Biogeosciences, 10, 929–944, 2013
www.biogeosciences.ne /10/929/2013/
doi:10.5194/bg-10-929-2013
© Au ho (s) 2013. CC A ibu ion 3.0 License.
EGU Jou nal Logos (RGB)
Ad ances in
Geosciences
Open Access
Na u al Haza ds
and Ea h Sys em
Sciences
Open Access
Annales
Geophysicae
Open Access
Nonlinea P ocesses
in Geophysics
Open Access
A mosphe ic
Chemis y
and Physics
Open Access
A mosphe ic
Chemis y
and Physics
Open Access
Discussions
A mosphe ic
Measu emen
Techniques
Open Access
A mosphe ic
Measu emen
Techniques
Open Access
Discussions
Biogeosciences
Open Access
Open Access
Biogeosciences
Discussions
Clima e
o he Pas
Open Access
Open Access
Clima e
o he Pas
Discussions
Ea h Sys em
Dynamics
Open Access
Open Access
Ea h Sys em
Dynamics
Discussions
Geoscien i ic
Ins umen a ion
Me hods and
Da a Sys ems
Open Access
Geoscien i ic
Ins umen a ion
Me hods and
Da a Sys ems
Open Access
Discussions
Geoscien i ic
Model De elopmen
Open Access
Open Access
Geoscien i ic
Model De elopmen
Discussions
Hyd ology and
Ea h Sys em
Sciences
Open Access
Hyd ology and
Ea h Sys em
Sciences
Open Access
Discussions
Ocean Science
Open Access
Open Access
Ocean Science
Discussions
Solid Ea h
Open Access
Open Access
Solid Ea h
Discussions
The C yosphe e
Open Access
Open Access
The C yosphe e
Discussions
Na u al Haza ds
and Ea h Sys em
Sciences
Open Access
Discussions
Clima e- ela ed changes in pea land ca bon accumula ion du ing
he las millennium
D. J. Cha man1, D. W. Beilman2, M. Blaauw3, R. K. Boo h4, S. B ewe 5, F. M. Chambe s6, J. A. Ch is en7,
A. Gallego-Sala8,9,10, S. P. Ha ison9,11, P. D. M. Hughes12, S. T. Jackson13, A. Ko hola14, D. Mauquoy15,
F. J. G. Mi chell16, I. C. P en ice11,17, M. an de Linden18, F. De Vleeschouwe 19, Z. C. Yu4, J. Alm20, I. E. Baue 21,
Y. M. C. Co ish16, M. Ga neau22, V. Hohl1, Y. Huang23, E. Ka o eld24, G. Le Roux19, J. Loisel4, R. Moschen25,
J. E. Nichols26, T. M. Nieminen27, G. M. MacDonald28, N. R. Phad a e29, N. Rausch30,¨
U. Sillasoo31, G. T. Swindles32,
E.-S. Tui ila14, L. Ukonmaanaho27, M. V¨
ali an a14, S. an Bellen15, B. an Geel33, D. H. Vi 34, and Y. Zhao35
1Depa men o Geog aphy, College o Li e and En i onmen al Sciences, Uni e si y o Exe e , EX4 4RJ, UK
2Depa men o Geog aphy, Uni e si y o Hawai‘i M¯
anoa, Honolulu, HI 96822, USA
3School o Geog aphy, A chaeology and Palaeoecology, Queen’s Uni e si y Bel as , Bel as BT7 1NN, UK
4Depa men o Ea h and En i onmen al Sciences, Lehigh Uni e si y, Be hlehem, PA 18015, USA
5Depa men o Geog aphy, Uni e si y o U ah, Sal Lake Ci y, UT 84112, USA
6Cen e o En i onmen al Change and Qua e na y Resea ch, SNSS, Uni e si y o Glouces e shi e,
Chel enham, GL50 4AZ, UK
7Cen o de In es igaci´
on en Ma em´
a icas, A.P. 402, 36000 Guanajua o, G o., Mexico
8QUEST, Depa men o Ea h Sciences, Uni e si y o B is ol, B is ol, BS8 1RJ, UK
9School o Geog aphical Sciences, Uni e si y o B is ol, B is ol, BS8 1SS, UK
10Depa men o Ea h and Ecosys em Sciences, Lund Uni e si e , S¨
ol ega an 12, 223 62 Lund, Sweden
11Depa men o Biological Sciences, Macqua ie Uni e si y, No h Ryde, NSW 2109, Aus alia
12Geog aphy and En i onmen , Uni e si y o Sou hamp on, High ield, Sou hamp on, Han s, SO17 1BJ, UK
13Depa men o Bo any, Uni e si y o Wyoming, La amie, WY 82071, USA
14Depa men o En i onmen al Sciences, P.O. Box 65, Uni e si y o Helsinki, 00014 Helsinki, Finland
15School o Geosciences, Uni e si y o Abe deen, Elphins one Road, Abe deen AB24 3UF, UK
16Bo any Depa men , T ini y College Dublin, Dublin 2, I eland
17G an ham Ins i u e o Clima e Change and Depa men o Li e Sciences, Impe ial College, Silwood Pa k,
Asco , SL5 7PY, UK
18BIAX Consul , Hogendijk 134, 1506 AL Zaandam, The Ne he lands
19CNRS and Uni e si ´
e de Toulouse, INP, UPS, EcoLab, ENSAT, A enue de l’Ag obiopole, 31326 Cas ane -Tolosan, F ance
20School o Fo es Sciences, Uni e si y o Eas e n Finland, P.O. Box 68, 80101 Joensuu, Finland
21Si Wil ed G en ell College, Memo ial Uni e si y o New oundland, Co ne B ook, New oundland A2H 6P9, Canada
22D´
epa emen de G´
eog aphie and GEOTOP, Uni e si ´
e du Qu´
ebec `
a Mon ´
eal, Mon ´
eal, Quebec, H3C 3P8, Canada
23Depa men o Geological Sciences, B own Uni e si y, P o idence, RI 02912, USA
24Uni e si y o Ta u, Ins i u e o Ecology and Ea h Sciences, Lai 40, Ta u 51005, Es onia
25Ins i u e o Bio- and Geosciences 3: Ag osphe e, Resea ch Cen e Juelich (Fo schungszen um J¨
ulich),
J¨
ulich 52428, Ge many
26NASA Godda d Ins i u e o Space S udies, 2880 B oadway, New Yo k, NY 10025, USA, and Lamon -Dohe y Ea h
Obse a o y a Columbia Uni e si y, Palisades, NY 10964, USA
27The Finnish Fo es Resea ch Ins i u e, P.O. Box 18, 01301 Van aa, Finland
28Ins i u e o he En i onmen and Sus ainabili y, and Depa men o Geog aphy, UCLA, Los Angeles, CA 90095, USA
29Wadia Ins i u e o Himalayan Geology, Deh a Dun, India
30Ins i u e o En i onmen al Geochemis y, Uni e si y o Heidelbe g, Im Neuenheime Feld 236, 69120 Heidelbe g, Ge many
31Ins i u e o Ecology, Tallinn Uni e si y, Uus-Sadama 5, Tallinn 10120, Es onia
Published by Cope nicus Publica ions on behal o he Eu opean Geosciences Union.
930 D. J. Cha man e al.: Clima e- ela ed changes in pea land ca bon accumula ion
32School o Geog aphy, Uni e si y o Leeds, Leeds, LS2 9JT, UK
33Ins i u e o Biodi e si y and Ecosys em Dynamics, P.O. Box 94248 1090 GE Ams e dam, The Ne he lands
34Depa men o Plan Biology, Sou he n Illinois Uni e si y, Ca bondale, IL 62901, USA
35Ins i u e o Geog aphic Science and Na u al Resou ces Resea ch, Chinese Academy o Sciences, Beijing 100101, China
Co espondence o: D. J. Cha man (d.j.cha man@exe e .ac.uk)
Recei ed: 2 Oc obe 2012 – Published in Biogeosciences Discuss.: 17 Oc obe 2012
Re ised: 11 Janua y 2013 – Accep ed: 11 Janua y 2013 – Published: 8 Feb ua y 2013
Abs ac . Pea lands a e a majo e es ial ca bon s o e and a
pe sis en na u al ca bon sink du ing he Holocene, bu he e
is conside able unce ain y o e he a e o pea land ca bon in
a changing clima e. I is gene ally assumed ha highe em-
pe a u es will inc ease pea decay, causing a posi i e eed-
back o clima e wa ming and con ibu ing o he global pos-
i i e ca bon cycle eedback. He e we use a new ex ensi e
da abase o pea p o iles ac oss no he n high la i udes o ex-
amine spa ial and empo al pa e ns o ca bon accumula ion
o e he pas millennium. Opposi e o expec a ions, ou e-
sul s indica e a small nega i e ca bon cycle eedback om
pas changes in he long- e m accumula ion a es o no he n
pea lands. To al ca bon accumula ed o e he las 1000y is
linea ly ela ed o con empo a y g owing season leng h and
pho osyn he ically ac i e adia ion, sugges ing ha a iabil-
i y in ne p ima y p oduc i i y is mo e impo an han de-
composi ion in de e mining long- e m ca bon accumula ion.
Fu he mo e, no he n pea land ca bon seques a ion a e de-
clined o e he clima e ansi ion om he Medie al Clima e
Anomaly (MCA) o he Li le Ice Age (LIA), p obably be-
cause o lowe LIA empe a u es combined wi h inc eased
cloudiness supp essing ne p ima y p oduc i i y. O he ac-
o s including changing mois u e s a us, pea land dis ibu-
ion, i e, ni ogen deposi ion, pe ma os haw and me hane
emissions will also in luence u u e pea land ca bon cycle
eedbacks, bu ou da a sugges ha he ca bon seques a ion
a e could inc ease o e many a eas o no he n pea lands in
a wa me u u e.
1 In oduc ion
Pea lands con ain a ound 600giga onnes o ca bon (G C)
ha has accumula ed since he las glacial maximum in
no he n mid–high la i udes, opical egions and empe a e
a eas o he Sou he n Hemisphe e, and he s eady accumu-
la ion o ca bon has been a small bu pe sis en sink o a -
mosphe ic CO2 h oughou he Holocene (Yu, 2011). The e-
la ionship be ween clima e change and he a e o ca bon se-
ques a ion is impo an o unde s anding he pas and u u e
global ca bon cycle, and i has gene ally been assumed ha
because empe a u e d i es inc easing decay (Ise e al., 2008;
Do epaal e al., 2009), pea lands could be pa o he posi-
i e eedback om he global ca bon cycle (F iedlings ein e
al., 2006). A key objec i e in imp o ing unde s anding o he
global ca bon cycle in clima e models is o be able o simu-
la e pas obse ed a mosphe ic CO2changes.
The e is g owing in e es in he las millennium as a
clima e-modelling a ge , and especially in he assessmen o
he sensi i i y o he global ca bon cycle o clima e wa m-
ing (Abe-Ouchi and Ha ison, 2009; Jungclaus e al., 2010).
In he No he n Hemisphe e, he ansi ion om he gene -
ally wa me Medie al Clima e Anomaly (MCA) o he coole
Li le Ice Age (LIA) (Mann e al., 2008, 2009; Jansen e al.,
2007) was associa ed wi h a ca. 7–10ppm decline in a mo-
sphe ic CO2concen a ion (Ahn e al., 2012). This pa e n
suppo s he exis ence o a posi i e global clima e–ca bon
cycle eedback, as sugges ed by coupled clima e–ca bon cy-
cle models (F iedlings ein e al., 2006; Denman e al., 2007).
Howe e , es ima es o he magni ude o he clima e sensi-
i i y o he global ca bon cycle based on da a om he las
millennium a y om 1.7–21.4ppmCO2K−1(F ank e al.,
2010) o 40–60ppmCO2K−1(Cox and Jones, 2008). Ca -
bon cycle models also a y g ea ly in hei assessmen o his
eedback (F iedlings ein e al., 2006), al hough ecen es i-
ma es (Jungclaus e al., 2010) sugges sensi i i y wi hin he
lowe end o his ange (3.2–12ppmCO2K−1). The causes
o he educ ion in CO2concen a ions du ing he MCA
o LIA ansi ion a e poo ly known, bu educed soil he -
e o ophic espi a ion is assumed o be impo an (Jungclaus
e al., 2010; Pong a z e al., 2009). Howe e , he models
do no speci ically ake in o accoun possible clima e- ela ed
a ia ions in he a e o pea land ca bon seques a ion.
Pea lands ha e seques e ed and exchanged a mosphe ic
ca bon o e millennia (MacDonald e al., 2006; F olking and
Roule , 2007), wi h he la ges s o e in no he n ex a opi-
cal pea lands, an es ima ed 545G C (Yu e al., 2010). The
annual up ake o CO2by pea lands, p e iously es ima ed
as 0.076G Cy −1(Go ham, 1991) o 0.088G Cy −1(Yu,
2011), wi hou conside ing long- e m decay (see below), is
a small bu empo ally pe sis en componen o land ca bon
up ake. This is equi alen o 36ppm a mosphe ic CO2o e
1000y , based on a simple con e sion om change in ca bon
pool o a mosphe ic CO2o 1G C=2.123ppm. Howe e ,
Biogeosciences, 10, 929–944, 2013 www.biogeosciences.ne /10/929/2013/
D. J. Cha man e al.: Clima e- ela ed changes in pea land ca bon accumula ion 931
0 10 31 50 82
C mass pe uni a ea
(kg m-2)
a
6
5
4
3
2
1
0
0 1,000 2,000 3,000 4,000 5,000 6,000
P ecipi a ion / equilib ium e apo anspi a ion
G owing deg ee days abo e 0°C
b
Fig. 1. Dis ibu ion o si es in geog aphic and clima e space, o e lain on soil ca bon s o age. Poin s show high- esolu ion pea eco ds (la ge
blue ci cles) used o calcula ing empo al a ia ion in ca bon accumula ion a es and low- esolu ion pea eco ds (small whi e ci cles) used
o es ima es o o al millennial ca bon. (a) The soil ca bon densi y in he op 1m mapped om 10-min IGBP soil da a; (b) clima e space
(de ined by g owing deg ee days and mois u e balance) o soil ca bon densi y in he same classes as in panel a, a 0.5×0.5◦g id cells o all
land no h o 40◦N. Pea lands gene ally occu in he a eas o >31kgCm−2(b own and black).
o e millennial imescales, ca bon up ake o his magni ude
would be compensa ed by ocean ou gassing p ocesses as a
esul o educed CO2in he a mosphe e and educ ion in
ai –sea CO2pa ial p essu e, so ha he ac ual e ec on he
a mosphe e is only 20–35% o his o al o e pe iods o 200–
2000y (A che e al., 2009), o 7–12ppm a mosphe ic CO2
o e a 1000y pe iod. Va ia ions in he size o he pea -
land sink could he e o e ha e a signi ican cumula i e e -
ec on global a mosphe ic CO2concen a ions o e he las
millennium, o he same o de o magni ude as he obse ed
changes.
In his s udy, we compiled pea co e da a om no he n
pea lands o es ima e changes in ca bon accumula ion o e
he las millennium and o explo e he spa ial ela ionship
be ween clima e and he o al size o he ca bon sink ac-
cumula ed o e his pe iod. Fu he analysis on a subse o
well-da ed co es allowed an analysis o empo al a ia ion in
ca bon accumula ion in ela ions o he MCA–LIA clima e
changes es ima ed om palaeoclima e eco ds. We use hese
da a o help unde s and he ela ionships be ween clima e and
pea land ca bon accumula ion and o assess he di ec ion and
s eng h o he pea land ca bon cycle eedback.
2 Me hods
2.1 Si e selec ion and ca bon measu emen
A lis o No he n Hemisphe e, ex a opical pea land p o-
iles wi h published and unpublished ca bon accumula ion
da a was compiled (Tables 1 and 2) o si es ha me he
ollowing c i e ia:
a. a leas 3 e enly spaced da es (including 210Pb, eph a,
sphe oidal ca bonaceous pa icles, pollen ma ke s o 14C
(p e- o pos da ing he pe iod o nuclea bomb es -
ing), and he uncu pea su ace) and spanning he las
ca. 1000y . Mos o he si es had mo e han 5 da es (Ta-
ble 2), bu we also ejec ed some si es whe e a sa is ac o y
age–dep h model could no be p oduced, because o age
e e sals o o he p oblems; and
b. con iguous bulk densi y measu emen s a <5cm esolu-
ion.
Applica ion o hese c i e ia esul ed in he selec ion o
24 si es ha we e used in subsequen analyses o he em-
po al changes in ca bon accumula ion h ough he las mil-
lennium (Table 2). A second ie o si es (Table 1) was used
o he millennium ca bon in en o y analysis agains clima e
indices. These si es did no mee c i e ia (a) and (b), bu did
mee he ollowing c i e ia:
c. a basic age–dep h model o he las millennium; and
d. con iguous bulk densi y measu emen s bu no necessa ily
a high esolu ion.
A o al o 90 si es me hese less s ingen c i e ia and
we e used in he clima e–ca bon in en o y analyses (Ta-
bles 1 and 2). The si es in bo h da a se s a e widely dis-
ibu ed geog aphically and b oadly ep esen a i e o he cli-
ma e space occupied by no he n pea lands (Fig. 1).
www.biogeosciences.ne /10/929/2013/ Biogeosciences, 10, 929–944, 2013
932 D. J. Cha man e al.: Clima e- ela ed changes in pea land ca bon accumula ion
Table 1. Cha ac e is ics o he low- esolu ion si es used in he analyses.
No. Si e name La i ude Longi ude Pea land ype Con ibu o Sou ce e e ence
1 E115 67.8095 75.4346 Omb o ophic D. Beilman Beilman e al. (2009)
2 E110 66.4698 76.9943 Omb o ophic D. Beilman Beilman e al. (2009)
3 E113 66.4497 79.3234 Omb o ophic D. Beilman Beilman e al. (2009)
4 D122 65.5831 73.0058 Omb o ophic D. Beilman Beilman e al. (2009)
5 E119 65.4998 75.5025 Omb o ophic D. Beilman Beilman e al. (2009)
6 D127 64.3068 70.2948 Omb o ophic D. Beilman Beilman e al. (2009)
7 G136 64.1476 75.3611 Omb o ophic D. Beilman Beilman e al. (2009)
8 G137 63.7504 75.7662 Omb o ophic D. Beilman Beilman e al. (2009)
9 N015 63.6501 74.2693 Omb o ophic D. Beilman Beilman e al. (2009)
10 N001 63.1611 74.8233 Omb o ophic D. Beilman Beilman e al. (2009)
11 S009 62.1229 73.8412 Omb o ophic D. Beilman Beilman e al. (2009)
12 V034 61.4675 79.4601 Omb o ophic D. Beilman Beilman e al. (2009)
13 V039 61.0895 79.3806 Omb o ophic D. Beilman Beilman e al. (2009)
14 SIB02 61.0553 70.0588 Omb o ophic D. Beilman Beilman e al. (2009)
15 V026 61.0286 76.4686 Omb o ophic D. Beilman Beilman e al. (2009)
16 S022 60.8401 71.2558 Omb o ophic D. Beilman Beilman e al. (2009)
17 V038 60.8039 74.5416 Omb o ophic D. Beilman Beilman e al. (2009)
18 SIB01 59.3601 68.9849 Omb o ophic D. Beilman Beilman e al. (2009)
19 SIB06 58.4358 83.4343 Omb o ophic D. Beilman Beilman e al. (2009)
20 SIB05 57.3541 81.1647 Omb o ophic D. Beilman Beilman e al. (2009)
21 SIB03 56.3552 79.0689 Omb o ophic D. Beilman Beilman e al. (2009)
22 C si e 01 60.167 72.8330 Omb o ophic Z. C. Yu Yu e al. (2009)
23 C si e 02 60.167 72.8330 Omb o ophic Z. C. Yu Yu e al. (2009)
24 C si e 03 56.833 78.4170 Omb o ophic Z. C. Yu Yu e al. (2009)
25 C si e 10 54.15 −130.2500 Omb o ophic Z. C. Yu Yu e al. (2009)
26 C si e 13 55.017 −114.1500 Omb o ophic Z. C. Yu Yu e al. (2009)
27 C si e 17 55.85 −107.6830 Omb o ophic Z. C. Yu Yu e al. (2009)
28 C si e 20 59.883 −104.2000 Omb o ophic Z. C. Yu Yu e al. (2009)
29 C si e 21 45.684 −74.0470 Omb o ophic Z. C. Yu Yu e al. (2009)
30 C si e 23 47.933 −64.5000 Omb o ophic Z. C. Yu Yu e al. (2009)
31 C si e 24 45.2 −60.2670 Omb o ophic Z. C. Yu Yu e al. (2009)
32 C si e 25 57.522 −5.1600 Omb o ophic Z. C. Yu Yu e al. (2009)
33 C si e 26 57.56 −5.3770 Omb o ophic Z. C. Yu Yu e al. (2009)
34 C si e 27 57.687 −5.6870 Omb o ophic Z. C. Yu Yu e al. (2009)
35 C si e 28 68.4 23.5500 Omb o ophic Z. C. Yu Yu e al. (2009)
36 C si e 30 60.817 26.9500 Omb o ophic Z. C. Yu Yu e al. (2009)
37 Kohlh¨
u en Moo 47.9269 8.1844 Omb o ophic G. Le Roux Le Roux e al. (2005)
38 P131 66.1664 73.9889 Mine o ophic D. Beilman Beilman e al. (2009)
39 C si e 04 56.333 84.5830 Mine o ophic Z. C. Yu Yu e al. (2009)
40 C si e 05 60.446 −151.2470 Mine o ophic Z. C. Yu Yu e al. (2009)
41 C si e 06 60.641 −151.0800 Mine o ophic Z. C. Yu Yu e al. (2009)
42 C si e 07 60.416 −150.9020 Mine o ophic Z. C. Yu Yu e al. (2009)
43 C si e 08 60.784 −150.8190 Mine o ophic Z. C. Yu Yu e al. (2009)
44 C si e 09 64.875 −147.7670 Mine o ophic Z. C. Yu Yu e al. (2009)
45 C si e 11 53.583 −118.0170 Mine o ophic Z. C. Yu Yu e al. (2009)
46 C si e 12 52.45 −116.2000 Mine o ophic Z. C. Yu Yu e al. (2009)
47 C si e 14 61.8 −121.4000 Mine o ophic Z. C. Yu Yu e al. (2009)
48 C si e 15 68.288 −133.2500 Mine o ophic Z. C. Yu Yu e al. (2009)
49 C si e 16 69.493 −132.6720 Mine o ophic Z. C. Yu Yu e al. (2009)
50 C si e 18 64.713 −105.5790 Mine o ophic Z. C. Yu Yu e al. (2009)
51 C si e 19 66.451 −104.8350 Mine o ophic Z. C. Yu Yu e al. (2009)
52 C si e 22 82.333 −68.2500 Mine o ophic Z. C. Yu Yu e al. (2009)
53 C si e 29 68.4 23.5500 Mine o ophic Z. C. Yu Yu e al. (2009)
54 C si e 31 65.65 27.3170 Mine o ophic Z. C. Yu Yu e al. (2009)
55 C si e 32 65.65 27.3170 Mine o ophic Z. C. Yu Yu e al. (2009)
56 C si e 33 65.65 27.3170 Mine o ophic Z. C. Yu Yu e al. (2009)
Biogeosciences, 10, 929–944, 2013 www.biogeosciences.ne /10/929/2013/
D. J. Cha man e al.: Clima e- ela ed changes in pea land ca bon accumula ion 933
Table 1 . Con inued.
No. Si e name La i ude Longi ude Pea land ype Con ibu o Sou ce e e ence
57 Old Black Sp uce Fen 53.9983 −105.1153 Mine o ophic I. Baue Baue e al. (2009)
58 Sandhill Fen 53.8261 −104.6250 Mine o ophic I. Baue Baue e al. (2009)
59 Dhaku i 30.0500 79.9333 Mine o ophic N. R. Phad a e Unpublished da a
60 Ae opo 4 54.1041 −72.5167 Mine o ophic M. Ga neau Unpublished da a
61 Ae opo 5 54.1041 −72.5167 Mine o ophic M. Ga neau Unpublished da a
62 Lac Le Ca on La e al Co e4 52.2945 −75.8408 Omb o ophic S. an Bellen an Bellen e al. (2011)
63 M179 60.5875 −149.5347 Mine o ophic Z. C. Yu Unpublished da a
64 ZB08-S4 33.0954 102.6650 Mine o ophic Y. Zhao Unpublished da a
65 OURS4 Pea land 54.0597 −72.4602 Mine o ophic M. Ga neau Unpublished da a
66 LG1 Pea land 54.0597 −78.4602 Mine o ophic M. Ga neau Unpublished da a
Bulk densi y was measu ed on ca e ully cu esh o ozen
ma e ial using eeze d ying o o en d ying o samples o
known olume. Sample sizes a ied depending on he sam-
pling me hod and co e size, and sample esolu ion a ied
om 0.5 o 5cm3(Table 2). In all cases samples we e la ge
enough o accu a ely measu e bulk densi y and we e aken
con iguously o enable eliable es ima es o d y mass accu-
mula ion o e ime. Ca bon densi y was de i ed om bulk
densi y mul iplied by he ca bon con en o each sample.
Whe e ca bon da a we e no a ailable, we assumed ha
50% o he o ganic ac ion (measu ed by s anda d loss-
on-igni ion analysis a 500◦C) was o ganic ca bon. A ca -
bon alue o app oxima ely 50% is ou inely used o pea
(Go ham, 1991; Vi e al., 2000) and is easonable com-
pa ed o he mean ca bon con en o he nine si es o which
we ha e measu ed alues in his s udy (46.6±0.33%), and
o he s udies in wes e n Canada (51.8%; Yu e al., 2009) and
Wes Sibe ia (50.7–56.3%; Beilman e al., 2009).
To p o ide an assessmen o hyd ological di e ences
among he pea lands in ou analyses, we classi ied si es as
ei he bogs o ens. Al hough di e ences be ween hese wo
pea land ypes a e ela ed o he ela i e in luence o di e -
en wa e sou ces (i.e. g oundwa e , su ace wa e , p ecipi a-
ion), h esholds used o dis inc ion be ween he wo ypes
a e egionally a ied. Fo ou si e classi ica ion, we used a
ela i ely conse a i e app oach, including only Sphagnum-
domina ed sys ems ha lacked ege a i e o mo phological
e idence o mine o ophic condi ions in ou “omb o ophic”
ca ego y. Si es cha ac e ised as omb o ophic included aised
bogs, blanke bogs, and he ex ensi e bog sys ems o wes e n
Sibe ia (K emene ski e al., 2003).
2.2 Ch onology and age modelling
All si es we e 14C da ed using selec ed abo eg ound plan e-
mains, excep o si e 68 whe e bulk pea was 14C da ed. We
ecalib a ed all he da es om he o iginal s udies. Fo mod-
e n (pos -AD 1950) 14C da es, he NH1 pos bomb calib a ion
cu e was used (Hua and Ba be i, 2004). Remaining da es
we e calib a ed using In Cal09 (Reime e al., 2009). Age
models o he empo al analysis we e based on he p og am
“Bacon”, a lexible Bayesian age–dep h modelling app oach
ha uses p io in o ma ion on plausible accumula ion a es
and hei a iabili y and au oco ela ion o e ime (Blaauw
and Ch is en, 2005, 2011). Pea co es we e di ided in o con-
iguous 2cm segmen s, and linea accumula ion a es we e
calcula ed o all indi idual segmen s sequen ially down he
co e. Age models we e de eloped based on se e al million
i e a ions, ollowed by hinning o emo e any au oco ela-
ion be ween indi idual model uns, yielding ca. 5000–8000
i e a ions o each si e (Fig. 2a).
2.3 Spa ial analysis o ca bon accumula ion
To al ca bon accumula ion o e he 1000y was es ima ed
based on a da a se o he 90 adioca bon-da ed pea p o-
iles (Tables 1 and 2). The pos -1000y ca bon pool is he
di e ence be ween o al ca bon addi ions om pho osyn he-
sis and cumula i e espi a i e ca bon elease o e his in-
e al, e lec ing ca bon seques a ion a a si e. We analysed
he ela ionship be ween o al ca bon accumula ion o e he
las 1000y and clima e pa ame e s using a 0.5◦g id, de-
i ed om he CLIMATE 2.2 da a (Kaplan e al., 2003).
Clima e pa ame e s included g owing deg ee days abo e
0◦C (GDD0), cumula i e pho osyn he ically ac i e adia ion
du ing he g owing season (PAR0), PAR o e he g owing
season, g owing season leng h (days) and he mois u e in-
dex P/Eq, whe e Pis annual p ecipi a ion and Eq is an-
nually in eg a ed equilib ium e apo anspi a ion calcula ed
om daily ne adia ion and empe a u e (P en ice e al.,
1993). PAR was calcula ed om la i ude and sunshine hou s
(P en ice e al., 1993; Ha ison e al., 2010).
2.4 Tempo al a ia ion in ca bon accumula ion
A composi e ca bon accumula ion cu e was cons uc ed
based on he subse o 24 well-da ed, high- esolu ion si es
wi h con inuous eco ds o he pas 1000y (Table 2). The
Bayesian age–dep h models allowed ch onological unce -
ain y o be included in ca bon accumula ion cu es (Fig. 2b).
www.biogeosciences.ne /10/929/2013/ Biogeosciences, 10, 929–944, 2013

934 D. J. Cha man e al.: Clima e- ela ed changes in pea land ca bon accumula ion
Table 2 . Cha ac e is ics o he high- esolu ion si es used in he analyses. BDs – bulk densi y sample size; BDi – bulk densi y inc emen dep h; da es (o he ): Pb – lead 210, S – sphe oidal
ca bonaceous pa icles, T – eph a, A – Amb osia pollen ise; scaling o ca bon es ima es based on C (ca bon analyses), LOI (loss on igni ion), o 50% o d y mass.
No. Si e La . Long. Al . Pea land BDs BDi 14C Da es C Main Si e Si e
name (m) ype (cm3) (cm) da es (o he ) scaling in es iga o code e e ence
67 Ballydu 53.0861 −7.9930 60 Omb o ophic 45–117 1 12 S C F. J. G. Mi chell i ballyd Unpublished da a
68 Uppe Pin o Fen 53.5833 −118.0167 1310 Mine o ophic 0.8 1 4 LOI Z. C. Yu ca uppin Yu e al. (2003)
69 M¨
annikj¨
a e Bog 58.8753 26.2556 79 Omb o ophic 200–250 5 10 Pb 50% E. Ka o eld es mannik Sillasoo e al. (2007)
70 Mis en 50.5631 6.1642 620 Omb o ophic 7–10 1 11 Pb LOI F. De Vleeschouwe be mis en De Vleeschouwe e al.
(2010, 2012)
71 Słowi´
nskie Bło a 54.3646 16.4902 30 Omb o ophic 7 1 7 Pb LOI F. De Vleeschouwe po slowbl De Vleeschouwe e al. (2009)
72 Wal on Moss 54.9833 −2.6670 109 Omb o ophic 2 1 30 C D. Mauquoy en wal on Mauquoy e al. (2002)
73 D¨
u es Maa 50.1228 6.8722 455 Omb o ophic 3–4 2 5 50% R. Moschen de du ma Moschen e al. (2009)
74 I win Smi h Bog 45.0319 −83.6181 223 Omb o ophic 3 1 3 A 50% R. Boo h us i wsmi Unpublished da a
75 Sidnaw Bog 46.5618 −88.7824 400 Mine o ophic 3 1 5 Pb, A 50% R. Boo h us sidnaw Unpublished da a
76 Pinhook Bog 41.6152 −86.8468 245 Mine o ophic 3 1 5 A 50% R. Boo h us pinhoo Unpublished da a
77 Colesdalen 78.0925 14.9787 8 Mine o ophic 1.2 0.5 9 50% D. Beilman no colesd Unpublished da a
78 Fen yes e ¨
o 47.6670 24.0330 1340 Omb o ophic 45 1 16 50% D. Cha man o en ye Unpublished da a
79 Saxn¨
as Mosse 56.8558 13.4610 161 Omb o ophic 10.5 1 25 C M. an de Linden se saxnas an de Linden (2006)
80 Dead Island Bog 54.8875 −6.5491 41 Omb o ophic 20 1 0 S, T(5) 50% G. Swindles ni deasis Swindles e al. (2010)
81 Sib-04 56.8039 78.7369 131 Omb o ophic 2–3 1 5 LOI D. Beilman SIB04 Beilman e al. (2009)
82 The G ea Hea h 44.7013 −67.8092 75 Omb o ophic 3.5 1 8 50% J. Nichols us g ea h Unpublished da a
83 Kon olan ahka 60.7830 22.7830 87 Omb o ophic 125–320 5 18 C M. V¨
ali an a i kon ol V¨
ali an a e al. (2007)
84 Lappmy an 64.1647 19.5830 295 Omb o ophic 10.5 1 20 C M. an de Linden se lappmy an de Linden e al. (2008)
85 Ou okumpu 62.6667 28.8500 108 Omb o ophic 3 1 4 Pb C N. Rausch i ou oku Rausch e al. (2005),
Ukonmaanaho e al. (2006)
86 Lac Le Ca on Cen al 52.2897 −75.4331 254 Omb o ophic 1 1 3 LOI S. an Bellen ca llcen an Bellen e al. (2011)
87 Hie aj¨
a i 63.1500 30.6667 168 Omb o ophic 3 1 3 Pb C N. Rausch i hie aj Rausch e al. (2005),
Ukonmaanaho e al. (2006)
88 Ma iana Lake 56.0167 −111.9333 700 Omb o ophic 0.8 1 6 LOI Z. C. Yu ca ma ian Unpublished da a
89 Ha ja al a 61.3500 22.1833 40 Omb o ophic 3 1 4 Pb C N. Rausch i ha ja Rausch e al. (2005),
Ukonmaanaho e al. (2006)
90 No dan’s Pond bog 49.1500 −53.5830 60 Omb o ophic 5 2 2 Pb 50% P. Hughes ca no dan Hughes e al. (2006)
Biogeosciences, 10, 929–944, 2013 www.biogeosciences.ne /10/929/2013/
D. J. Cha man e al.: Clima e- ela ed changes in pea land ca bon accumula ion 935
b
a
Fig. 2. Age–dep h and ca bon accumula ion es ima es o indi idual p o iles. The example shown he e is Mis en Bog, Belgium. (a) An age–
dep h plo showing he calib a ed ages (blue shapes) and age–dep h model (g ey-scale g aph) (Blaauw and Ch is en, 2005, 2011). (b) Ca bon
accumula ion de i ed om age–dep h models, and bulk densi y and C measu emen s. Cu es a e i ed o each o 10000 possible age models
based on Bayesian analysis. Poin s ep esen indi idual samples on di e en age models, and g ey lines a e i ed cu es o indi idual
models. Only cu es i ed inco po a ing long- e m decay and ecosys em ma u i y (Yu e al., 2003) a e shown he e.
All age dep h models we e con e ed o ca bon accumula ion
using bulk densi y and ca bon o LOI measu emen s.
We de i ed di e en es ima es o a iabili y in ca bon ac-
cumula ion a es based on di e en assump ions abou au-
ogenic p ocesses o long- e m decay (Clymo, 1984) and
ecosys em ma u i y (Yu e al., 2003). Ca bon accumula ion
a es calcula ed om ou age–dep h models and ca bon den-
si y do no ake accoun o au ogenic pea accumula ion p o-
cesses, mos impo an ly he e ec o long- e m decay. Dead
plan ma e ial decays apidly in he su ace laye s, as he
mos labile o ganic ma e is b oken down quickly by mic o-
bial ac i i y. Decomposi ion a es a e much slowe ( hough
no ze o) in he pe manen ly sa u a ed zone, which con ains
mo e ecalci an o ganic ma e (Clymo, 1984; Belyea and
Bai d, 2006). I p oduc i i y and decay a e cons an , mea-
su ed appa en accumula ion a es will be highe o mo e
ecen pea , and he long- e m ca bon s o age will appea o
inc ease. We accoun ed o his ecological p ocess by i -
ing decay cu es o each p o ile (Clymo, 1984). We also
es ed he e ec o “ecosys em ma u i y”, ha is he slow-
ing o pea g ow h unde s able condi ions because o au o-
genic limi s on he heigh o he pea su ace (Yu e al., 2003).
We excluded ca bon accumula ion changes in he uppe mos
pea (conse a i ely app oxima ed he e as pea o med a -
e 1850) whe e ela i ely apid ae obic decay is s ill aking
place. We used AD 1850 o his because his is likely ou side
o he ae obic decay zone o all co es.
The changes in accumula ion a es o each si e we e ex-
p essed as di e ences be ween obse ed accumula ion and
hose de i ed om h ee models: (1) linea decay model
(i.e. no au ogenic p ocesses); (2) he Clymo model, which
includes long- e m decay only (Clymo, 1984):
M=Pc
ac(1−e−ac ), (1)
whe e Mis he accumula ed ca bon, Pcis he pea added o
he ca o elm each yea (gCcm−2),acis he ca o elm decay
cons an and is ime; and (3) he ex ended pea accumula-
ion a e (Ex PAR) model (Yu e al., 2003), which includes
long- e m decay and ecosys em ma u i y:
M=Pc
ac−bc(e−bc −e−ac ), (2)
whe e he pa ame e s a e he same as hose lis ed abo e, wi h
he addi ion o bc, a coe icien ha allows he accumula ion
a e o be modi ied. Each cu e i ing exe cise p oduces es i-
ma ed alues o Pc,acand bc. The e a e se e al o he mo e
complex models ha could be applied o accoun o long-
e m decay, bu i is o en di icul o impossible o de e mine
he mos app op ia e one, gi en he sub le a ia ions in he
ca bon accumula ion cu es (Belyea and Bai d, 2006). Ou
in en ion he e is o es whe he obse ed a ia ions in he
aw ca bon accumula ion da a could be explained by long-
e m decay and ecosys em ma u i y.
Decay models we e i indi idually o each ca bon accu-
mula ion cu e de i ed om he Bacon ou ine. Fo he lin-
ea model, he i ing was ca ied ou using o dina y leas -
squa es. Fo he Clymo and Ex PAR models, op imiza ion
was ca ied ou using an i e a i e o hogonal sampling ech-
nique ha samples he en i e pa ame e space, hen uses a
leas squa es i o ob ain a subse o he pa ame e space.
This subse is hen sampled in he nex i e a ion o p oduce
www.biogeosciences.ne /10/929/2013/ Biogeosciences, 10, 929–944, 2013
936 D. J. Cha man e al.: Clima e- ela ed changes in pea land ca bon accumula ion
Fig. 3. S eps in ol ed in de i ing a non-au ogenic accumula ion
cu e om a single age p o ile a e i ing o he pea accumula-
ion model wi h long- e m decay and ecosys em ma u i y (Yu e al.,
2003). See ex o de ails.
an inc easingly well de ined pa ame e space. All ollow-
ing analyses we e applied o esul s om all h ee e sions
(Fig. 3).
The numbe o age models a ied by o e an o de o mag-
ni ude be ween si es ( om ca. 2300 o ca. 44000). To a oid
biases owa ds si es wi h a highe numbe o age models,
a single age model was andomly sampled om each si e.
The accumula ion a es we e in e pola ed on o a egula ime
s ep by aking he median alue in a mo ing window (hal -
window o 25y ) wi h a s ep o 10y o a oid bias owa d
si es wi h highe sampling esolu ion. Finally, a ime se ies
o accumula ion a es was calcula ed as he median o he
24 in e pola ed accumula ion a es (one pe si e). This was
epea ed 10000 imes o p o ide a ai sampling o he a ail-
able age models, and ga e a ma ix o median ime se ies on
a egula ime s ep. Finally, his ma ix was used o calcula e
he median and pe cen ile alues o he accumula ion a es
o each ime.
The e ec o his Mon e Ca lo esampling o he possi-
ble age models (and associa ed accumula ion a e cu es) is
o gi e g ea e weigh o he si es wi h he bes -cons ained
ch onologies. In each i e a ion o he esampling, we ook
one age model and se o accumula ion a es om each si e.
Si es ha a e well-cons ained will p o ide age models ha
a e simila in each i e a ion, and poo ly cons ained si es will
p o ide age models ha a e widely di e en . The end esul
o his is ha well-cons ained si es will e ec i ely ha e a
g ea e weigh in he o e all composi e.
To a oid any bias owa d si es wi h gene ally e y high
accumula ion a es, a second composi e was made, based
on ans o med alues. This ollowed he me hodology used
p e iously o cha coal da a (Ma lon e al., 2008): (1) min-
imax ans o ma ion o he o iginal accumula ion a e ime
se ies; (2) Box–Cox ans o ma ion o no malise he ime
se ies; and (3) z-sco e calcula ion. The composi e z-sco es
we e es ima ed using he same p ocedu e as o compila ion
o he un ans o med alues desc ibed abo e. The inal com-
posi e cu es a e shown in Fig. 4.
3 Resul s and discussion
3.1 Spa ial ela ionships be ween ca bon accumula ion
and clima e
Wa ming would be expec ed o inc ease ne p ima y p oduc-
i i y (NPP) in high-la i ude ecosys ems because o inc eased
g owing season leng h. The g owing season o no he n
pea lands is app op ia ely de ined as he pe iod o he yea
wi h ai empe a u es abo e eezing, because b yophy es be-
gin pho osyn hesis a his h eshold, and a e he dominan
pea - o me in mos o ou si es. PAR, de e mined by la i-
ude and cloudiness, is he d i e o pho osyn he ic ca bon
ixa ion and may also be an impo an con ol on NPP. How-
e e , highe empe a u es could also inc ease pea decompo-
si ion a es h ough accele a ed mic obial ac i i y (Ise e al.,
2008; Do epaal e al., 2009).
Linea eg ession o o al ca bon accumula ed o e he las
1000y (C) agains PAR0 yielded he s onges ela ionship:
C=0.0055 PAR0−3.82,(3)
wi h an R2o 0.33 (Fig. 5a). In single-p edic o eg essions,
C showed a weake ela ionship wi h GDD0 (R2=0.13,
Fig. 5b) and no signi ican ela ionship wi h P/Eq (P=0.19,
Fig. 5c). Residuals om Eq. (3) showed no sys ema ic e-
la ion o ei he GDD0 o P/Eq and inclusion o hese addi-
ional p edic o s in a mul iple linea eg ession yielded non-
signi ican eg ession coe icien s. The co ela ion be ween
PAR0 and GDD0 is high (0.83), owing o he g owing sea-
son leng h ha is sha ed by bo h a iables. We checked he
in luence o wo appa en ou lie s wi h highe PAR0 alues
on ou conclusions. These a e he wo sou he nmos si es
om Dhaku i (India) and Pinhook (USA). Remo ing hese
wo si es does no a ec he signi icance o he ela ionship
be ween PAR0 and 1kaC (P < 0.0001) bu changes he R2
alues om 0.33 o 0.24 and sligh ly changes he slope om
0.0055 o 0.0049. Thus, i s ill explains mo e o he a ia-
ion han GDD0. The in luence o hese wo si es is no in-
signi ican , bu emo ing hem does no impac ou main con-
clusions conce ning PAR. Wi hou he wo “ou lie s” o al
C s ill shows a posi i e signi ican ela ionship (P < 0.001)
wi h GDD0 bu wi h a change in R2 om 0.18 o 0.13 and a
Biogeosciences, 10, 929–944, 2013 www.biogeosciences.ne /10/929/2013/
D. J. Cha man e al.: Clima e- ela ed changes in pea land ca bon accumula ion 937
05−95%
10−90%
15−85%
20−80%
25−75%
30−70%
35−65%
40−60%
45−55%
1000 1200 1400 1600 1800 2000
Yea s
g m−2 y −1
15
10
5
0
Non−au ogenic C Accumula ion Ra e (Linea model)
a
05−95%
10−90%
15−85%
20−80%
25−75%
30−70%
35−65%
40−60%
45−55%
1000 1200 1400 1600 1800 2000
Yea s
z–sco e
0.4
0.2
0
-0.2
-0.4
Non−au ogenic C Accumula ion Ra e Z−sco es (Linea model)
b
05−95%
10−90%
15−85%
20−80%
25−75%
30−70%
35−65%
40−60%
45−55%
1000 1200 1400 1600 1800 2000
Yea s
g m−2 y −1
15
10
5
0
Non−au ogenic C Accumula ion Ra e (Clymo model)
05−95%
10−90%
15−85%
20−80%
25−75%
30−70%
35−65%
40−60%
45−55%
1000 1200 1400 1600 1800 2000
Yea s
z–sco e
0.4
0.2
0
-0.2
-0.4
Non−au ogenic C Accumula ion Ra e Z−sco es (Clymo model)
c
05−95%
10−90%
15−85%
20−80%
25−75%
30−70%
35−65%
40−60%
45−55%
1000 1200 1400 1600 1800 2000
Yea s
g m−2 y −1
15
10
5
0
Non−au ogenic C Accumula ion Ra e (Ex PAR model)
05−95%
10−90%
15−85%
20−80%
25−75%
30−70%
35−65%
40−60%
45−55%
1000 1200 1400 1600 1800 2000
Yea s
z–sco e
0.4
0.2
0
-0.2
-0.4
Non−au ogenic C Accumula ion Ra e Z−sco es (Ex PAR model)
Fig. 4. Composi e ca bon accumula ion cu es o he las millennium based on di e en assump ions conce ning au ogenic p ocesses. Le
panels show un ans o med da a, igh panels show z-sco es. (a) Linea accumula ion wi hou conside ing au ogenic p ocesses, (b) wi h
long- e m decay a es, and (c) including long- e m decay and ecosys em ma u i y (le panel also shown in Fig. 6). The da a a e AD 1850
a e shown only in ou line because he appa en up u n in ca bon accumula ion is due o incomple e decay o ecen ly accumula ed o ganic
ma e ials.
b
G owing deg ee days abo e 0°C
0
10
20
30
40
50
60
70
0 1,000 2,000 3,000 4,000
0.22
0.12
0.15
c
P ecipi a ion /
equilib ium e apo anspi a ion
0
10
20
30
40
50
60
70
0 1 2 43 5 6
C accumula ed (kg m-2)
a
Pho osyn he ically ac i e adia ion
days abo e 0°C
(mol pho ons m-2 season-1)
0
10
20
30
40
50
60
70
2,000 4,000 6,000 8,000 10,000
0.48
0.33
0.25
Fig. 5. Rela ionships be ween clima e a iables and pea ca bon accumula ion. The o al ca bon accumula ed o e he las 1000y (1ka)
a each si e compa ed o PAR0 (a), GDD0 (b) and he a io o p ecipi a ion o equilib ium e apo anspi a ion (c). Bog (omb o ophic) and
en (mine o ophic) si es (see Tables 1 and 2) a e shown in blue and g een, espec i ely, and sepa a e eg essions (R2 alues a e shown)
ha e been calcula ed o each si e ype, wi h a black eg ession line o he me ged da a se s. Ve ical e o ba s ep esen ch onological
unce ain ies (2σ) in es ima ing AD 1000 in each p o ile (n=90).
www.biogeosciences.ne /10/929/2013/ Biogeosciences, 10, 929–944, 2013
944 D. J. Cha man e al.: Clima e- ela ed changes in pea land ca bon accumula ion
an Bellen, S., Dallai e, P.-L., Ga neau, M. and Be ge on,
Y.: Quan i ying spa ial and empo al Holocene ca bon ac-
cumula ion in omb o ophic pea lands o he Eas main e-
gion, Quebec, Canada, Global Biogeochem. Cy., 25, GB2016,
doi:10.1029/2010GB003877, 2011.
an de Linden, M. and an Geel, B.: La e Holocene clima e change
and human impac eco ded in a Sou h Swedish omb o ophic
pea bog, Palaeogeog . Palaeoecol., 240, 649–667, 2006.
an de Linden, M., Ba ke, J., Vicke y, E., Cha man, D. J.,
and an Geel, B.: La e Holocene human impac and clima e
change eco ded in a no h Swedish pea deposi , Palaeogeog .
Palaeoecol., 258, 1–27, 2008.
Vi , D. H., Halsey, L. A., Baue , I. E., and Campbell, C.: Spa ial
and empo al ends in ca bon s o age o pea lands o con inen al
wes e n Canada h ough he Holocene, Can. J. Ea h Sci., 37,
683–693, 2000.
Wania, R., Ross, I., and P en ice, I. C.: In eg a ing pea lands and
pe ma os in o a dynamic global ege a ion model: 1. E alua-
ion and sensi i i y o physical land su ace p ocesses, Global
Biogeochem. Cy., 23, GB3014, doi:10.1029/2008GB003412,
2009.
Young, G. H. F., McCa oll, D., Loade , N. J., and Ki chhe e , A. J.:
A 500-yea eco d o summe nea -g ound sola adia ion om
ee- ing s able ca bon iso opes, Holocene, 20, 315–324, 2010.
Yu, Z. C.: Holocene ca bon lux his o ies o he wo ld’s pea lands:
Global ca bon-cycle implica ions, Holocene, 21, 761–774, 2011.
Yu, Z. C., Vi , D. H., Campbell, I. D., and Apps, M. J.: Unde -
s anding Holocene pea accumula ion pa e n o con inen al ens
in wes e n Canada, Can. J. Bo ., 81, 267–282, 2003.
Yu, Z., Beilman, D. W., and Jones, M. C.: Sensi i i y o no he n
pea land ca bon dynamics o Holocene clima e change, in: Ca -
bon cycling in no he n pea lands, edi ed by: Bai d, A. J., Belyea,
L. R., Comas, X., Ree e, A., and Sla e , L., Ame ican Geophys-
ical Union, Washing on DC, Geoph. Monog. Se ies, 184, 55–69,
doi:10.1029/2008GM000822, 2009.
Yu, Z. C., Loisel, J., B osseau, D. P., Beilman, D. W., and Hun , S.
J.: Global pea land dynamics since he Las Glacial Maximum,
Geophys. Res. Le ., 37, L13402, doi:10.1029/2010GL043584,
2010.
Biogeosciences, 10, 929–944, 2013 www.biogeosciences.ne /10/929/2013/