he da a in he WMO s a ion o Apukka
showed ha as an a e age, he snow co e in ou iled si e was abou hal ha in he open a ea
in Apukka (54.0 ± 8.7%, n = 5). Reco ding o ai empe a u e (2-m heigh ) showed ha du ing
No embe 2013–July 2014 he empe a u e was 0.61°C colde a ou o es si e han in he clos-
es WMO s a ion in Apukka du ing he same pe iod.
Fi e s udy plo s wi h simila exposu e, hyd ological s a us and ege a ion we e selec ed
wi hin each o he 10 blocks and andomly assigned o he ollowing ea men s: 1) ambien
(AMB) wi h no ea men , 2) p e en ion o IE o ma ion du ing win e (NoICE, plas ic shel e s
we e placed on a wood s uc u e o co e he plo in case o a na u ally occu ing ain-on-snow
e en ), 3) induc ion o IE (IE, a i icially c ea ed by snow wa e ing), 4) no snow co e (NoS-
NOW: oo s and low walls made o anslucen whi e plas ic ( hickness 0.2mm) we e posi-
ioned o e he plo s o p e en na u al snow all o d i ed snow) and 5) compac ion o snow
co e (COMP: snow compac ion by hand o educe he insula ing capaci y o snow). Plo s
we e loca ed below ee canopies ha we e as simila as possible. Plo s (1 m x 3 m) we e di ided
in o h ee sec ions: wo subplo s (A and B, bo h 1 m x 1 m) sepa a ed by a bu e zone (1 m x 1
m; Fig 1). The ea ed a ea ex ended 0.5 m beyond he edges o he plo . A subplo s we e used
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 3/18
o ake soil and biochemical samples om he seedlings in he beginning o he 2014 g owing
season o u he analyses. B subplo s we e used o eco d seedling g ow h and heal h du ing
he 2014 g owing season ( his s udy).
Snow wa e ing was done by applying cold ap wa e using plas ic wa e ing cans o e he
ex ended plo a ea (in o al 8 m
2
,Fig 1). Snow wa e ing was conduc ed h ee imes o e he
win e . The i s wa e ing was pe o med on Decembe 13, 2013 (3.75 l m
-2
)bu awa m
spell p e en ed ice o ma ion and hawed mos o snow and ice away. A second wa e ing
was pe o med on Decembe 17, 2013 (7.5 l m
-2
). A hi d wa e ing was done on Janua y 9,
2014 (7.5 l m
-2
). Wa e ing co esponded o a o al o 18.75 mm o ain. Snow compac ion
was applied by hand on Decembe 13, 2013, Janua y 9 and Janua y 29, 2014 wi h ca e no o
damage he seedlings. Ou aim being o mimic mainly ea ly win e wa ming, he snow
manipula ion ea men s we e done no la e han Janua y and na u al snow was co e ing
he IE and COMP plo s a e snow manipula ion. Plas ic shel e s on NoICE plo s we e pu
up se e al imes o e he win e du ing na u al ain-on-snow e en s: No embe 14–20,
Decembe 11–12, Decembe 25, 2013, Janua y 3, Ma ch 7–8and13–14, 2014. The plas ic
oo s o e he NoSNOW plo s we e se in place on Oc obe 24, 2013 and emo ed on Ap il
25, 2014. To limi any possible g eenhouse e ec , he open sides we e widened by cu ing
he uppe pa o plas ic walls in ea ly Ma ch when i was conside ed unlikely ha wind
would blow any snow in o he ea ed a ea. The whi ecolo o he anslucen plas ic oo s
allowed main aining he albedo o he NoSNOW plo s simila o he snow-co e ed plo s
un il he end o Ap il. Tempe a u e logge s we e ins alled 30 cm abo e he g ound on
Decembe 17, 2013 in AMB and NoSNOW plo s in h ee blocks o check o any e ec o
he plas ic oo s on ai empe a u e. Tempe a u e was eco ded e e y 10 min; he di e ence
be ween AMB and NoSNOW ea men s was compu ed o each ime poin and a e aged
pe hou . The esul s showed ha he p esence o he plas ic oo s inc eased he mean daily
ai empe a u e on a e age by 0.144 ± 0.004°C (n = 18008 measu emen s).
Ambien ai (2-m heigh ), g ound su ace and humus (2-cm dep h) empe a u e we e
eco ded a 1.5-h in e als using empe a u e logge s (EL-USB-1, Lasca elec onics) ins alled
in e e y plo o h ee andomly selec ed blocks. Soil os dep h was ollowed by ins alling os
ubes illed wi h me hylene blue solu ion in silicone ubes (2 m long) placed in o igh PVC
guide pipes inse ed e ically in o 2-m deep holes dug in he soil [15]. The soil os p obes
we e placed in he cen e o he bu e zone, close o he ai sampling ube and empe a u e log-
ge s (Fig 1). F os dep hs we e eco ded on a weekly basis om ea ly Decembe o mid-June.
Pic u es om all subplo s A (n = 50) we e aken on Ap il 24, 2014 wi h a digi al came a
(Sony α-300). The pe spec i e o he plan ed a ea was co ec ed in Pho oshop (CS3 e sion
Fig 1. Plo design. Red do = empe a u e logge ; ube = ai -collec ing silicon ube inse ed in humus laye ;
iangle = No way sp uce seedling; can = Sco s pine seedlings; blue a ea = ea ed a ea.
doi:10.1371/jou nal.pone.0156620.g001
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 4/18
10.0.1). Snow co e ed-a eas we e manually selec ed in ImageJ1.49B and he p opo ion o
snow co e was calcula ed as he a io o snow-co e ed pixels o he o al pixel numbe .
Soil gas sampling and analysis
O
2
and CO
2
concen a ions in soil we e measu ed egula ly o e he win e om Decembe 4,
2013 o May 14, 2014 when soil was ully hawed. The me hod o soil gas di usion in o silicon
ubes was used [16]. In au umn 2013, silicon ubes (in e nal diame e 1.0 cm, wall hickness 0.3
cm, leng h 100 cm, V = 78.5 cm
3
) closed on bo h ends wi h silicon s oppe s we e ins alled ho i-
zon ally in he middle o he bu e zone a a 2 cm dep h in he humus laye . Bended s ainless
s eel ubes (in e nal diame e 1 mm) inse ed in o a silicon s oppe in one side o he silicon
ubing allowed connec ing i o he su ace. The s eel ubes we e equipped wi h a h ee-way
s opcock o allow gas sampling 1 m abo e he soil su ace and ou side o he plo o a oid am-
pling when sampling. Gas samples o 30 ml we e aken wi h polyp opolene sy inges equipped
wi h a h ee-way s opcock, s o ed a -20°C o a maximum o 24 h and analyzed by gas ch o-
ma og aphy (Agilen 6890N) using a ShinCa bon ST 100/120 mesh 2 m x 1mm ID mic o-
packed column (Res ek). The e ec o s o age a -20°C in a polyp opolene sy inge on O
2
and
CO
2
concen a ions was es ed using s anda d gas mix u es (N
2
,CO
2
) o soil samples wi h
high CO
2
concen a ions bu no e ec was ound o s o age up o 48 h (da a no shown). Sam-
ples we e injec ed wi h a 1-ml loop. Running condi ions we e as ollows: 45°C 3 min, 45°C o
150°C a 120°C min
-1
, low 15ml min
-1
(He), TCD (200°C). In addi ion o soil gas sampling,
wo a mosphe ic ai samples we e collec ed e e y ime and used as echnical con ols o qual-
i y o GC analyses. A gas mix u e o 1% CO
2
, 19.8% O
2
in N
2
was used as a s anda d. Gas con-
cen a ions we e exp essed as % pe olume.
Soil sampling and analysis
Ten o wel e o ganic laye samples we e collec ed using a soil co e (diame e 3.5 cm) om
e e y subplo A on Ap il 24, June 4 and Augus 30, 2014. Soil co es we e pooled o o m one
composi e soil sample pe each subplo pe sampling da e, esul ing in 50 soil samples pe da e.
On a e age, ac oss all blocks and ea men plo s (n = 50), he humus laye was 2.55 ± 0.08
(mean ± SE) cm hick and no s a is ically signi ican di e ences we e obse ed be ween he
blocks o he ea men s (F
4, 36
= 1.9, p = 0.132). F esh soil samples we e anspo ed o he lab-
o a o y in an icebox immedia ely a e sampling. As soil was s ill ozen a he ime o Ap il
sampling, samples we e s o ed a +4°C o e nigh o allow slow hawing be o e p ocessing. To
analyze soil and mic obial biomass N, one sub-sample o abou 3 g soil was ex ac ed wi h 50
mL o 0.5 M K
2
SO
4
. Ano he subsample was ex ac ed using he same me hod ollowing chlo-
o o m umiga ion o 18 h [17]. The concen a ion o NH
4
-N was de e mined om esh soil
ex ac s acco ding o he s anda d p o ocol (SFS 3032, Shimadzu UV-1700 spec opho ome e )
and he concen a ion o NO
3
-N ia low analysis (FIA Pe s o p). The o al ex ac able N in
bo h soil and umiga ed ex ac s was oxidized o NO
3
[18] and hen analyzed as NO
3
-N (FIA,
Pe s o p). Mic obial N was calcula ed by sub ac ing he o al ex ac able N o he soil ex ac s
om ha o he umiga ed ex ac s. Soil mois u e was de e mined by d ying he samples
(105°C, 12 h). Soil pH was measu ed in 3:5 / soil:wa e suspensions (Den e Ins umen
Model 220).
Seedling ma e ial and in en o y o seedling condi ion and g ow h
Con aine seedling ma e ial o he expe imen was g own by Fin Fo elia Oy (Ro aniemi,
No he n Finland). Seedlings o local o igin we e g own in PL121F (121 cells pe ay, 816
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 5/18

cells m
-2
, cell olume 50 cm
3
) ays illed wi h e ilized (N, P, K and mic o nu ien s) and
limed sphagnum pea .
One-yea -old Sco s pine (Pinus syl es is, a e age heigh 7–12 cm) and wo-yea -old No -
way sp uce (Picea abies, a e age heigh 27 cm) seedlings we e plan ed on ea men plo s in
Sep embe 2013 as desc ibed in Fig 1 (10 seedlings o bo h species in each o he 100 subplo s,
in o al 2000 seedlings) using plan ing ubes o place he plug in mine al soil.
An in en o y o seedling heal h and su i al was done one mon h a e plan ing on Oc obe
22, 2013 o eco d seedling s a us be o e win e : only 4 seedlings we e isibly quali ied as dead
(3 pine, 1 sp uce seedling, in AMB plo s (2 seedlings) and COMP plo s (2 seedlings)).
Seedlings in AMB, IE, COMP and NoSNOW plo s we e in en o ied du ing he 2014 g ow-
ing season. As soil and seedling samples in A subplo s we e collec ed in he beginning o he
2014 g owing season o u he biochemical analysis, only seedlings in B subplo s could be
in en o ied. Because shel e s in he NoICE plo s p e en ed ain-on-snow bu no snow com-
pac ion occu ing du ing wa m spells, he NoICE plo s we e conside ed o be simila as con-
ols and we e no used o seedling in en o y. The s a us o seedlings was eco ded on June 6
(beginning o g owing season), July 11, (a e cessa ion o he annual main shoo heigh
g ow h) and Sep embe 30 (end o g owing season), 2014 o he ollowing pa ame e s: leng h
o he cu en (2014) annual shoo , heal h class o he cu en (2014) annual shoo (0: heal hy,
1: <50% b own needles, 2: >50% b own needles, and 3: dead) and seedling heal h class (0:
heal hy, 1: <50% b own needles, 2: >50% b own needles, 3: dead).
S a is ical analysis
Da a was analyzed by means o a linea mixed model, whe e ea men and sampling da e we e
ixed ac o s and block was a andom ac o . Sampling da e was a epea ed ac o wi h he ea -
men plo as a subjec . Seedlings ha we e dead al eady in Oc obe 2013 we e excluded om
he analysis. Also seedlings wi h b oken shoo s (likely due o snow manipula ion) we e
excluded om he da ase : a o al o 10 seedlings we e b oken, wi h 8 in COMP and 2 in IE
plo s. Fo g ow h, seedling heal h and shoo win e su i al, ea men -wise means pe ea -
men plo we e i s compu ed o s a is ical analysis (i.e. n = 10 pe ea men pe sampling
da e). Shoo s om all li ing seedlings, wha e e hei heal h condi ions, we e included when
compu ing he mean heigh g ow h (a li ing seedling can ha e a dead shoo and a new shoo
may g ow he ollowing yea ). Seedlings dead be o e win e we e excluded om he da ase as
well as seedlings dead in June (June g ow h da a) and July (July g ow h da a). Seedlings dead in
July we e also excluded om he Sep embe g ow h da a as no seedling heal h in en o y was
done in Sep embe . To es whe he seedlings in snow manipula ion plo s we e s a is ically di -
e en om seedlings in AMB plo es ima ed ma ginal means we e compu ed and compa ed
unde he linea mixed model. In case o he leng h o he cu en (2014) annual shoo and he
p opo ion o dead shoo s he da a o he inal s a us (Sep . 30) only was included in he analy-
sis (i.e. da e was excluded om he mixed model). Co ela ions we e compu ed by means o
Pea son co ela ion coe icien s ( ). All s a is ical analyses we e ca ied ou wi h SPSS 22.0 o
Windows (SPSS, Inc., Chicago, IL, USA).
Resul s
Wea he condi ions du ing he expe imen and snow co e da a
The 2013–2014 win e was cha ac e ized by unusually wa m empe a u es in Decembe , Feb-
ua y and Ma ch (S1A Fig) wi h empe a u es 4.1, 9.4 and 4.0°C highe han he long- e m
a e ages (1981–2010), espec i ely. This led o a high numbe o days wi h daily ai empe a-
u es abo e eezing (7, 2, 7 and 11 days in Decembe , Janua y, Feb ua y and Ma ch,
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 6/18
espec i ely, in Apukka). Decembe 2013 was also cha ac e ized by high p ecipi a ion,
mainly in he o m o ain due o wa m spells in ea ly Decembe (S1B Fig). In la e win e
2014, he snow co e was hinne han he 30-yea a e age in open a eas (S1C Fig). P esence
o ees and s and s uc u e in luences snow co e dis ibu ion and he mal p ope ies due o
in e cep ion o ligh and p ecipi a ion [14].In hespeci iccaseo ou expe imen al o es ,we
es ima ed ha he snow co e is abou hal ha inopena eainApukka(S1C Fig). Ne e he-
less, he snow dep h was lowe han a e age in win e 2013–2014 a he expe imen al si e.
Snow manipula ion a ec ed he snowpack hickness, pa icula ly he IE ea men whe e a
s ong dec ease in snow dep h was measu ed in Feb ua y (S2A Fig). The e ec was lesse
la e in Ma ch due o he na u al snow ha co e ed he plo s a e he snow manipula ion.
The ligh e snow in he NoICE plo s (p e en ion o ain-on-snow) was he i s snow o mel
by he end o Ap il (S2B Fig).
Soil os and empe a u e
The win e 2013–2014 was cha ac e ized by wo cold spells (ai empe a u e below -20°C a he
ield si e: Decembe 6–9, 2013 and Janua y 14–19, 2014) and wo wa m spells (ai empe a u e
abo e 0°C o mo e han 5 consecu i e days, Decembe 25–30, 2013 and Ma ch 5–14, 2014; Fig
2A). The beginning o he 2013–2014 win e was unusually wa m wi h e y li le snow in ea ly
win e due o a wa m spell and ain om Decembe 25, 2013 o Janua y 6, 2014. Due o he
low snow co e in ea ly Decembe , he soil os wen deepe han 50 cm du ing he i s cold
spell in AMB plo s (S1 Table).
The soil hen comple ely hawed du ing he wa m spell in ea ly Janua y 2014 o eeze again
la e in Janua y. The wa m spell in mid-Ma ch was ollowed by a dec ease o he soil os
dep h a week la e (as measu ed on Ma ch 20, S1 Table), howe e no leading o comple e soil
hawing as he su ace oze a ew days a e he wa m spell. Subsequen eezing empe a u es
led o deepening o soil os o le els simila o hose be o e he wa m spell. Finally, he soil
s a ed o haw om he su ace in he second pa o Ap il and comple e soil hawing was
obse ed in AMB plo s du ing he hi d week o May (S1 Table).
Fig 2. Ai empe a u e and e ec o snow manipula ion on soil empe a u e. Mean daily ai (2-m heigh ) and soil (humus laye ) empe a u es du ing
win e 2013–2014 (a) and close-up a he ime o snow mel (b). Values a e mean o 3 blocks. Snow wa e ing occasions a e indica ed. Blue a ows = cold
spells, ed a ows = wa ms spells.
doi:10.1371/jou nal.pone.0156620.g002
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 7/18
Du ing wa m spells, he os dep h in NoSNOW plo s dec eased as e han in o he plo s
and in la e sp ing, he NoSNOW plo s we e all among he i s o be comple ely hawed (S1
Table).
Being independen o soil physical p ope ies, he g ound su ace and humus empe a u es
showed less he e ogenei y be ween he blocks han soil os . Du ing he cold pe iod o win e
2013–2014, snow emo al, bu also IE o a smalle ex en , led o lowe soil empe a u e han in
AMB, NoICE and COMP plo s in he humus laye (Fig 2A). Howe e , plo ing he di e ence
in empe a u e be ween he ea ed and AMB plo s as a unc ion o ai empe a u e (S3 Fig), i
appea ed ha IE and snow compac ion did no signi ican ly dec ease soil empe a u e unde
he snowpack. The na u al snow ha co e ed he plo s a e ea men likely p o ided su icien
insula ion. Soil empe a u e in NoSNOW plo s mi o ed g ound os dep h, and a s ong co -
ela ion be ween he eezing ai empe a u e and empe a u es o bo h g ound and humus
we e de ec ed ( = 0.73 and 0.65, espec i ely; S3E Fig).
Snow-co e ed a ea
The calcula ion o snow-co e ed a ea a he end o Ap il sugges ed a la e comple e snow mel
due o IE compa ed o ambien condi ions (S2B Fig). The empe a u e eco ds suppo a la e
da e o snowmel in IE compa ed o o he plo s (appea ance o diu nal luc ua ions in he em-
pe a u e in he humus laye a e linked wi h he da e o snow mel ; Fig 2B).
Soil gas composi ion
Analysis o gas samples collec ed in he humus laye (2-cm deep) h oughou he win e
showed la ge seasonal changes in CO
2
and O
2
concen a ions and, abo e all, s ong ea men
e ec s (Fig 3). In AMB plo s, he CO
2
concen a ion s a ed o inc ease om he beginning o
Ma ch, up o a maximum concen a ion o 2.3 ± 0.5% on Ap il 16, 2014. Simila changes we e
measu ed in he NoICE plo s. Snow manipula ion had a clea and s a is ically signi ican e ec
on soil gas composi ion om ea ly Ma ch o ea ly May (Table 1 and Fig 3).
IE led o de elopmen o hypoxia and o a s ong accumula ion o CO
2
unde he snowpack
wi h concen a ions o 7.3%, close o 200 imes he a mosphe ic concen a ion. An in e medi-
a e si ua ion was obse ed in COMP plo s wi h concen a ions be ween hose in AMB and IE
plo s (Fig 3). No CO
2
accumula ion was measu ed in he NoSNOW plo s. The a bi a y h esh-
old o 4% (100 imes he a mosphe ic concen a ion) was measu ed du ing 0, 12 and 45 conse-
cu i e days in he AMB, COMP and IE plo s, espec i ely. Al e na i ely, he CO
2
concen a ion was abo e 5% [19] du ing 0, 1 and 17 consecu i e days in he AMB, COMP and
IE plo s, espec i ely.
The O
2
concen a ion dec eased in close co ela ion wi h accumula ion o CO
2
( = -0.93).
In AMB plo s, he lowes O
2
concen a ion was 13.7 ± 2.2%. I was lowe han 16% o 9 and 41
days in he COMP and IE plo s, espec i ely. No O
2
dec ease was obse ed in he NoSNOW
plo s. Seasonal a ia ion in soil gas concen a ion unde he snowpack was pa ly ela ed o
changes in ai empe a u e: wa m spells we e ollowed by inc eases in CO
2
concen a ion
(dec ease in O
2
concen a ion) and in e sely, cold pe iods we e ollowed by dec eases in CO
2
concen a ion (inc eases in O
2
concen a ion; Fig 3). The ex eme alues in Ap il we e ela ed
o he seasonal inc ease in ai empe a u e in sp ing.
Soil p ope ies
Fo es soil a he s udy si e is acidic wi h a ypical pH anging om 3.6 and 3.9 depending on
he season. Time o sampling had a signi ican e ec on pH (S3 Table): highe pH alues we e
measu ed in ea ly June (3.85 ± 0.04) compa ed o Ap il (3.77 ± 0.05) and Augus (3.66 ± 0.03)
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 8/18
in AMB plo s (mean ± SE, n = 10). Al hough ea men had no signi ican e ec on pH (S3
Table), a endency o lowe pH was obse ed in IE compa ed o AMB plo s in Ap il and June
(S4 Fig).
The snow-manipula ion ea men s had s a is ically signi ican e ec s on soil mois u e and
N con en (S3 Table), bu only he NoSNOW plo s we e signi ican ly di e en ha AMB plo s.
A dec ease in soil mois u e due o p e en ion o snow all (-20% compa ed o AMB plo s) as
well as a dec ease in o al ex ac able N ( he sum o NH
4
-N, NO
3
-N, and ex ac able o ganic N;
S2 and S3 Tables) we e de ec ed in Ap il bu no la e du ing he g owing season. All o ms o
N we e simila ly a ec ed by snow emo al.
Fig 3. E ec o snow manipula ion on CO
2
and O
2
concen a ions 2-cm deep in humus laye . Mean daily ai empe a u e a ield si e is depic ed.
Snow wa e ing occasions a e indica ed by blue diamonds. Fo be e cla i y, one-sided e o ba s a e depic ed. Values a e means ±SE (n = 10).
doi:10.1371/jou nal.pone.0156620.g003
Table 1. S a is ical signi icance es s (linea mixed model o ea men , da e and hei in e ac ion) o CO
2
(log ans o med) and O
2
concen a-
ions (%).
log CO
2
O
2
d F Sig. d F Sig.
T ea men 5 / 104 40.7 <0.001 5 / 162.1 19.4 <0.001
Da e 22 / 520 23.3 <0.001 22 / 497.7 10.4 <0.001
T ea men *da e 77 / 517 3.57 <0.001 76 /504.2 2.95 <0.001
doi:10.1371/jou nal.pone.0156620. 001
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 9/18
Acknowledgmen s
We a e g a e ul o o es enginee s Pekka Nä hi, Ta mo Aal o, Aa no Ni a, Raimo Pikkupeu a
and Ee o Sii ola o he main enance o he ield expe imen and help in sampling, and o
Si kka Aakkonen, Leena Sei amäki and Mau i Heikkinen o hei echnical help. The au ho s
a e g a e ul o he Ro aniemi pa ish o allowing us o se up he ield expe imen on hei land
and hei collabo a ion. We u he hank he anonymous e iewe s o hei commen s on he
manusc ip .
Au ho Con ibu ions
Concei ed and designed he expe imen s: FM PR. Pe o med he expe imen s: FM JV AO PR.
Analyzed he da a: FM AO SS PR. Con ibu ed eagen s/ma e ials/analysis ools: FM SS PR.
W o e he pape : FM JV SS PR.
Re e ences
1. Mikkonen S, Laine M, Mäkelä HM, G egow H, Tuomen i a H, Lah inen M, e al. T ends in he a e age
empe a u e in Finland, 1847–2013. S och En i on Res Risk Assess. 2014:1–9.
2. Lis on GE, Hiems a CA. The changing c yosphe e: Pan-A c ic snow ends (1979–2009). J Clima e.
2011; 24(21):5691–712.
3. IPCC. Clima e Change 2013: The physical science basis (Wo king G oup I). 22.XII.2013: Camb idge
Uni e si y p ess; 2013.
4. Rasmus S, Räisänen J, Lehning M. Es ima ing snow condi ions in Finland in he la e 21s cen u y using
he SNOWPACK-model wi h egional clima e scena io da a as inpu . Ann Glaciol. 2004; 38:238–44.
5. Sch ade J, Moyle R, Bhale ao R, He zbe g M, Lundebe g J, Nilsson P, e al. Cambial me is em do -
mancy in ees in ol es ex ensi e emodelling o he ansc ip ome. Plan J. 2004; 40(2):173–87.
PMID: 15447645
6. Al a ez-U ia P, Kö ne C. Low empe a u e limi s o oo g ow h in deciduous and e e g een empe a e
ee species. Func Ecol. 2007; 21(2):211–8.
7. Chomba BM, Guy RD, Wege HG. Ca bohyd a e ese e accumula ion and deple ion in Engelmann
sp uce (Picea engelmannii Pa y): e ec s o cold s o age and p e-s o age CO
2
en ichmen . T ee Phy-
siol. 1993; 13(4):351–64. PMID: 14969991
8. Be gjo d AK, Bonesmo H, Skjel åg AO. Modelling he cou se o os ole ance in win e whea : I. Model
de elopmen . Eu J Ag on. 2008; 28(3):321–30.
9. Gudlei sson B, Bja nado i B. Es ima ing ice encasemen ole ance o he bage plan s. In: Hincha DK,
Zu he E, edi o s. Plan Cold Acclima ion. Sp inge New Yo k; 2014. p. 225–40.
10. Höglind M, Bakken AK, Jø gensen M, Øs em L. Tole ance o os and ice encasemen in cul i a s o
imo hy and pe ennial yeg ass du ing win e . G ass Fo age Sci. 2010; 65(4):431–45.
11. C aw o d RMM, B aendle R. Oxygen dep i a ion s ess in a changing en i onmen . J Exp Bo . 1996; 47
(295):145–59.
12. Solomon DK, Ce ling TE. The annual ca bon dioxide cycle in a mon ane soil: Obse a ions, modeling,
and implica ions o wea he ing. Wa e Resou Res. 1987; 23(12):2257–65.
13. Pi inen P, Simola H, Aal o J, Kauko an a J, Ka lsson P, Ruuhela R. Clima ological s a is ics o Finland
1981–2010. Helsinki, Finland: Finnish Me eo ological Ins i u e; 2012.
14. Va hola A, Coops NC, Weile M, Moo e RD. Fo es canopy e ec s on snow accumula ion and abla ion:
An in eg a i e e iew o empi ical esul s. J Hyd ol. 2010; 392(3–4):219–33.
15. Rica d JA, Tobiasson W, G ea o ex A. The ield assembled os gage. Cold Regions Resea ch and
Enginee ing Labo a o y, Technical No e ed. Hano e , New Hampshi e: U.S., A my Co ps o Engi-
nee s; 1976.
16. Kammann C, G ünhage L, Jäge H-. A new sampling echnique o moni o concen a ions o CH4, N2O
and CO2 in ai a well-de ined dep hs in soils wi h a ied wa e po en ial. Eu J Soil Sci. 2001; 52
(2):297–303.
17. B ookes PC, Landman A, P uden G, Jenkinson DS. Chlo o o m umiga ion and he elease o soil ni o-
gen: A apid di ec ex ac ion me hod o measu e mic obial biomass ni ogen in soil. Soil Biol Biochem.
1985; 17(6):837–42.
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 16 / 18

18. Williams BL, Shand CA, Hill M, O'Ha a C, Smi h S, Young ME. A p ocedu e o he simul aneous oxida-
ion o o al soluble ni ogen and phospho us in ex ac s o esh and umiga ed soils and li e s. Com-
mun Soil Sci Plan Anal. 1995; 26(1–2):91–106.
19. G eenway H, A ms ong W, Colme TD. Condi ions leading o high CO
2
(>5 kPa) in wa e logged–
looded soils and possible e ec s on oo g ow h and me abolism. Ann Bo . 2006; 98(1):9–32. PMID:
16644893
20. Repo T, Roi o M, Su inen S. Does he emo al o snowpack and he consequen changes in soil os
a ec he physiology o No way sp uce needles? En i on Exp Bo . 2011; 72(3):387–96.
21. Decke KLM, Wang D, Wai e C, Sche ba skoy T. Snow emo al and ambien ai empe a u e e ec s on
o es soil empe a u es in no he n Ve mon . Soil Sci Soc Am J. 2003; 67(4):1234–42.
22. Gudlei sson BE. Ice encasemen damage on g ass c ops and alpine plan s in Iceland—impac o cli-
ma e change. In: Plan cold ha diness: om he labo a o y o he ield. Gus a L. V., Wisniewski M. E.,
Tanino K. K. ed. CAB eBooks; 2009. p. 163.
23. Bje ke JW, Tømme ik H, Zielke M, Ma i J. Impac s o snow season on g ound-ice accumula ion, soil
os and p ima y p oduc i i y in a g assland o sub-A c ic No way. En i on Res Le . 2015; 10
(9):095007.
24. Come o d D, Schabe g P, Temple P, Socci A, Campbell J, Wallin K. In luence o expe imen al snow
emo al on oo and canopy physiology o suga maple ees in a no he n ha dwood o es . Oecologia.
2013; 171(1):261–9. doi: 10.1007/s00442-012-2393-x PMID: 22752211
25. Ha dy JP, G o man PM, Fi zhugh RD, Hen y KS, Welman AT, Deme s JD, e al. Snow dep h manipula-
ion and i s in luence on soil os and wa e dynamics in a no he n ha dwood o es . Biogeochem.
2001; 56(2):151–74.
26. Rixen C, F eppaz M, S oeckli V, Huo inen C, Huo inen K, Wip S. Al e ed snow densi y and chemis y
change soil ni ogen mine aliza ion and plan g ow h. A c An a c Alp Res. 2008; 40:568–75.
27. K üge M, Jones D, F e ichs J, Oppe mann BI, Wes J, Coombs P, e al. E ec s o ele a ed CO
2
con-
cen a ions on he ege a ion and mic obial popula ions a a e es ial CO
2
en a Laache See, Ge -
many. In J G eenh Gas Con ol. 2011; 5(4):1093–8.
28. Al-T aboulsi M, Sjöge s en S, Colls J, S e en M, Black C. Po en ial impac o CO
2
leakage om Ca bon
Cap u e and S o age (CCS) sys ems on g ow h and yield in maize. Plan Soil. 2013; 365(1–2):267–81.
29. Clein JS, Schimel JP. Mic obial ac i i y o und a and aiga soils a sub-ze o empe a u es. Soil Biol Bio-
chem. 1995; 27(9):1231–4.
30. Mikan CJ, Schimel JP, Doyle AP. Tempe a u e con ols o mic obial espi a ion in a c ic und a soils
abo e and below eezing. Soil Biol Biochem. 2002; 34(11):1785–95.
31. Schimel JP, Bilb ough C, Welke JM. Inc eased snow dep h a ec s mic obial ac i i y and ni ogen min-
e aliza ion in wo A c ic und a communi ies. Soil Biol Biochem. 2004; 36(2):217–27.
32. Sulli an BW, Do e S, Mon es-Helu M, Kolb TE, Ha SC. Pulse emissions o ca bon dioxide du ing
snowmel a a high-ele a ion si e in No he n A izona, U.S.A. A c An a c Alp Res. 2012; 44(2):247–54.
33. Musselman RC, Massman WJ, F ank JM, Ko mache JL. The empo al dynamics o ca bon dioxide
unde snow in a high ele a ion ocky moun ain subalpine o es and meadow. A c An a c Alp Res.
2005; 37(4):527–38.
34. Yanai Y, Hi o a T, Iwa a Y, Nemo o M, Naga a O, Koga N. Accumula ion o ni ous oxide and deple ion
o oxygen in seasonally ozen soils in no he n Japan–Snow co e manipula ion expe imen s. Soil Biol
Biochem. 2011; 43(9):1779–86.
35. F ey M. The in luence o snow on g ow h and su i al o plan ed ees. A c An a c Alp Res. 1983;
2:241–51.
36. P eece C, Callaghan TV, Phoenix GK. Impac s o win e icing e en s on he g ow h, phenology and
physiology o sub-a c ic dwa sh ubs. Physiol Plan a um. 2012; 146(4):460–72.
37. P eece C, Phoenix GK. Responses o sub-a c ic dwa sh ubs o low oxygen and high ca bon dioxide
condi ions. En i on Exp Bo . 2013; 85:7–15.
38. K asowski MJ, Simpson DG. F os - ela ed p oblems in he es ablishmen o coni e ous o es s. In:
Big as FJ, Colombo SJ, edi o s. Coni e Cold Ha diness. Do d ech : Sp inge Ne he lands; 2001. p.
253–85.
39. Su inen M, Repo T, Su inen S, Lasa o H, Al ila L, Pakkanen TT. Physiological changes in Pinus syl-
es is needles du ing ea ly sp ing unde sub-a c ic condi ions. Fo Ecol Manage. 2000; 135(1–3):217–
28.
40. Su inen S, Roi o M, Leh o T, Repo T. Simula ed snowmel and in il a ion in o ozen soil a ec ed oo
g ow h, needle s uc u e and physiology o Sco s pine saplings. Bo eal En i onm Res. 2014; 19
(4):281–94.
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 17 / 18
41. D esche M, Thomas SC. Snow co e manipula ions al e su i al o ea ly li e s ages o cold- empe a e
ee species. Oikos. 2013; 122(4):541–54.
42. Gaul D, He el D, Leuschne C. E ec s o expe imen al soil os on he ine- oo sys em o ma u e No -
way sp uce. J Plan Nu Soil Sci. 2008; 171(5):690–8.
43. Su inen M, Mäki alo K, Su inen R. F eezing dehyd a ion damages oo s o con aine ized Sco s pine
(Pinus syl es is) seedlings o e win e ing unde suba c ic condi ions. Can J Fo Res. 1996; 26
(9):1602–9.
44. Tie ney G, Fahey T, G o man P, Ha dy J, Fi zhugh R, D iscoll C. Soil eezing al e s ine oo dynamics
in a no he n ha dwood o es . Biogeochem. 2001; 56(2):175–90.
45. Repo T, Si kiä S, Heinonen J, La igné A, Roi o M, Koljonen E, e al. E ec s o ozen soil on g ow h
and longe i y o ine oo s o No way sp uce. Fo Ecol Manage. 2014; 313(0):112–22.
46. Heiskanen J, Su inen S, Hy önen J. E ec s o ligh exposu e in eezing empe a u es on win e dam-
age o oliage o No way sp uce con aine seedlings in mid and la e win e : Pilo expe imen s in an open
ield. J Fo Sci. 2015; 61:35–44.
47. K asowski MJ, Le ch o d T, Capu a A, Be ge ud WA. Desicca ion o whi e sp uce seedlings plan ed in
he sou he n bo eal o es o B i ish Columbia. Wa e Ai Soil Pollu . 1995; 82(1–2):133–46.
48. C aw o d RM. Seasonal di e ences in plan esponses o looding and anoxia. Can J Bo . 2003; 81
(12):1224–46.
49. Voesenek LACJ, Bailey-Se es J. Flood adap i e ai s and p ocesses: an o e iew. New Phy ol. 2015;
206(1):57–73. doi: 10.1111/nph.13209 PMID: 25580769
50. Smi h KL, S e en MD, Jones DG, Wes JM, Coombs P, G een KA, e al. En i onmen al impac s o CO
2
leakage: ecen esul s om he ASGARD acili y, UK. Ene gy P ocedia. 2013; 37:791–9.
51. Qi J, Ma shall JD, Ma son KG. High soil ca bon dioxide concen a ions inhibi oo espi a ion o Doug-
las i . New Phy ol. 1994; 128(3):435–42.
52. Gonzalez-Mele MA, Tane a L, T ueman RJ. Plan espi a ion and ele a ed a mosphe ic CO
2
concen-
a ion: Cellula esponses and global signi icance. Ann Bo . 2004; 94(5):647–56. PMID: 15355864
53. Cas onguay Y, Thibaul G, Roche e P, Be and A, Roche o S, Dionne J. Physiological esponses o
annual blueg ass and c eeping Ben g ass o con as ed le els o O
2
and CO
2
a low empe a u es.
C op Sci. 2009; 49(2):671–89.
54. Su inen R, Vajda A, Hänninen P, Su inen M. Signi icance o snowpack o oo -zone wa e and empe -
a u e cycles in suba c ic Lapland. A c An a c Alp Res. 2009; 41(3):373–80.
55. Wang D, Hecka ho n S, Wang X, Philpo S. A me a-analysis o plan physiological and g ow h
esponses o empe a u e and ele a ed CO2. Oecologia. 2012; 169(1):1–13. doi: 10.1007/s00442-011-
2172-0 PMID: 22037993
56. Bö che K, Au ela M, Ke inen M, Ma kkanen T, Ma ila O, Kola i P, e al. MODIS ime-se ies-de i ed
indica o s o he beginning o he g owing season in bo eal coni e ous o es —A compa ison wi h CO
2
lux measu emen s and phenological obse a ions in Finland. Remo e Sens En i on. 2014; 140
(0):625–38.
57. Repo T, Kalliokoski T, Domisch T, Leh o T, Manne koski H, Su inen S, e al. E ec o iming o soil os
hawing on Sco s pine. T ee Physiol. 2005; 25:1053–62. PMID: 15929936
58. Repo T, Leh o T, Finé L. Delayed soil hawing a ec s oo and shoo unc ioning and g ow h in Sco s
pine. T ee Physiol. 2008; 28(10):1583–91. PMID: 18708340
59. Ba bei o I, Dawes MA, Rixen C, Senn J, Bebi P. Fac o s d i ing mo ali y and g ow h a eeline: a 30-
yea expe imen o 92 000 coni e s. Ecology. 2012; 93(2):389–401. PMID: 22624320
60. Yli alo E. The s a is ical yea book o o es y. Finnish Fo es Resea ch Ins i u e; 2013.
Changing Snow Co e Impac s Fo es Ecosys em
PLOS ONE | DOI:10.1371/jou nal.pone.0156620 June 2, 2016 18 / 18