Biological activity and calcium carbonate dynamics in Greenland sea ice – Implication for the inorganic carbon cycle. Technical Report No. 92

BIOLOGICAL ACTIVITY AND CALCIUM CARBONATE DYNAMICS IN GREENLAND SEA ICE – IMPLICATION FOR THE INORGANIC CARBON CYCLE
Do e H. Søgaa d
PhD hesis 2014
BIOLOGICAL ACTIVITY AND
CALCIUM CARBONATE DYNAMICS
IN GREENLAND SEA ICE
– IMPLICATION FOR THE INORGANIC CARBON CYCLE
ISBN: 87-91214-68-8
Do e_omslag_Ok obe .indd 1 24-10-2014 07:30:13
BIOLOGICAL ACTIVITY AND CALCIUM CARBONATE
DYNAMICS IN GREENLAND SEA ICE
– IMPLICATION FOR THE INORGANIC CARBON CYCLE
PhD hesis 2014
Do e H. Søgaa d
Ti le: Biological ac i i y and calcium ca bona e dynamics in G eenland sea ice
– Implica ion o he ino ganic ca bon cycle
Sub i le: PhD hesis
Au ho : Do e H. Søgaa d
A i li a ions: G eenland Clima e Resea ch Cen e (c/o G eenland Ins i u e o Na u al Resou ces),
Ki ioq 2, Box 570, 3900 Nuuk, G eenland
Uni e si y o Sou he n Denma k, Campus ej 55, 5230 Odense M, Denma k
Publishe : G eenland Ins i u e o Na u al Resou ces, Nuuk G eenland
Yea o publica ion: 2014
PhD supe iso s
In e nal: P o esso Ronnie N Glud
Uni e si y o Sou he n Denma k
Campus ej 55
5230 Odense M, Denma k
Ex e nal: P o esso Sø en Rysgaa d
A c ic Resea ch Cen e, Depa men o Bioscience
Ny Munkegade 116, building 1540
8000 Aa hus C, Denma k
Please ci e as: Søgaa d D.H. (2014) Biological ac i i y and calcium ca bona e dynamics in G eenland
sea ice – Implica ion o he ino ganic ca bon cycle. PhD hesis. G eenland Clima e Re-
sea ch Cen e and Depa men o Biology, Uni e si y o Sou he n Denma k. G eenland
Ins i u e o Na u al Resou ces, 148 pp.
Keywo ds: Suba c ic sea ice • High A c ic sea ice • Ai -sea CO2 exchange • G eenland • Spa ial
a iabili y • Calcium ca bona e • Ne au o ophic ac i i y • Ne he e o ophic ac i i y
Rep oduc ion pe mi ed p o ided he sou ce is explici ly acknowledged
Layou : Tinna Ch is ensen
Co e pho o: Jakob Sie e s
Numbe o pages: 148
P in ed by: Rosendahls – Schul z G a i sk a/s
ISBN: 87-91214-68-8
EAN: 9788791214684
Ci cula ion: 160
Elec onic e sion: www.na u .gl
Da a shee
Con en s
Lis o publica ions 5
P e ace 6
Acknowledgemen s 7
Abs ac 8
Dansk abs ak (abs ac in Danish) 9
Eqikkaaneq (abs ac in G eenlandic) 10
CHAPTER 1 – INTRODUCTION 11
1.1 Se ing he scene 12
1.1.1 Gene al oceanog aphy o he s udy a eas 13
1.2 Phase I: T ansi ion om open wa e o ice-co e ed oceans 14
1.2.1 Sea ice o ma ion – abio ic p ocesses 15
1.2.2 Mic oo ganisms in newly o med sea ice 18
1.2.3 F os l owe s and b ine skim o ma ion – abio ic p ocesses 19
1.3 Phase II: G owing win e sea ice 20
1.3.1 Abio ic p ocesses in g owing win e ice 21
1.3.2 G ow h limi a ion o mic oo ganisms in win e sea ice 22
1.4 Phase III: T ansi ion om ice-co e ed ocean o open wa e s 23
1.4.1 Abio ic p ocesses in mel ing sea ice 23
1.4.2 Bio ic p ocesses in sp ing/summe sea ice 24
1.4.3 The open wa e pe iod 28
1.5 Ocean and sea ice in he con ex o global change 29
1.6 Conclusions and pe spec i es 30
1.7 Glossa y 31
1.8 Li e a u e ci ed 34

CHAPTER 2 – PUBLICATIONS 41
Pape I The ela i e con ibu ions o biological and abio ic 43
p ocesses o ca bon dynamics in suba c ic sea ice
Pape II Ikai e c ys al dis ibu ion in win e sea ice and 61
implica ions o CO2 sys em dynamics
Pape III F os l owe s on young A c ic sea ice: The clima ic, chemical 75
and mic obial signi i cance o an eme ging ice ype
Pape IV Au o ophic and he e o ophic ac i i y in A c ic i s -yea 95
sea ice: seasonal s udy om Malene Bigh , SW G eenland
Pape V G ow h limi a ion o h ee A c ic sea ice algal species: 111
e ec s o salini y, pH, and ino ganic ca bon a ailabili y
Pape VI Sho - e m a iabili y in bac e ial abundance, cell p ope ies, 121
and inco po a ion o leucine and hymidine in suba c ic
sea ice
Pape VII High ai -sea CO2 up ake a es in nea sho e and shel a eas 139
o Sou he n G eenland: Tempo al and spa ial a iabili y
5
PhD hesis by Do e Haubje g Søgaa d
Lis o publica ions
Pa o he disse a ion
Pape I: Søgaa d DH, Thomas DN, Rysgaa d S, Glud RN,
No man L, Kaa okallio H, Juul-Pede sen T, Geil us N-X
(2013) The ela i e con ibu ions o biological and abio ic
p ocesses o ca bon dynamics in suba c ic sea ice. Pola
Biol 36:1761 – 1777, doi:10.1007/s003-013-1396-6
Pape II: Rysgaa d S, Søgaa d DH, Coope M, Pucko M, Lenne
K, Papaky iakou TN, Wang F, Geil us NX, Glud RN, Ehn J, Mc-
Ginnis DF, A a d K, Sie e s J, Deming JW, Ba be D (2013) Ikai e
c ys al dis ibu ion in win e sea ice and implica ions o CO2
sys em dynamics. TC 7:707 – 718, doi:10.5194/ c7-707-2013
Pape III: Ba be DG, Ehn JK, Pucko M, Rysgaa d S,
Papaky iakou T, Deming J, Galley R, Søgaa d DH (2014)
F os l owe s on young A c ic sea ice: The clima ic, chemical
and mic obial signi i cance o an eme ging ice ype. J Geo-
phys Res-A mos, doi:10.1002/2014JD021736
Pape IV: Søgaa d DH, K is ensen M, Rysgaa d S, Glud RN,
Hansen PJ, Hilligsøe KM (2010) Au o ophic and he e o-
ophic ac i i y in A c ic i s -yea sea ice: seasonal s udy
om Malene Bigh , SW G eenland. Ma Ecol P og Se
419:31 – 45, doi:10.3354/meps08845*
Pape V: Søgaa d DH, Hansen PJ, Rysgaa d S, Glud RN (2011)
G ow h limi a ion o h ee A c ic sea ice algal species: e ec s
o salini y, pH, and ino ganic ca bon a ailabili y. Pola Biol
34:1157 – 1165, doi:10.1007/s00300-011-0976-3**
Pape VI: Kaa okallio H, Søgaa d DH, No man L, Rysgaa d
S, Tison JL, Delille B, Thomas DN (2013) Sho - e m a ia-
bili y in bac e ial abundance, cell p ope ies, and inco po-
a ion o leucine and hymidine in suba c ic sea ice. Aqua
Mic ob Ecol 71:57 – 73, doi:10.3354/ame01667
Pape VII: Rysgaa d S, Mo ensen J, Juul-Pede sen T, Sø-
ensen LL, Lenne K, Søgaa d DH, A end KE, Bliche ME,
Sej MK, Bend sen J (2012) High ai -sea CO2 up ake a es
in nea sho e and shel a eas o Sou he n G eenland: Tem-
po al and spa ial a iabili y. Ma Chem 128-129:26 – 33,
doi:10.1016/j.ma chem.2011.11.002
Rela ed wo k no included in he disse a ion
Long MH, Koopmans D, Be g P, Rysgaa d S, Glud RN,
Søgaa d DH (2012) Oxygen exchange and ice mel mea-
su ed a he ice-wa e in e ace by eddy co ela ion. BG
9:1957 – 1967, doi:10.5194/bg-9-1-2012
Geil us N-X, Galley RJ, Coope M, Halden N, Ha e A, Wang F,
Søgaa d DH, Rysgaa d S (2013b) Gypsum c ys als obse ed
in expe imen al and na u al sea ice. Geophys Res Le 40:
1 – 6, doi:10.1002/2013GL058479
Sø ensen LL, Jensen B, Glud RN, McGinnis DF, Sej MK, Sie-
e s J, Søgaa d DH, Tison JL, Rysgaa d S (2014) Pa ame e i-
za ion o a mosphe ic-su ace exchange o CO2 o e sea
ice. TC 8:853 – 866, doi:10.5194/ c-8-853-2014
Juul-Pede sen, A end KE, Mo ensen J, Bliche M, Søgaa d
DH, Rysgaa d S (submi ed) Seasonal and in e annual phy-
oplank on p oduc ion in a sub-a c ic j o d (God håbs j o d)
connec ed o he G eeland Ice Shee . Ma Ecol P og Se
Søgaa d DH, Glud RN, Rysgaa d S, Jody Deming (in p ep)
A compa a i e s udy o bio ic- and abio ic-induced oxygen
and ino ganic ca bon dynamics in me e hick win e and
young hin polynya ice. P epa ed o Ma Ecol P og Se
Mei e L, Søgaa d DH, Juul-Pede sen T, Bliche M, Rysgaa d
S, Glud RN, Sej M, A end K, Lenne K, Mo ensen J (in
p ep) In l uence o glacie uno and biology on he CO2
up ake in a Suba c ic G eenlandic j o d (God håbs j o d, SW
G eenland). P epa ed o Ma Chem
*) The da a ma e ial was a pa o my mas e ´s hesis, bu he da a p ocessing,
w i ing he pape and he e iew p ocess was done du ing my Ph.D. s udy.
**) A small pa o he da a ma e ial was a pa o a model p ojec , bu he
es o he da a, he da a p ocessing, w i ing he pape and he e iew p o-
cess was done du ing my Ph.D. s udy.
Pho o: Jakob Sie e s..
6PhD hesis by Do e Haubje g Søgaa d
P e ace
This disse a ion is he esul o a 3-yea Ph.D. p ojec con-
duc ed a he G eenland Clima e Resea ch Cen e (GCRC)
and he G eenland Ins i u e o Na u al Resou ces (GINR).
The p ojec was unded by he Commission o Scien-
i i c Resea ch in G eenland (KVUG). Much o he wo k has
been conduc ed wi h logis ical suppo by GCRC, GINR,
he Depa men o Educa ion, Chu ch, Cul u e and Equa-
li y (IIKNN), he Ma ine Basis Moni o ing P og ammes o
G eenland Ecosys em Moni o ing (GEM; www.g-e-m.dk),
he Canada Excellence Resea ch Chai (CERC) p og am and
he Danish Agency o Science, Technology and Inno a ion.
One o he i s desc ip ions o he anspo o CO2 ac oss
he sea ice-ocean in e ace (i.e., he sea ice CO2 pump) was
made by E.P. Jones and A.R. Coo e (1981) in he yea I was
bo n. Now 33 yea s la e he unde s anding o he seasonal
e en s con olling he ino ganic ca bon dynamics in ice-
co e ed seas is s ill a challenging subjec . A e y impo an
ac o in clima e change is he global ca bon budge ; and,
in o de o e i ne i , we need de ailed measu emen s o he
seasonal ino ganic ca bon dynamics in ice-co e ed seas.
My disse a ion consis s o wo chap e s: Chap e 1 is s uc-
u ed as a e iew wi h a de ailed discussion o he p ocesses
d i ing he ino ganic ca bon cycle in high la i ude wa e s
du ing all phases o he sea ice g ow h and decay cycle (sec-
ions 1.2 – 1.4). In addi ion, he e is a discussion o he A c ic
Ocean and sea ice co e age in he con ex o global change
(sec ion 1.5) and, i nally an ou line and discussion o u u e
s udies needed o imp o e ou unde s anding o he sea-
sonal e en s con olling he dynamics o ino ganic ca bon in
high la i ude oceans (sec ion 1.6); Chap e 2 consis s o se en
published pee - e iewed pape s. The i ndings in his hesis
p o ide s ong e idence o he idea o a sea-ice-d i en ca -
bon pump in Suba c ic and High A c ic sea ice, imp o e he
s a e o knowledge on he ela i e con ibu ion o bio ic and
abio ic p ocesses o ca bon dynamics wi hin sea ice, and p e-
sen a compila ion o he cu en knowledge o he seasonal
ino ganic ca bon dynamics in ice-co e ed seas du ing sea ice
g ow h and decay. Fu he mo e, i p o ides new knowledge
o he in l uence o biological p ocesses and glacie uno on
he pCO2 l ux in A c ic coas al wa e s.
Pho o: Jakob Sie e s.
7
PhD hesis by Do e Haubje g Søgaa d
Acknowledgemen s
I owe emendous hanks o my supe iso s Sø en Rysgaa d
(Aa hus Uni e si y) and Ronnie N. Glud (Uni e si y o Sou he n
Denma k) o excellen supe ision and suppo , o making
his p ojec possible in he i s place as well as in oducing me
o he i eld wo k in Malene Bigh , SW G eenland and Young
Sound, NE G eenland and o many quali y hou s spend in he
A c ic win e .
Thanks o John Mo ensen and Lo enz Mei e o a ui ul co-
ope a ion on he subjec o he in l uence o glacie uno
and biology on he CO2 up ake in a Suba c ic G eenlandic
j o d. Nicolas-Xa ie Geil us (ARC) is hanked o cons uc i e
ideas and discussions du ing ou s udy o ikai e in sea ice.
Thanks o he wa m welcome om GINR when I a i ed in
G eenland in 2007. Special hanks o di ec o Klaus Nygaa d
(GINR) o suppo ing me i nancially and logis ically. Thanks
o my colleagues a GINR o an inspi ing and pleasan
wo king en i onmen – especially Ma in E. Bliche , K is ine
E. A end , Thomas Juul Pede sen, John Mo ensen, Ka ine
Raund up and Rasmus Hedeholm o scien i i c ad ice and
o co ec ion o my manusc ip s.
This p ojec could no ha e been comple ed wi hou help in
he i eld: Thomas K ogh, Michael S. Sch øde , Thomas Juul-
Pede sen, Paul Ba y, K is ine E. A end , Ma in E. Bliche ,
Mo en K is ensen, Lo en z Mei e, Kunuk Lenne , I ali Len-
ne and Egon F andsen.
Thanks o all my co-au ho s o commen s on manusc ip s
bu especially o b oadening my scien i i c ho izon by chal-
lenging my wo k.
A special hanks o my bes iend Cha lo e F. Michelsen
o helping me whene e needed and o p oo eading my
hesis.
I am g a e ul o he suppo o my amily: my pa en s Ben
and Hanne Søgaa d, my sis e Susanne Søgaa d, my g and-
pa en s Ma gi and Bø ge Sø ensen and my in-laws Ulla and
Pe e Sch øde o always encou aging me o explo e new
ho izons and o p o iding unwa e ing suppo .
Deep el hanks o Michael S. Sch øde , my wonde ul hus-
band, who has pa icipa ed igh om he beginning – in he
i eld wi h s a is ical suppo , o ha ing pa ience wi h me a
busy imes, o engaging in in e es ing discussions, o p o-
iding a p o ound sense o s abili y and o gi ing me my wo
daugh e s Si and Naja. You ollowed me o G eenland and
made his p ojec possible. I will o e e be g a e ul o his.
Ilaqu akka asasakka uannu pingaa ne paa usi!
Nuuk, Augus 2014
Do e H. Søgaa d
o
co ec ion
o
my
manusc ip s
.
Do e H Søgaa d
14 PhD hesis by Do e Haubje g Søgaa d
ec s he wa e mass p ope ies and con ibu e o di e ences
in he ma ine sys ems as well as in sea ice p oduc ion and dis-
ibu ion in G eenland coas al a eas and j o ds.
Wi h espec o he Young Sound A ea in NE G eenland (Fig.
2; Rysgaa d e al. (II), Ba be e al. (III)), he Eas G eenland
Cu en (i.e., a con inua ion o he T anspola Cu en ) has
an impo an in l uence on he wa e mass p ope ies in his
a ea – especially due o he la ge amoun s o sea ice i ca ies
along wi h i . Ano he oceanog aphic ea u e in his a ea is
he la ge seasonal inpu o eshwa e om he G eenland Ice
Shee , e es ial uno and mel wa e om bo h sea ice and
cal ed glacial ice (Rysgaa d and Glud 2007). Fu he mo e, he
a ea ou side Young Sound is a polynya si e, whe e new sea
ice is p oduced and equen ly blown away, he eby allowing
new ice o o m again and again (Pede sen e al. 2010). The
ma ine sys em in he God håbs j o d a ea in SW G eenland
(Fig. 2; Søgaa d e al. (I), Søgaa d e al. (IV), Kaa okallio e
al. (VI), Rysgaa d e al. (VII)) is a ec ed by h ee p incipal wa-
e masses, a mosphe ic hea exchange and mel / eeze p o-
cesses (Mo ensen e al. 2011, 2013). Two o he h ee p in-
cipal wa e masses a e ound ou side he j o d – sub-pola
mode wa e and coas al wa e ; whe eas he hi d, eshwa e ,
comes om mel wa e uno om he G eenland Ice Shee ,
e es ial uno , mel wa e om sea ice and cal ed glacial
ice (Mo ensen e al. 2011, 2013). The eshwa e inpu in his
a ea om he G eenland Ice Shee induces a seasonal s a i i -
ca ion o he uppe pa o he wa e column and is a sou ce o
la ge amoun s o bioa ailable nu ien (A end 2011, Calbe
2011, Lyde sen e al. 2014).
1.2 Phase I: T ansi ion om open
wa e o ice-co e ed oceans
In his sec ion, he mechanisms behind sea ice o ma ion,
phase I, a e desc ibed as well as he mic oo ganisms in
newly o med sea ice and hei con ibu ion o he CO2 dy-
namics in his phase (sec ions 1.2.1 and 1.2.2). In addi ion,
he mechanisms behind b ine skim and os l owe o ma-
ion a e b ie l y desc ibed (1.2.3). The p ocesses d i ing CO2
exchange in he ansi ion pe iod om open wa e o ice-
co e ed seas a e discussed and whe he phase I ac s as ne
sink o sou ce o CO2 o he a mosphe e.
G ease ice:
hin laye o azil ice
Pancake ice:
consolida ion o
azil in o la ge uni s
F azil c ys als:
3 o 4 mm in diame e
New ice:
ecen ly ozen sea wa e
Nilas: designa es a sea ice c us
up o 10 cm in hickness
Young ice:
om 10 o 30 cm
Fi s -yea ice: mel s away du ing
he sp ing and summe mon hs
Mul i-yea ice
2) Calm condi ions1) Agi a ed condi ions
G
e
a
s
e
i
c
e
:
hin la
y
e o
azil
ice
Pan
cak
e i
ce
:
c
onsolida ion o
azil in o la
ge
un
i s
F azi
l
c
y
s a
l
s:
3
o 4
mm
in
di
ame
e
N
ew
i
c
e:
ecen l
y
ozen sea wa e
N
il
a
s
:
d
esi
g
na es a sea ice c us
up
o 10 cm in hickn
ess
Fi s -
y
ea ice
:
m
el s awa
y
du i
ng
he sp ing and summe mon hs
2)
Ca
l
m con
d
i ions1) Agi a e
d
con
d
i ions
Figu e 3. Sea ice de elopmen
s ages 1) agi a ed condi ions
wi h ice g ow h in wa e- i eld
and 2) calm condi ions wi h ice
g ow h h ough quiescen bo -
om eezing.

15
PhD hesis by Do e Haubje g Søgaa d
1.2.1 Sea ice o ma ion – abio ic p ocesses
In au umn and win e , when he ocean su ace laye is
cooled down o empe a u es close o –1.86° C ( eezing
poin o seawa e wi h a salini y o 34), ice c ys al s a o
g ow on he su ace (Weeks 2010) and o m a soupy suspen-
sion known as azil ice c ys als (i.e., 3 o 4 mm in diame e ;
Fig. 3).
Unde calm condi ions, he ice c ys als eeze oge he o
o m a shee o new ice called Nilas, which designa es a sea
ice c us up o 10 cm in hickness wi h andomly-o ien ed
ice c ys als (i.e., g anula ice ex u e; Timco and Weeks 2010;
Fig. 3 and Fig. 4). I is a his poin ha os l owe s a e
some imes o med om he b ine skim on he ice su ace
(desc ibed in mo e de ail in sec ion 1.2.4). As he shee ice
hickens h ough congela ion a he ice-wa e in e ace, he
ansi ional g anula /columna ice laye o ms (Eicken and
Lange 1989). This laye is a ew cen ime es hick ansi ion
zone ha is mainly cha ac e ised by elonga ed g ains (Fig. 4).
Below he ansi ional g anula /columna laye , he sea
ice consis s o columna ice, cha ac e ised by e ically-
elonga ed ice c ys als (Fig. 4 and Fig. 5). The sea ice is classi-
i ed as young ice when he ice shee becomes hicke han
10 cm and i s -yea sea ice when he ice shee becomes
hicke han 30 cm (Fig. 3; Weeks 2010). I he i s -yea ice
su i es a leas one mel ing season (i.e., one summe ), i is
called mul i-yea ice (Fig. 3).
Unde mo e agi a ed condi ions wi h ice g ow h in wa e-
i elds, azil c ys als consolida e in o g ease. The g ease ice
laye consolida es in o ice discs known as pancakes (Fig. 3).
As hey g ow om a ew cen ime es o a ew me e s ac oss,
hey solidi y and hicken mechanically by a ing on op
o each o he . Pancakes eeze oge he o o m cakes and
l oes, which con ain a la ge amoun o ice wi h a g anula
ex u e (Fig. 4). Wi h ongoing eezing, he pancakes adhe e
in o a con inuous ice shee by bo om eezing a he ice-
wa e in e ace.
S a ig aphy
classi ica ion
G ow h condi ions
Snow deposi ion
Flooding
Tu bulen mixing
Quiscen
(g ow h a e,
cu en shea )
Gene ic
Snow ice
F azil ice
T ansi ion
zone
Congela ion
ice
Ice wa e in e ace
Tex u al
G anula
Mixed
columna /
g anula
Columna
Skele al laye
Figu e 4. Schema ic summa-
izing he main ice ex u e and
g ow h condi ions o i s -yea
sea ice. The i gu e is adap ed
om Eicken (2003). The ho i-
zon al hin-sec ion pho o-
g aphs is o i s -yea sea ice
om Young Sound in NE G een-
land o de ailed desc ip ion
see Rysgaa d e al. (II) (pho o
cou esy o M. Pucko).
16 PhD hesis by Do e Haubje g Søgaa d
The azil – congela ion sea ice g ow h p ocess desc ibed
abo e is he p ocess ha occu s mos equen ly, bu o he
sea ice ypes exis s, e.g., snow ice and supe imposed ice.
Snow ice is o med h ough sea wa e l ooding due o nega-
i e eeboa d and hick snow co e (F is sen e al. 1998),
and i is qui e po ous (Eicken 2003, Rysgaa d e al. (II),
Søgaa d e al. (IV)). Supe imposed ice occu s du ing sp ing
and summe when he snow mel s in e nally and mel wa-
e e eezes ei he in he snow o a he snow-ice in e ace,
when he empe a u e g adien s wi hin snow and ice a e
nega i e (Hass e al. 2001, Nicolaus e al. 2009).
As soon as sea wa e solidi i es, some o he sal s and gases
p esen in he seawa e a e ejec ed, whe eas he es a e
apped wi hin he b ine pocke s, channels and ubes (Fig.
5; Weeks and Ackley 1982, Pe ich and Eicken 2010). A e-
duc ion in sea ice empe a u e dec eases he b ine olume
and, concu en ly, inc eases he b ine salini y and b ine
pCO2 h ough a dec ease in b ine CO2 solubili y (Cox and
Weeks 1983, Papadimi iou e al. 2004). Thus, a his poin ,
he pCO2 in he sea ice b ine is highe han ha in he ai
abo e he sea ice; and, he e o e, he sea ice b ine has he
po en ial o elease CO2 o he a mosphe e. Howe e , when
he sea ice empe a u e eaches –5˚ C, he b ine olume de-
c eases; a b ine olume o 5 % is gene ally conside ed he
h eshold a which he sea ice ma ix becomes impe me-
able, hus p e en ing ai -sea ice gas exchange (Golden e
al. 1998, Zhou e al. 2013). This pe cola ion h eshold a -
ies wi h changes in he ice c ys al s uc u es – e.g., g anula
ice shows a highe pe cola ion h eshold han columna ice
(Fig. 4; Weeks 2010). In addi ion, as empe a u e dec eases
and solu e concen a ion inc eases, calcium ca bona e
p ecipi a es (Ma ion e al. 2001). On he basis o he mody-
namic equilib ium calcula ion, calcium ca bona e p ecipi a-
ion was p edic ed o occu du ing na u al sea ice o ma ion
(Assu 1958), which was la e con i med i s in eezing sea
wa e by Richa dson (1976), hen in a i i cial sea ice (Tison e
al. 2002) and, i nally, in An a c ic and A c ic sea ice as ikai e
(Dieckmann e al. 2008; 2010, Rysgaa d e al. 2011; 2012,
Fische e al. 2013, Geil us e al. 2013a, Nomu a e al. 2013b,
Rysgaa d (II)). A p esen , i is s ill no clea whe he ikai e
is he only calcium ca bona e phase o med in sea ice
(Dieckmann e al. 2010).
The calcium ca bona e o ma ion inc eases he amoun o
CO2 in he b ine beyond ha a ibu ed solely o he solu-
bili y e ec . The calcium ca bona e c ys als a e apped
wi hin he in e s ices be ween he ice c ys als (Rysgaa d
(II)), whe eas he CO2 eleased h ough calcium ca bona e
p oduc ion wi hin he b ine can be los om he sea ice.
The e o e, as sea ice g ows, b ine d ainage leads o an ex-
po o gases om he sea ice, lea ing sea ice deple ed in
CO2 compa ed o ambien seawa e (Rysgaa d e al. 2007,
C abeck e al. 2014, Søgaa d e al. (I)). B ine d ainage om
sea ice causes he o ma ion o highly saline dense cold
1 cm
65 o 75 cm sec ion
2 mm
Figu e 5. Thin sec ion o i s yea sea ice showing ice pla ele s
and he b ine pocke s along he g ain bounda ies. Fo de ailed de-
sc ip ion see Rysgaa d e al. (II) (Pho o cou esy o M. Pucko).
g m
-2
FF
FYYS
POLY
FYKB
SC70
NI-MY
GIS
BS
SC17
SC7
FIFS
FIAN
0
10
20
30
40
Figu e 6. Calcium ca bona e concen a ion in di e en ice ypes in
he A c ic and An a c ic: FYKB= Fi s -yea sea ice in Kapisigdli
Bigh , SW G eenland (Søgaa d e al. (I)), FYYS=Fi s -yea sea ice in
Yound Sound, NE G eenland and POLY=newly o med polynya ice
in Young Sound, NE G eenland (Rysgaa d e al. (II)), FF= os l owe s
and BS=b ine skim in Young Sound, NE G eenland (Ba be e al. (III)),
NI-MY=Nilas o mul i-yea ice in he An a c ic om Dieckmann
e al. (2008), FIAN=Fas ice in An a c ic om Fische e al. (2013),
FIFS=Fas ice in F am S ai om Rysgaa d e al. (2012), GIS=Sea ice
nea he G eenland Ice Shee (unpublished da a D.H. Søgaa d),
SC70=70 cm hick snow co e and SC17=17 cm hick snow co e in
Young Sound, NE G eenland (unpublished da a D.H. Søgaa d) and
SC7=7 cm hick snow co e in Kapisigdli Bigh , SW G eenland
(Søgaa d e al. (I)).
17
PhD hesis by Do e Haubje g Søgaa d
wa e ha sinks o deepe laye s and con ibu es o he
global ocean ci cula ion. Fu he mo e, obse a ions in he
A c ic sugges ha TCO2 can be anspo ed below he
pycnocline and, subsequen ly, be inco po a ed in o in-
e media e and deep-wa e masses (Rysgaa d e al. 2007;
Rysgaa d e al. 2011).
Knowledge on p ecipi a ion o calcium ca bona e in newly
o med sea ice and i s e ec on ino ganic ca bon dynamics
a e s ill no well desc ibed. This issue is add essed in wo o
ou pape s (Søgaa d e al. (I), Rysgaa d e al. (II)), whe e we
epo measu emen s o calcium ca bona e concen a ion
o 1 g m-2 in a 2 weeks old Suba c ic sea ice in Kapisigdli
Bigh in SW G eenland (Fig. 2 and Fig. 6) and 10 g m-2 in a
less han one-week-old High A c ic sea ice in Young Sound,
NE G eenland (Fig. 2 and Fig. 6). The amoun o calcium
ca bona e in he newly o med High A c ic sea ice (Rysgaa d
e al. (II)) was 2 o 5 imes highe han hose measu ed in
o he sea ice ypes in bo h A c ic and An a c ic wa e s
and also 10 imes highe han concen a ions measu ed in
newly o med Suba c ic sea ice in Kapisigdli Bigh in SW
G eenland (Fig. 2 and Fig. 6; Søgaa d e al. (I)). Howe e ,
he calcium ca bona e concen a ion in he newly o med
High A c ic sea ice was lowe han concen a ions obse ed
in i s -yea sea ice and in 70 cm- hick snow co e a he
same sampling loca ion in NE G eenland (Fig. 2 and Fig. 6;
Rysgaa d e al. (II)) as well as in land- as ice in F am S ai
(Fig. 6.; Rysgaa d e al. 2012). These esul s indica e e y
dynamic condi ions o calcium ca bona e o ma ion e en
on sho imescales bu also indica e conside able spa ial
dis ibu ion o calcium ca bona e (Søgaa d e al. (I)).
The po en ial in l uence o mel ing he en i e newly o med
ice co e in Kapisigdli Bigh in SW G eenland (Fig. 2;
Søgaa d e al. (I)) and Young Sound in NE G eenland (Fig. 2;
Rysgaa d e al. (II)) in o a 20 m hick mixed laye ( ypical o
summe condi ions in hese loca ion) on he CO2 l ux can be
de e mined using he measu ed ice ca bon chemis y (Table
1) and he ini ial mixed laye cha ac e is ics (Table 2) om
he wo di e en egions. Assuming ha mel occu s o e
one mon h, he esul an ai -sea CO2 l ux o e u n o p e-
mel condi ions would be – 3.4 mmol C m-2 d-1 in Kapisigdli
Bigh in SW G eenland (Fig. 2) and – 2.3 mmol C m-2 d-1 in
Young Sound in NE G eenland (Fig. 2). The po en ial ai -sea
CO2 l ux a y by a ac o o 0.7 wi h he highes po en ial
ai -sea CO2 l ux es ima ed in Suba c ic sea ice (Søgaa d e
al. (I), Rysgaa d e al. (II)). The eason o his di e ence is
unclea bu indica es ha he age o he sea ice plays an im-
po an ole o his po en ial l ux. In addi ion, his i nding
sugges s ha TA and TCO2 ea u e high a iabili y be ween
ice loca ions and clea ly emphasizes he impo ance o
s udies co e ing he spa ial a iabili y in hese pa ame e s
in o de o make eliable ca bon budge s.
In hese calcula ions, we assume ha sea ice o ma ion oc-
cu s only once. Howe e , in polynya a eas, sea ice wi h low
bulk concen a ions o CO2 and high alkalini ies is p oduced
(Rysgaa d e al. (II)) and equen ly blown away om he
a ea, which he eby allows new ice o o m again and again.
The unc ion o a polynya and he in l uence on CO2 exchange
a e no well unde s ood, bu he p oduc ion o new sea ice in
hese a eas may con ibu e o a signi i can CO2 elease o he
a mosphe e, which is balanced by he dissolu ion o calcium
A ea and Si e Da e Type TCO
2
(μmol kg
-1
)
TA
(μmol kg
-1
)
TA:TCO
2
Bulk
salini y
Re e ence
G eenland
Kapisigdli Bigh 17 Feb ua y 2010 Newly o med sea ice 281 359 1.28 4.6 Søgaa d e al. (I)
Kapisigdli Bigh 11 o 15 Ma ch 2010 Win e sea ice 233 280 1.20 5.6 Søgaa d e al. (I)
Kapsigdli Bigh 8 Ap il o 1 May 2010 Sp ing/summe sea ice 282 356 1.26 2.8 Søgaa d e al. (I)
Young Sound 17 Ma ch 2012 Win e sea ice 406 516 1.27 6.5 Rysgaa d e al. (II)
Young Sound 20 Ma ch 2012 Newly o med polynya sea ice 502 605 1.21 7.8 Rysgaa d e al. (II)
G eenland/S alba d
F am S ai 25 o 29 June 2010 Sp ing/summe sea ice 221 420 1.90 3.9 Rysgaa d e al. (2012)
Table 1. Sea ice bulk condi ions o TCO2, TA and salini y du ing di e en i eld campaigns.
A ea and Si e Da e TCO
2
(μmol kg
-1
)
TA
(μmol kg
-1
)
TA:TCO
2
Bulk
salini y
Re e ence
G eenland
Kapisigdli Bigh 17 Feb ua y 2010 2110 2180 1.03 33 Søgaa d e al. (I)
Kapisigdli Bigh 11 o 15 Ma ch 2010 2095 2230 1.06 32.7 Søgaa d e al. (I)
Kapsigdli Bigh 8 Ap il o 1 May 2010 2015 2138 1.06 32.7 Søgaa d e al. (I)
Young Sound 17 Ma ch 2012 2101 2276 1.08 31.7 Rysgaa d e al. (II)
Young Sound 20 Ma ch 2012 2070 2205 1.07 31.7 Rysgaa d e al. (II)
G eenland/S alba d
F am S ai 25 o 29 June 2010 1987 2203 1.11 32.6 Rysgaa d e al. (2012)
Table 2. Su ace wa e condi ions below sea ice o TCO2, TA and salini y du ing di e en i eld campaigns.
18 PhD hesis by Do e Haubje g Søgaa d
ca bona e when he sea ice mel s. On he o he hand, i is
possible ha polynya a eas ac as a downwa d e ical ans-
po mechanism o emo e CO2 ejec ed om sea ice away
om he su ace laye and, he e o e, ensu e a ne CO2 l ux
in o he ocean o e he en i e yea (Rysgaa d e al. 2011). Ou
i ndings o high calcium ca bona e concen a ions in newly
o med polynya ice sugges ha polynya o ma ion inc eases
he po en ial o seawa e up ake o CO2 (Rysgaa d e al. (II)).
Ano he scena io is ha he o ma ion o calcium ca bona e
and he concen a ion o solu es in newly o med polynya
sea ice lead o CO2 de-gassing. I all CO2 p oduced du ing
he p ecipi a ion o calcium ca bona e (1) we e eleased o
he a mosphe e, hen he consump ion o CO2 du ing dis-
solu ion o he calcium ca bona e mine al in he mel ing
phase would balance he e l ux du ing p ecipi a ion (Rys-
gaa d e al. 2011; 2012). Thus, calcium ca bona e would no
con ibu e o he pola ca bon cycle. Howe e , as soon as
an excess o CO2 is ejec ed oge he wi h b ine o he un-
de lying wa e column and anspo ed away om he sea
ice o ma ion egion, hen he mine al may po en ially ha e
an impo an ole in he pola ca bon cycle as p oposed by
Rysgaa d e al. (2007; 2011 and 2012). Howe e , i is s ill c i i-
cal ha he su ace sea wa e laye is exposed o a su i cien
ime so ha he mixed laye has ime o equilib a e wi h he
a mosphe e, which may no be he case in High A c ic a eas
wi h sho open wa e pe iods (F ansson e al. 2009).
In he newly o med sea ice in Young Sound in NE G een-
land, we obse ed high concen a ions o calcium
ca bona e, and a his ea ly s age, he sea ice was s ill pe -
meable wi h a b ine olume o e 5 % (Søgaa d e al. (I),
Rysgaa d e al. (II)). As men ioned ea lie , p ecipi a ion
o calcium ca bona e inc eases he amoun o CO2 in he
b ine beyond ha a ibu ed solely o he solubili y (1). As
a consequence, CO2 may di use o he a mosphe e (Fig.
1; Papadimi iou e al 2004, Nomu a e al. 2006; 2010a;
2010b, Loose e al. 2011a, Rysgaa d e al. 2011). Since high
ice pe meabili y is usually encoun e ed in newly o med
sea ice, we would expec some ice-a mosphe e l uxes du-
ing sea ice o ma ion. Indeed, we measu ed a CO2 e l ux
o 3.67 ± 1.99 mmol m-2 d-1 abo e newly o med sea ice in
Young Sound in NE G eenland (Fig. 2), using chambe l ux
measu emen s (Ba be e al. (III)), which is consis en wi h
labo a o y expe imen s by Nomu a e al. (2006; 2010) and
o i ndings by Geil us e al. (2013), who es ima ed a CO2 e-
lease om young g owing sea ice o 4.2 – 9.9 mmol m-2 d-1.
This sugges ed a elease o CO2 om newly o med sea ice.
Howe e , ecen s udies ha e shown ha he la ges l ux
o TCO2 and CO2 is d i en by b ine d ainage o he unde -
lying wa e column and subsequen ly inco po a ion in o
deep-wa e masses (Rysgaa d e al. 2007, Sej e al. 2011).
1.2.2 Mic oo ganisms in newly o med sea ice
Mic oo ganisms in sea ice ha e been epo ed o mo e
han 160 yea s (Ho ne 1985 and e e ences he ein); bu ,
o ou bes knowledge, he biological p ocesses du ing
he ini ial s ages o sea ice o ma ion ha e no been well
desc ibed. The ini ial s ages o sea ice o ma ion gene ally
begin when he e a e s ill subs an ial mic obial popula ions
le in he wa e column (desc ibed in sec ion 1.4.3). As a
esul , pa icles such as i uses, bac e ia and he e o ophic
(e.g., l agella es and cilia es) and au o ophic (e.g., dia oms)
p o is s a e o en sca enged om he wa e column as he
newly o med azil ice ise o he su ace (Fig. 3; Ga ison e
al. 1990, Reimni z e al. 1992, G ossmann and Glei z 1993,
G adinge and Ikä alko 1998, Kaa okallio e al. 2006). I is
also possible ha l oa ing ice algal agg ega es can be e o-
zen in o he sea ice du ing au umn and, he e o e, ac as a
seeding s ock (Assmy e al. 2013 and e e ence he ein).
O ganisms la ge han 10 μm a e selec i ely sca enged om
he wa e column in o he sea ice and can accumula e a
concen a ions highe han ha in he unde lying sea wa e
(G adinge and Ikä alko 1998, Riedel e al. 2007, Mikkelsen e
al. 2008, Rózanska e al. 2008). Bac e ia seem o become en-
ained in he sea ice along wi h mic o-algae (e.g., Riedel e al.
2007). The bac e ial cells may a ach o he ou e su ace o he
algae (e.g., epiphy ic a achmen ), o o pa icles in he wa e
column, which can subsequen ly anspo hem in o he sea
ice. The sea ice bac e ial communi y closely esembles he
sea wa e bac e ial communi y in oligo ophic sys ems; hus,
selec ion p ocesses du ing sea ice o ma ion seem o play a
mino ole (Collins e al. 2010). By con as , in mo e p oduc i e
egions, selec ion p ocesses due o i al lysis o he he e o-
ophic bac e ial communi y is possible (Collins e al. 2011).
Once inco po a ed in o he sea ice, he mic oo ganisms a e
challenged wi h changes in space, ligh a ailabili y, salini y,
nu ien s, TCO2 and O2 concen a ions, empe a u e and pH
(G adinge and Ikä alko 1998, Søgaa d e al. (I), Søgaa d
e al. (IV – V), Kaa okallio e al. (VI)). These mic o-en i-
onmen al di e ences can lead o d ama ic di e ences in
he composi ion and magni ude o he mic obial commu-
ni y li ing he e. Sea ice en i onmen s a e domina ed by
psych o ole an and/o psych ophilic o ganisms (Cameo a
and Makka 1998), and adap a ion o low empe a u es in
all cellula componen s is o g ea impo ance o algae and
bac e ia li ing in cold en i onmen s.
Fo some sea ice algae and p o ozoans, one ep oduc i e
s a egy is he o ma ion o obus s ess esis an cys s, which
can lie do man un il sui able condi ions o g ow h a e p e-
sen . Sea ice algae ha a e s essed by ex eme ice salini ies
a e likely o exhibi inc eased halo ole ance and some s ud-
ies show ha sea ice dia oms emain physiologically ac i e
a salini ies abo e 100 (S oecke e al. 1997). A ecen s udy
19
PhD hesis by Do e Haubje g Søgaa d
showed ha dia oms a e he mos success ul colonise s o
newly o med sea ice (G adinge and Ikä alko 1998). In ad-
di ion, we ha e shown ha dia oms a e less a ec ed by in-
c easing salini ies, which migh explain he high abundance
o dia oms in sea ice (Søgaa d e al. (V)). The di e ences in
ole ance be ween sea ice algal species ha e been asc ibed
o a ious abili ies o osmo ic acclima ions, e.g., p oduc ion
o osmoly es (such as dime hylsulphoniop opiona e; DMSP),
which balances he ionic p essu e du ing changes in salini y
(Glei z and Thomas 1992, A igo e al. 2010). In addi ion o
high salini ies, sea ice algae mus also adap o he low ligh
condi ions in he sea ice, whe e acul a i e he e o ophy is an
impo an su i al s a egy (Ho ne and Alexande 1972). Sea
ice algae a e known o be adap ed o low ambien ligh le els
and a e able o g ow benea h se e al me e s o ice and snow
only ecei ing < 1 % o he sola i adiance (Ho ne & Sch ade
1982, Gosselin e al. 1990, G adinge and Ikä alko 1998, Mock
and G adinge 1999, Lazza a e al. 2007).
Cold adap a ion by sea ice bac e ia include main aining
memb ane l uidi y (Gouno & Russell 1999), modi i ca ion
o amino acid composi ion o he p o eome (Deming 2010),
i.e., con e ing l exibili y o p o eins o hei enzyma ic
unc ions, and s o age o in e cellula ese es in he o m
o la ge polyme s o polyhyd oxyalkanoa es (PHA; Deming
2010, Kaa okallio e al. (VI)). Sea ice bac e ia also elease
an i eeze p o eins, cold-ac i e enzymes and exopolyme ic
subs ances (EPS; Felle and Ge day 2003, Ma x e al. 2009,
Deming 2010). In addi ion o low empe a u es, he sea ice
bac e ia also need o adap o ex emely high salini ies and
o p o ec hemsel es agains osmo ic shock. These adap a-
ion p ocesses include he p oduc ion o accumula ion o
in acellula , compa ible solu es ( ypically suga s and amino
acids), changes in memb ane a y acid composi ion and
by he p oduc ion o sal - ole an enzymes (Thomas and
Dieckmann 2002, Bowman 2008).
High concen a ions o EPS ha e been measu ed in A c ic
sea ice h oughou he sea ice season, and EPS is eleased
bo h by bac e ia and algae in sea ice. The ole o EPS in sea
ice is o p o ec he algal and bac e ial cells agains he
ha sh en i onmen al condi ions, assis in cell locomo ion,
se e as a ca bon- ich subs a e, p o ide a de ence agains
g azing and c ea e a mic ohabi a in which bac e ial a ach-
men is a ou ed he eby inc easing bac e ia-media ed p o-
cesses (Deming 2010).
The p oduc i i y o he mic oo ganisms inhabi ing sea ice
du ing he i s s ages o ice o ma ion and de elopmen
is no well desc ibed. Howe e , in Søgaa d e al. (I) and
Søgaa d e al. (IV), we showed ha he ice-associa ed bio-
logical communi y in newly o med sea ice was ne he e o-
ophic wi h a bac e ial ca bon demand o 0.50 mg C m-2 d-1
in Kapisigdli Bigh in SW G eenland (Fig. 2) and 0.20 mg C
m-2 d-1 in Malene Bigh in SW G eenland (Fig. 2). An ob ious
ques ion is whe he he biological p ocesses a ec he TCO2
concen a ion wi hin he sea ice in his phase. The ela i e
e ec s o he biological ac i i y and p ecipi a ion o calcium
ca bona e on he ai -sea exchange o CO2 can be es ima ed.
A bo h loca ions in he newly o med ice (i.e., Kapisigdli
Bigh and Malene Bigh ; Fig. 2) he ice-associa ed biological
communi ies we e ne he e o ophic wi h a CO2 p oduc ion
a e o 4.10 o 2.10 mg m-2 which we e e y low compa ed o
he in eg a ed calcium ca bona e concen a ion o 1,000 mg
m-2 (Søgaa d e al. (I)). In ou s udies he highes bac e ial
p oduc ion was obse ed a he base o he newly o med
sea ice (Søgaa d e al. (I), Søgaa d e al. (IV)). Consequen ly,
his indica es ha al hough he ela i e con ibu ion o he
biological p ocesses o he ca bon dynamics is low, i is s ill
possible ha he obse ed CO2 e l ux abo e newly o med
sea ice in some a eas (see desc ip ion in sec ion 1.2.1) is
d i en, in pa , by he espi a ion o he he e o ophic com-
muni y a he base o he sea ice.
1.2.3 F os l owe s and b ine skim o ma ion
– abio ic p ocesses
Du ing he ini ial hou s o sea ice o ma ion, a highly saline
laye o b ine skim is o en obse ed on he su ace o he
new and young ice (D inkwa e and C ocke 1988, Pe o ich
and Rich e -Menge 1994, Islei son e al. 2014). The skim laye
is ypically 1 – 2 mm hick and is highly saline, yielding
salini ies in he o de o 30 o 120 (Douglas e al. 2012,
Geil us e al. 2013a, Ba be e al. (III)). I is sugges ed ha
his skim laye is o med by b ine anspo as he b ine
channels cons ic du ing sea ice g ow h (Ma in e al. 1995).
In calm wind condi ions (< 5 m s-1), os l owe s may o m
on op o he newly o med ice and b ine skim as dis inc
nodules and, hen, expand om hei nuclea ion si es a-
dially ou wa ds in all di ec ions (S yle and Wo s e 2009,
Ba be e al. (III)). In he li e a u e, he e is s ill an ongoing
deba e whe he os l owe s o m due o sublima ion o
e apo a ion om he sea ice su ace a he han om de-
posi ion om he a mosphe e (Domine e al. 2005). In he
mean ime, ou measu emen s show ha he ini ial δ18O
alues in he os l owe s a e close o 0 ‰, which is simi-
la o he signa u e o he su ace slush laye (Ba be e al.
(III)), sugges ing ha newly o med os l owe s a e com-
posed p ima ily o b ine. This is also suppo ed by p e ious
wo k showing ha b ine can be wicked in o os l owe s
om b ine skim (Pe o ich and Rich e -Menge 1994, Ma in
e al. 1995, Roscoe e al. 2011, Islei son e al. 2012).
F os l owe s a e modi i ed signi i can ly wi hin a ew days,
since hey a e ex emely e ec i e collec o s o blowing
snow, which sugges s a empo al inc ease in he a mos-
phe ic ac ion in he os l owe s (Ba be e al. (III)). As a
consequence, hey a e in eg a ed in o he snow laye on op
o he sea ice (Pe o ich and Rich e -Menge 1994).

20 PhD hesis by Do e Haubje g Søgaa d
Al hough he mechanisms behind os l owe o ma ion
a e ela i ely well desc ibed, he biological componen s
in os l owe s a e no well-known (Bowman and Deming
2010, Bowman e al. 2013, E onen-Rasimus e al. 2014). In
Ba be e al. (III), we showed ha he bac e ial concen a-
ions gene ally inc ease wi h salini y in os l owe s. I also
con i ms ha he bac e ial communi y in os l owe s is
signi i can ly di e en om ha in he unde lying sea wa e .
In ou s udy, we ound high calcium ca bona e concen a-
ions o 2.00 – 2.12 g m-2 (Fig. 6; Ba be e al. (III)) in hese
newly o med saline laye s (< 1 h old), which is simila o
calcium ca bona e concen a ions ound in os l owe s and
b ine skim in an ou doo pool o he Sea-Ice En i onmen-
al Resea ch Facili y (SERF; Rysgaa d e al. 2014), bu hey
a e se e al imes highe han concen a ions epo ed om
Ba ow, Alaska (Geil us e al. 2013a). I we compa e he
amoun o calcium ca bona e in he os l owe s and b ine
skim (Fig. 6; Ba be e al. (III)) wi h he o al amoun o
calcium ca bona e in he en i e sea ice column in Young
Sound in NE G eenland (Fig. 2 and Fig. 6; Rysgaa d e al. (II)),
i accoun ed o app oxima ely 8 %. This is in gene al ag ee-
men wi h o he s udies in which os l owe s and b ine
skim in newly o med sea ice ully co e ed by os l owe s
accoun ed o 1-5 % o he o al calcium ca bona e (Domine
2005, Rysgaa d e al. 2014).
An explana ion o he occu ence o calcium ca bona e
in he os l owe s and b ine skim could be he upwa d
mig a ion o b ine wi h calcium ca bona e om he unde -
lying sea ice (Geil us e al. 2013a) and, subsequen ly, he in-
co po a ion o his b ine in o os l owe s and b ine skim on
he ice su ace. The obse a ion o high calcium ca bona e
concen a ions in he newly o med sea ice wi hin he same
season and a he same sampling s a ion suppo s his idea
(Fig. 6; Rysgaa d e al. (II)). Fu he mo e, we obse ed
ha he newly o med os l owe s we e composed p i-
ma ily o b ine om he unde lying sea ice (Ba be e al.
(III)). Howe e , i is possible ha p ecipi a ion o calcium
ca bona e con inued a e he b ine skim and os l owe s
we e o med due o low ai empe a u e a he sampling
s a ion, allowing he p ecipi a ion o sal s o occu wi hin
hese s uc u es (Ba be e al. (III)).
The upwa d mig a ion o b ine om he ice column o he
os l owe s and b ine skim (Ba be e al. (III)) acili a es sal
anspo o he a mosphe e and inc eases he speci i c su -
ace a ea o he ice, which may po en ially p omo e CO2 ex-
change be ween he ice and he a mosphe e (Rankin e al.
2000; 2002, Al a ez-A iles e al. 2008, Bowman and Deming
2010, Geil us e al. 2013a, Ba be e al. (III)). Ou esul s indi-
ca e an e l ux o CO2 a he b ine we ed newly o med sea
ice su ace (3.35 ± 0.86 mmol m-2 d-1, n = 8), bu he e we e no
s iking di e ences in he CO2 l ux i os l owe s we e inside
he chambe oo p in (3.02 ± 0.76 mmol m-2 d-1, n = 4) o ou -
side he oo p in (3.67 ± 1.99 mmol m-2 d-1, n = 4), sugges -
ing ha he o ma ion o os l owe s only p omo es mino
CO2 de-gassing (Ba be e al. (III)). Howe e , ano he s udy
showed ha os l owe o ma ion eleases high amoun s o
CO2, p oduced du ing he p ecipi a ion o calcium ca bona e
(Geil us e al. 2013a). Ne e heless, when he p ecipi a ed
calcium ca bona e om os l owe o ma ion is dissol ed
la e du ing summe haw, his leads o a simila CO2 up ake
om he a mosphe e; and, he e o e, his will, mos likely,
no ha e a ne e ec on he CO2 exchange be ween he
a mosphe e and ocean (Fig. 1; phase II). We ound high
amoun s o calcium ca bona e (9 – 20 g m-2; Fig. 6) in he snow
co e abo e he sea ice in Young Sound, NE G eenland (Fig.
2), which has also been shown o snow on op o sea ice in
o he A c ic loca ions (Søgaa d e al. (IV)). I he p ecipi a ed
calcium ca bona e is inco po a ed in o he snow co e o-
ge he wi h he CO2 eleased h ough calcium ca bona e
p oduc ion (1), hen i is possible ha he CO2 is los om
he snow co e o he a mosphe e unde condi ions wi h
high wind speed (Fig. 1; wind-d i ). The o e all ou come is
ha he dissolu ion o calcium ca bona e du ing he summe
haw, mos likely, will ha e no ne e ec on he CO2 exchange
be ween he a mosphe e and ocean. As a esul o he way
os l owe s a e o med and hei la e in eg a ion in o he
snow co e , os l owe s p esen a unique way o he ocean
and sea ice o in e ac wi h he a mosphe e.
The e en s con olling he dynamics o ino ganic ca bon
in he ansi ion om open wa e o an ice-co e ed ocean
in au umn and win e ha e been ou lined in he sec ions
abo e (Fig. 1; phase I). Al oge he , my esul s om his
phase I indica e ha CO2 is eleased o he a mosphe e
while, a he same ime, TCO2 is ejec ed oge he wi h b ine
o he unde lying sea wa e (Fig. 1). Fo ma ion o new ice
may lead o a de-gassing o CO2 in pa , because o he in-
c ease in concen a ion o solu es bu also because o he
o ma ion o calcium ca bona e. Howe e , i may also be a
esul o espi a ion by he ne he e o ophic communi y a
he ice base in newly o med sea ice. High concen a ions
o calcium ca bona e occu bo h in os l owe s and b ine
skim, accoun ing o app oxima ely 5 % o he o al calcium
ca bona e in he en i e sea ice column. Howe e , he a e
o he p ecipi a ed calcium ca bona e s ill emains unclea .
1.3 Phase II: G owing win e sea ice
In his sec ion, he p ocesses d i ing CO2 exchange in he
g owing win e sea ice a e discussed (1.3.1) and whe he
his phase ac s as a ne sink o sou ce o CO2 o he a mos-
phe e. A sho desc ip ion o he mic oo ganisms in win e
sea ice and hei con ibu ion o he CO2 dynamics du ing
phase II is also p o ided (1.3.2).
21
PhD hesis by Do e Haubje g Søgaa d
1.3.1 Abio ic p ocesses in g owing win e ice
Du ing win e (Fig. 1; phase II), he la ges l ux o TCO2 and
CO2 is d i en by b ine d ainage o he unde lying wa e
column, lea ing sea ice deple ed in CO2 compa ed o ambien
seawa e (Fig. 1). This is suppo ed by esul s in Søgaa d e al.
(I) indica ing ha he TCO2 deple ion in sea ice in Kapisigdli
Bigh in SW G eenland (Fig. 2) was mainly con olled by
b ine d ainage o he unde lying wa e . The e a e h ee
ypes o desalina ion mechanisms: 1) b ine expulsion, 2)
g a i y d ainage and 3) b ine pocke mig a ion. The i s wo
desalina ion mechanisms a e he only ones o quan i a i e
impo ance (Eicken 2003). B ine expulsion is he mig a ion
o b ine d i en by he cooling o he sea ice, which esul s in
p essu e buildup in he b ine pocke . This allows he b ine
o escape om he b ine pocke and mig a e owa d he
wa me end o he ice shee (Weeks 2010). G a i y d ainage
includes all p ocesses in which he b ine, unde he in l u-
ence o g a i y, d ains ou o he sea ice in o he unde lying
wa e column (Cox and Weeks 1974). These wo desalina-
ion mechanisms esul in he ejec ion o la ge amoun s o
TCO2 and CO2 o he unde lying wa e column along wi h
he expelled b ine du ing sea ice g ow h (Killawee e al.
1998, Ande son e al. 2004, Rysgaa d e al. 2007), which sub-
sequen ly inc eases he pCO2 concen a ion below he sea
ice (Fig. 1; Gibson and T ull 1999, Semile o e al. 2004; 2007,
Delille 2006; 2010). TCO2 is assumed o be emo ed om he
su ace oceanic mixed laye o deepe wa e masses ia he
sinking o he expelled dense b ine (Nansen 1906, Rysgaa d
e al. 2011). Howe e , he a e o he ejec ed TCO2 in he
wa e column is s ill poo ly unde s ood.
Du ing win e calcium ca bona e p ecipi a ion con inues in
sea ice (Søgaa d e al. (I), Rysgaa d e al. (II)). Ou measu e-
men om g owing Suba c ic win e sea ice in Kapisigdli
Bigh (Fig. 2) showed high calcium ca bona e concen a ions
(i.e., 9 g m-2; Søgaa d e al. (I)), which was 4 imes highe han
he a e age seasonal calcium ca bona e concen a ion in sea
ice measu ed in his a ea (i.e., 2.13 g m-2; Søgaa d e al. (I)).
In addi ion, in win e ice in he High A c ic in Young Sound,
NE G eenland (Fig. 2), we obse ed he highes calcium
ca bona e concen a ions e e measu ed o ou knowledge
in na u al sea ice (i.e., 25 g m-2; Rysgaa d e al. (II)), sugges -
ing ha win e (Fig. 1; phase II) is ex emely impo an o he
annual calcium ca bona e p ecipi a ion (Fig. 6). Fu he mo e,
i indica es ha he calcium ca bona e concen a ion in he
b ine inc eases wi h dec easing empe a u es as he highes
calcium ca bona e concen a ions a e obse ed in cold win-
e High A c ic sea ice (Rysgaa d e al. (II)).
Calcium ca bona e is mainly ound as ikai e in he na u al sea
ice en i onmen , and we con i med ha he calcium ca bona e
ound in he win e sea ice in he Young Sound a ea in NE
G eenland (Fig. 2) was, indeed, ikai e (Rysgaa d e al. (II)). P e-
cipi a ion o ikai e equi es nea - eezing empe a u es and
condi ions o high alkalini y (Bischo e al. 1993, Bucha d
e al. 2001, Selleck e al. 2007, Hu e al. 2014). P e iously, i
was also pos ula ed ha ele a ed phospha e concen a ions
we e c i ical o ikai e p ecipi a ion (Bischo e al. 1993); how-
e e , his has been shown no o be he case (Hu e al. 2014).
In he win e sea ice in Young Sound, NE G eenland and in
Kapisigdli Bigh , SW G eenland (Fig. 2), we did obse e low
sea ice empe a u es and condi ions o high alkalini y wi h
he pH ollowing a C-shaped pH p o i le, i.e., high pH (> 9) in
he su ace and bo om sea ice laye s and sligh ly lowe pH
condi ions (8.5) in he in e nal sea ice laye s (Søgaa d e al.
(I), Rysgaa d e al. (II)). The C-shaped pH p o i le in he sea ice
is c ea ed when he b ine inclusions in he uppe ice laye s
a e closed due o low ice pe meabili y. This will es ic u -
he CO2 expulsion om he in e io ice, which hen esul s
in CO2 deple ion ela i e o sea wa e and, hus, highe pH
le els in he uppe ice laye s (Ha e e al. 2013). The in e io ice
laye con inues o ecei e CO2 om calcium ca bona e p e-
cipi a ion and CO2 anspo due o b ine mo emen , which
explains he low pH alues encoun e ed he e. The bo om ice
laye has open b ine channels ha con inuously expo b ine
o he unde lying wa e esul ing in CO2-deple ion and high
pH in his laye (Ha e e al. 2013).
The calcium ca bona e concen a ions measu ed a wo di -
e en loca ions ollowed he C-shaped pH p o i le wi h high
calcium ca bona e concen a ions in he su ace and bo om
sea ice laye s (Fig. 7; Søgaa d e al. (I)). This sugges s ha
he calcium ca bona e p ecipi a ion in na u al sea ice is also
con olled by pH, as sugges ed in a labo a o y s udy by Hu e
al. (2014). This con adic s p e ious s udies in which calcium
ca bona e is p ima ily ound in he uppe mos laye s o sea
ice (e.g., Fische e al. 2013, Nomu a e al. 2013b, Rysgaa d e
al. 2014).
Calcium ca bona e concen a ion (μmol l-1)
Ice dep h (cm)
1 2 3 400 800 1200 1600
0
20
40
60
80
100
120
Figu e 7. Ve ical dis ibu ion o he calcium ca bona e concen a-
ion in i s -yea win e sea ice in Kapisigdli , SW G eenland (black
line; Søgaa d e al. (I)) and in i s -yea win e sea ice in Young
Sound, NE G eenland (dashed black line; unpublished da a D. H.
Søgaa d).
22 PhD hesis by Do e Haubje g Søgaa d
The po en ial in l uence o mel ing he en i e win e ice co e
in Kapisigdli Bigh in SW G eenland (Fig. 2; Søgaa d e al. (I))
and Young Sound in NE G eenland (Fig. 2; Rysgaa d e al. (II))
in o a 20 m hick mixed laye ( ypical o summe condi ions in
hese loca ion) on he CO2 l ux can be de e mined using he
measu ed ice ca bon chemis y (Table 1) and he ini ial mixed
laye cha ac e is ics (Table 2) om he wo di e en egions.
Assuming ha mel occu s o e one mon h, he esul an
ai -sea CO2 l ux o e u n o p e-mel condi ions would be –
4.8 mmol C m-2 d-1 in Kapisigdli Bigh in SW G eenland and
– 5.9 mmol C m-2 d-1 in Young Sound in NE G eenland (Fig. 2).
The po en ial ai -sea CO2 l ux a y by a ac o o 0.8 wi h he
highes po en ial ai -sea CO2 l ux es ima ed in High A c ic sea
ice (Søgaa d e al. (II), Søgaa d e al. (IV)). Ou obse a ions
o calcium ca bona e concen a ions in sea ice a y by se e al
o de s o magni ude depending on he sea ice ype and he
locali y (Fig. 6; Søgaa d e al. (I), Rysgaa d e al. (II)), sugges -
ing ha TA, TCO2 and calcium ca bona e concen a ions ea-
u e high ho izon al a iabili y. This issue is add essed in one
o my pape s (Søgaa d e al. (I)) in which we ound high spa-
ial a iabili y on he scale o me es o hund eds o me es
o calcium ca bona e concen a ion and o he sea ice bio-
geochemical p ope ies. Ano he impo an i nding in hese
s udies is ha he p esence o calcium ca bona e in sea ice
almos doubles he ai -sea CO2 l ux as compa ed o mel ing
o calcium ca bona e- ee sea ice, which emphasizes ha he
p ecipi a ion and dissolu ion o calcium ca bona e is c i ical
o he e i ciency o he sea ice-d i en ca bon pump in hese
a eas (Søgaa d e al. (I), Rysgaa d e al. (II)).
Al hough he ejec ion o TCO2 om g owing sea ice o he
unde lying wa e column du ing win e is he dominan
p ocess in his phase, he sea ice-a mosphe e l uxes should
also be aken in o accoun when desc ibing he sea ice
ca bon budge . As men ioned abo e, ou da a showed ha
calcium ca bona e concen a ion ollows a C-shaped p o i le
(Fig. 6; Søgaa d e al. (I)). Since mos o he calcium ca bona e
is ound in he uppe and lowe sea ice laye s, i sugges s
ha he CO2 p oduced du ing p ecipi a ion o calcium
ca bona e is bo h in close con ac wi h he a mosphe e
and he unde lying wa e column. The amoun o CO2 e-
leased o he a mosphe e depends on he ice pe meabili y,
while CO2 in he lowe ice laye s is ejec ed oge he wi h
he b ine o he unde lying wa e column when he e is
an excess o CO2 due o high pe meabili y in his ice laye .
The low empe a u es du ing win e esul in lowe b ine
olumes and, subsequen ly, a dec ease in pe meabili y,
which impedes he ice-a mosphe e gas exchange (Loose e
al. 2011a, Rysgaa d e al. 2011) and esul s in insigni i can
l uxes abo e cold sea ice (Heinesch e al. 2010, Mille e al.
2011b, Geil us e al. 2012b, Sø ensen e al. 2014). Howe e ,
small l uxes o CO2 ha e been obse ed abo e cold win e
sea ice unde condi ions wi h high wind speed (Heinesch e
al. 2010, Mille e al. 2011a, Papaky iakou and Mille 2011,
Else e al. 2011), indica ing ha CO2 de-gassing migh oc-
cu occasionally abo e win e sea ice (Fig. 1; snow-d i ). The
CO2 de-gassing abo e win e sea ice is mainly due o he
loss o s o ed CO2 om he snow pack (Rysgaa d e al. 2014).
This is suppo ed by a ecen s udy showing low pH alues
in he snow co e abo e win e sea ice, indica ing high CO2
concen a ions (Ha e e al. 2013). In addi ion, we also ound
high amoun s o calcium ca bona e (0.12 – 20 g m-2; Fig.
6) in he snow co e abo e he sea ice in Young Sound, NE
G eenland and in Kapisigdli Bigh in SW G eenland (Fig. 6;
Søgaa d e al. (IV)). I he p ecipi a ed calcium ca bona e is
inco po a ed in o he snow co e oge he wi h he CO2 e-
leased h ough calcium ca bona e p oduc ion (1), his p o-
cess may be esponsible o he obse ed snow-d i en CO2
degassing unde condi ions wi h high wind speeds abo e
win e sea ice (Fig. 1; phase II). The snow co e also has an in-
sula ing e ec (S u m and Massom 2010, Fische e al. 2013),
main aining sea ice empe a u es high enough o allow a
small CO2 e l ux om he ice column in o he snow co e
(Golden e al. 2007; Nomu a e al. 2010a; 2010b, Ha e e al.
2013). In addi ion, new snow all o a edis ibu ion o snow
due o high winds o he same si e may wa m sea ice co e
locally du ing win e , which may dissol e calcium ca bona e
(Rysgaa d e al. 2014). This indica es dynamic condi ions o
calcium ca bona e p ecipi a ion/dissolu ion, CO2 and pH
du ing win e .
1.3.2 G ow h limi a ion o mic oo ganisms in
win e sea ice
In win e (Fig. 1; phase II), ligh a ailabili y in he sea ice is
e y low and he sympagic communi y emains ne he e o-
ophic, coun e ac ing he a mosphe ic CO2 d awdown as
espi a ion eleases CO2 (Søgaa d e al. (I), Søgaa d e al.
(IV)).
Ou s udies show ha , al hough bo h bio ic and abio ic
p ocesses can in l uence ai -sea CO2 exchange, he ela i e
e ec s o bio ic p ocesses h oughou he examined win e
sea ice a e e y low (Søgaa d e al. (I)). In addi ion, ou s udies
e ealed ha he ice-associa ed biological communi y was
ne he e o ophic in his win e phase wi h high bac e ial
ca bon demands (i.e., a ies be ween 19.40 mg C m-2 – 40.50
mg C m-2) compa ed o es ima ed p ima y p oduc ion al-
ues (i.e., a ies be ween 12.20 – 20.70 mg C m-2; Søgaa d
e al. (I), Søgaa d e al. (IV)). The bac e ial ca bon demand
in win e sea ice a ies be ween loca ions wi h eco ded
alues o 0.80 – 26 mg C m-2 d-1 in G eenland sea ice (Long e
al. 2012, Søgaa d e al. (I), Søgaa d e al. (IV)), 0.02 – 1.5 mg
C m-2 d-1 in he Bal ic Sea (Kaa okallio 2004) and 5.96 mg C
m-2 d-1 in sea ice in Resolu e (Smi h and Clemen 1990). Re-
ga dless o loca ion and bac e ial ca bon demand, minima
in bac e ial abundance ha e been obse ed du ing win e
(Collins e al. 2008, Deming 2010 and e e ences he ein).
The low bac e ial abundance in sea ice du ing win e migh
be due o i ally-media ed cell dea h (Collins e al. 2008),
23
PhD hesis by Do e Haubje g Søgaa d
g azing by bac e i o ous p o is s (Rózanska e al. 2008), a
educ ion in habi able space o he o ma ion o in acel-
lula ice c ys als and cell-punc u ing by ice c ys als (Col-
lins and Deming 2011). Ou s udies indica e ha high ice
salini y is a key pa ame e in l uencing he cons i u ion
and ac i i y o he bac e ial communi y in win e sea ice
(Kaa okallio al. (VI)). Ou i ndings also sugges ha he
sea ice is ac ing as a bio i lm-like sys em a he han being
analogous o open-wa e sys ems, which may be a su i al
s a egy o he he e o ophic communi y in his hos ile en-
i onmen (Kaa okallio al. (VI)).
As men ioned in sec ion 1.3.1, he pH in win e sea ice ol-
lows a C-shape p o i le i.e., high pH (> 9) in he su ace and
bo om sea ice laye s and sligh ly lowe pH condi ions (8.5)
in he in e nal sea ice laye s. This may ha e an e ec on he
g ow h o he mic oo ganisms wi hin he sea ice. Limi ed
knowledge exis s on how high pH a ec s g ow h a es o
sea ice algae and bac e ia. Howe e , p e ious s udies indi-
ca e ha high ex acellula pH may cause g oss al e a ions
in memb ane anspo p ocesses and me abolic unc ions
in ol ed in in e nal pH egula ion (Ra en 1980) o cause al-
e a ions o cellula con en o amino acids and hei com-
posi ion, which migh a ec cellula g ow h (e.g., Ta alds ik
& Mykles ad 2000). Changes o pH in l uence he in e -
specia ion o ino ganic ca bon (CO2 (aq), HCO3-, CO32-). A
pH 8 in sea wa e (TCO2 app oxima ely 2mM in ma ine wa-
e s), app oxima ely 1 % o TCO2 is p esen as CO2 while, a
pH 9, only 0.1 % o TCO2 is p esen in his o m (Hinga 2002).
Po en ially, he limi a ion o he CO2 supply due o ele a-
ed pH may es ic pho osyn hesis and g ow h o phy o-
plank on (Hansen 2002). Howe e , some phy oplank on
species ha e ac i e anspo sys ems by which hey
u ilize HCO3- in o de o a oid TCO2 limi a ion a ele a ed
pH (Ko b e al. 1997, Hue as e al. 2000, Hansen 2002).
We ha e shown ha he g ow h a es o he sea ice dia om
species (i.e., F agila iosis nana and F agila iopsis sp.) we e sig-
ni i can ly educed a pH > 9.0; and, a pH=9.5, hey s opped
g owing i espec i e o TCO2, indica ing ha pH had a di ec
e ec on algal g ow h (Søgaa d e al. (V)). In addi ion, ou
expe imen s e ealed ha a chlo ophy e species commonly
encoun e ed in sea ice (i.e., Chlamydomonas sp.) had an ex-
eme pH- ole ance since only a small educ ion in i s g ow h
a e was obse ed abo e pH=9.5, bu i was sensi i e o low
TCO2 concen a ions (Søgaa d e al. (V)). Fu he mo e, he
chlo ophy es species ou -g ew wo species o sea ice dia oms
in a succession expe imen , sugges ing ha an ele a ed pH
played a po en ial ole in he succession o A c ic sea ice algae
in win e sea ice (Søgaa d e al. (V)).
Sea ice algae a e known o be adap ed o low ligh le els
and a e able o g ow a ligh in ensi ies down o 0.36 – 20
μmol pho on m-2 s-1 (Ho ne and Sch ade 1982, Gosselin e
al. 1990, G adinge and Ikä alko 1998, Mock and G adinge
1999). Howe e , much lowe ligh in ensi ies a e obse ed in
he win e sea ice co e ed wi h hea y snow (Søgaa d e al.
(IV)). In ou s udy, ligh was iden i i ed as he majo limi ing
ac o o algal p oduc i i y – wi h snow co e dep h la gely
con olling ligh ansmission in Suba c ic sea ice in Malene
Bigh in SW G eenland du ing win e (Fig. 2; Søgaa d e al.
(IV)). This sugges s ha he low algal biomass and p oduc-
i i y in he win e sea ice in Malene Bigh in SW G eenland
is caused by poo ligh condi ions (Søgaa d e al. (IV)).
In he sec ions abo e, he sequence o e en s con olling
he dynamics o ino ganic ca bon in g owing win e sea ice
(Fig. 1; phase II) is ou lined. Ve y high amoun s o calcium
ca bona e we e obse ed in win e he High A c ic sea ice
in Young Sound in NE G eenland (Fig. 2), and he concen a-
ion ollowed a C-shaped p o i le, sugges ing ha he CO2
p oduced in his phase is bo h in close con ac wi h he
a mosphe e and he unde lying wa e column. In addi ion,
he sea ice pe meabili y was e y low, sugges ing ha b ine
d ainage we e he dominan p ocesses leading o TCO2
deple ion in he ice column du ing he win e mon hs.
The e o e, CO2 de-gassing plays only a ole in a eas wi h
high ice pe meabili y, high wind speeds and/o hea y snow
co e whe e ice empe a u es a e main ained high enough
o allow o CO2 e l ux. Al oge he , my esul s indica e ha
he ole o biology in modula ing ino ganic ca bon dyna-
mics in his phase is mino , and is appa en ly delayed un il
sp ing and summe .
1.4 Phase III: T ansi ion om ice-
co e ed ocean o open wa e s
In his sec ion, he abio ic p ocesses d i ing CO2 exchange
in mel ing sea ice a e discussed (1.4.1) and whe he his
phase ac s as a ne sink o sou ce o CO2 o he a mosphe e.
In addi ion, he mic oo ganisms in mel ing sea ice and hei
con ibu ion o he CO2 dynamics in his phase a e de-
sc ibed (1.4.2) wi h a ocus on he p ocesses de e mining
he magni ude o he p ima y p oduc ion in G eenland sea
ice and he conside able geog aphic di e ences in p ima y
p oduc ion. Finally, a sho discussion o he balance o CO2
sinks and sou ces in he open wa e pe iod in sec ion 1.4.3
is p o ided.
1.4.1 Abio ic p ocesses in mel ing sea ice
The wa ming o sea ice is accompanied by a educ ion in
ice salini y, app oaching ze o salini y, because o in e nal
ice mel and b ine l ushing due o he d aining o mel wa e
om su ace mel ponds (Un e s eine 1968, Cox and Weeks
1974, Fe e e and Un e s eine 1998). B ine inclusions
and channels enla ge upon wa ming and o m new pa h-
ways o b ine and mel wa e . These new pa hways o m
longe and bigge channels han he p ima y pa hways in
30 PhD hesis by Do e Haubje g Søgaa d
In addi ion, mel ponds a e a widesp ead and inc easing
su ace ea u e o A c ic sea ice du ing sp ing and summe
(Rösel and Kaleschke 2012), and hei impac on ino ganic
ca bon anspo h ough sea ice migh , he e o e, inc ease
in he u u e (see desc ip ion in sec ion 1.4.1).
The mo e anspa en sea ice co e and ea lie ice mel will
also esul in an inc easing po en ial o pelagic p ima y p o-
duc ion bo h below he hinne sea ice co e and in he open
wa e s (A igo e al. 2012, Nicolaus e al. 2012, Mundy e al.
2013, Ba be e al. (III)). Recen s udies also indica e ha he
cu en sea ice hinning may enhance ice-algal expo due o
algal agg ega ion (Boe ius e al. 2013). As a consequence, he
biological d awdown o CO2 is expec ed o inc ease as sea
ice co e is educed, which will lead o inc eased ne oceanic
up ake o CO2 (Ba es e al. 2006, A igo e al. 2008, Ba es and
Ma his 2009, MacGilch is e al. 2014). Howe e , ecen s udies
show ha he up ake capaci y o he A c ic Ocean o a -
mosphe ic CO2 is limi ed as a esul o su ace wa ming and
inc eased s a i i ca ion (F ansson e al. 2009, Cai e al. 2010,
B en e al. 2013, Else e al. 2013) and ha hese changes will
also s imula e bac e ial p oduc ion and, consequen ly, limi
he sink unc ion o he A c ic Ocean when he sea ice co e
is educed (Xie e al. 2009). Fu he mo e, obse a ion has
shown ha he A c ic i e uno has inc eased, which would
lead o an e en s onge s a i i ca ion in he A c ic Ocean
(F ansson e al. 2009). This will mos likely esul in a dec ease
in he annual biological p oduc ion in hese a eas and, con-
sequen ly, also limi he sink unc ion o he A c ic Ocean in
summe (F ansson e al. 2009).
The mel om he G eenland Ice Shee has also inc eased as
a esponse o global wa ming, adding mo e eshwa e o
he su ace wa e in he coas al a eas and j o ds in G eenland
(Inall e al. 2014 and e e ences he ein, Khan e al. 2014).
Wha is he e ec o glacial uno on he ai -sea CO2 ex-
change in coas al a eas and j o ds in G eenland? Ou s udies
e ealed ha he God håbs j o d sys em in SW G eenland
was, indeed, a sink o CO2 (Rysgaa d e al. (VII)) and ha he
CO2 dynamics we e con olled by bo h he biological p o-
cesses and mixing be ween glacial mel wa e and coas al
wa e s (Rysgaa d e al. (VII)). The s eng h o he sink was
pa icula ly s ong nea he G eenland Ice Shee , indica -
ing spa ial a ia ions in ai -sea CO2 up ake. High amoun s
o CO2 a e aken up in he God håbs j o d (annual l ux o –
83 o – 108 g C m-2 y -1; Rysgaa d e al. (VII)). This es ima e
is highe han l ux es ima es om o he si es in he G een-
land Sea (i.e., – 52 g C m-2 y -1; Nakaoka e al. 2006), in Young
Sound in NE G eenland (– 32 g C m-2 y -1; Sej e al. 2011;
Fig. 2) and da a om o he A c ic shel sys ems (Ba es and
Ma his 2009), unde lining he impo ance o his Suba c ic
j o d sys em ac ing as a s ong sink o CO2 (Rysgaa d e
al. (VII)). I his up ake is ypical o simila Suba c ic j o d
sys ems in G eenland, hen he coas al a eas o G eenland
cons i u e a la ge sink han an icipa ed. Fu he mo e, a e-
cen s udy om he God håbs j o d indica es ha he glacial
eshwa e uno is likely o be impo an o he s imula-
ion e ec on p ima y p oduc ion ( o al annual p oduc ion
o 84.6 – 139.1 g C m-2 y -1). This sugges s ha he iming,
du a ion and magni ude o he glacial eshwa e uno a e
likely o be impo an o he CO2 up ake in his Suba c ic
j o d sys em (Rysgaa d e al. (VII), Juul-Pede sen e al. sub-
mi ed). The e o e, an inc eased abla ion o he G eenland
Ice Shee due o u u e wa ming could esul in an inc ease
in he annual biological p oduc ion, which could po en ially
inc ease he ai -sea CO2 l ux in hese a eas.
1.6 Conclusions and pe spec i es
So, wha ha e we lea ned since he sea ice CO2 pump was
i s sugges ed by Jones and Coo e (1981) 33 yea s ago
and la e con i med in na u al sea ice by, e.g., Rysgaa d
e al. 2007? The s udies o sea ice ca bon dynamics o e a
comple e seasonal cycle o sea ice o ma ion and decay as
p esen ed in his hesis ha e, indeed, imp o ed ou unde -
s anding o he d i e s o he sea ice CO2 pump.
Collec i ely, he esul s in his hesis showed ha he o -
ma ion o sea ice esul s in anspo o TCO2 ou o he
ice column and, he e o e, s eng hens he idea o an e -
ec i e sea-ice-d i en ca bon pump in bo h Suba c ic and
High A c ic sea ice in G eenland (Fig. 1). Howe e , he ex-
en o which TCO2 is anspo ed o he unde lying wa e
column and, subsequen ly, en e s he in e media e and
deep wa e masses has ye o be de e mined. The highes
concen a ions o calcium ca bona e e e epo ed in na u-
al sea ice was measu ed in app oxima ely 5-mon h-old
High A c ic land- as sea ice, ollowed by high concen a-
ions in newly o med High A c ic polynya sea ice; whe eas
he lowes concen a ions obse ed du ing ou s udies we e
in Suba c ic land- as sea ice. Va ia ions in sea ice p ope ies
such as empe a u e, salini y, pH, ice ex u e and eshwa e
inpu a e likely esponsible o some o he di e ences
ound in calcium ca bona e concen a ions be ween si es.
I seems ha he di ec ion and magni ude o he ai -ice l ux
appea o be de e mined by he s age o ice de elopmen ,
he p ope ies o he sea ice ca bona e sys em (i.e., TA, TCO2,
pCO2) as well as sea ice geophysical (i.e., salini y and pe me-
abili y) and he modynamic (i.e., empe a u e) p ope ies.
The con ibu ion o p ima y p oduc ion o he TCO2 deple-
ion was mino compa ed o he con ibu ion o calcium
ca bona e p ecipi a ion/dissolu ion in my s udy a eas; how-
e e , in a eas o high p ima y p oduc ion he con ibu ion
o he TCO2 deple ion migh be signi i can ly highe . As a
esul , he e alua ion o he sea ice sink desc ibed in his
hesis is no ep esen a i e o he A c ic as a whole since he
up ake o CO2 by biological ac i i y seems o be much lowe
in G eenland sea ice compa ed o o he egions.

31
PhD hesis by Do e Haubje g Søgaa d
These esul s e ealed la ge a ia ion in calcium ca bona e
concen a ion and o he biogeochemical p ope ies a di -
e en empo al and spa ial scales in sea ice, emphasising
he impo ance o ull-season s udies co e ing he hun-
d ed-me e spa ial scale in o de o make eliable egional
and e en ually global ca bon budge s.
Many ques ions s ill emain unanswe ed wi hin his i eld o
esea ch, some o which I would like o add ess in he u u e.
The i s s ep would be o pe o m ull-season s udies o he
ino ganic ca bon dynamics in di e en sea ice ypes and in
di e en ice loca ions. I is impo an o pe o m con inu-
ous measu emen s o all componen s in ol ed in he sea-
ice-d i en ca bon cycle du ing a comple e seasonal cycle, in
di e en sea ice ypes and om di e en geog aphic a eas.
Fu he mo e, seasonal measu emen s o he esul ing CO2
l uxes ac oss he a mosphe e-ice-ocean bounda y laye a e
needed. In addi ion o measu ing hese pa ame e s wi h
al eady es ablished echniques, i is also impo an o de-
elop new me hods o measu emen s including in si u mea-
su emen s in he mic oen i onmen o b ine channels and
pocke s.
This i s s ep equi e in i sel a ocused e o and s ill o he
basic ques ions will undoub edly p esen hemsel es: 1) Will
he u u e changes in sea ice co e a ec he capaci y o he
A c ic Ocean o ake up a mosphe ic CO2? 2) Wha is he a e
o he ejec ed TCO2 in he wa e column a di e en A c ic
a eas, and is i anspo ed below he pycnocline? 3) Wha
is he a e o he calcium ca bona e and CO2 in os l owe s
and b ine skim? 4) Wha is he impo ance o polynya
a eas o he ai -sea CO2 l ux? 5) How impo an is he sea
ice ca bon pump compa ed o he solubili y pump and bio-
logical pump? He e, I ou line wha a e o me he mos in-
e es ing unanswe ed ques ions o ad ance ou knowledge
o he e en s con olling he ino ganic ca bon dynamics in
high la i ude oceans.
The empo al and spa ial a ia ions in he ai -sea CO2 l ux
we e discussed o he coas al seawa e in God håbs j o d,
SW G eenland (Fig. 2). This a ea is de e mined o be a s ong
sink o CO2, which was highly egula ed bo h by he biologi-
cal p ocesses and by mixing be ween glacial mel wa e and
coas al wa e s. The s eng h o his sink is pa icula ly s ong
nea he G eenland Ice shee .
The nex s ep will be o pu hese esul s in o a global con-
ex o unde s and how impo an he ole o G eenland
coas al wa e s is in ela ion o he global ca bon budge .
Th ee basic ques ions p esen hemsel es: 1) Wha is he
in e -annual pa e n o he CO2 sink and biological p o-
cesses in G eenland in he Suba c ic and High A c ic j o ds?
2) Wha is he spa ial pa e n o he CO2 sink and biologi-
cal p ocesses in G eenland? 3) Will hese pa e ns change as
a consequence o ongoing global wa ming, which is mo e
p onounced a hese high la i udes? I is a non i ial ask o
answe hese ques ions as he coas al CO2 l ux and biologi-
cal p ocesses show high spa ial and empo al he e ogenei y.
Howe e , G eenland p esen s a unique oppo uni y o
s udy changes in coas al biological and physical cha ac e is-
ics along a clima e g adien om he Suba c ic o he High
A c ic.
The e o e, he e a e s ill a lo o impo an u u e esea ch
ques ions in e ms o unde s anding he in l uence o he
sea-ice-d i en CO2 pump and he ole o G eenland coas al
wa e s in ela ion o he global ca bon cycle, and i is needed
u gen ly since clima e change h ea ens o ake hese o-
zen high la i ude en i onmen s om us.
1.7 Glossa y
Agg ega es: a e a ached o ee- l oa ing ma s o agg e-
ga es o , e.g., he cen ic dia om Melosi a a c ica benea h
he sea ice, in mel ponds, ozen in o he sea ice o sunk o
he bo om o he deep-sea l oo .
Ai -sea CO2 exchange: is p ima ily con olled by he ai -sea
di e ence in gas concen a ions and he exchange coe i -
cien . I akes abou one yea o equilib a e CO2 in he su -
ace ocean wi h a mosphe ic CO2. The e o e, in some a eas
la ge ai -sea di e ences in CO2 concen a ions can be ob-
se ed. In my PhD hesis nega i e l ux indica es sea ice o
sea wa e up ake o CO2.
Algae: gene al e m o euka yo ic o ganisms anging om
unicellula gene a, e.g., dia oms o mul icellula o ms, such
as gian kelp, o which mos a e au o ophic and non- ascula
o ganisms ha li e almos exclusi ely in aqua ic en i on-
men s.
Au o ophic: abili y o con e ene gy om ligh (pho o-
au o ophic) o om oxida ion o ino ganic compounds
(chemoau o ophic) o o ganic ma e ial by u ilising ino -
ganic ca bon (usually CO2).
Bac e ia: cons i u e a la ge domain o p oka yo ic mic oo -
ganisms, mos o which a e he e o ophic and, ypically, a
ew mic ome es in leng h.
Biological pump, he: is d i en by he sinking o pa icula e
ma e ial -ei he o ganic ca bon (i.e., dead algal cell) o pa -
icula e ino ganic ca bon (i.e., calcium ca bona e om cal-
ci ying o ganisms such as coccoli hopho es, o amini e ans
o p e opods).
B ine skim: a highly saline skim o b ine ha is o med on
he su ace o newly o med sea ice.
32 PhD hesis by Do e Haubje g Søgaa d
Calcium ca bona e: Exis s in six phases, namely, amo -
phous calcium ca bona e, calcium ca bona e monohyd a e,
calcium ca bona e hexahyd a e (ikai e) and h ee anhyd-
ous phases: a e i e, a agoni e and calci e. In bo h A c ic
and An a c ic sea ice, p ecipi a ion o calcium ca bona e in
he o m o ikai e has been obse ed. A p esen , i is no
clea whe he ikai e is he only calcium ca bona e phase
o med in sea ice. Howe e , p ecipi a ion o ikai e in sea ice
is an impo an p ocess as i ca alyses chemical p ocesses
such as bounda y laye ozone deple ion e en s (ODEs) and
he o ma ion and subsequen d aw-down o CO2 ia b ine
d ainage.
Ca bon cycle: is he biogeochemical cycle by which ca bon
is exchanged among he a mosphe e, hyd osphe e, li ho-
sphe e, c yosphe e and pedosphe e o he ea h.
Ca bon dioxide (CO2): nex o wa e apou CO2 is he
mos abundan g eenhouse gas on ea h. Mos global
CO2 is dissol ed in wa e , and CO2 eac s wi h wa e and
o ms bica bona e (HCO3-) and ca bona e ions (CO32-). The
concen a ion o he di e en ions depends on he mo-
dynamic equilib iums ha a e ela ed o empe a u e, pH,
salini y and p essu e. Howe e , a ypical sea wa e condi-
ions, HCO3- is dominan (86.5 %), whe eas CO2 (0.5 %) and
CO32- (13 %) a e only p esen in small concen a ions.
Cells: a e he smalles uni o li e ha can eplica e indepen-
den ly.
Cell memb ane: is a biological memb ane ha sepa a es
he in e io o all cells om he ou side en i onmen (called
cell wall). I is made o a lipid bilaye in e spe sed wi h p o-
eins, which makes i selec i ely pe meable o ions and
o ganic molecules.
Chlo ophyll: a g oup o g een pigmen s in pho osyn he ic
o ganisms ha aps he ene gy o sunligh o pho osyn-
hesis and exis s in se e al o ms o which he mos abun-
dan is chlo ophyll a.
Chlo ophyll a: a ype o chlo ophyll ha is common and p e-
dominan in all oxygen-e ol ing pho osyn he ic o ganisms.
I is abb e ia ed Chl a.
Chlo ophy a: is a di ision o g een algae.
Cilia es: a e a g oup o p o ozoans cha ac e ized by he
p esence o hai -like o ganelles called cilia.
Deep-wa e masses: wa e loca ed below he in e media e
wa e s. I gene ally has low empe a u es (2° C) and high
salini y (34.9) and, he e o e, a high densi y.
Dia oms: a e a majo g oup o algae and a e among he
mos common ypes o phy oplank on o which mos a e
unicellula . They can exis as colonies and wi h a cell wall o
amo phous silica.
Dime hylsulphoniop opiona e (DMSP): is a widely used
osmoly e used by mic oalgae o acclima e o changes in
salini y. DMSP is a p ecu so o dime hylsulphide (DMS),
which is a clima e-ac i e gas.
Dissol ed o ganic ca bon (DOC): a e o ganic molecules o
a ied o igin and composi ion wi hin aqua ic sys ems. DOC
in ma ine sys ems is gene ally a esul o decomposi ion p o-
cesses om dead o ganic ma e such as plan s. DOC is a ood
supplemen suppo ing he g ow h o mic oo ganisms and
plays an impo an ole in he global ca bon cycle h ough
he mic obial loop.
Dissol ed o ganic ni ogen (DON): is a mix u e o com-
pounds anging om simple amino acids o complex humic
subs ances.
Exponen ial g ow h: G ow h o mic oo ganisms whe eby
he cell numbe doubles wi hin a i xed ime pe iod.
Ex acellula polyme ic subs ances (EPS): a e high-molecula
weigh compounds sec e ed by mic oo ganisms in o hei
en i onmen . EPS can unc ion as c yop o ec ion, as ex e nal
ese es o hyd olysable o ganic compounds, o dep ess he
eezing poin and o p o ide a physical bu e agains en-
c oaching ice c ys als.
Fi s -yea sea ice: is sea ice o no mo e han one win e ´s
g ow h. I de elops om young ice and ha e a hickness >
30 cm.
Flagella e: is an o ganism wi h one o mo e o ganelles
called l agella. Flagella-bea ing species a e common in
all algal classes excep Cyanophyceas, Rhodophyceae,
Phaeophyceae and Bacilla iophyceae.
F os l owe s: clus e s o saline ice c ys als ha ha e a den-
d i ic and b anched s uc u e. F os l owe s o m a he in-
e ace be ween a wa m ice su ace and a cold a mosphe e
a condi ions wi h low su ace wind condi ions.
Global ca bon budge : is he sum o all exchanges (in l ows
and ou l ows) o ca bon compounds be ween he ea h´s
ca bon ese oi s o he ca bon cycle.
Halo ole an : he abili y o wi hs and la ge changes in
salini y.
He e o ophic: abili y o ob ain ca bon o o ganic syn hesis
by me abolising o ganic ma e ial.
Ikai e (CaCO3 · 6H2O): is an uns able hexahyd a e poly-
mo ph o CaCO3, which begins o p ecipi a e a -2.2˚ C and
dissol es a empe a u es abo e 4˚ C. P ecipi a ion o ikai e
has been con i med in sea ice om bo h hemisphe es.
In si u: means on si e.
In acellula : occu s o unc ions wi hin a cell.
33
PhD hesis by Do e Haubje g Søgaa d
Lipid bilaye : is a l a and hin pola memb ane ha con-
sis s o wo laye s o lipid molecules. This shee o ms a
con inuous ba ie a ound all cells.
Mel ponds: esul om an accumula ion o mel wa e
on sea ice – mainly, due o he mel ing o snow, bu he
unde lying sea ice co e also con ibu es o he mel pond
o ma ion. Mel ponds abso b sola adia ion a he han
e l ec ing i as ice does and, he eby, ha e a signi i can in-
l uence on he ea h’s adia ion balance.
Memb ane l uidi y: is he iscosi y o he lipid bilaye o
a cell memb ane ha can a ec he o a ion and di u-
sion o p o eins and, he e o e, also a ec s he unc ion o
hese molecules. The lipid bilaye has p o eins embedded
in hem, and lipid packing can in l uence he l uidi y o he
cell memb ane.
Mul i-yea ice: ice o mo e han one yea ´s g ow h.
Osmoly es: a e dissol ed ion o o ganic solu es wi hin a cell
ha p e en osmo ic shock by main aining osmo ic p es-
su e wi hin he cell o a oid cell lysis ( oo much in e nal
p essu e) o sh inkage ( oo li le in e nal p essu e).
Osmo ic shock: is a sudden change in he solu e concen a-
ion a ound a cell ha causes a change in he mo emen o
wa e ac oss i s cell memb ane. In en i onmen s wi h high
concen a ions o sal s, wa e is d awn ou o he cells. This
is a oided by he inco po a ion o osmoly es.
pCO2: he pa ial p essu e o CO2.
Pelagic: desc ibes o ganisms ha swim o d i in a sea.
Pelagic o ganisms: a e plank on and nek on.
pH: pH o seawa e plays an impo an ole in he ocean’s
ca bon cycle. pH measu emen in sea wa e is complica ed by
i s chemical p ope ies, and se e al dis inc pH scales exis ,
i.e., o al scale, sea wa e scale and ee scale.
Pho osyn hesis: he p ocess by which g een plan s and
some unicellula o ganisms con e incoming sunligh in o
o ganic ma e ial om CO2 and wa e .
Pho osyn he ically ac i e adia ion (PAR): is he spec-
al ange (wa e band) o sola adia ion om 400 o 700
nanome es in which pho osyn he ic o ganisms a e able o
use o pho osyn hesis. PAR is no mally quan i i ed asμmol
pho onsm-2 s-1.
Polyhyd oxyalkanoa es (PHA): p oduced in na u e by bac-
e ia o s o e ca bon and ene gy.
Polynya: is an a ea o open wa e su ounded by sea ice.
In his a ea, new sea ice is p oduced and equen ly blown
away, he eby allowing new ice o o m again and again
P ecipi a ion: o ma ion o a solid om solu ion by chemi-
cal o physical p ocesses.
P o eome: is he en i e complemen o p o eins ha a e o
can be exp essed by a cell, issue o o ganism.
Psych ophilic o ganisms: a e o ganisms ha ha e op imal
g ow h a es a empe a u es usually below 15° C and can-
no g ow abo e 20° C.
Psych o ole an o ganisms: a e o ganisms ha ha e op i-
cal g ow h a es a empe a u es abo e 20° C bu a e able o
ole a e and, o bac e ia, g ow unde cold condi ions.
Pycnocline: is a bounda y in oceanog aphy sepa a ing wo
wa e laye s o di e en densi ies. The o ma ion o a pycno-
cline may esul om changes in salini y o empe a u e.
Because he pycnocline laye is ex emely s able, i ac s as
a ba ie o su ace p ocesses; and, he e o e he changes
in salini y and empe a u e a e e y small below he pycno-
cline bu a e seasonal in su ace wa e .
Respi a ion: bac e ia pe o m wo majo unc ions in he
ans o ma ion o o ganic ma e ial: 1) hey p oduce new
bac e ial biomass (bac e ial p oduc ion), and 2) hey espi e
o ganic ca bon o ino ganic ca bon (bac e ial espi a ion).
Salini y: he o al g ams o sal s in 1 kg o sea wa e .
Solubili y pump, he: is d i en by wo p ocesses in he
ocean: 1) CO2 solubili y s ongly ela ed o sea wa e em-
pe a u e whe e CO2 is mo e soluble in cold wa e s, and 2)
he mohaline ci cula ion, d i en by he o ma ion o cold,
dense wa e masses a high la i udes
S a i i ca ion: occu s when wa e masses wi h di e en
p ope ies – e.g., o salini y (halocline), densi y (pycnocline)
and empe a u e ( he mocline) – o m laye s ha ac as ba -
ie s o wa e mixing and, he e o e, c ea e ba ie s o nu i-
en -mixing be ween laye s.
Sympagic: desc ibes o ganisms ha li e whe e wa e exis s
mos ly as a solid, i.e., sea ice.
TA: o al alkalini y is ela ed o he cha ge balance in sea
wa e and in na u al sea wa e a pH > 8 in μmol kg-1 o sea-
wa e .
TCO2: in sea wa e , CO2 exis s in h ee ino ganic o ms: CO2
(aq), HCO3- and CO3-2. The sum o he concen a ion o hese
o ms is called o al dissol ed ino ganic ca bon, abb e ia ed
TCO2, in μmol kg-1 o seawa e .
34 PhD hesis by Do e Haubje g Søgaa d
1.8 Li e a u e ci ed
Alou-Fon E, Mundy J-C, Roy S, Gosselin M, Agus i S (2013) Snow
co e a ec s ice algal pigmen composi ion in he coas al A c ic
Ocean du ing sp ing. Ma Ecol P og Se 474:89 – 104
Al a ez-A iles L, Simpson WR, Douglas TA, S u m M, Pe o ich D,
Domine F (2008) F os l owe chemical composi ion du ing g ow h
and i s implica ions o ae osol p oduc ion and b omine ac i a ion.
J Geophys Res-A mos 113:D21304, doi:10.1029/2008JD010277
Ande son LG, Falck E, Jones EP, Ju e s öm S, Swi JH (2004) En-
hanced up ake o a mosphe ic CO2 du ing eezing o seawa e :
a i eld s udy in S o j o d, S alba d. J Geophys Res 109:C06004,
doi:10.1029/2003JC002120
A end KE, Nielsen TG, Rysgaa d S, Tönnesson K (2010) Di e ences
in plank on communi y s uc u e along he God håbs j o d, om
he G eenland Ice Shee o o sho e wa e s. Ma Ecol P og Se
401:49 – 162
A end K, Du z J, Jónasdó i SH, Jung-Madsen S, Mo ensen J,
Mølle EF, Nielsen TG (2011) E ec s o suspended sedimen s on
copepods eeding in a glacial in l uenced sub-A c ic j o d. Jou
Plank Res 33:1526 – 11537
A igo KR, Dijken GL, Pabi S (2008) Impac o a sh inking A c ic ice
co e on ma ine p ima y p oduc ion. Geophys Res Le 35:L19603,
doi:10.1029/2008GL035028
A igo KR, Mock T, Lizo e MP (2010) P ima y p oduc ion and sea
ice. In: Thomas DN, Dieckmann GS (eds) Sea ice, 2nd Edi ion. Wiley-
Blackwell Publishing Ox o d, pp 283 – 1326
A igo KR, Pe o ich DK, Picka RS, B own ZW, an Dijken GL, Low y
KE, Mills MM, Palme MA, Balch WM, Bah F, Ba es NR, Beni ez-
Nelson C, Bowle B, B ownlee E, Ehn JK, F ey KE, Ga ley R, Laney SR,
Lubelczyk L, Ma his J, Ma suoka A, Mi chell BG, Moo e GWK, O ega-
Re ue a E, Pal S, Polashenski CM, Reynolds RA, Schiebe B, Sosik
HM, S ephens M, Swi JH (2012) Massi e Phy oplank on Blooms
Unde A c ic Sea Ice. Science 336:1408 – 11408
Assmy P, Ehn JK, Fe nández-Mendez M, Hop H, Ka lein C, Sund-
j o d A, Bluhm K, Daase M, Engel A, F ansson A, G anskog MA,
Hudson SR, K is iansen S, Nicolaus M, Peeken I, Renne AHH,
Sp een G, Ta a ek A, Wik o J (2013) Floa ing ice-algal agg ega es
below mel ing A c ic sea ice. PLOS 8:1 – 113
Assu A (1958) Composi ion o sea ice and i s ensile s eng h, in
A c ic Sea ice Con e ence P oceesings, pp. 106-138, NAS-NRC Publ.
598, Na . Academy o Sci., Washing on D.C
Assu A (1960) Composi ion o sea ice and i s ensile s eng h,
Res. Rep. 44:49 pp. U.S. A my Snow Ice and Pe ma os Res. Es abl.,
Wilme e, III
Ba es NR (2006) Ai -sea CO2 l uxes and he con inen al shel pump
o ca bon in he Chukchi Sea adjacen o he A c ic Ocean. J Geo-
phys Res 111:C10013, doi:10.1029/2005JC003083
Ba es NR, Ma his JT (2009) The A c ic Ocean ma ine ca bon cycle:
e alua ion o ai -sea CO2 exchanges, ocean acidi i ca ion impac s
and po en ial eedbacks. Biogeosciences 6:2433 – 12459
Bischo JL, Fi zpa ick JA, Rosenbaue RJ (1993) The solubili y and
s abiliza ion o ikai e (CaCO3 · H2O) om 0° o 25°C: En i onmen al
and paleaclima ic implica ions o hinoli e u a. J Geol 101:21 – 133
Boe ius A, Alb ech S, Bakke K, Bienhold C, Felden J, Fe nandez-
Mendez M, Hend icks A, Ka lein C, Lalande C, K umpen T, Nicolaus
M, Peeken I, Rabe B, Rogache a A, Rybako a E, Soma illa R, Wen-
zhö e F (2013) Expo o algal biomass om he mel ing A c ic
sea ice. Science 339:1430 – 11432
Bowman JP (2008) Genomic analysis o psysc ophillic p oka y-
o es. In: Psych ophiles: om Biodi e si y o Bio echnology (Eds.
R.Ma gesin, F. Schinne , J-C. Ma x and C. Ge day) Sp inge -Ve lag,
Be lin, pp. 265 – 1284
Bowman JS, Deming JW (2010) Ele a ed bac e ial abundance
and exopolyme s in saline os l owe s and implica ions o a -
mosphe ic chemis y and mic obial dispe sal. Geophys Res Le
37:L13501, doi:10.1029/2010GL043020
Bowman JS, La osa C, Vogel TM, Deming JW (2013) Selec i e oc-
cu ence o Rhizobiales in os l owe s on he su ace o young
sea ice nea Ba ow, Alaska and dis ibu ion in he pola ma ine
a e biosphe e. En i on Mic obiol Rep 5: 575 – 1582
B en G, Else T ,Galley RJ ,Lansa d B ,Ba be DG ,B own K ,Mille LA
,Mucci A,Papaky iakou TN,T emblay J-E,Rysgaa d S (2013) Fu he
obse a ions o a dec easing a mosphe ic CO2 up ake capaci y in
he Canada Basin (A c ic Ocean) due o sea ice loss. Geophys Res
Le 40:1132 – 11137, doi:10.1002/g l.50268
Bucha d B, Is aelson C, Seaman P, S ockmann G (2001) Ikai e
u a owe s in Ikka Fjo d, sou hwes G eenland: Thei o ma ion
by mixing o seawa e and alkaline sp ing wa e . J Sedimen Res
71:176 – 1189
Buck KR, Nielsen TG, Hansen BW, Gas up-Hansen D, Thomsen HA
(1998) In i l a ion phy o- and p o ozooplank on assemblages in
he annual sea ice o Disko Island, Wes G eenland, sp ing 1996.
Pola Biol 20:377 – 1381
Cai W-J, Chen L, Chen B, Gao Z, Lee SH, Chen J, Pie o D, Sulli an
K, Wang Y, Hu X, Huang W-J, Zhang Y, Xu S, Mu a a A, G ebmeie
JM, Jones EP, Zhang H (2010) Dec ease in he CO2 up ake capaci y
in an ice- ee A c ic Ocean Basin. Science 329:556 – 1559
Calbe A, Riisgaa d K, Saiz E, S edmon C, Nielsen TG (2011) Phy o-
plank on g ow h and mic ozooplank on g azing along a sub-A c ic
j o d (God håbs j o d, Wes G eenland) Ma Ecol P og Se 442:11 –
122, doi:10.3354/meps09343
Cameo a SS, Makka RS (1998) Syn hesis o biosu ac an s in ex-
eme condi ions. Appl Mic obiol Bio echnol 50: 520 – 1529
Ca alie i DJ, Pa kinson CL (2012) A c ic sea ice a iabili y and
ends, 1979 – 2010. TC 6: 881-889, doi: 10.5194/ c-6-881 – 12012
Collins RE, Ca pen e SD, Deming JW (2008) Spa ial he e ogenei y
and empo al dynamics o pa icles and pEPS in A c ic win e sea
ice. J Ma Sys 74:902 – 1917
Collins RE, Rocap G, Deming JW (2010) Pe sis ence o bac e ial and
a chaeal communi ies in sea ice h ough an A c ic win e . En i on
Mic obiol 12:1828 – 11841
35
PhD hesis by Do e Haubje g Søgaa d
Collins RE, Deming JW (2011) Abundan dissol ed gene ic ma e-
ial in A c ic sea ice, Pa II: Vi us dynamics du ing au umn eeze-
up. Pola Biol 34:1831- – 11841
Comiso JC (2003) La ge-scale cha ac e is ics and a iabili y o he
global sea ice co e . In: Thomas DN, Dieckmann GS (eds), 1s Edi-
ion. Sea ice: an in oduc ion o i s physics, chemis y, biology and
geology. Blackwell Science, Ox o d, pp 112 – 1142
Comiso JC, Pa kinson CL, Ge s en R, S ock L (2008) Accele a ed
decline in he A c ic sea ice co e . Geophys Res Le 35:L01703,
doi:10.1029/2007GL031972
Comiso JC (2012) La ge Decadal Decline o he A c ic Mul iyea Ice
Co e . J Clima e 25:1176 – 11193
Co a GF, Ho ne EPW (1989) Physical con ol o a c ic ice algal p o-
duc ion. Ma Ecol P og Se 52:111 – 1121
Co a GF, Anning JL, Ha is LR, Ha ison WG, Smi h REH (1990) The
impac o ice algae on ino ganic nu ien s in seawa e and sea ice
in Ba ow S ai , NWT, Canada du ing sp ing. Can J Fish Aqua Sci
47:1402 – 11415
Cox GFN, Weeks WF (1974) Salini y a ia ions in sea ice. J Glaciol
13:109 – 1120
Cox GFN, Weeks WF (1983) Equa ions o de e mining he gas and
b ine olumes in sea-ice samples. J Glaciol 29:306 – 1316
C abeck O, Delille B, Thomas DN, Geil us NX, Rysgaa d S, Tison JL
(2014) CO2 and CH4 in sea ice om a suba c ic j o d. Biogeosciences
Discuss 11:4047 – 14088, doi:10.5194/bgd-11-4047-2014
Delille B (2006) Ino ganic ca bon dynamics and ai -ice-sea CO2
l uxes in he open and coas al wa e s o he Sou he n Ocean, PhD
hesis, 296 pp. Uni . de Liège, Liège, F ance.
Delille B, Jou dain B, Bo ges AV, Tison J-L, Delille D (2007) Biogas
CO2, O2, dime hylsul i de dynamics in sp ing An a c ic as ice. Lim-
nol Oceanog 52:1367 – 11379
Deming J (2010) Sea ice bac e ia and i uses. In: Thomas DN,
Dieckmann GS (eds) Sea ice, 2nd Edi ion. Wiley-Blackwell
Publishing Ox o d, pp 247 – 1282
Dieckmann GS, Neh ke G, Papadimi iou S, Gö liche J, S eininge
R, Kennedy H, Wol -Glad ow D, Thomas DN (2008) Calcium
ca bona e as ikai e c ys als in An a c ic sea ice. Geophys Res Le
35:L08501, doi:10.1029/2008GL033540
Dieckmann GS, Neh ke G, Uhlig C, Gö liche J, Ge land S, G an-
skog MA, Thomas DN (2010) B ie communica ion: ikai e CaCO3·
6H2O disco e ed in A c ic sea ice. TC 4:227 – 1230, doi:10.5194/
c-4-227-2010
Domine F, Taillandie AS, Simpson WR, Se e in K (2005) Speci i c
su ace a ea, densi y and mic os uc u e o os l owe s. Geophys
Res Le 32: L13502, doi: 10.1029/2005GL023245
Douglas TA, Domine F, Ba e M, Anas asio C, Beine HJ, Bo enheim
J, G annas A, Houdie S, Ne che a S, Rowland G, S aeble R, S e en
A (2012) F os l owe s g owing in he A c ic ocean-a mosphe e-
sea ice-snow in e ace: 1. Chemical composi ion. J Geophys Res
117:D00R09, doi: 10.1029/2011JD016460
D inkwa e MR, C ocke GB (1988) Modelling changes in he
dielec ic and sca e ing p ope ies o young snow-co e ed sea
ice a GHz equencies. J Glaciol 34:274 – 1282
Eicken H, Lange MA (1989) De elopmen and p ope ies o sea ice
in he coas al egime o he Sou heas e n Weddel Sea. J Geophys
Res-Ocean 94:8193 – 18206, doi:10.1029/JC094iC06p08193
Eicken H (2003) F om he mic oscopic o he mac oscopic o he
egional scale, g ow h, mic os uc u e and p ope ies o sea ice. In:
Thomas DN, Dieckmann GS (eds), 1s Edi ion. Sea ice: an in oduc-
ion o i s physics, chemis y, biology and geology. Wiley-Blackwell
Publishing Ox o d, pp 22 – 181
Eicken H, G e ell TC, PE o ich DK, Rich e -Menge JA, F ey K (2004)
Hyd aulic con ols o summe A c ic pack ice albedo. J Geophys
Res-Oceans 109:C08007, doi:10.1029/2003JC001989
Eicken H, Dmi enko I, Tyshko K, Da o skikh A, Die king W, Blahak U,
G o es J, Kassens H (2005) Zona ion o he Lap e Sea land as ice
co e and i s impo ance in a ozen es ua y. Global Plane Change
48: 55 – 183, doi:10.1016/j.gloplacha.2004.12.005
Else BGT, Papky iakou T, Asplin M, Ba be D, Galley R, Mille L, Mucci
A (2013a) Annual cycle o ai -sea CO2 exchange in an A c ic polynya
egion. Glob Biogeochem Cycles 27, doi:10.1002/gbc.20016
Else BGT, Galley RJ, Lansa d B, Ba be DG, B own K, Mille LA,
Mucci A, Papky iakou TN, Temblay JE, Rysgaad S (2013b) Fu he
obse a ions o a dec easing a mosphe ic CO2 up ake capaci y in
he Canada Basin (A c ic Ocean) due o sea ice loss. Geophys Res
Le 40:1132 – 11137, doi:10.1002/g l.50268
Else BGT, Papky iakou TN, Galley RJ, D ennan WM, Mille LA, Mucci
A, Thomas H (2011) Win e ime CO2 l uxes in an A c ic polynya
using eddy co a iance: e idence o enhanced ai -sea gas
ans e du ing ice o ma ion. J Geophys Res 116:C00G03, doi:
10.1029/2010JC006760
E onen-Rasimus W, Kaa okallio H, Ly a C, Au io C, Au io R, Kuosa
H, Dieckmann GS, Thomas DN (2014) Bac e ial communi y
dynamics and ac i i y in ela ion o dissol ed o ganic ma e
a ailabili y du ing sea ice o ma ion in a mesocosm expe imen .
Mic obiol 3:139 – 1156
Felle G, Ge day C (2003) Psych ophillic enzymes: ho opics in
cold adap a ion. Na Re Mic obiol 1: 200 – 1208
Fe e e F, Un e s eine N (1998) Obse a ions o mel ponds on
A c ic sea ice. J Geophys Res-Oceans 103:24821 – 124835
Fische M (2013) Sea ice and he ai -sea exchange o CO2. PhD
hesis, Uni e si y B emen
Fische M, ThomasDN, K ellA, Neh keG, Gö liche J, No manL,
Meine s KM, Riaux-GobinC, DieckmannGS (2013) Quan i i ca ion
o ikai e in An a c ic sea ice. An Sci 25:421 – 1432
F ansson A, Chie ici M, Noji i Y (2006) Inc eased ne CO2 ou -
gassing in he upwelling egion o he sou he n Be ing Sea in
a pe iod o a iable ma ine clima e be ween 1995 and 2001. J
Geophys Res 111: C08008, 10.1029/2004JC002759
F ansson A, Chie ici M, Noji i Y (2009) New insigh s in o he spa ial
a iabili y o he su ace wa e ca bon dioxide in a ying sea ice
condi ions in he A c ic Ocean. Con inen shel Res 29:1317 – 11328

36 PhD hesis by Do e Haubje g Søgaa d
F ansson A, Chie ici M, Yage PL, Smid J . Wo (2011) An a c ic
sea ice ca bon dioxide sys em and con ols. J Geophys Res
116:C12035. doi:10.1029/2010JC006844
F i sen CH, Ackley SF, K eme JN, Sulli an CW (1998) Flood- eeze
and mic oalgal dynamics in An a c ic pack ice, in An a c ic sea ice:
Biological p ocesses, in e ac ions and a iabili y. An a c Res Se
73:1 – 122
Ga ison DL, Buck KR (1986) O ganisms losses du ing mel ing: A
se ious bias in sea ice communi y s udies. Pola Biol 6:237 – 239
Ga ison DL, Close AR, Reimni z E (1990) Mic oo ganisms concen-
a ed by azil ice: e idence om labo a o y expe imen s and
i eld measu emen s. In Sea Ice P ope ies and P ocesses (eds) SF
Ackley and WF Weeks, U.S. A my Co ps o Enginee s, Cold o
Enginee s, Cold Regions Resea ch and Enginee ing Labo a o y,
Monog aph 90-1, Pp. 92 – 196
Geil us N-X, Ga na G, Papaky iakou T, Tison JL, Else B, Thomas H,
Shadwick E, Delille B (2012b) Dynamics o pCO2 and ela ed ai -ice
CO2 l uxes in he A c ic coas al zone (Amundsen Gul , Beau o
Sea). J Geophys Res 117:C00G10, doi:10.1029/2011JC007118
Geil us NX, Ca na G, Dieckmann GS, Halden N, Neh ke G,
Papaky iakou T, Tison JL, Delille B (2013a) Fi s es ima es o he
con ibu ion o CaCO3 p ecipi a ion o he elease o CO2 o he
a mosphe e du ing young sea ice g ow h. J Geophys Res 188:244
– 1255, doi:10.1029/2012JC007980
Geil us NX, Galley RJ, C abeck O, Papky iakou T, Landy J, Tison
JL, Rysgaa d S (2014) Ino ganic ca bon dynamics o mel pond-
co e ed i s yea sea ice in he Canadian A c ic. Biogeosciences
Discuss 11:7485 – 17519, doi: 10.5194/bgd-11-7485-2014
Gibson J, T ull T (1999) Annual cycle o CO2 unde sea-ice and in
open wa e in P ydz Bay, Eas An a c ica. Ma Chem 66:187 – 1200
Glei z M, Thomas DN (1992) Physiological esponses o a small
An a c ic dia om (Chae oce os sp.) o simula ed en i onmen al
cons ain s associa ed wi h sea ice o ma ion. Ma Ecol P og Se
88:271 – 1278
Glei z G, Ru ge s VD, Thomas DN, Dieckmann GS, Mille o FJ (1995)
Compa ison o summe and win e ino ganic ca bon, oxygen and nu-
ien concen a ions in An a c ic sea ice b ine. Ma Chem 51:81 – 191
Glud RN, Rysgaa d S, Kühl M, Hansen JW (2007) The sea ice in Young
Sound: Implica ions o C cycling. In: Ca bon cycling in A c ic ma ine
Ecosys ems: Case s udy Young Sound. Chap e 4 p: 62 – 185. Rysgaa d
S & Glud RN (Eds), Ca bon cycling in A c ic ma ine Ecosys em: Case
s udy Young Sound. Meddelse om G ønland, Bioscience Vol 58. Co-
penhagen he Commission o Scien i i c Resea ch in G eenland 2007
Glud RN, Rysgaa d S, Tu ne G, McGinnis DF, Leakey RJG (2014)
Biological and physical-induced oxygen dynamics in mel ing sea
ice o he F am S ai . Limnol Oceanog 59:1097 – 1111
Golden K, Ackley SF, Ly le V (1998) The Pe cola ion Phase
T ansi ion in Sea Ice. Science 282:2238 – 12241, doi:10.1126/
science.282.5397.2238
Golden KM, Eicken H, Hea on L, Mine J, P ingle D, Zhu J (2007)
The mal e olu ion o pe meabili y and mic os uc u e in sea ice.
Geophys Res Le 34:L16501, doi:10.1029/2007GL030447
Gosselin M, Legend e L, The iaul J-C, Deme s S, Roche M (1986)
Physical con ol o he ho izon al pa chiness o sea-ice mic oalgae.
Ma Ecol P og Se 29:289 – 1298
Gosselin M, Legend e L, The iaul J-C, Deme s S (1990) Ligh and
nu ien limi a ion o sea ice mic oalgae (Hudson Bay, Canadian
A c ic). J Phycol 26:220 – 1232
Gouno , A.M., Russell, N.J., 1999. Physiology o cold-adap ed
mic oo ganisms. In: Ma gesin, R., Schinne , F. (Eds.), Cold-adap ed
O ganisms. Ecology, Physiology, Enzymology and Molecula Bio-
logy. Sp inge , Be lin, pp. 33 – 155
Gowing MM, Ga ison DL, Gibson AH, K upp JM, Je ies MO,
F i sen CH (2004) Bac e ial and i al abundance in Ross Sea sum-
me pack ice communi ies. Ma Ecol P og Se 279:3 – 112
G adinge R, Ikä alko J (1998) O ganism inco po a ion in o
newly o ming A c ic sea ice in he G eenland Sea. J Plank on Res
20:871 – 1886
G adinge R, F ied ich C, Spindle M (1999) Abundance, biomass
and composi ion o he sea ice bio a o he G eenland Sea pack-
ice. Deep-Sea Res Pa Li Top S ud Oceanog 46:1457 – 11472
G adinge R (2009) Sea ice algae: Majo con ibu o s o p ima y
p oduc ion and algal biomass in he Chukchi and Beau o Seas
du ing May/June 2002. Deep-Sea Res PT II 17: 1201 – 11212,
doi:10.1016/j.ds 2.2008.10.016
G ossmann S, Glei z M (1993) Mic obial esponses o expe imen al
sea-ice o ma ion: implica ions o he es ablishmen o An a c ic
sea-ice communi ies. J Exp Ma Biol Ecol 173:273 – 1289
Haas C, Hend icks S, Eicken H, He be A (2010) Synop ic ai bo ne
hickness su eys e eal s a e o A c ic sea ice co e . Geophys Res
Le 37:L09501, doi:10.1029/2010GL042652
Hansen PJ (2002) The e ec o high pH on he g ow h and su i al
o ma ine phy oplank on: implica ions o species succession.
Aqua Mic ob Ecol 28:279 – 1288
Ha e AA, Wang F, Ba be D, Geil us N-X, Galley R, Rysgaa d S (2013)
pH e olu ion in sea ice g own a an ou doo expe imen al acili y.
Ma Chem 154:46 – 164, doi:10.1016/j.ma chem.2013.04.007
Heinesch B, Tison JL, Ca na G, Eicken H, Geil us N-X, Goosens T,
Papaky iakou T, Ye naux M, Delille B (2010) Mic ome eo ological
su ey o ai -sea ice CO2 l uxes in a c ic coas al wa e s. In EGU
Gen al Assembly, 2 – 17 May, 2010, Vienna, Aus ia
Hinga KR (2002) E ec o pH on coas al ma ine phy oplank on.
Ma Ecol P og Se 238:281 – 1300
Ho ne R, Alexande V (1972) Algal popula ions in a c ic sea ice: an
in es iga ion o he e o ophy. Limnol Oceanog 17:454 – 1458
Ho ne R, Sch ade GC (1982) Rela i e con ibu ions o ice algae,
phy oplank on, and ben hic mic oalgae o p ima y p oduc ion in
nea sho e egions o he Beau o Sea. A c ic 35:485 – 1503
Ho ne RA (1985) Ecology o sea ice mic oalgae. In: Ho ne RA
(Ed.), Sea Ice Bio a. CRC P ess, Boca Ra on, FL, pp. 83 – 1103
Hu Y-B, Wol -Glad ow DA, Dieckmann GS, Völke C, Neh ke G
(2014) A labo a o y s udy o ikai e (CaCO3 · 6H2O) p ecipi a ion as
a unc ion o pH, salini y, empe a u e and phospha e concen a-
ion. Ma Chem 162:10 – 118
37
PhD hesis by Do e Haubje g Søgaa d
Hue as IE, Colman B, Espie GS; Lubian LM (2000) Ac i e anspo o
CO2 by he h ee species o ma ine mic oalgae. J Phycol 36:314 – 1320
Inall ME, Mu ay T, Co ie FR, Scha e K, Boyd TJ, Heywo d KJ,
Be an SL (2014) Oceanic hea deli e y ia Kange dlugssuaq Fjo d
o he sou h-eas G eenland ice shee . J Geophys Res-Ocean 119:1
– 115, doi:10.1002/2013JC009295
Islei son D, Galley RJ, Ba be DG, Landy JC, Koma o AS, Sha ai L
(2013) A s udy on he C-band pola ime ic sca e ing and physical
cha ac e is ics o os l owe s on expe imen al sea ice. IEEE T ans
Geosci Remo e Sensing 52:1787 – 11798
Jakobsson, M G an z A, K is o e sen Y, Macnab R (2004) Ba hyme y
and physiog aphy o he A c ic Ocean and i s cons i uen seas. In
S ein R & MacDonald RW (eds) The o ganic ca bon cycle in he
A c ic Ocean. Sp inge Ve lag Be lin Heidelbe g New Yo k, pp. 1 – 16
Jones EP, Coo e AR (1981) Oceanic CO2 p oduced by he p ecipi a-
ion o CaCO3 om b ines in sea ice. J Geophys Res 86: 11041 –
111043, doi:10.1029/JC086iC11p11041
Juul-Pede sen, A end KE, Mo ensen J, Bliche M, Søgaa d DH,
Rysgaa d S (submi ed) Seasonal and in e annual phy oplank on
p oduc ion in a sub-a c ic j o d (God håbs j o d) connec ed o he
G eeland Ice Shee . Ma Ecol P og Se
Kaa okallio H (2004) Food web componen s, and physical and
chemical p ope ies o Bal ic Sea ice. Ma Ecol P og Se 273:49 – 163
Kaa okallio H, Kuosa H, Thomas DN, G anskog MA, Ki i K (2006)
Biomass, composi ion and ac i i y o o ganism assemblages along
a salini y g adien in sea ice subjec ed o i e discha ge in he
Bal ic Sea. Pola Biol 30:183 – 1197
Khan SA, Kjæ KH, Be is M, Bambe JL, Wah J, Kjeldsen KK, Bjø k
AA, Ko sgaa d NJ, S ea ns LA, B oeke MR, Liu L, La sen NK, Mu e-
san LS (2014) Sus ained mass loss o he no heas G eenland
ice shee igge ed by egional wa ming. Na u e Clima e Change
4:292 – 1299, doi:10.1038/NCLIMATE2161
Killawee JA, Fai child IJ, Tison JL, Janssens L, Lo ain R (1998) Se-
g ega ion o solu es and gases in expe imen al eezing o dilu e
solu ions: Implica ions o na u al glacial sys ems. Geochim
Cosmochim Ac a 62:3637 – 13655
Ko b RE, Sa ille PJ, Johns on AM, Ra en J (1997) Sou ces o ino -
ganic ca bon o pho osyn esis by h ee species o ma ine dia-
oms. J Phycol 33:433 – 1440
K ell A, Funck D, Ple ne I John U, Dieckmann G (2007) Regula-
ion o p oline me abolism unde sal s ess in he psych ophillic
dia om F agila iopsis cylind us (Bacilla iophyceas). J Phycol 43: 753
– 1762
Kuosa H, Bo man B, Ki i K, B andini F (1992) E ec s o An a c ic
sea ice bio a on seeding as s udied in aqua ium expe imen s.
Pola Biol 12:333 – 1339
Kühl M, Glud RN, Bo um J, Robe s R, Rygaa d S (2001) Pho osyn-
he ic pe o mance o su ace-associa ed algae below sea ice as
measu ed wi h a pulse-ampli ude-modula ed (PAM) l uo ome e
and O2 mic osenso s. Ma Ecol P og Se 223:1 – 114
Lannuzel D, Schoe mann V, Jong JD, Tison J-L, Chou L (2007) Dis i-
bu ion and biogeochemical beha iou o i on in he Eas An a c ic
sea ice. Ma Chem 106:18 – 132
La oie D, Denman K, Michel C (2005) Modeling ice algal g ow h
and decline in a seasonally ice-co e ed egion o he A c ic (Reso-
lu e Passage, Canadian A chipelago). J Geophys Res-Ocean 110:
C11009, doi:10.1029/2005JC002922
Lazza a L, Na dello I, E manni C, Mangoni O, Saggiomo V (2007)
Ligh en i onmen and seasonal dynamics o mic oalgae in he
annual sea ice a Te a No a Bay, Ross Sea, An a c ica. An a c Sci
19: 83 – 192
Lee SH, S ockwell DA, Joo H-M, Son YB, Kang C-K, Whi ledge TE
(2012) Phy oplank on p oduc ion om mel ing ponds on A c ic
sea ice. J Geophys Res 117:C04030, doi:10.1029/2011JC007717
Legend e L, Ackley SF, Dieckmann GS, Gulliksen B, Ho ne R,
Hoshiai T, Melniko IA, Reebu gh WS, Spindle M, Sulli an CW
(1992) Ecology o sea ice bio a. Pola Biol 12:429 – 1444
Leu E, Wik o J, Sø eide JE, Be ge J, Falk-Pe e sen S (2010) In-
c eased i adiance educed ood quali y o sea ice algae. Ma Ecol
P og Se 411:49 – 160
Leu E, Sø eide JE, Hessen DO, Falk-Pede sen S, Be ge J (2011) Con-
sequences o chaning sea ice co e o p ima y p oduc ion and
seconda y p oduce s in he Eu opean A c ic Shel seas: Timing,
quan i y and quali y. P og Oceanog aphy 90:18 – 132
Leu E, Mundy CJ, Assmy P, Campbell K, Gab ielsen TM, Gosselin M,
Juul-Pede sen T, G adinge R (submi ed) A c ic sp ing awakening –
s ee ing p inciples behind he phenology o e nal ice algae blooms
Long MH, Koopmans D, Be g P, Rysgaa d S, Glud RN, Søgaa d DH
(2012) Oxygen exchange and ice mel measu ed a he ice-wa e
in e ace by eddy co ela ion. BG 9:1957 – 11967, doi:10.5194/
bg-9-1-2012
Loose B, Mille LA, Ellio S, Papaky iakou T (2011a) Sea Ice Bio-
geochemis y and Ma e ial T anspo Ac oss he F ozen In e ace.
Ocean 24:203 – 1218
Loose B, Schlosse P, Pe o ich DK, Ringelbe g D, Ho D, Takahashi
T, Rich e -Menge JA, Reynolds C, Mcgillis W, Tison JL (2011b) Gas
di usion h ough columna labo a o y sea ice: implica ions o
mixed-laye en ila ion o CO2 in he seasonal ice zone. Tellus
63B:23 – 139, doi:10.1111/j.1600-0889.2010.00506.x
Lü hje M, Fel ham Taylo PD, Wo s e MG (2006) Modeling he
summe ime e olu ion o sea ice mel ponds. J Geophys Res
111:C02001, doi:10.1029/2004JC002818
Lyde sen C, Assmy P, Falk-Pe e sen S, Kohle J, Ko acs KM, Reigs ad
M, S een H, S øm H, Sund j o d A, Va pe Ø, Walczowski W,
Weslawski JM, Zajaczkowski M (2014) The impo ance o idewa e
glacie s o ma ine mammals and seabi ds in S alba d, No way. J
Ma Sys 129:452 – 1471
MacGilch is GA, Ga aba o CAN, Tsubouchi T, Bacon S, To es-
Valdés S, Aze su-Sco K (2014) The A c ic Ocean ca bon sink.
Deep-Sea Res I 86:39 – 155
Ma ion GM (2001) Ca bona e mine al solubili y a low empe a-
u es in he Na-K-Mg-Ca-H-Cl-SO4-OH-HCO3-CO3-CO2-H2O sys em.
Geochim Cosmochimica Ac 65:1883 – 11896
Ma in S, D ucke R, Fo M (1995) A labo a o y s udy o os l owe
g ow h on he su ace o young sea ice. J Geophys Res 100:7027 –
17036, doi:10.1029/94JC03243
38 PhD hesis by Do e Haubje g Søgaa d
Ma x JG, Ca pen e SD, Deming JW (2009) P oduc ion o c yo-
p o ec an ex acellula polysaccha ide subs ances (EPS) by he
ma ine psych ophilic bac e iumColwellia psych e y h aeas ain
34H unde ex eme condi ions. Can J Mic obiol 55: 63 – 172, doi:
10.1139/W08-130
Mille L, Ga na G, Else B, Su he land N, Papaky iakou T
(2011a) Ca bona e sys em e olu ion a he A c ic Ocean su -
ace du ing au umn eeze-up. J Geophys Res 116:C00G04,
doi:10.1029/2011JC007143
Mille LA, Papaky iakou TN, Collinsn RE, Deming JW, Ehn JK, Mac-
donald RW, Mucci A, Owens O, Raudsepp M, Su he land N (2011b)
Ca bon dynamics in sea ice: A win e l ux ime se ies. J Geophys
Res 116:C02028, doi:10.1029/2009JC006058
Mille o FJ (1995) The modynamic o he ca bon dioxide sys em in
he ocean. Geochim Cosmochim Ac a 59:661 – 1677
Mikkelsen DM, Rysgaa d S, Glud RN (2008) Mic oalgal composi ion
and p ima y p oduc ion in A c ic sea ice: a seasonal s udy om
Kobbe j o d (Kange lua sunnguaq), Wes G eenland. Ma Ecol P og
Se 368:65 – 174
Mikkelsen DM, Wi kowski A (2010) Mel ing sea ice o axonomic
analysis: a compa ison o ou mel ing p ocedu es. Pola Res,
doi:10.1111/j.1751-8369.2010.00162.x
Mock T, G adinge R (1999) De e mina ion o A c ic ice algal p o-
duc ion wi h a new in si u incuba ion echnique. Ma Ecol P og
Se 177:15 – 126
Mo ensen J, Lenne K, Bend sen J, Rysgaa d S (2011) Hea
sou ces o glacial mel in a sub-A c ic j o d (God håbs j o d) in
con ac wi h he G eenland Ice Shee . J Geophys Res 116, C01013,
doi:10.1029/2010JC006528
Mo ensen J, Bend sen J, Mo yka RJ, Lenne K, T u e M, Fahne-
s ock M, Rysgaa d S (2013) On he seasonal eshwa e s a i i ca-
ion in he p oximi y o as - l owing idewa e ou le glacie s in
a sub-A c ic sill j o d. J Geophys Res-Oceans 118:1382 – 11395,
doi:10.1002/jg c.20134
Mundy CJ, Gosselin M, Ehn J, G a on Y, Rossnagel A, Ba be DG,
Ma in J, T emblay J-E, Palme M, A igo KR, Da nis G, Fo ie L,
Else B, Papaky iakou T (2009) Con ibu ion o unde -ice p ima y
p oduc ion o an ice-edge upwelling phy oplank on bloom
in he Canadian Beau o Sea. Geophys Res Le 36:L17601,
doi:10.1029/2009GL038837
Mundy CJ, Gosselin M, G a on Y, B own K, Galindo V, Campbell K,
Le asseu M, Ba be D, Papaky iakou T, Bélange S (2013) Role o
en i onmen al ac o s on phy oplank on bloom ini ia ion unde
land as sea ice in Resolu e Passage, Canada. Ma Ecol P og Se
497:39 – 149
Mun o DR, Dunba RB, Muccia one DA, A igo KR, Long MC (2010)
S able iso ope composi ion o dissol ed ino ganic ca bon and
pa icula e o ganic ca bon in sea ice om he Ross Sea, An a c ica.
J Geophys Res-Ocean 115:C09005, doi:10.1029/2009JC005661
Nansen F (1906) No he n Wa e s: Cap ain Roald Amundsen’s
oceanog aphic obse a ions in he A c ic Seas in 1901. In commis-
sion by Jacob Dyb ad, Ch is iania.
Nakaoka S-I, Aoki S, Nakazawa T, Hashida G, Mo imo o S,
Yamanouchi T, Yoshikawa-Inoue H (2006) Tempo al and spa ial
a ia ions o oceanic pCO2 and ai -sea CO2 l ux in he G eenland
Sea and he Ba en s Sea. Tellus 58B:148 – 1161
Nedashko sky A, Kh edynich S, Pe o sky T (2009). Alkalini y o
sea ice in he High-La i udinal A c ic acco ding o he su eys pe -
o med a No h Pole D i ing S a ion 34 and cha ac e iza ion o he
ole o he A c ic ice in he CO2 exchange. Ma Chem 49:61 – 169
Nicolaus M, Haas C, Willmes S (2009) E olu ion o i s -yea and
second-yea snow p ope ies on sea ice in he Weddell Sea du-
ing sp ing-summe ansi ion. J Geophys Res-A mos 114:1 – 117,
doi:10.1029/2008JD011227
Nicolaus M, Ka lein C, Maslanik J, Hend icks S (2012) Changes in
A c ic sea ice esul in inc easing ligh ansmi ance and abso p-
ion. Geophys Res Le 39:L24501, doi:10.1029/2012GL053738
Nomu a DH, Yoshikawa-Inoue, Toyo a T (2006) The e ec o sea-ice
g ow h on a ai -sea CO2 l ux in a ank expe imen . Tellus B Chem
Phys Me eol 58:418 – 1426
Nomu a D, Eicken H, G adinge R, Shi asawa K (2010a) Rapid
physically d i en in e sion o he ai -sea ice CO2 l ux in he sea-
sonal land as ice o Ba ow, Alaska a e onse o su ace mel .
Con inen Shel Res 30:1998 – 12004, doi:10.1016/j.cs .2010.09.014
Nomu a D, Yoshikawa-Inoue H, Toyo a T, Shi asawa K (2010b)
E ec s o snow, snowmel ing and e eezing p ocesses on ai -sea-
ice CO2 l ux. J Glaciol 56:262 – 1270
Nomu a D, G anskog MA, Assemy P, Simizu D, Hashida G (2013a)
A c ic and An a c ic sea ice ac s as a sink o a mosphe ic CO2
du ing pe iods o snowmel and su ace l ooding. J Geophys Res-
Ocean 118:6511 – 16524, doi:10.1002/2013JC009048
Nomu a D, Assmy P, Neh ke G, G anskog MA, Fische M, Dieckmann
GS, F ansson A, Hu Yubin, Schne ge B (2013b) Cha ac e iza ion
o ikai e (CaCO3· 6H2O) c ys als in i s -yea A c ic sea ice no h
o S alba d. Annals Glaciology 62:125 – 1131, doi:10.3189/
2013AoG62A034
Pa kinson CL (2000) Va iabili y o A c ic sea ice: The iew om
space and 18-yea eco d. A c ic 53:341 – 1358
Papadimi iou S, Kennedy H, Ka ne G, Dieckmann GS, Thomas
DN (2004) Expe imen al e idence o ca bona e p ecipi a ion and
CO2 degassing du ing sea ice o ma ion. Geochim Cosmochim
Ac a 68:1749 – 11761
Papadimi iou S, Thomas DN, Kennedy H, Haas C, Kuosa H, K ell A,
Dieckmann GS (2007) Biogeochemical composi ion o na u al sea
ice b ines om he Weddell Sea du ing ea ly aus al summe .
Limnol Oceanog ap 52:1809 – 11823
Papadimi iou S, Thomas D, Kennedy H, Kuosa H, Dieckmann
GS (2009) Ino ganic ca bon emo al and iso opic en ichmen in
An a c ic sea ice gap laye s du ing ea ly aus al summe . Ma Ecol
P og Se 386:15 – 127
Papadimi iou S, Kennedy H, No man L, Kennedy DP, Dieckmann
GS, Thomas DN (2012) The e ec o biological ac i i y, CaCO3 mine-
al dynamics and CO2 degassing in he ino ganic ca bon cycle in
sea ice in la e win e -ea ly sp ing in he Weddell Sea, An a c ica. J
Geophys Res 117:C08011, doi:10.1029/2012JC008058
39
PhD hesis by Do e Haubje g Søgaa d
Papaky iakou T, Mille L (2011) Sp ing ime CO2 exchange o e
seasonal sea ice in he Canadian A c ic A chipelago. Annals Glaciol
52:215 – 1224
Pede sen JTP, Kau mann LH, K oon A, Jakobsen BH (2010) The
no heas G eenland Si ius Wa e Polynya dynamics and a iabili y
in e ed om sa elli e image y. Geog a i sk Tidsk i 110:131 – 1142
Pe o ich DK, Rich e -Menge JA (1994) Su ace cha ac e is ics o lead
ice. J Geophys Res Le 99:16341 – 116350, doi:10.1029/94JC01194
Pe o ich DK (1996) The op ical p ope ies o sea ice, CRREL Mono-
g aph, 96 – 11, 25 pp
Pe o ich DK, Jones KF, Ligh B, Eicken H, Ma kus T, S oe e J, Lindsay
R (2011) Sola pa i ioning in a changing A c ic sea-ice co e . Ann
Glaciol 52:192 – 1196, doi:10.3189/172756411795931543
Pe ich C, Eicken H (2010) G ow h, S uc u e and P ope ies o sea
ice. In: Thomas DN, Dieckmann GS (eds) sea ice, 2nd Edi ion. Wiley-
Blackwell Publishing Ox o d, pp 23 – 178
Polashenski C, Pe o ich D, Cou ille Z (2012) The mechanisms o
sea ice mel pond o ma ion and e olu ion. J Geophys Res-Oceans
117:C01001, doi:10.1029/2011JC007231
Ralph PJ, Ryan KG, Ma in A, Fen on G (2007) Mel ing ou o sea ice
causes g ea e pho osyn he ic s ess han eezing in. J Phycol 43:
948 – 1956
Ramaswamy V, Chanin M-L, Angell J, Ba ne J, Ga en D, Gelman
M, Keckhu P, Koshelko Y, Labi zke K, Lin J-JR, O`Neill A, Nash J,
Randel W, Rood R, Shine K, Shio ani M, Swinbank R (2001) S a o-
sphe ic empe a u e ends: Obse a ion and model simula ions.
Re iew Geophys 39:71 – 1122, doi:10.1029/199RG000065
Rankin AM, Auld V, Wol EW (2000) F os l owe s as a sou ce o
ac iona ed sea sal ae osol in he pola egions. Geophys Res Le
27:3469 – 13472
Rankin AM, Wol EW, Ma in S (2002) F os l owe s: Implica ions
o oposphe ic chemis y and ice co e in e p e a ion. J Geophys
Res-A mos 107:D23, doi:10.1029/2002JD002492
Ra en J (1980) Limi s on g ow h a e. Na u e 361:209 – 1210
Reimni z E, Ma inco ich L J , McCo mick M, B iggs WM (1992) Sus-
pension eezing o bo om sedimen and bio a in he No hwes
Passage and implica ions o A c ic Ocean sedimen a ion. Can J
Ea h Sci 29:693 – 1703
Richa dson C (1976) Phase ela ionships in sea ice as a unc ion o
empe a u e. J Glaciol 17:507 – 519
Riedel A, Michel C, Gosselin M, LeBlanc B (2007) En ichmen
o nu ien s, exopolyme ic subs ances and mic oo ganisms in
newly o med sea ice on he Mackenzie Shel . Ma Ecol P og Se
342:55 – 67
Roscoe HK, B ooks B, Jackson AV, Smi h MH, Walke SJ, Obba d RW,
Wol EW (2011) F os l owe s in he labo a o y: G ow h,
cha ac e is ics, ae osol and he unde lying sea ice. J Geophys Res
116: D12301, doi:10.1029/20010JD 015144
Rösel A, Kaleschke L (2012) Excep ional mel pond occu ence
in he yea s 2007 and 2011 on he A c ic sea ice e ealed om
MODIS sa elli e da a. J Geophys Res-Oceans 117: C05018,
doi:10.1029/2011JC007869
Ròzanska M, Poulin M, Gosselin M (2008) P o is en apmen in newly
o med sea ice in he coas al A c ic Ocean. J Ma Sys 74:887 – 1901
Ròzanska M, Gosselin M, Poulin M, Wik o MP, Michel C (2009) In-
l uence o en i onemen al ac o s on he de elopmen o bo om
ice p o is communi ies du ing he win e -sp ing ansi ion. Ma
Ecol P og Se 386: 43 – 159
Ryan KG, Ralph PJ, McMinn A (2004) Acclima ion o An a c ic bo -
om-ice algal communi ies o lowe ed salini ies du ing mel ing.
Pola Biol 27:679 – 186
Rysgaa d S, Kühl M, Glud RN, Hansen JW (2001) Biomass, p oduc-
ion and ho izon al pa chiness o sea ice algae in a high-A c ic j o d
(Young Sound, NE G eenland). Ma Ecol P og Se 223:15 – 26
Rysgaa d S, Glud R (Eds.) (2007) Ca bon cyclin in A c ic ma ine
ecosys em: Case s udy Young Sound. Medd G ønland, Bioscience
58. Copenhagen, he Commision o Scien i i c Resea ch in
G eenland, 214 pp
Rysgaa d S, Glud RN, Sej MK, Bend sen J, Ch is ensen PB (2007)
Ino ganic ca bon anspo du ing sea-ice g ow h and de-
cay: A ca bon pump in pola seas. J Geophys Res 112:C03016,
doi:10.1029/2006JC003572
Rysgaa d S, A end K, F ede iksen M. e al. (2008) Nuuk Basic: The
Ma ine Basic P og am 2005-2006. In Kli gaa d AB, Rasch M and
Caning K (eds.), Nuuk Ecological Resea ch Ope a ions, 1s Annual
Repo 2008. Danish Pola Cen e , Copenhagen
Rysgaa d S, Bend sen J, Pe e sen LT, Ramlø J, Glud RN (2009) In-
c eased CO2 up ake due o sea ice g ow h and decay in he No dic
Seas. J Geophys Res 114:C09011, doi:10.1029/2008JC005088
Rysgaa d S, Bend sen J, Delille B, Dieckmann GS, Glud RN,
Kennedy H, Mo ensen J, Papadimi iou S, Thomas DN, Tison JL
(2011) Sea ice con ibu ion o he ai -sea CO2 exchange in he
A c ic and Sou he n Oceans. Tellus B 63:823 – 830, doi:10.1111/
J.1600-0889-2011.005471.x
Rysgaa d S, Glud RN, Lenne K, Coope M, Halden N, Leaky R,
Haw ho ne FC, Ba be D (2012) Ikai e c ys als in mel ing sea ice
– implica ions o pCO2 and pH le els in A c ic su ace wa e s. TC
6:1015 – 11035, doi:10.51947 cd-6-015-2012
Rysgaa d S, Wang F, Galley RJ, G imm R, No z D, Lemes M, Geil us NX,
Chaulk A, Ha e AA, C abeck O, Else BGT, Campbell K, Sø ensen LL,
Sie e s J, Papaky iakou T (2014) Tempo al dynamics o ikai e in ex-
pe imen al sea ice. TC 8:1469 – 11478, doi:10.5194/ c-8-1469-2014
Sabine CL, Feely RA, G ube N, Key RM, Lee K, Bullis e JL,
Wanninkho R, Wong CS, Wallace DWR, Tilb ook B, Mille o FJ, Peng
T-H, Kozy A, Ono T, Rios AF (2004) The oceanic sink o an h opo-
genic CO2. Science 305: 367 – 1371, doi:10.1126/science.1097403
Sakshaug E (2004) P ima y and seconda y p oduc ion in he A c ic seas.
In: S ein R, Macdonald RW (eds) The o ganic ca bon cycle in he A c ic
Ocean, 2nd Edi ion. Sp inge -Ve lag Be lin Heidelbe g. Pp: 57 – 181
Sej MK, K ause-Jensen D, Rysgaa d S, Sø ensen LL, Ch is ensen
PB, Glud RN (2011) Ai -sea l ux o CO2 in a c ic coas al wa e s in-
l uenced by glacial mel wa e and sea ice. Tellus B 63:815 – 1822,
doi:10.1111/j.1600-0889.2011.00540.x
46 PhD hesis by Do e Haubje g Søgaa d
in Ma ch. The wa e column in he Kapisigdli Bigh was
ully mixed and he salini y unde he sea ice a ied
be ween 33 and 33.5. The wa e dep h unde he sea ice
was 30–40 me e . Mo e de ails on he oceanog aphic
condi ions in he God ha
˚bs jo d a e p o ided by Mo ensen
e al. (2011).
Two sampling designs we e used:
1. Spa ial a iabili y was in es iga ed o e a pe pendic-
ula x–y ansec co e ing 0.07 km
2
ha was in es i-
ga ed on Ma ch 11 and Ap il 8, 2010 (See Fig. 1 o
loca ion o hese). The pa ame e s sampled in his pa
o he wo k we e CaCO
3
, dissol ed o ganic ca bon and
ni ogen (DOC, DON), ino ganic nu ien s (PO
43-
,
Si(OH)
4
,NO
2
-
,NO
3
-
, and NH
4
?
), empe a u e,
salini y, and snow and sea ice hickness.
2. The empo al de elopmen o CaCO
3
, TCO
2
, TA,
ino ganic nu ien s, DOC, DON, p ima y p oduc ion,
bac e ia p oduc ion, bulk salini y, snow and sea ice
hickness, and empe a u e was also in es iga ed a a
single loca ion in he s udy a ea (Fig. 1) e e y
2–3 weeks om Feb ua y o May 2010 (i.e. 17
Feb ua y, 10, 11, 12 and 15 Ma ch, 8 Ap il and 1
May).
Spa ial a iabili y
In bo h o he ansec su eys (i.e., Ma ch 11 and Ap il 8,
2010), 25 sea ice co es we e aken along a 266-m-long
pe pendicula x–y ansec . The co es we e collec ed a
dis ances o 0, 0.2, 0.4, 0.6, 0.8, 1, 3, 9, 20, 42, 64, 128, and
266 m in bo h x and y di ec ions. A each sampling poin , a
sea ice co e was collec ed using a MARK II co ing sys em
(Ko acs En e p ises L d) and an o e lying snow sample
was collec ed using a small sho el. The ai empe a u e
was measu ed 2 m abo e he snow, and e ical p ofiles o
empe a u es wi hin he ice we e measu ed using a cali-
b a ed he mome e (Tes o
Ò
). Sea ice empe a u e mea-
su emen s we e complemen ed by a cus om-buil s ing o
he mis o s ha we e ozen in o he sea ice om 11 Ma ch
un il 8 Ap il. The he mis o s ing da a we e eco ded
e e y 6 h a a spa ial esolu ion o 4 cm.
The e ie ed ice co es we e cu in o 12 cm sec ions
wi h a s ainless s eel saw and placed in plas ic con aine s
and anspo ed back o he labo a o y in da k he mo-
insula ed boxes. Sea ice and snow samples we e slowly
mel ed in he da k a 3 ±1°C, which ook be ween 2 and
3 days. We measu ed all he pa ame e s in all sea ice
sec ions. Howe e , we only epo da a om he op and
bo om sec ion as hey a e he mos impo an .
To de e mine he amoun o CaCO
3
wi hin he ice and
snow, be ween 300 and 500 ml o mel ed ice o snow was
di ided in o h ee subsamples and fil e ed (3 ±1°C)
h ough p e-combus ed (450 °C, 6 h) Wha man
Ò
GF/F
fil e s. The exac olume o he fil e ed mel wa e was
measu ed. The fil e s we e ans e ed o ubes (12 ml
Exe aine
Ò
) con aining 20 ll HgCl
2
(5 % w/ , sa u a ed
solu ion) o a oid mic obial ac i i y du ing s o age and
12 ml deionized wa e wi h a known TCO
2
concen a ion.
The ubes we e hen spiked wi h 300 ll o 8.5 %
Fig. 1 Map showing he sea ice a ea in Kapisigdli in SW G eenland and he empo al de elopmen and spa ial a iabili y sampling s a ions
Pola Biol (2013) 36:1761–1777 1763
123

47
PhD hesis by Do e Haubje g Søgaa d
phospho ic acid o con e CaCO
3
on he fil e s o CO
2
,
and a e coulome ic analysis (Johnson e al. 1993)o CO
2
he CaCO
3
concen a ion was calcula ed.
A each sampling occasion, samples om di e en ice
dep h ho izons we e inspec ed unde he mic oscope o
check o he p esence o mic oo ganisms wi h calcium
ca bona e ex e nal s uc u es such as coccoli hopho es and
o amini e s. These inspec ions showed ha no mic oo -
ganisms wi h calcium ca bona e ex e nal s uc u es we e
p esen in any o he samples.
The emaining mel wa e was fil e ed h ough 25-mm
Wha man
Ò
GD/X disposable sy inge fil e s (po e size
0.45 lm). A subsample o he fil a e was ans e ed o p e-
combus ed glass ials; 100 ll o 85 % phospho ic acid was
added; and he ials we e ozen o la e analyses o DOC.
The emaining fil a e was ans e ed o p e-combus ed
(450 °C, 6 h) and alkali-washed glass ials and ozen o
la e analysis o DON, PO
43-
,NO
3
-
,NO
2
-
, Si(OH)
4
, and
NH
4
?
. The DOC, DON, and nu ien samples we e ozen a
-19 °C un il analysis. DOC was measu ed by high- em-
pe a u e ca aly ic oxida ion, using a MQ 1001 TOC Analyze
(Qian and Moppe 1996). The concen a ions o NO
3
,NO
2
-
,
PO
43-
, and Si(OH)
4
we e de e mined by s anda d colo i-
me ic me hods (G assho e al. 1983) as adap ed o flow
injec ion analysis (FIA) on a LACHAT Ins umen s Quick-
Chem 8000 au oanalyze (Hales e al. 2004). The PO
43-
samples om he fi s sampling was con amina ed, and
he e o e we did no include hem. The concen a ion o
NH
4
?
was de e mined wi h he fluo ome ic me hod o
Holmes e al. (1999) using a HITACHI F2000 fluo escence
spec opho ome e . The concen a ion o DON was de e -
mined by sub ac ion o he concen a ion o DIN
(DIN =[NO
3
-
]?[NO
2
-
]?[NH
4
?
]) om ha o he
o al dissol ed ni ogen de e mined by FIA on he LACHAT
au oanalyze , using online pe oxodisulpha e oxida ion cou-
pled wi h UV adia ion a pH 9.0 and 100 °C (K oon 1993).
The conduc i i y o he mel ed sea ice sec ions was
measu ed (The mo O ion 3-s a wi h an O ion 013610MD
conduc i i y cell), and alues we e con e ed o bulk
salini y (G assho e al. 1983). The b ine olumes o he
o iginal sea ice samples we e calcula ed om he measu ed
bulk salini y and empe a u e and a fixed densi y o
0.917 g cm
-3
acco ding o Leppa
¨ an a and Manninen
(1988) o empe a u es[-2°C and acco ding o Cox and
Weeks (1983) o empe a u es -2°C.
Spa ial au oco ela ion (Legend e and Legend e 1998)
was used o analyze he co ela ion o he ho izon al and
e ical dis ibu ion o CaCO
3
concen a ion, DOC, DON,
ino ganic nu ien s, empe a u e, and salini y as well as he
snow and sea ice hickness. Au oco ela ion was es ima ed
by Mo an’s I coe ficien s (Mo an 1950; Legend e and
Legend e 1998). This coe ficien was calcula ed o each o
he ollowing in e als along he ansec (classes o
dis ance): 0–0.25, 0.25–0.50, 0.50–1.5, 1.5–2.5, 2.5–5.0,
5.0–10, 10–50, 100–150, 150–200, 200–250, and [250 m.
The au oco ela ion coe ficien s es ima ed by he Mo an’s I
coe ficien we e es ed o significance acco ding o he
me hod desc ibed in Legend e and Legend e (1998). A
2- ailed es o significance was used. Posi i e (?) indica es
posi i e au oco ela ion (co ela ion) and nega i e (-)
indica es nega i e au oco ela ion. A ze o (0) alue indi-
ca es a andom spa ial pa e n. We applied a significance
le el o P 0.05. Pea son’s co ela ion was used o find
he co ela ion be ween he pa ame e s. Fu he mo e, a ull
ac o ial gene alized linea model (GLM) including ime,
dep h, and posi ion as explana o y a iables, which was
educed based on Akaike’s In o ma ion C i e ion (AIC),
was applied. The same model was applied o se e al
dependen a iables: posi ion (ho izon al), dep h ( e ical),
and ime on CaCO
3
concen a ion, bulk salini y, DOC, and
DON.
Tempo al de elopmen
On each o he 7 sampling occasions o he empo al
s udy, iplica e ice co es and en i onmen al pa ame e s
we e collec ed om a defined a ea (5 m
2
), and samples
we e p ocessed as desc ibed abo e.
P ima y p oduc ion was measu ed (Søgaa d e al. 2010)
on 4 occasions (i.e.17 Feb ua y, 11 Ma ch, 8 Ap il, and 1
May). In sho , p ima y p oduc ion was de e mined on
mel ed sea ice samples (mel ed wi hin 48 h in he da k a
3±1°C). The po en ial p ima y p oduc ion in he sea ice
a di e en sea ice dep hs (i.e. 12 cm sec ions) was mea-
su ed in he labo a o y cold oom a 3 i adiances (72,
50,14 lmol pho ons m
-2
s
-1
) and co ec ed wi h one da k
incuba ion, using he H
14
CO
3
-
incuba ion echnique
(incuba ion ime was 5 h). The po en ial p ima y p oduc-
ion measu ed in he labo a o y a di e en sea ice dep hs
was plo ed agains he h ee labo a o y ligh in ensi ies 42,
21, and 9 lmol pho on m
-2
s
-1
and fi ed o he ollowing
unc ion desc ibed by Pla e al. (1980)
PPðlgCl
1h1Þ¼Pm1exp aEPAR
Pm

ð1Þ
whe e PP is he p ima y p oduc ion, P
m
(lgCl
-1
h
-1
)is
he maximum pho osyn he ic a e a ligh sa u a ion, a
(lgCm
2
slmol pho ons
-1
l
-1
h
-1
) is he ini ial slope o
he ligh cu e, and E
PAR
(lmol pho ons m
-2
s
-1
) is he
labo a o y i adiance. The pho oadap a ion index, E
k
(lmol pho ons m
-2
s
-1
), was calcula ed as P
m
/a.
In si u down-welling i adiance was measu ed a g ound
le el (Kipp & Zonen py ome e , CM21, spec um ange o
305–2,800 nm) once e e y 5 min, and hou ly a e ages
we e p o ided by Asiaq (G eenland Su ey). Hou ly down-
welling i adiance was con e ed in o hou ly
1764 Pola Biol (2013) 36:1761–1777
123
48 PhD hesis by Do e Haubje g Søgaa d
pho osyn he ically ac i e adia ion (PAR; ligh spec um
300–700 nm) a e in e calib a ion (R
2
=0.99, P 0.001,
n=133) wi h a Li-Co quan um 2 pi senso connec ed o a
LI-1400 da a logge (Li-Co Biosciences
Ò
). The in si u
hou ly PAR i adiance was calcula ed a di e en dep hs,
depending on sea ice and snow hickness, using he a en-
ua ion coe ficien s measu ed du ing he sea ice season.
In si u p ima y p oduc ion was calcula ed o each hou
a di e en sea ice dep hs using hou ly in si u PAR i adi-
ance (Eq. 1). To al daily (24 h) in si u p ima y p oduc ion
was calcula ed as he sum o hou ly in si u p ima y p o-
duc ion o each dep h. The dep h-in eg a ed ne p ima y
p oduc ion was calcula ed using apezoid in eg a ion.
Ligh a enua ion o he sea ice samples was measu ed
wi h a Li-Co quan um 2 pi senso connec ed o a LI-1400
da a logge (Li-Co Biosciences
Ò
) in a da k, empe a u e-
egula ed oom (a in si u empe a u es o a oid b ine loss)
using a fibe lamp wi h a spec um close o na u al sunligh
(15 V, 150 W, fibe -op ic ungs en–halogen bulb). The
senso was placed unde he sea ice sec ion and he fibe
lamp was placed abo e. Ligh a enua ion was measu ed in
his way o each sea ice sec ion. We assume dep h-inde-
penden a enua ion in he sea ice. To measu e ligh
a enua ion o he snow co e , we gen ly emo ed he snow
and placed he senso on he ice su ace and hen we placed
he snow on op o he senso . Thus, down-welling i adi-
ance was measu ed di ec ly abo e and below he snow
(Søgaa d e al. 2010).
The bac e ial p oduc ion p ocedu es employed (i.e., 17
Feb ua y, 11 Ma ch, 8 Ap il, and 1 May) ha e been
desc ibed by Søgaa d e al. (2010), excep hose be ween
13 and 16 Ma ch, when he measu emen s we e made using
an ice-c ushing me hod desc ibed by Kaa okallio (2004)
and Kaa okallio e al. (2007). The wo me hods used o
bac e ial p oduc ion measu emen s yield compa able
esul s ( he mean alues om Ma ch using he ice-c ushing
me hod we e 2.4 lgCl
-1
day
-1
and he mean alues
om Ap il using he mel ing sea ice app oach we e
2.5 lgCl
-1
day
-1
).
Bac e ial p oduc ion in mel ed sea ice samples was
de e mined by measu ing he inco po a ion o [
3
H] hy-
midine in o DNA. T iplica e samples ( olume =0.01 L)
we e incuba ed in da kness a 3 ±1°C wi h 10 nmol l
-1
o labeled [
3
H] hymidine (New England Nuclea
Ò
, specific
ac i i y 10.1 Ci mmol
-1
). T ichlo oace ic acid (TCA)-
ea ed con ols we e made o measu e he abio ic adso p-
ion. A he end o incuba ion pe iod (T=6 h), 1 ml o
50 % cold TCA was added o all he samples. The samples
we e fil e ed and coun ed using a liquid scin illa ion ana-
lyze (T icCa b 2800, Pe kinElme
Ò
).
Fo he ice-c ushing me hod, samples we e p epa ed by
c ushing each in ac 5 o 10 cm ice co e sec ion, fi s using
a spike ool, and hen g inding ice chunks in an elec ical
ice cube c ushe . App oxima ely 10 ml o c ushed ice was
placed in a scin illa ion ial and weighed. To ensu e e en
dis ibu ion o labeled subs a e, 2–4 ml o s e ile-fil e ed
(0.2 lm minisa fil e s, Sa o ius
Ò
) seawa e was added o
he scin illa ion ials. All ice-p ocessing wo k was done in
a cold on-deck labo a o y a nea -ze o empe a u e. Two
aliquo s and a o maldehyde-killed abso p ion blank we e
amended wi h [me hyl-3H] hymidine (New England
Nuclea
Ò
; specific ac i i y 20 Ci mmol
-1
). Concen a ions
o 20 nmol l
-1
o hymidine we e used o all samples.
Samples we e incuba ed in he da k a -0.2 °C in a sea-
wa e /ice-c ush ba h o 17–18 h and incuba ion s opped
wi h he addi ion o 200 ll o 37 % s e ile-fil e ed o m-
aldehyde. Samples we e p ocessed using s anda d cold-
TCA ex ac ion and fil a ion p ocedu e (using Ad an ec
Ò
MFS 0.2 lm MCE fil e s). A Wallac Win-Spec al 1414
coun e (Pe kinElme
Ò
) and Ins aGel (Pe kinElme
Ò
)
cock ail we e used o scin illa ion coun ing.
Fo bo h me hods, he bac e ial ca bon p oduc ion was
calcula ed, using he con e sion ac o s p esen ed in Smi h
and Clemen (1990).
Bac e ial ca bon demand (BCD) o g ow h was calcu-
la ed as:
BCDðlgCl
1h1Þ¼ BP
BGE ð2Þ
whe e BP is he bac e ia p oduc ion and BGE is a bac e ial
g ow h e ficiency es ima e o 0.5 measu ed in pola oceans
(Ri kin and Legend e 2001).
To in es iga e he empo al dis ibu ion o TCO
2
and
TA, an addi ional sea ice co e was collec ed on each
sampling occasion. The co e was cu in o 12-cm sec ions
and placed in lamina ed anspa en NEN/PE plas ic bags
(Hansen e al. 2000) fi ed wi h a gas- igh Tygon ube and
a al e o sampling. These sec ions we e b ough back o
he labo a o y cold oom (3 ±1°C). Cold (1 °C) deion-
ized wa e o known weigh and TA and TCO
2
concen-
a ion was added (10–30 ml) o each NEN/PE bag
(Hansen e al. 2000). The bags we e closed, and excess ai
quickly ex ac ed h ough he al e. Then, he ice was
mel ed ( 48 h) in he da k. Gas bubbles eleased om he
mel ing sea ice we e ans e ed o ubes (12 ml Exe aine
Ò
). Sea ice mel wa e was likewise ans e ed o simila
ubes con aining 20 ll HgCl
2
(5 % w/ sa u a ed solu ion;
Rysgaa d and Glud 2004). S anda d me hods o analysis
we e used: TCO
2
concen a ions we e measu ed on a
coulome e , TA by po en iome ic i a ion (Ha aldsson
e al. 1997), and gaseous O
2
,CO
2
,N
2
by gas ch oma og-
aphy (SRI 8610C; FID/TCD de ec o ; Lee e al. 2005).
A Wilcoxon ank sum es was used o es whe he
CaCO
3
concen a ion was significan ly di e en ly dis ib-
u ed wi hin he sea ice. We applied a significance le el o
95 %.
Pola Biol (2013) 36:1761–1777 1765
123
49
PhD hesis by Do e Haubje g Søgaa d
Following he bulk de e mina ion o TCO
2
and TA, he
bulk pCO
2
and pH (on he o al scale) we e compu ed using
he empe a u e and salini y condi ions in he field and a
s anda d se o ca bona e sys em equa ions (See Rysgaa d
e al. 2013), excluding nu ien s, wi h he CO2SYS p og am
o Lewis and Wallace (2012). We used he equilib ium
cons an s o Meh bach e al. (1973), efi ed by Dickson and
Mille o (1987,1989). We assumed a conse a i e beha io
o CO
2
dissocia ion cons an s a subze o empe a u es since
Ma ion (2001) and Delille e al. (2007) sugges ed ha a
he modynamic cons an ele an o he ca bona e sys em
can be assumed o be alid a subze o empe a u es.
Resul s
Figu e 1shows a map o he in es iga ed sea ice a ea wi h
he di e en sampling s a ions in Kapisigdli , SW G een-
land (Fig. 1).
The ai empe a u e du ing he s udy pe iod anged om
-17 °C in Feb ua y o ?16 °C in May jus be o e he sea
ice b eak-up (Fig. 2). Tempe a u es wi hin he snow and
sea ice a ied om -6.0 ±0.1 o 0 ±0.02 °C, wi h
minimum empe a u es measu ed in Feb ua y and Ma ch,
ollowed by a g adual inc ease o maximum alues in la e
Ap il (Fig. 3a).
The bulk salini y o he sea ice samples a ied om 1.7
o 6 (Fig. 3b). The b ine olume a ied om 5 o 32 % a
he op o he sea ice and om 12 o 40 % a he bo om o
he sea ice (Fig. 3c), an indica ion ha he ice was pe -
meable o mos o he s udy (Golden e al. 1998). In Ap il,
when ai empe a u es a ied be ween -2 and ?16 °C, he
ice began o mel , which esul ed in high ela i e b ine
olumes and low bulk salini ies (Fig. 3b, c).
Spa ial a iabili y
Mo an’s I (Table 1) was used o es ima e he spa ial
au oco ela ion wi hin he da ase s collec ed o he wo
ansec samplings in Ma ch and Ap il. All pa ame e s had
a andom spa ial dis ibu ion pa e n wi h no appa en
pa ches, indica ing ha he dis ibu ion o he pa ame e s
in es iga ed was highly he e ogeneous on he scale o
me e s o hund eds o me e s (Table 1).
Despi e his he e ogenei y, he e was e idence o a
co ela ion be ween se e al pa ame e s: CaCO
3
had sig-
nifican co ela ion wi h PO
43-
, bulk salini y, Si(OH)
4
and
NO
3
-
. CaCO
3
and DON we e significan ly co ela ed
(nega i e) only du ing he second sampling. The e was no
-20
-15
-10
-5
0
5
10
15
20
17/02 03/03 17/03 31/03 14/04 28/04 12/05
Ai empe a u e (
o
C)
Ai empe a u e
Fig. 2 Me eo ological da a on ai empe a u e (°C) a he Asiaq
me eo ological s a ion in Kapisigdli
(a)
Sea ice and snow empe a u e [ºC]
-20
0
20
-40
-60
-80
Sea ice and snow dep h [cm]
(c)
Rela i e b ine olume [%]
-20
0
20
-40
-60
-80
-20
0
20
-40
-60
-80
Feb Ma ch A
p
il Ma
y
30 30
30
40
20
20
20
10
10
5
-1.2
-0.8
-3
-4
-2.8
-2.8
-2.0
-2.0
-7.0
-5.0
-5.0
-1.2
-4
-2.0
-6.0 -6.0
(b)
Bulk salini y
5
4 6
4
3
3
Fig. 3 Tempo al de elopmen in (a) sea ice and snow empe a u e
[°C] N.B. The empe a u e da a collec ed om e ie ed ice co es a e
supplemen ed by he mis o s ing da a om 11 Ma ch un il 8 Ap il,
(b) bulk salini y, (c) ela i e b ine olume ac ion [%]. The black
do s ep esen iplica e measu emen s
1766 Pola Biol (2013) 36:1761–1777
123
50 PhD hesis by Do e Haubje g Søgaa d
co ela ion be ween CaCO
3
and DOC, NH
4
?
,NO
2
-
, and
empe a u e (Table 2).
A GLM whe e dep h ( e ical), ime, and posi ion
(ho izon al) as explana o y a iables is used o es whe he
he e was a significan e ec on se e al dependen a i-
ables: CaCO
3
, bulk salini y, DOC, and DON. The GLM
es was applied on he da ase s collec ed o he wo
ansec samplings in Ma ch and Ap il. The e was a sig-
nifican e ec o dep h o all he pa ame e s. The signifi-
can e ec o dep h was la gely dependen on he ime o
sampling o CaCO
3
(F
5,261
=15.9784, P 0.001), bulk
salini y (F
5,248
=5.6322 P 0.001), and DON (F
5,239
=
3.6861 P 0.001) (da a no shown). Fu he mo e, when
he wo s udies in Ma ch and Ap il we e compa ed, a
significan e ec o ime was ound o bulk salini y, DOC,
and DON (Table 3). No significan e ec o posi ion
(ho izon al) was ound o CaCO
3
, bulk salini y, and DON
(Table 3).
Tempo al dis ibu ion
In he empo al su ey, he e we e no e ical di e ences
in TCO
2
and TA (Fig. 4), al hough he concen a ions o
TCO
2
and TA dec eased wi h ime (Fig. 4). The highes
TCO
2
and TA concen a ions we e measu ed in Feb ua y
(293 ±6.1 and 410 ±40 lmol kg
-1
in mel ed sea ice),
which dec eased o 181 ±1.3 and 190 ±1.2 lmol kg
-1
in mel ed sea ice in he beginning o May, espec i ely
(Fig. 4). Howe e , high concen a ions o TA and TCO
2
we e also measu ed in Ap il (Fig. 4). No small-scale
a iabili y was obse ed o TCO
2
and TA
concen a ions.
Table 1 Mo an’s I as a unc ion o dis ance class (m) be ween si es
o CaCO
3
, bulk salini y, DOC, DON, NH
4
?
,NO
2
-
,NO
3
-
,
Si(OH)
4
and PO
43-
in snow (S), op sea ice (T) and bo om sea ice
(B). Fi s sampling pe iod was on Ma ch 11 and second sampling
pe iod was on Ap il 8, 2010. Posi i e (?) indica es posi i e
au oco ela ion (co ela ion) o nega i e (-) nega i e au oco ela ion
(dispe sion). A ze o alue indica es a andom spa ial pa e n
Fi s sampling
Dis ance class (m) CaCO
3
Bulk salini y DOC DON NH
4
?
NO
2
-
NO
3
-
Si(OH)
4
S TBS T B STBS T BS TBS T BS T BS T B
0.25 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0.5 0 0 0 0 -0 000000000000000000
1.0 0 0 0 0 --0000 -00000000---00
2.5 0 0 0 -00000??0-00-????00-0
5.0 0 0 0 0 -00000-0000000000-00
10 0 0 0 -0 0 000000000000000000
50 0000 0 0 000000-000 -0000000
100 0 0 0 0 0 0 0 0 0 -0-000000000000
200 0 0 0 0 0 0 0 0 0 -000000000000-0
250 -000 ?0000-00?000 0 0 -00000
[250 -000 0 0 000000000000000000
Second sampling
Dis ance class (m) CaCO
3
Bulk salini y DOC DON NH
4
?NO
2
-
NO
3
-
Si(OH)
4
PO
43-
STBS T B STBSTBSTBSTBSTBSTBSTB
0.25 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -000000000
0.5 0000 0 0 0-00-0000000000000000
1.0 0000 -0 0000000000--00-000000
2.5 ?-0???0?0-00-0?00 0 ?00000000
5.0 0000 0 0 00000000000--00-00?00
10 0000 0 0 0000000000?0000000000
50 0000 0 0 0000000000-0000000000
100 0000 0 0 0-00000-00000000000-0
200 0000 0 0 000000000000000000000
250 0 0 -0000?000000000?0--000000
[250 0000 0 0 000000000000000000000
Pola Biol (2013) 36:1761–1777 1767
123
51
PhD hesis by Do e Haubje g Søgaa d
In he wa e column, he highes TCO
2
concen a ion, o
2,101 ±7.7 lmol kg
-1
, was measu ed in Feb ua y, which
dec eased o 2,085 ±27 lmol kg
-1
in May (Fig. 4). The
highes TA concen a ion o 2,253 ±2.5 lmol kg
-1
was
measu ed in May and he lowes o 2,220 ±2.2 lmol kg
-1
in Ma ch (Fig. 4).
The a e age TA:TCO
2
a io wi hin he sea ice was [1
du ing Feb ua y (a e age 1.25) and Ma ch (a e age 1.20),
and highe han ha in he wa e column (Fig. 4). The
highes TA:TCO
2
a io (1.75) was calcula ed o he
uppe mos ho izons o he sea ice in Ap il. In Ap il, he
a e age a io was 1.32, while he alue in he beginning o
May was 1.1.
The CaCO
3
concen a ions a ied e ically wi hin he ice
co es: in Feb ua y he highes concen a ion o
2.4 ±0.4 lmol CaCO
3
l
-1
was measu ed in he uppe mos
laye o he sea ice (Wilcoxon ank sum es ; P 0.05;
Fig. 5). Con e sely, jus be o e ice b eak-up in ea ly May,
he highes CaCO
3
concen a ion o 4.4 ±0.1 lmol Ca-
CO
3
l
-1
was measu ed in he lowe mos ice ho izon (Wil-
coxon ank sum es ; P 0.05; Fig. 5). In Ma ch and Ap il,
CaCO
3
was e enly dis ibu ed wi hin he sea ice wi h an
a e age concen a ion o 1.8 ±0.40 lmol CaCO
3
l
-1
in
Ma ch and 2.0 ±0.30 lmol CaCO
3
l
-1
in Ap il (Wilcoxon
ank sum es ; P[0.05). No small-scale a iabili y was
obse ed o CaCO
3
concen a ion in he sea ice (Fig. 5).
CaCO
3
concen a ions in he snow dec eased h oughou he
sea ice season: om 5.2 ±1.5 lmol CaCO
3
l
-1
in Feb ua y
o 3.0 ±1.5 lmol CaCO
3
l
-1
in Ap il (Fig. 5). Sedimen
aps we e deployed unde he sea ice du ing he s udy pe iod
bu no CaCO
3
c ys als we e ound (da a no shown).
The DOC concen a ions inc eased o e ime wi h a
maximum concen a ion o 160 lmol l
-1
being measu ed
in May in he bo ommos laye o he ice (Fig. 6). DON
concen a ions did no a y e ically wi hin he sea ice
du ing win e . Howe e , in Ap il and May, when he sea
ice began o mel , he DON concen a ions inc eased, and
maximum DON alues o 15 lmol l
-1
we e measu ed in
he bo ommos pa o he sea ice (Fig. 6). The DOC:DON
a ios anged om 5 o 20 (a e age 12).
The highes olume-specific p ima y p oduc ion (bulk)
o 25 lgCl
-1
day
-1
was measu ed in Ma ch in he bo -
om o he sea ice (Fig. 6). Bac e ial ca bon demand a ied
e ically wi hin he sea ice, wi h maximum alues mea-
su ed in he op and bo om sec ion o he sea ice in Ma ch
and May (Fig. 6). In Feb ua y and Ap il, he maximum
BCD was measu ed in he in e nal sea ice laye s. The
highes BCD o 5.8 lgCl
-1
day
-1
was es ima ed in he
bo ommos sec ions o he sea ice in Ma ch.
Bulk nu ien concen a ions o each sampling da e
we e plo ed as a unc ion o bulk salini y and compa ed
wi h he expec ed dilu ion line (Cla ke and Ackley 1984).
To calcula e he dilu ion line, we used he a e age nu ien
concen a ion and salini y measu ed a 17 Feb ua y, 11
Table 2 Pai wise compa ison o CaCO
3
, bulk salini y, empe a u e,
DOC, DON, NH
4
?
,NO
2
-
,NO
3
-
, Si(OH)
4
and PO
43-
.(-) indica es
significan ly nega i e (-) and (?) indica es significan ly posi i e (?)
co ela ed a iables (Pea sons) a a 5 % significance le el. Fi s
sampling pe iod was in Ma ch and second sampling pe iod in Ap il
2010
12
bulk salini y
12
empe a u e
12
DOC
12
DON
12
NH
4+
12
NO
2-
12
NO
3-
12
Si(OH)
4
12
PO
43-
CaCO
3
bulk salini y - -
empe a u e · · · ·
DOC · · · · · ·
DON · - · - · · · ·
·· · - · · ·-··
·· · · + · · ····+
++ - - + · · - · -++·+
-- · + · · · · ·+--··- -
-- · · · · · · · --+··-+· ·
NH
4+
NO
2-
NO
3-
Si(OH)
4
PO
43-
CaCO
3
Table 3 Resul s om a GLM model whe e dep h, ime, and posi ion as explana o y a iables we e used o es whe he he e was a significan
e ec on se e al dependen a iables: CaCO
3
, bulk salini y, DOC, and DON
Va iable Posi ion Time Dep h
CaCO
3
F
24,261
=1.2685 P=0.19 F
1,261
=2.6087 P=0.11 F
5,261
=53.7673 P 0.001
Bulk salini y F
24,248
=0.8054 P=0.73 F
1,248
=18.4728 P 0.001 F
5,248
=14.2994 P 0.001
DOC F
24,217
=2.13 P 0.001 F
1,217
=224.6497 P 0.001 F
5,217
=3.1986 P 0.001
DON F
24,239
=0.5919 P=0.94 F
1,239
=162.4025 P 0.001 F
5,239
=6.6925 P 0.001
1768 Pola Biol (2013) 36:1761–1777
123

52 PhD hesis by Do e Haubje g Søgaa d
Ma ch, 8 Ap il, and 1 May in he wa e column (i.e.
0–10 m; PO
43-
=0.94 lmol l
-1
, Si(OH)
4
=7.0 l-
mol l
-1
,NO
2
-
?NO
3
-
=8.6 lmol l
-1
,NH
4
?
=0.28
lmol l
-1
, DOC =62.6 lmol l
-1
, DON =1.2 lmol l
-1
and a a e age salini y o 33). I alues a e below he line,
deple ion o nu ien s has aken place, and i abo e he
dilu ion line p oduc ion, o ne -deposi ion, o he solu e has
occu ed. Plo s o salini y e sus PO
43-
, Si(OH)
4
,NO
2
-
?
NO
3
-
,NH
4
?
, DOC, and DON in sea ice we e gene ally all
abo e he dilu ion line implying accumula ion o nu ien s
and o ganic ma e wi hin he ice (Fig. 7a– ). Howe e ,
deple ion o PO
43-
was obse ed in Feb ua y and Ma ch.
Fu he mo e, NO
2
-
?NO
3
-
was deple ed in Ap il and
May.
The e was a nega i e co ela ion be ween PO
43-
and
CaCO
3
and Si(OH)
4
and CaCO
3
, while a posi i e co ela-
ion was obse ed be ween NO
3
-
and CaCO
3
(Table 2).
0 500 1000 1500 2000 2500
1,0 1,2 1,4 1,6 1,8
0 500 1000 1500 2000 2500
1,0 1,2 1,4 1,6 1,8
TCO
2
and TA (μmol kg
-1
)
0
20
40
60
Wa e
Feb ua y Ma ch
Ta:TCO
2
a io Ta:TCO
2
a io
Ap il
Ta:TCO
2
a io Ta:TCO
2
a io
May
TCO
2
and TA (μmol kg
-1
)
0 500 1000 1500 2000 2500
1,0 1,2 1,4 1,6 1,8
Sea ice and wa e dep h (cm)
0 500 1000 1500 2000 2500
0,8 1,0 1,2 1,4 1,6 1,8
0
20
40
60
Wa e
0,8
0
20
40
60
Wa e
0,8
0
20
40
60
Wa e
0,8
Sea ice and wa e dep h (cm)
Fig. 4 Tempo al de elopmen o he e ical concen a ion p ofiles
o TCO
2
(black ba s) and TA (g ay ba s) and he TA:TCO
2
a io
(ci cles) in bulk mel ed sea ice du ing he 2010 sea ice season. No e
ha TCO
2
and TA below 60 cm a e wa e column alues. Ho izon al
do ed line ep esen s he sea ice–wa e column in e ace. Da a poin s
ep esen ea men mean ±SE (n=3)
μmol CaCO3 l-1
0246810
Snow and sea ice dep h (cm)
0
-20
-40
-60
Feb ua y
Ma ch
Ap il
May
Snow
Fig. 5 Tempo al de elopmen o he CaCO
3
concen a ion [lmol
CaCO
3
l
-1
] in bulk sea ice and snow in Feb ua y, Ma ch, Ap il, and
May 2010
Pola Biol (2013) 36:1761–1777 1769
123
53
PhD hesis by Do e Haubje g Søgaa d
Figu e 8shows nTA and nTCO
2
(TA and TCO
2
alue
being no malized o a salini y o 33 o emo e co ela ion
o salini y) ela ionships in seawa e samples and bulk ice
samples. The di e en lines ep esen he heo e ical
e ec s o p ecipi a ion–dissolu ion o CaCO
3
,CO
2
elease–up ake and pho osyn hesis– espi a ion on he a io
nTCO
2
:nTA. The p ecipi a ion o CaCO
3
dec eases bo h
TCO
2
and TA in a a io o 2:1. An exchange o CO
2
has no
impac on TA, while TCO
2
will be a ec ed. Biological
ac i i y has an almos negligible e ec on TA, wi h a a io
TA:TCO
2
=-0.16 (Zeebe and Wol -Glad ow 2001).
Conside ing ha sea ice was o med om seawa e wi h a
known TA:TCO
2
a io (Fig. 4), we a e able o deciphe
which p ocess ook place in he ice: in Feb ua y and
Ma ch, he sea ice samples we e aligned on he heo e ical
line o CaCO
3
p ecipi a ion (Fig. 8). In Ap il and May, he
ice samples we e well aligned (slope 0.75; R
2
=0.86)
be ween he heo e ical end o CaCO
3
p ecipi a ion/dis-
solu ion and he one o CO
2
elease/up ake (Fig. 8). In
Ap il wo sea ice samples we e aligned on he heo e ical
line o CaCO
3
dissolu ion. This implies ha bo h CaCO
3
p ecipi a ion/dissolu ion and CO
2
elease/up ake had
occu ed in he ice.
Discussion
Biological ac i i y
A maximum BCD o 5.8 lgCl
-1
day
-1
(Fig. 6) was
es ima ed, which is low compa ed wi h maximum a es o
27 lgCl
-1
day
-1
es ima ed in he neighbo ing jo d
Malene Bigh in Ap il 2008 (60 km o he SW om he
p esen s udy si e; Søgaa d e al. 2010).
Pai wise co ela ions be ween p ima y p oduc ion and
BCD e ealed significan posi i e ela ionships. Fu he -
mo e, accumula ion o DOC was obse ed (Fig. 6e– ) as
well as DOC/DON a ios anging om 5 o 20 indica ing a
p obable p oduc ion o ca bon- ich ex acellula polyme ic
subs ances (EPS) by he sea ice algae and bac e ia
(Unde wood e al. 2010; K embs e al. 2011; Juhl e al.
2011). Co ela ion be ween p ima y p oduc ion and BCD,
high DOC/DON a io and accumula ion o DOC poin s o
he e being a low-quali y subs a e esul ing in a low bac-
e ia p oduc ion. P e ious s udies ha e shown ha EPS is a
low-quali y subs a e o he e o ophic bac e ia (Pome oy
and William 2001), which migh explain he low BCD and
he obse ed DOC accumula ion (Fig. 6). Howe e , se e al
DON (μmol l-1)
Bac e ial ca bon demand (μg C l-1
d-1 )
P ima y p oduc ion (μg C l-1
d-1 )
0
10
20
30
40
50
60
0
10
20
30
40
50
60
DOC (μmol l-1)
Sea ice dep n (cm)
Sea ice dep n (cm)
0
10
20
30
40
50
60
810121416
20 40 60 80 100 120 140 160 180 0246
0123456
0 5 10 15 20 25 30
0
10
20
30
40
50
60
Feb ua y
Ma ch
Ap il
May
Feb ua y
Ma ch
Ap il
May
Ma ch
Ap il
May
Ma ch
Ap il
May
Fig. 6 Ve ical p ofiles o DOC, DON, p ima y p oduc ion, and bac e ial ca bon demand in bulk sea ice in Feb ua y o May. N.B. he e we e no
da a a ailable on he e ical p ofiles o DOC and DON in Feb ua y
1770 Pola Biol (2013) 36:1761–1777
123
54 PhD hesis by Do e Haubje g Søgaa d
s udies sugges he opposi e ha EPS se e as high-quali y
subs a e o bac e ia (e.g. Junge e al. 2004; Meine s e al.
2008).
Ano he explana ion o he low BCD migh be he alue
o bac e ial g ow h e ficiency used. We used a bac e ial
g ow h e ficiency o 0.50 (Ri kin and Legend e 2001).
Howe e , g ow h e ficiency is an in e se unc ion o
empe a u e, and small changes in empe a u e would
influence g ow h e ficiency (*2.5 % dec ease in g ow h
e ficiency pe 1 °C inc ease) and he eby he BCD (Ri kin
and Legend e 2001). Using a lowe g ow h e ficiency
( 0.15; e.g. Middelboe e al. 2012), he seasonal ne
au o ophic sea ice would change o a ne he e o ophic sea
ice, which compa es o esul ound by Long e al. (2011)
in Kapisigdli Bigh in Ma ch 2010. The eby he biological
ac i i y would no con ibu e o he a mosphe ic CO
2
up ake a all. Howe e , we belie e he use o a g ow h
e ficiency o 0.50 o be mos alid since i also ag ees wi h
he g ow h e ficiency o 0.41 measu ed by Kupa inen e al.
(2011) and Del Gio gio and Cole (1998).
The highes es ima es o p ima y p oduc ion (Fig. 6)
we e measu ed in he bo om ice ho izons in Ma ch.
A e age a es o sea ice algal p ima y p oduc ion du ing
Ma ch (4.03 mg C m
-2
day
-1
) a e a he lowe end o he
llomμ(noi a necnoc nei uN
-1
)
0
2
4
6
8
10
0
1
2
3
4
0
20
40
60
80
100
120
140
llomμ(noi a necnoc nei uN
-1
)
(a) PO43-
llomμ(noi a necnoc nei uN
-1
)
(b) Si(OH)4
0
2
4
6
8(c) NO2- + NO3-
0,0
0,5
1,0
1,5
2,0
2,5
3,0 (d) NH4+
(e) DOC
0246810
0246810
0246810
02468100246810
0246810
0
2
4
6
8
10
12
14
16
Bulk salini y
Bulk salini y
( ) DON
Feb ua y
Ma ch
Ap il
May
Feb ua y
Ma ch
Ap il
May
Feb ua y
Ma ch
Ap il
May
Feb ua y
Ma ch
Ap il
May
Feb ua y
Ma ch
Ap il
May
Feb ua y
Ma ch
Ap il
May
Fig. 7 Tempo al de elopmen o (a) PO
43-
,(b) Si(OH)
4
,(c)
NO
2
-
?NO
3
-
,(d)NH
4
?
,e) DOC, and ) DON concen a ions
e sus bulk salini y in sea ice om Feb ua y o May. The solid line
indica es he expec ed dilu ion line p edic ed om salini y and
nu ien concen a ions in seawa e (0–10 m, a e age salini y o 33
unde he sea ice)
Pola Biol (2013) 36:1761–1777 1771
123
55
PhD hesis by Do e Haubje g Søgaa d
scale o alues epo ed om he A c ic (0.2–463
mg C m
-2
day
-1
; A igo e al. 2010 and e e ences
he ein). A dec ease in TCO
2
was measu ed in he bo -
ommos ice a he ime o he algae g ow h, sugges ing ha
p ima y p oduc ion was esponsible o he dec ease in
TCO
2
(Figs. 4,6). Howe e , he a e age ne biological
p oduc ion in he bo om o he sea ice in Ma ch was only
1.6 lmol C day
-1
(Fig. 6) and he a e age TCO
2
loss in
he bo om o he sea ice in Ma ch was 6.4 lmol day
-1
,
indica ing ha p ocesses o he han p ima y p oduc ion
influence he ino ganic ca bon dynamics in he sea ice.
TCO
2
in sea ice is con olled by p ima y p oduc ion and
espi a ion by bo h au o ophic and he e o ophic o gan-
isms, CaCO
3
p ecipi a ion/dissolu ion, and CO
2
elease/
up ake. The magni ude o he p ima y p oduc ion and hus
he po en ial ole o he p oduc ion in con olling he sea
ice ino ganic ca bon cycle depend p ima ily on ligh
a ailabili y and he size o he ino ganic nu ien pool
(Papadimi iou e al. 2012). The highes ligh a enua ion
coe ficien o 12 m
-1
was measu ed in Ma ch in he snow,
co esponding o coe ficien s epo ed in snow co e in
bo h he A c ic and An a c ic (Thomas 1963; Welle and
Schwe d ege 1967; Søgaa d e al. 2010). High snow
eflec ion causes low ligh condi ions in he sea ice.
Howe e , he highes p ima y p oduc ion was ound in he
bo om o he sea ice in Ma ch (Fig. 6), whe e he ligh
a ailabili y was low compa ed o sp ing, indica ing ha
ligh was no he main ac o con olling he p ima y p o-
duc ion. Howe e , he low p ima y p oduc ion in he bo -
om o he sea ice a e Ma ch sugges s ha he sea ice
algae we e nu ien -limi ed la e in he sea ice season.
Assuming ha nu ien up ake by he ice algae ollows he
Redfield-B zezinski a io o 106C:16N:15Si:1P ( om
Redfield e al. 1963; B zezinski 1985), hen he ni ogen
(N:P a io 2) and silica e (Si:P a io 6) appea o ha e
limi ed he sea ice algal p ima y p oduc ion in Ap il and
May, while phospha e was ound a ela i ely highe con-
cen a ions. This is suppo ed by he nu ien -salini y plo
o NO
2
-
?NO
3
-
in Fig. 7, which indica es deple ion o
NO
2
-
?NO
3
-
a he end o he sea ice season. A u he
ac o known o influence he sea ice algal communi ies is
g azing (G adinge e al. 1999; Bluhm e al. 2010); how-
e e , g azing was no measu ed du ing he p esen s udy.
CaCO
3
p ecipi a ion
The obse ed CaCO
3
o ma ion was expec ed (Ande son
and Jones 1985; Ma ion 2001; Papadimi iou e al. 2004)
and is consis en wi h he ele a ed TA:TCO
2
a ios
(Fig. 4). We measu ed much lowe TCO
2
and TA con-
cen a ions wi hin he sea ice compa ed o concen a ions
ound in he unde lying seawa e (a e age TA:TCO
2
a io =1.06 in seawa e ; Fig. 4). Likewise, we ound
ele a ed TA:TCO
2
a ios in he sea ice wi h a maximum o
1.75, as compa ed o he unde lying seawa e (Fig. 4).
Howe e , he TA:TCO
2
a ios ound in he sea ice in he
p esen s udy a e low compa ed o alues ound in o he
sea ice s udies (Rysgaa d e al. 2007,2012; Papadimi iou
e al. 2012; Geil us e al. 2012; Rysgaa d e al. 2013).
Ikai e p ecipi a ion is a o ed by nea - eezing empe -
a u e, alkaline condi ion, and ele a ed phospha e concen-
a ions ([5lmol l
-1
; Bischo e al. 1993; Bucha d e al.
2001; Selleck e al. 2007).
In his s udy, bulk phospha e concen a ions be ween
0.2 and 3.1 lmol l
-1
we e measu ed in he sea ice, and
he e o e, he b ine phospha e concen a ions we e occa-
sionally abo e 5 lmol l
-1
. Fu he mo e, he phospha e
concen a ions obse ed in p esen s udy we e 5–20 imes
highe han concen a ions ound in p e ious s udies in
he A c ic (e.g. K embs e al. 2002; Mikkelsen e al.
2008; Søgaa d e al. 2010). Sea ice empe a u es du ing
ou s udy anged om -6 o0°C, which is he empe -
a u e whe e ikai e will o m. Alkalini y condi ion was
also sa isfied as a C-shaped pH p ofile wi h high pH ([9)
in su ace, and bo om sea ice laye s, and sligh ly lowe
pH condi ions (8.5) in he in e nal sea ice laye s we e
calcula ed o Feb ua y, using empe a u e and bulk
salini y (Fig. 3), TA and TCO
2
concen a ions (see
‘‘Ma e ials and me hods’’ sec ion; Fig. 4). A C-shaped pH
p ofile was also obse ed in a ecen s udy on expe i-
men al sea ice (Ha e e al. 2013).
In he ea ly pa o he s udy, we measu ed he highes
CaCO
3
concen a ions in he su ace o he sea ice, and he
concen a ions dec eased wi h dep h (Fig. 5). This is sim-
ila o ha desc ibed by Geil us e al. (2013) and Rysgaa d
e al. (2013).
nTCO2
0 1000 2000 3000 4000 5000 6000 7000
0
2000
4000
6000
CaCO3 p ecipi a ion
CaCO3 dissolu ion
CO2 in asion
CO2 elease
Respi a ion
Pho osyn hesis
nTA
Sea ice - Feb.
Sea ice - Ma .
Sea ice - Ap.
Sea ice - May
Seawa e
Fig. 8 nTA:nTCO
2
( alues no malized o a salini y o 33) ela ion-
ship in seawa e samples and bulk ice samples om Feb ua y o May.
The di e en lines ep esen he heo e ical e olu ion o TCO
2
:TA
ollowing p ecipi a ion/dissolu ion o calcium ca bona e (dashed
line), a elease o up ake o CO
2
(g) (do ed line) and impac o
biology (solid line)
1772 Pola Biol (2013) 36:1761–1777
123
62 PhD hesis by Do e Haubje g Søgaa d
The C yosphe e, 7, 707–718, 2013
www. he-c yosphe e.ne /7/707/2013/
doi:10.5194/ c-7-707-2013
© Au ho (s) 2013. CC A ibu ion 3.0 License. The C yosphe e
Open Access
Ikai e c ys al dis ibu ion in win e sea ice and implica ions o CO2
sys em dynamics
S. Rysgaa d1,2,3,4, D. H. Søgaa d3,6,M.Coope
2,M.Pu´
cko1, K. Lenne 3, T. N. Papaky iakou1,F.Wang
1,5,
N. X. Geil us1,R.N.Glud
3,6,7,J.Ehn
1,D.F.McGinnis
6, K. A a d3,6,J.Sie e s
4,J.W.Deming
8, and D. Ba be 1
1Cen e o Ea h Obse a ion Science, Depa men o En i onmen and Geog aphy, Uni e si y o Mani oba, Winnipeg,
MB R3T 2N2, Canada
2Depa men o Geological Sciences, Uni e si y o Mani oba, Winnipeg, MB R3T 2N2, Canada
3G eenland Clima e Resea ch Cen e, G eenland Ins i u e o Na u al Resou ces, 3900 Nuuk, G eenland
4A c ic Resea ch Cen e, Aa hus Uni e si y, 8000 Aa hus, Denma k
5Depa men o Chemis y, Uni e si y o Mani oba, Winnipeg, MB R3T 2N2, Canada
6Uni e si y o Sou he n Denma k and No dCEE, Odense M, Denma k
7Sco ish Associa ion o Ma ine Science, Oban, UK
8Uni e si y o Washing on, School o Oceanog aphy, Sea le, WA, USA
Co espondence o: S. Rysgaa d ([email p o ec ed])
Recei ed: 18 No embe 2012 – Published in The C yosphe e Discuss.: 6 Decembe 2012
Re ised: 30 Ma ch 2013 – Accep ed: 2 Ap il 2013 – Published: 23 Ap il 2013
Abs ac . The p ecipi a ion o ikai e (CaCO3·6H2O) in pola
sea ice is c i ical o he e iciency o he sea ice-d i en ca bon
pump and po en ially impo an o he global ca bon cycle,
ye he spa ial and empo al occu ence o ikai e wi hin he
ice is poo ly known. We epo unique obse a ions o ikai e
in unmel ed ice and e ical p o iles o ikai e abundance and
concen a ion in sea ice o he c ucial season o win e . Ice
was examined om wo loca ions: a 1m hick land- as ice
si e and a 0.3m hick polynya si e, bo h in he Young Sound
a ea (74◦N, 20◦W) o NE G eenland. Ikai e c ys als, ang-
ing in size om a ewμm o 700μm, we e obse ed o con-
cen a e in he in e s ices be ween he ice pla ele s in bo h
g anula and columna sea ice. In e ical sea ice p o iles
om bo h loca ions, ikai e concen a ion de e mined om
image analysis, dec eased wi h dep h om su ace-ice alues
o 700–900μmolkg−1ice (∼25×106c ys alskg−1) o al-
ues o 100–200μmolkg−1ice (1–7×106c ys alskg−1)nea
he sea ice–wa e in e ace, all o which a e much highe (4–
10 imes) han hose epo ed in he ew p e ious s udies. Di-
ec measu emen s o o al alkalini y (TA) in su ace laye s
ell wi hin he same ange as ikai e concen a ion, whe eas
TA concen a ions in he lowe hal o he sea ice we e wice
as high. This dep h- ela ed disc epancy sugges s in e io ice
p ocesses whe e ikai e c ys als o m in su ace sea ice laye s
and pa ly dissol e in laye s below. Mel ing o sea ice and
dissolu ion o obse ed concen a ions o ikai e would esul
in mel wa e wi h a pCO2o <15μa m. This alue is a be-
low a mosphe ic alues o 390μa m and su ace wa e con-
cen a ions o 315μa m. Hence, he mel wa e inc eases he
po en ial o seawa e up ake o CO2.
1In oduc ion
As sea ice o ms om seawa e , dissol ed sal s a e apped
in in e s i ial liquid b ine inclusions. Because phase equi-
lib ium mus be main ained be ween hese inclusions and
he su ounding ice, he b ine becomes inc easingly concen-
a ed as empe a u es dec ease. Solid sal s begin o p ecip-
i a e ou o solu ion, s a ing wi h ikai e (CaCO3·6H2O) a
−2.2◦C, mi abili e (NaSO4·10H2O) a −8.2◦C and hyd o-
hali e (NaCl·2H2O) a −23◦C (Assu , 1960). The mine al
ikai e was ecen ly disco e ed in sp ing ime sea ice in bo h
hemisphe es (Dieckmann e al., 2008, 2010). Ikai e c ys als
appea ed o be p esen h oughou he sea ice, bu wi h la ge
c ys als and highe abundance in su ace laye s (Dieckmann
e al., 2010; Rysgaa d e al., 2012; Geil us e al., 2013). Ikai e
c ys als o a ious sizes and mo phologies ha e been isola ed
Published by Cope nicus Publica ions on behal o he Eu opean Geosciences Union.

63
PhD hesis by Do e Haubje g Søgaa d
708 S. Rysgaa d e al.: Ikai e c ys al dis ibu ion in win e sea ice
om sea ice. They ange in size om a ew μm o la ge mm-
size c ys als; all a e highly anspa en wi h ounded hombic
mo phology and show uni o m ex inc ion unde c ossed po-
la ized ligh , sugges ing simple, single well-c ys alline c ys-
als. The speci ic condi ions p omo ing ikai e p ecipi a ion
in sea ice a e poo ly unde s ood, bu i p ecipi a ion occu s
du ing he ice-g owing season in he po ous lowe sea ice
laye , whe e he b ine olume is g ea e han 5% (Weeks and
Ackley, 1986; Golden e al., 1998, 2007; Ehn e al., 2007),
hen he esul ing CO2-en iched b ine will exchange wi h
seawa e ia g a i y d ainage (No z and Wo s e , 2009). Ea -
lie wo k has led o he sugges ion ha ikai e c ys als may
emain apped wi hin he skele al laye whe e hey ac as a
s o e o TA, becoming a sou ce o excess TA o he ocean wa-
e upon subsequen mine al dissolu ion du ing summe mel
(Nedashko sky e al., 2009; Rysgaa d e al., 2011). This ex-
cess would lowe he pa ial p essu e o CO2(pCO2)o su -
ace wa e s a ec ed by mel ing sea ice and cause an inc ease
in he ai –sea CO2 lux.
A his poin , he spa ial and empo al occu ence o ikai e
p ecipi a es wi hin sea ice is poo ly cons ained. The e is
u gency o inc easing he knowledge base, gi en ha he
p ecipi a ion o CaCO3is implica ed in many p ocesses o
global signi icance, including he sea ice-d i en ca bon pump
and global ca bon cycle (Delille e al., 2007; Rysgaa d e
al., 2007, 2011; Papadimi iou e al., 2012) and pH condi-
ions (acidi ica ion) in su ace wa e s (Rysgaa d e al., 2012;
Ha e e al., 2013). Quan i ica ion o CaCO3c ys als in sea
ice in he ew p e ious s udies has been made on mel ed
samples assuming ha ikai e will no dissol e i empe a u e
is main ained below 4◦C. In p inciple, howe e , dissolu ion
may also be ela ed o he eac ion o ikai e wi h CO2in he
mel wa e o wi h he a mosphe e du ing he mel ing p oce-
du e, which can las o days a low empe a u e allowing he
possibili y o changing pH in he su oundings o he ikai e
c ys al and unde es ima ion o ikai e concen a ion. Examin-
ing ikai e in such mel ed samples also makes i impossible
o e alua e spa ial dis ibu ion wi hin he sea ice ma ix on
he mic ome e o millime e scale and de e mine whe he o
no he mine al is en apped in he ice c ys al la ice and hus
sepa a ed om he solu ion.
P ecipi a ion o ikai e in s anda d seawa e condi ions is
desc ibed by
Ca2++2HCO−
3+5H2O↔CaCO3·6H2O+CO2.(1)
B ine d ainage om sea ice is expec ed o esul in a
emo al o dissol ed CO2along wi h sal s. I p ecipi a ed
ikai e c ys als become apped wi hin sea ice in e s ices, hen
he eac ion in Eq. (1) is pushed o he igh in b ine, p o-
iding u he po en ial o CaCO3g ow h. O e ime, he
concen a ion o apped ikai e c ys als could inc ease. He e
we p o ide no el es s o hese ideas du ing c ucial win e
condi ions by examining ikai e in in ac (unmel ed) na u al
sea ice and de e mining de ailed e ical ikai e dis ibu ions
in he sea ice om wo loca ions in Young Sound (no heas
G eenland). Combined wi h o he measu emen s and model
calcula ions, he esul s allow us o ela e sea ice o ma ion
and mel o he obse ed pCO2condi ions in su ace wa e s,
and hence, he ai –sea CO2 lux.
2 Me hods
2.1 S udy si e and sampling
Sampling was pe o med a wo loca ions in he Young
Sound a ea (74◦N, 20◦W: Rysgaa d and Glud, 2007), NE
G eenland, in Ma ch 2012 (Fig. 1). The land- as ice s a ion,
ICE I (74◦18.5764N, 20◦18.2749W), was in he jo d wi h
110–115cm hick sea ice co e ed by 70cm o snow. F ee-
boa d a ICE I was nega i e, e.g. when a hole was d illed
h ough he ice, he ice su ace looded. Howe e , a si es
wi hou d illed holes, we did no obse e looding du ing ou
expe imen and he e was a dis inc snow–ice in e ace. On
op o he ice o ICE I, abou 8cm o slush snow a he snow–
ice in e ace was obse ed. The new-ice s a ion, POLY I
(74◦13.905N, 20◦07.701W), was si ua ed in a polynya e-
gion abou 3km o he sha p land- as ice edge, whe e sea
ice egula ly b eaks up in win e . A he ime o sampling,
POLY I was 15–30cm hick and co e ed by 17cm o snow.
A POLY I he e was nega i e eeboa d oo – wi h abou
2cm o slush snow a he snow–ice in e ace. A POLY I he
snow was dense up o ∼8cm om he ice in e ace, hen a
bi ligh e (newe snow) abo e. Ai empe a u e du ing sam-
pling anged om −20 o −25◦C. The polynya si e is ep-
esen a i e o hin A c ic sea ice, whe eas he ICE I loca ion
is ep esen a i e o jo d ice in he A c ic whe e he e is a
la ge sou ce o win e p ecipi a ion (snow). Snow was e y
d y and e y cold (unlike An a c ic ice). The e was also no
su ace looding. ICE I is ypical o sea ice in jo ds o whe e
la ge mois u e sou ces a e a ailable o he A c ic win e cli-
ma e sys em.
Sea ice co es we e ex ac ed using a MARK II co ing
sys em (Ko acs En e p ises). Ve ical empe a u e p o iles
we e measu ed wi h a he mome e (Tes o O ion 3-s a wi h
an O ion 013610MD conduc i i y cell) in he cen e o he
co es immedia ely a e co ing. Sea ice was hen cu in o
5–10cm sec ions, kep cold and b ough o he ield labo a-
o y wi hin 1h o p ocessing. In he 20◦C labo a o y, sea ice
sec ions we e mel ed o measu emen o bulk salini y wi h
a conduc i i y p obe (The mo O ion 3-s a wi h an O ion
013610MD conduc i i y cell, UK) calib a ed agains a 15N
KCLsolu iona 20◦C. B ine olume in sea ice was calcu-
la ed acco ding o Cox and Weeks (1983) and Lepp¨
a an a
and Manninen (1988). B ine salini y was calcula ed om he
measu ed sea ice empe a u es and eezing poin o seawa e
(Unesco, 1978).
Al hough he e was a lo o snow on he hick as ice in
he jo d he e was no na u al looding o he su ace. When
we ins alled ocean ins umen s (d illing h ough he ice) we
The C yosphe e, 7, 707–718, 2013 www. he-c yosphe e.ne /7/707/2013/
64 PhD hesis by Do e Haubje g Søgaa d
S. Rysgaa d e al.: Ikai e c ys al dis ibu ion in win e sea ice 709
Godha n
Thule
God håb Angmagssalik
Sco esby Sund
G eenland
Canada
ABWollas on
Fo land
Cla e ing
island
Fig. 1. S udy si e. (A) G eenland showing he s udy a ea (box) and (B) he Young Sound jo d be ween Wollas on Fo land and Cla e ing
Island. Sea ice co ing si es o he land- as ice loca ion (ICE I) is shown as a ed s a and he polynya co ing si e (POLY I) as a yellow
s a . The sa elli e image showing land- as ice in he jo d and hin ice–open wa e condi ions in he polynya ou side he jo d is om ea ly
Ma ch 2012.
c ea ed some a i icial looding. A widely dis ibu ed snow-
ice in e ace su ey, howe e , showed ha looding was lim-
i ed o only he a ea immedia ely su ounding he ins alled
ins umen s. The same was he case o he Polynya si e.
2.2 Analysis
In he cold labo a o y (−20 o −25◦C), co espond-
ing sea ice co es we e cu in o e ical hick sec-
ions (10cm×6cm×1cm) and ho izon al hick sec ions
(6cm×6cm×1cm), hen moun ed on o sligh ly wa med
glass pla es and hinned o 1–3mm hickness using a mi-
c o ome (Leica SM 2010R). Each sec ion was hen pho-
og aphed (Nikon D70) using pola ized il e s o documen
ice ex u e. Subsequen ly, each hin sec ion was inspec ed un-
de a s e eomic oscope (Leica M125 equipped wi h a Leica
DFC 295 came a and Leica Applica ion Sui e e . 4.0.0. so -
wa e) o documen he e ical and ho izon al posi ion o
ikai e c ys als in sea ice.
To documen he abundance and concen a ions o ikai e
c ys als in sea ice, 20–90mg o sea ice we e cu o , using a
s ainless s eel kni e, a h ee andom places wi hin each 5–
10cm sea ice sec ion. These subsamples we e weighed and
placed on o a glass slide ha es ed on a chilled aluminum
block wi h a 1cm cen al iewing hole, hen b ough in o he
20◦C labo a o y. The e hey we e inspec ed in ac unde a
mic oscope (Leica DMiL LED) unde 100–400 magni ica-
ion as hey we e also allowed o mel . A ew seconds a e
he sea ice had mel ed he i s image was aken (same cam-
e a and so wa e as desc ibed abo e). This image was used
o documen he amoun and concen a ion o ikai e c ys als
( u he de ails gi en below). A e 2–5min he second image
was aken and compa ed wi h he i s one. I c ys als we e
dissol ing, hey we e assumed o be ikai e. Th ee andom
samples, each co e ing 1.07mm2o he coun ing a ea, we e
imaged in his ashion. The a ea o he mel ed sea ice sample
was de e mined a e hawing o calcula e he coun ed a ea
o o al a ea a io. Ikai e c ys als and o he p ecipi a es we e
obse ed o se le o he glass slide apidly a e ice c ys al
mel due o hei high densi y. In o al, h ee 20–90mg sea
ice sub-samples we e p ocessed om each sea ice sec ion in
iplica es, esul ing in 9 eplica e images o each 5–10cm
e ical sea ice sec ion.
The abundance and concen a ions o c ys als we e calcu-
la ed om he images using he so wa e ImageJ (1.45s). In-
di idual images we e b igh ness/con as adjus ed p io o bi-
na y ile con e sion. A “close- unc ion” ou ine hen ensu ed
ha c ys als we e closed be o e “ illing-holes” wi h black.
An “analyze” ou ine was hen applied o coun he c ys als
and analyze hei a ea ela i e o coun ing a ea. This a io
was mul iplied by he a ea o he mel ed ice subsample. Fo
concen a ion es ima es o ikai e a cubic o m was assumed
o he mine al. Ikai e concen a ion in sea ice was calcu-
la ed om i s olume, densi y (1.78gcm−3), mola weigh
(208.18gmol−1), and weigh o subsample and con e ed
in o μmolkg−1mel ed sea ice uni s.
A sea ice co e (en i e co e) om each si e was kep a
−20◦C o h ee weeks and b ough o he x- ay labo a-
o y a he Depa men o Geological Sciences a he Uni-
e si y o Mani oba, Canada. The e, 20–90mg subsamples
o sea ice we e cu andomly om each sea ice sec ion (5–
10cm e ical sec ions) and moun ed on o a cold glass slide
es ing on a chilled aluminum block con aining a 1cm cen-
al iewing hole. The c ys als we e i s examined wi h a
pola ized ligh mic oscope o assess hei op ical p ope ies
and hen moun ed o x- ay s udy using a s e eo binocula
mic oscope. Ikai e c ys als we e selec ed om each 10cm
sec ion o he sea ice co es om bo h ICE I and POLY I,
d agged ac oss he cold glass slide using a me al p obe and
imme sed in o a d op o special pu pose sampling oil ha e-
s ic ed sublima ion. Each c ys al was hen scooped up wi h
a low x- ay sca e ing mic o-loop and ins an ly ans e ed
o he ni ogen cold s eam (−10◦C) on he x- ay di ac ion
ins umen wi h a magne ic coupling goniome e head. The
www. he-c yosphe e.ne /7/707/2013/ The C yosphe e, 7, 707–718, 2013
65
PhD hesis by Do e Haubje g Søgaa d
710 S. Rysgaa d e al.: Ikai e c ys al dis ibu ion in win e sea ice
x- ay di ac ion ins umen consis ed o a B uke D8 h ee-
ci cle di ac ome e equipped wi h a o a ing anode gene -
a o (MoKαX- adia ion), mul i-laye op ics, APEX-II CCD
de ec o , and an Ox o d 700 Se ies liquid-N C yos eam. The
in ensi ies o mo e han 100 e lec ions we e ha es ed om
six ame se ies (each spanning 15◦in ei he ωo ϕ) col-
lec ed o 60◦2θusing 0.6s pe 1◦ ame wi h a c ys al-
o-de ec o dis ance o 5cm. In o al, 12 c ys als (one om
all ice laye s in es iga ed om bo h ICE I and POLY I)
we e iden i ied h ough success ul indexing o obse ed x-
ay di ac ion maxima on o known cha ac e is ic uni cells.
To de e mine TA and o al dissol ed ino ganic ca bon
(TCO2)concen a ions in sea ice, h ee sea ice co es we e
cu in o 5–10cm sec ions and b ough o he ield labo a o y.
He e, he ice segmen was placed immedia ely in a gas- igh
lamina ed (Nylon, e hylene inyl alcohol, and polye hylene)
plas ic bag (Hansen e al., 2000) i ed wi h a 50cm gas-
igh Tygon ube and a al e o sampling. The weigh o he
bag con aining he sea ice sample was eco ded. Cold (1◦C)
deionized wa e (25–50mL) o known weigh and TA and
TCO2concen a ion was added. The plas ic bag was closed
immedia ely and excess ai and deionized wa e quickly e-
mo ed h ough he al e and weighed. The weigh o he
deionized wa e accoun ed o <5% o he sea ice weigh .
This p ocessed sea ice was subsequen ly mel ed, and he
mel wa e mix u e ans e ed o a gas- igh ial (12ml Ex-
e aine , Labco High Wycombe, UK). Any CaCO3c ys als
p esen in hese ice co e sec ions a e expec ed o ha e dis-
sol ed du ing sample p ocessing and, hus, be included in
he measu ed TA and TCO2concen a ions. S anda d me h-
ods o analysis we e used: TCO2concen a ions we e mea-
su ed on a coulome e (Johnson e al., 1987), TA by po en-
iome ic i a ion (Ha aldsson e al., 1997), and gaseous CO2
by gas ch oma og aphy. Rou ine analysis o Ce i ied Re e -
ence Ma e ials (p o ided by A. G. Dickson, Sc ipps Ins i u-
ion o Oceanog aphy) e i ied ha TCO2and TA concen-
a ions (n=3) could be analyzed wi hin ±1μmolkg−1and
±4μmolkg−1, espec i ely. Bulk concen a ions o TA and
TCO2in he sea ice (Ci)we e calcula ed as Ci=([CmWm−
CaWa]/Wi), whe e Cmis he TA o TCO2concen a ion in
he mel wa e mix u e, Wmis he weigh o he mel wa e
mix u e, Cais he TA o TCO2concen a ion in he deion-
ized wa e , Wais he weigh o he deionized wa e , and Wi
is he weigh o he sea ice (Rysgaa d e al., 2008).
Following he bulk de e mina ion o TCO2and TA, he
bulk pCO2and pH (on he o al scale) we e compu ed using
he empe a u e and salini y condi ions in he ield and a s an-
da d se o ca bona e sys em equa ions, excluding nu ien s,
wi h he CO2SYS p og am o Lewis and Wallace (2012). We
used he equilib ium cons an s o Meh bach e al. (1973) e-
i ed by Dickson and Mille o (1987, 1989).
In summa y, iplica e co es we e collec ed 18, 20 and
24 Ma ch o snow and ice hickness de e mina ion, ice em-
pe a u es and bulk ice sal concen a ions. Measu emen s
we e pe o med on he same co es. T iplica e sepa a e co es
we e sampled o de e mina ion o TCO2and TA concen a-
ions. Ice ex u e, densi y and pho o documen a ion o ikai e
loca ion in in ac ice co e s we e pe o med on iplica e sep-
a a e co es (ICE I, 17 Ma ch: POLY I, 20 Ma ch). Ikai e
image analysis o concen a ion de e mina ion was done on
h ee sepa a e co es collec ed 18, 20 and 24 Ma ch. X- ay
di ac ion analyses we e pe o med on a sepa a e co e col-
lec ed 24 Ma ch. Samples o bulk sal , TA, TCO2, ikai e
c ys al abundance and concen a ion we e p ocessed imme-
dia ely in he ield labo a o y. C ys als o ikai e de e mi-
na ion on x- ay we e kep ozen (−20◦C) un il analysis a
mon h la e . Pho o documen a ion showed ha ikai e c ys al
mo phology did no change due o s o age. Thus, p olonged
eezing pe iod p io analysis should no a ec he esul s
ob ained. Howe e , i is impo an ha p ese a ion o TA,
TCO2and quan i ica ion o ikai e is done in he ield. These
samplesshould no be s o ed a low empe a u es be o eanal-
ysis as CaCO3could be p oduced du ing he s o age.
2.3 FREZCHEM modeling
The p oduc ion o ikai e in he sea ice co es was mod-
eled by FREZCHEM ( e sion 10), an equilib ium chem-
ical he modynamic model pa ame e ized o concen a ed
solu ions (up o 20molkg−1(H2O)) and sub-ze o empe -
a u es ( o −70◦C) (Ma ion e al., 2010). The model uses
he Pi ze app oach co ec ing o ac i i y coe icien s o so-
lu es in concen a ed solu ions. Ou calcula ion was done by
ollowing he eezing p ocess o seawa e wi h he same
chemical composi ion as he local seawa e unde he local
pCO2 alues; he he modynamic cons an s used we e he
de aul alues p o ided wi h he model. I should be no ed
ha FREZCHEM modeling is based on he assump ion ha
chemical species in he sea ice en i onmen (ice, b ine, and
ai ) ha e eached he modynamic equilib ium, and ha mos
o he he modynamic cons an s used in he model we e ex-
apola ed o low empe a u es a he han being di ec ly de-
e mined expe imen ally. Ne e heless, he model has shown
p omising applica ions in explo ing cold geochemical p o-
cesses associa ed wi h seawa e eezing among many o he s
(Ma ion e al., 1999, 2010).
3 Resul s
Sea ice co e ed he jo d on ou a i al, bu he e was a
polynya ou side he jo d wi h a dis inc ice edge unning
ac oss he jo d be ween Wollas on Fo land and Cla e ing
Island (Fig. 1). Se e al eezing and opening inciden s we e
obse ed in his a ea om sa elli e images p io o ou a -
i al; howe e , due o low empe a u es o −20◦C o−36◦C
and calm condi ions, he polynya began o e- eeze p io
o ou sampling. Due o he insula ing e ec o he hick
(70cm) snow co e a ICE I, ou land- as s a ion in he jo d,
sea ice su ace empe a u e was ela i ely wa m, −10◦C,
The C yosphe e, 7, 707–718, 2013 www. he-c yosphe e.ne /7/707/2013/
66 PhD hesis by Do e Haubje g Søgaa d
S. Rysgaa d e al.: Ikai e c ys al dis ibu ion in win e sea ice 711
-100
-80
-60
-40
-20
0
Sea ice, cm
-12 -8 -4 0
Tempe a u e, ºC
1210864
Bulk salini y
-100
-80
-60
-40
-20
0
Sea ice, cm
-12 -8 -4 0
Tempe a u e, ºC
1210864
Bulk salini y
AB
CD
-100
-80
-60
-40
-20
0
0.300.250.200.150.100.050.00
B ine olume ( / )
2001601208040
B ine salini y
-100
-80
-60
-40
-20
0
0.300.250.200.150.100.050.00
B ine olume ( / )
2001601208040
B ine salini y
Fig. 2. Ve ical p o iles o sea ice ea u es. (A) Tempe a u e (-o-)
and bulk salini y (-•-) and (B) b ine olume (-o-) and b ine salin-
i y (-•-) a ICE I. (C) Tempe a u e (-o-) and bulk salini y (-•-) and
(D) b ine olume (-o-) and b ine salini y (-•-) a POLY I. Ve ical
do ed line ep esen s he b ine olume whe e sea ice becomes pe -
meable.
wi h a g adien o −2◦C owa ds he sea ice–wa e in e ace
(Fig. 2a). Bulk salini ies anged om 10–12 in he op laye s
o 4 a he bo om. Calcula ed b ine olumes anged om less
ha 5% in su ace laye s o 12% nea he bo om (Fig. 2b);
b ine salini ies, om 150 in he su ace sea ice o 33 nea he
wa e column. A POLY I, ou new-ice s a ion in he polynya
whe e snow co e was hinne (17cm), su ace sea ice em-
pe a u e was only −5◦C (Fig. 2c) due o he apid eezing
wi h concu en hea elease and p oximi y o he wa e col-
umn ( hin ice o 15–30cm). Bulk salini ies anged om 10 in
su ace ice laye s o 7 in he lowe ice laye s (Fig. 2c). Gi en
high bulk salini ies and empe a u es, b ine olumes anged
om 10% in he op laye s o 20% in he lowe ice laye s;
b ine salini ies we e lowe a ICE I, anging om 78 a he
su ace o 33 nea he wa e column (Fig. 2d).
The uppe 35cm o sea ice a ICE I was composed o
polygonal g anula ice, o med h ough pe cula ion and e-
eezing o b ine and seawa e in o snow, i.e. snow ice
(Fig. 3a). The snow ice o ma ion was likely p omo ed by
nega i e eeboa d. The ac ha we do no obse e lood-
ing is p obably due o low empe a u es (b ine o seawa e
quickly e eezes in he cold snow). Daily images wi h au o-
ma ic came a sys ems (Ma inBasis P og am, c/o G eenland
Ins i u e o Na u al Resou ces) show ha sea ice s a ed o
o m locally ou side he jo d in ea ly Oc obe , and ha hose
ice loes d i ed in o he jo d and consolida ed wi h local
ice o o m a uni o m ice co e ha pe sis ed h ough ou
s udy. Thus, he op laye s o ice a ICE I may ha e been
p oduced ou side he jo d. The ice laye om 35–112cm
consis ed o columna sea ice (Fig. 3d). Mic oscopic exam-
ina ion o di e en e ical and ho izon al ice hin sec ions
showed he p esence o ikai e c ys als h oughou he ice col-
umn (Fig. 3b, c and e, ). Ikai e c ys als we e loca ed in he
in e s ices be ween he ice pla ele s.
The sea ice a POLY I was less han one week old when
i s sampled. The uppe 5cm o he ice consis ed o o -
bicula g anula ice c ys als, ollowed by ansi ional g an-
ula /columna ex u e om 5 o 12cm (Fig. 4a). This uppe -
mos laye closely esembles he disc-like g anula ice ob-
se ed in nilas (Ehn e al., 2007) and may be ela ed o eez-
ing o b ine expulsed o he sea ice su ace. The ansi ion
in o mainly columna -like sea ice occu ed a a ound 10–
12cm om he su ace and con inued o he sea ice–wa e
in e ace (29cm); howe e , he diso de ly s uc u e implies
in usion o azil ice and pla ele ice o ms, and hus supe -
cooling a he ice–ocean in e ace was a signi ican ac o
in de e mining he ice s uc u e. Simila ly o ICE I, mic o-
scopic examina ion o hin ice sec ions om POLY I e-
ealed he p esence o ikai e c ys als h oughou he ice col-
umn (Fig. 4b, c and e, ). Ikai e c ys als we e again obse ed
p ima ily in he in e s ices be ween he ice pla ele s.
Ikai e c ys als become easily isible when he sea ice
mel s. As an example om POLY 1, 19 c ys als we e ob-
se ed a ew seconds a e mel ing 50mg o andomly sub-
sampled sea ice (10–20cm sec ion; Fig. 5a and b). Allowing
he c ys als o dissol e o a ew minu es be o e aking an-
o he image allowed us o iden i y ikai e c ys als (as he ones
ha had dissol ed; Fig. 5c). In his case, he c ys al a ea co -
e ed 0.48% o he coun ing a ea.
A ICE I, he numbe o ikai e c ys al pe kg o mel ed
ice anged om ∼25×106kg−1in he uppe laye s o
∼1×106kg−1nea he ice–wa e in e ace (Fig. 6a). A
POLY I, simila ikai e abundances we e obse ed in up-
pe ice laye s, whe eas abundances nea he wa e column
eached ∼7×106kg−1(Fig. 6b). The mola concen a ion o
ikai e pe kg mel ed ice a ICE I dec eased wi h dep h om
su ace alues o 900μmolkg−1 o 100μmolkg−1nea he
ice–wa e in e ace. A POLY I, he highes concen a ions
(700μmolkg−1)we e obse ed a 5–10cm om he ice su -
ace, dec easing o ∼200μmolkg−1nea he wa e column.
The c ys als (a ew μm o 700μm in size) obse ed in he
sea ice we e highly anspa en wi h a hombic mo phol-
ogy and showed uni o m ex inc ion unde c oss-pola ized
www. he-c yosphe e.ne /7/707/2013/ The C yosphe e, 7, 707–718, 2013
67
PhD hesis by Do e Haubje g Søgaa d
712 S. Rysgaa d e al.: Ikai e c ys al dis ibu ion in win e sea ice
1 cm
10-20 cm sec ion
250 μm
250 μm1 cm
49-59 cm sec ion
A B
D E
100 μm
100 μm
C
F
Fig. 3. Images o sea ice wi h ikai e c ys als, om ICE I. (A, D) Sea
ice ex u e (pola ized ligh ); (B, E) mic oscopic images o sea ice.
Yellow a ows poin o ice c ys al bo de s, blue a ows o b ine
pocke s, g een a ows o ai bubbles and ed a ows o ikai e c ys-
als. (C, F) Ikai e c ys als a highe magni ica ion. (A),(B) and (C)
om 10–20cm sec ion; (D),(E) and (F) om 49–59cm sec ion.
ligh , sugges ing ha hey we e simple single c ys als. All
x- ay e lec ions i ed well o a monoclinic C-cen e ed cell
wi h e ined cell pa ame e s shown in Table 1. F om he
gene al shape, op ical p ope ies and uni -cell de e mina-
ion, he c ys als examined we e ikai e (Hesse and K¨
uppe s,
1983). The c ys als iden i ied as ikai e had a e y dis inc
250 μm
100 μm
250 μm
100 μm
1 cm
10-20 cm sec ion
1 cm
0-10 cm sec ion
A B
C
D E
F
Fig. 4. Images o sea ice wi h ikai e c ys als, om POLY I. (A,
D) Sea ice ex u e (pola ized ligh ); (B, E) mic oscopic image o
sea ice. Yellow a ows poin o ice c ys al boa de s, blue a ows o
b ine pocke s, and ed a ows o ikai e c ys als. (C, F) Ikai e c ys als
a highe magni ica ion. (A),(B)and (C) om 0–10cm sec ion; (D),
(E) and (F) om 10–20cm sec ion.
mo phology, and we e easily ecognized wi h wo signi ican
a ia ions: he mo e common hicke hombs (Fig. 7a) and
a he a e hinne pla es (Fig. 7b). Whe eas ikai e c ys als in
in ac sea ice some imes we e assembled in agg ega es such
as he ones illus a ed in Fig. 4e and , hey consis ed mos ly
o isola ed hombic o ms and hin pla es as obse ed unde
The C yosphe e, 7, 707–718, 2013 www. he-c yosphe e.ne /7/707/2013/

68 PhD hesis by Do e Haubje g Søgaa d
S. Rysgaa d e al.: Ikai e c ys al dis ibu ion in win e sea ice 713
100 μm
100 μm
A
100 μm
B
C
Fig. 5. Mic oscopic images o ikai e c ys als (A) a ew seconds a -
e mel ing 50mg sea ice (10–20cm sec ion) om POLY I. Image
ep esen s a e y small ac ion (2μg) o he sample. (B) So wa e
(ImageJ 1.45s) p ocessing o image A o ind he numbe (19) and
a ea (0.48% o coun ing a ea) o he c ys als. (C) Image aken a e
2min showing ha c ys als a e dissol ing.
he mic oscope immedia ely a e ice pla ele s had mel ed
(Fig. 5).
Su ace TA concen a ions o 600–800μmolkg−1mel ed
sea ice we e measu ed a ICE I, wi h dec easing concen a-
ions o 380μmolkg−1mel ed sea ice in ice laye s close o
he wa e column (Fig. 8a). TCO2 alues a ICE I had simila
e ical dis ibu ion wi h lowe concen a ions. A POLY I
si e, TA alues we e measu ed a ∼800μmolkg−1mel ed
sea ice nea he su ace, and dec eased o ∼500μmolkg−1
mel ed sea ice nea he wa e column. As a ICE I, he e i-
cal TCO2concen a ions a POLY I we e simila o TA, bu
wi h lowe alues.
40x106
3020100
Ikai e c ys al numbe s/kg
-100
-80
-60
-40
-20
0
Sea ice, cm
160012008004000
Ikai e c ys al conc., μmol/kg
40x106
3020100
Ikai e c ys al numbe s/kg
-100
-80
-60
-40
-20
0
160012008004000
Ikai e c ys al conc., μmol/kg
AB
Fig. 6. Ve ical dis ibu ion o ikai e c ys als de e mined om image
analysis. Abundance o c ys als (-o-) and concen a ion o ikai e
(- -•--)a ICEI(A) and POLY I (B).
4 Discussion
4.1 Spa ial and empo al a iabili y o ikai e
occu ence and concen a ion
Ikai e concen a ions om his s udy a e highe han hose
epo ed p e iously: 10 imes highe han hose measu ed in
sp ing sea ice om An a c ica (Dieckmann e al., 2008), and
4 imes highe han in su ace summe ice om F am S ai
(160–240μmolkg−1mel ed sea ice; Rysgaa d e al., 2012).
The di e ences may be explained in pa by di e ences in
quan i ica ion p ocedu e, as ou s udy is unique in employing
immedia e analysis in he ield wi hou p olonged mel ing.
The highe alues we epo may also e lec eal di e ences
be ween si es and seasons, as we p esen esul s o win e
sea ice, while o he s udies we e conduc ed in sp ing and
summe wi h mel al eady appa en . As a esul , a ac ion
o ikai e c ys als in he p e ious s udies may al eady ha e
been dissol ed due o ice wa ming and mel ing o los ia
b ine d ainage. In addi ion, ikai e c ys als we e obse ed
in he week-old POLY I sea ice and e en wi hin 1 hou
in os lowe s and hin ice in an a i icially opened lead
(da a o be p esen ed elsewhe e). This obse a ion indica es
dynamic condi ions o ikai e o ma ion e en on sho
imescales. Unde s anding he dynamics o hose p ocesses
is an impo an objec i e o u u e s udies. He e, we no e
ha highe concen a ions o ikai e in su ace sea ice a e
p edic ed by he FREZCHEM model (Ma ion e al., 2010).
Assuming ha a s anda d seawa e (S=35, [Na+]=0.4861,
[K+]=0.01058, [Ca2+]=0.01065, [Mg2+]=0.05475,
[Cl−]=0.56664, [SO2−
4]=0.02927, [HCO−
3]=0.0023;
ions concen a ion uni : molekg−1wa e ) eezes in an open
sys em wi h pCO2=320μa m, he FREZCHEM model
p edic s an ikai e concen a ion o up o 620μmolkg−1sea
ice in he cold su ace laye o ICE I, dec easing exponen-
ially downwa ds (Fig. 9). Bo h he concen a ion ange
www. he-c yosphe e.ne /7/707/2013/ The C yosphe e, 7, 707–718, 2013
69
PhD hesis by Do e Haubje g Søgaa d
714 S. Rysgaa d e al.: Ikai e c ys al dis ibu ion in win e sea ice
Table 1. X- ay di ac ion da a. Re ined uni -cell pa ame e s o
ikai e c ys als iden i ied by x- ay di ac ion analysis. ( )=S anda d
de ia ion, R= hombs, TP= hin pla e.
ICE I
Sec ion
cm a( ˚
A) b( ˚
A) c( ˚
A) β(◦)V( ˚
A3)
0–10 R 8.808(2) 8.313(2) 11.031(2) 110.59(1) 756.1(8)
10–20 R 8.820(2) 8.330(2) 11.050(3) 110.57(2) 761.7(4)
20–30 TP 8.823(3) 8.326(3) 11.049(4) 110.64(2) 759.6(7)
30–40 R 8.813(2) 8.312(2) 11.034(3) 110.60(1) 756.7(4)
40–50 R 8.817(3) 8.312(2) 11.027(3) 110.58(2) 756.6(5)
50–60 R 8.812(2) 8.311(1) 11.040(2) 110.58(1) 756.9(6)
60–70 R 8.831(3) 8.320(2) 11.047(3) 110.55(2) 760.1(6)
70–80 R 8.812(2) 8.318(2) 11.031(3) 110.56(2) 757.0(8)
80–90 R 8.819(2) 8.315(2) 11.025(3) 110.57(2) 758.3(5)
90–100 TP 8.820(2) 8.323(2) 11.049(2) 110.64(1) 759.2(5)
POLY I
Sec ion
cm a( ˚
A) b( ˚
A) c( ˚
A) β(◦)V( ˚
A3)
0–10 TP 8.816(3) 8.333(2) 11.043(3) 110.68(2) 759.1(6)
10–20 TP 8.810(1) 8.320(1) 11.033(1) 110.58(1) 757.1(2)
and dis ibu ion pa e n ag eed well wi h he empi ical da a
a ICE I. Al hough he FREZCHEM modeling assumes
ha he sys em eaches he modynamic equilib ium and
is always open o a cons an pCO2, assump ions ha a e
no ul illed unde na u al condi ions, he modeling esul s
ne e heless suppo he obse a ion ha ikai e concen a ion
inc eases wi h dec easing empe a u e. Seasonally a iable
ikai e concen a ion, wi h highes alues in win e , is hus
expec ed.
I is a key poin as o exac ly whe e he c ys als a e lo-
ca ed. I hey a e in he b ine channels hen hey can po en-
ially mo e wi h he ci cula ion o b ine as he empe a u es
change in e nally in he ice. I hey a e isola ed om la ge
b ine ne wo ks and a e loca ed a he in e s ices hen hey
may emain apped in he ice as con ec ion occu s. This will
make a big di e ence on he exchange h ough win e and
well in o sp ing. Ou images documen ha ikai e c ys als
a e loca ed be ween he in e s ices o he sea ice ma ix be-
ween he pu e ice pla ele s. As hey a e pa icles hey can be
apped be ween he small in e s i ial po e spaces and he e-
o e e ained in he ice. In con as , solu es and gases (CO2),
as well as o ganic pa icles including mic oo ganisms (Junge
e al., 2001; K embs e al., 2011), can be anspo ed wi hin
he b ine sys em. I was possible o see b ine mo ion in he
mic oscope as small pa icles and ai bubbles mo ed be ween
he in e s ices o ice pla ele s. Inc easing he empe a u e by
a ew deg ees signi ican ly accele a ed he anspo eloci y.
The TA: bulk salini y a io in sea ice was 84±4 (ICE I) and
78±0.8 (POLYI)as compa edwi h he wa e column 67±3
(ICE I) and 61±0.2 (POLY I). Thus, he highe TA-S a io
in sea ice han in seawa e shows ha CaCO3 elease adds a
20 μm
20 μm
A
B
Fig. 7. Two mos common o ms o ikai e in eezing sea ice.
(A) hombs, (B) hin pla es.
-100
-80
-60
-40
-20
0
10008006004002000
TA & TCO2 μmol/kg
-100
-80
-60
-40
-20
0
Sea ice, cm
10008006004002000
TA & TCO2 μmol/kg
AB
Fig. 8. Ve ical p o iles o concen a ion o o al alkalini y (- -o- -)
and dissol ed ino ganic ca bon (-•-) in (A) a ICE I and (B) a
POLY I.
u he alkalini y con ibu ion o he simple mel o ice. These
obse a ions con i m p e ious hypo heses (Rysgaa d e al.,
2007, 2009, 2011) ha ikai e c ys als a e apped wi hin he
sea ice ma ix, whe eas CO2 eleased h ough ikai e p oduc-
ion (Eq. 1) and dissol ed wi hin he b ine can be los om
he sea ice. As a esul , CaCO3s o es wice as much TA as
TCO2, hence TA o he mel wa e inc eases ela i e o TCO2
in sea ice. When ikai e c ys als dissol e du ing sea ice mel ,
su ace wa e pCO2will dec ease. This is impo an as low
The C yosphe e, 7, 707–718, 2013 www. he-c yosphe e.ne /7/707/2013/
70 PhD hesis by Do e Haubje g Søgaa d
S. Rysgaa d e al.: Ikai e c ys al dis ibu ion in win e sea ice 715
-100
-80
-60
-40
-20
0
Sea ice, cm
160012008004000
Ikai e concen a ion μmol/kg
Fig. 9. Ve ical p o iles o ikai e concen a ion. Measu ed (-•-) and
FREZCHEM calcula ed (-o-) a ICE I.
pCO2 alues in su ace wa e s will lead o a la ge CO2 lux
om he a mosphe e in o he ocean.
Bulk TA concen a ions in su ace ice laye s (including
ikai e c ys als) we e wi hin he same ange as ikai e concen-
a ions, implying ha mos o TA is p esen in he c ys al
o m o ikai e and, hus, apped wi hin he su ace sea ice
ma ix. In con as , bulk TA concen a ions we e highe han
ikai e c ys al concen a ions in he in e nal and bo om laye
o he sea ice a bo h s a ions, implying ha no all TA in he
lowe ice laye s o igina e om ikai e c ys als. One explana-
ion o his inding in okes dissolu ion o educed g ow h
a e o ikai e in in e io ice laye s due o exposu e o ex-
cess CO2o igina ing om cold uppe ice laye s, whe e el-
e a ed b ine concen a ions o Ca2+and HCO−
3a e highe
due o lowe b ine olumes. The CO2 eleased in uppe ice
could be anspo ed o in e io ice laye s by downwa d b ine
d ainage. Low pH condi ions in in e io ice laye s ha e e-
cen ly been epo ed o expe imen al sea ice (Ha e e al.,
2013). He e, e ical pH p o iles o bulk ice, as measu ed a
nea - eezing empe a u es, e ealed a consis en C-shaped
pa e n du ing columna ice g ow h, wi h he highes pH al-
ues (>9) in bo h he ex e io ( op and bo om) ice sec ions
and lowes pH (∼7) in he in e io ice sec ions (Ha e e al.,
2013). Calcula ing he e icalpH p o ile a ICE I, using em-
pe a u e and bulk salini y (Fig. 2) and TA and TCO2con-
di ions (Fig. 8) om ou win e ice s udy, yielded a simi-
la C-shaped pH p o ile wi h high pH (>9) in su ace and
bo om ice laye s, and lowe pH (∼8) in in e io laye s. A
POLY I, a C-shaped pH p o ile was also es ima ed, al hough
he pH alues in ex e io laye s we e sligh ly lowe (∼8.5).
The sum o hese indings sugges s ha downwa d anspo
o pH equi alen s could be esponsible o dissol ing ikai e
in he in e io ice laye s.
Ano he mechanism ha could a ec he dissolu-
ion/p ecipi a ion dynamics o apped ikai e c ys als in-
ol es con ec i e solu es (Wo s e and We lau e , 1997) en-
su ing con ac be ween he in e io b ine sys em and he un-
de lying wa e column. CO2-en iched b ine can exchange
wi h seawa e ia g a i y d ainage (No z and Wo s e , 2009)
i he b ine olume is abo e 5% o allow e ical ice pe -
meabili y (Cox and Weeks, 1975; No z and Wo s e , 2009).
In his s udy, b ine olume ac ions we e abo e 5% in he
lowe 50cm o sea ice a ICE I, and h oughou he ice col-
umn a POLY I. Thus, he inc ease in pH in he sea ice laye s
close o he wa e column could be caused by eplenishmen
o b ine wi h su ace seawa e (pH 8.1–8.3).
4.2 Role o ikai e in seawa e CO2sys em and
gas exchange
The obse a ion ha ikai e c ys als a e apped wi hin he
sea ice ma ix, while CO2dissol ed in he b ine h ough
ikai e p oduc ion (Eq. 1) can be mobile (los om he ice
wi h b ine), means ha ikai e c ys al o ma ion inc eases he
amoun o CO2a ailable o expo beyond ha a ibu ed
solely o he solubili y e ec . The implica ion is ha , in an
open sys em, TA is p e e en ially s o ed in he sea ice as
ikai e, aising he bu e ing capaci y o sea ice (and su ace
wa e ) upon mel ing and subsequen c ys al dissolu ion. Wi h
apped ikai e c ys als, ice mel will he e o e lead o bo h
lowe su ace wa e salini y and pCO2. As he low pCO2
mel wa e emains a he ocean su ace due o densi y s a -
i ica ion g adien , CO2 lux om he a mosphe e o he su -
ace will be enhanced (Rysgaa d e al., 2012). Based on da a
om his s udy, he mel ing o sea ice (salini y om Fig. 2,
TA and TCO2 om Fig. 8) a 0◦C and dissolu ion o ob-
se ed concen a ions o ikai e (Fig. 6) would esul in mel -
wa e wi h a pCO2o <15μa m a bo h si es. This alue is
a below a mosphe ic alues o 390μa m and su ace wa e
concen a ions o 315μa m measu ed in his s udy. Hence,
he mel wa e can inc ease he ai –sea CO2up ake.
Du ing ice g ow h, pCO2in he b ine sys em will in-
c ease and educe pH in in e io ice laye s. In he su ace
ice laye , eleased CO2may escape o he a mosphe e o be
anspo ed o deepe ice laye s by downwa d b ine d ainage.
Few measu emen s o ai –sea CO2 luxes ha e been e-
po ed o win e condi ions. Geil us e al. (2012) de ec ed
no CO2 elease o ice su ace empe a u es below −10◦C,
ins ead obse ing a small nega i e lux (in o he sea ice)
o 0.23mmolm−2d−1o e sea ice in he Amundsen Gul ,
Beau o Sea. Mille e al. (2011) measu ed luc ua ing CO2
luxes o e sea ice wi h s ong downwa d luxes in Feb u-
a y in he Sou he n Beau o Sea. Sej e al. (2011) obse ed
ha 99% o TCO2in newly o ming ice o e d illed holes
www. he-c yosphe e.ne /7/707/2013/ The C yosphe e, 7, 707–718, 2013
71
PhD hesis by Do e Haubje g Søgaa d
716 S. Rysgaa d e al.: Ikai e c ys al dis ibu ion in win e sea ice
in me e hick sea ice was ejec ed o he unde lying seawa-
e in Young Sound, G eenland. The sugges ion ha mos o
he CO2 eleased h ough ikai e c ys al o ma ion will sink
owa ds he seawa e in dense b ine is consis en wi h hese
se e al indings.
The a e o CO
2expelled om he sea ice o seawa e e-
mains unclea . Dense b ine p oduc ion in he polynya egion
(Ande son e al., 2004) may p o ide a mechanism o deli e
CO2below he mixed laye , making ikai e p oduc ion in sea
ice a “ca bon pump” ha emo es CO2 om he su ace
ocean o deepe wa e laye s (Rysgaa d e al., 2009, 2011).
Low-densi y ice mel wa e emaining a he su ace will a-
cili a e a mosphe ic CO2deposi ion as a esul o ikai e dis-
solu ion. An ai –sea CO2 lux o −10.6mmolm−2d−1has
been epo ed du ing sp ing and summe o he egion p e-
iously (Rysgaa d e al., 2012).
Geil us e al. (2012) calcula ed, based on he ca bon chem-
is y in he sea ice b ine in equilib ium pa i ioning wi h he
a mosphe e, ha mel ing o a 1.3m hick sea ice co e wi hin
one mon h in he Beau o Sea could lead o CO2 luxes
anging om −1.2 o −3.1mmolm−2d−1, which a e com-
pa able o CO2 luxes measu ed o e mel ponds. A ICE I,
we calcula ed he po en ial in luence o mel ing he en i e
ice co e in o a 20m hick mixed laye ( ypical o sum-
me condi ions a his loca ion) on he CO2 lux in he e-
gion, using he measu ed ice ca bon chemis y (salini y o
6.5, TA o 516μmolkg−1,TCO2o 406μmolkg−1,anda -
e age empe a u e o 0◦C) and he ini ial mixed laye cha -
ac e is ics (a e age empe a u e o 0◦C, salini y o 31.7, TA
o 2276μmolkg−1,andTCO2o 2101μmolkg−1). The e-
sul an condi ions in a 20m mixed laye ( empe a u e o
0◦C, salini y o 30.4, TA o 2184μmolkg−1,andTCO2o
2013μmolkg−1)would cause a 14μa m dec ease in pCO2.
Assuming he mel occu s o e one mon h, he esul an
ai –sea CO2 lux o e u n o p e-mel condi ions would be
−5.9mmolm−2d−1, which alls wi hin he ange epo ed
p e iously.
This calcula ion is based on a single mel e en and does
no accoun o he nume ous cycles o sea ice g ow h
and mel cha ac e is ic o polynya sys ems (Tamu a and
Ohshima, 2011; D ucke e al., 2011). The polynya o ma ion
a POLY I is p edominan ly go e ned by mechanical o cing
caused by no he ly gales; i has been classi ied as a wind-
d i en shel wa e sys em (Pede sen e al., 2010), whe e sea
ice o ma ion is con inuous and ejec ion o CO2 o deepe
wa e laye wi h dense b ine occu s (Ande son e al., 2004).
Ikai e c ys als apped in he o ming sea ice will be expo ed
wi h he ice o mel elsewhe e. Such polynya sys ems a e hus
likely o expo CO2 o dep h e ec i ely. Based on esul s
p esen ed he e, enhanced ice p oduc ion in win e polynyas
would add conside able amoun s o TA o he su ace wa e s
in he o m o ikai e c ys als om sea ice, lowe ing su ace
wa e pCO2upon ice mel and c ys al dissolu ion, and in-
c easing he po en ial o seawa e up ake o CO2.
An in e es ing obse a ion in he polynya egion we ha e
s udied (Sej e al., 2011; Papaky iakou e al., 2013) is ha
he pCO2le els in he wa e column a e e y low, g adu-
ally inc easing wi h dep h om su ace alues o 315μa m
o 360μa m a 80m. Du ing he p esen s udy, he a e age
pCO2concen a ion o he uppe 20m wa e column was
335μa m compa ed o 390μa m in he a mosphe e. As mel -
ing 1m o sea ice will educe pCO2le els by 14μa m in
a 20m wa e column acco ding o he calcula ions abo e,
∼4m o sea ice would need o mel locally o explain he
low pCO2concen a ions in he wa e column. In polynya
a eas he sea ice p oduc ion is usually much g ea e han in-
dica ed by he annual sea ice hickness due o he con in-
ued p oduc ion o ice (McLa en, 2006). The local ice p o-
duc ion a POLY I is hus likely esponsible o much mo e
la ge ice g ow h han he Ap il–May ice hickness o 1.4–
1.6m usually obse ed a ICE I (Rysgaa d and Glud, 2007).
Fu he mo e, he s udy si e is loca ed on he NE G eenland
Sea shel whe e la ge amoun s o sea ice expo ed om he
A c ic Ocean h ough F am S ai subsequen ly mel du -
ing sou hwa d anspo owa d he Denma k S ai (Vinje,
2001). Sea ice con aining ikai e c ys als as obse ed in he
p esen s udy may well explain a la ge pa o he low pCO2
le els in he uppe pa o he wa e column due o mel ing o
sea ice om p e ious summe /au umn in he a ea. Biological
CO2 ixa ion will also con ibu e o he a mosphe ic d aw-
down, bu cons aining his ac o is beyond he scope o he
p esen s udy. The cu en wo k ne e heless suppo s p e i-
ous model calcula ions om he a ea es ima ing ha mel -
ing o sea ice expo ed om he A c ic Ocean in o he Eas
G eenland cu en and he No dic Seas g ea ly inc eases he
seasonaland egionalCO2up akein he a ea (Rysgaa d e al.,
2009). Mo e wo k is equi ed o de e mine how applicable
ou esul s a e o o he egions o he A c ic (o An a c ic),
gi en he a iable na u e o CaCO3among he ew a ailable
ikai e s udies (abo e e e ences) and he seasonal na u e o
pCO2in ice- ee A c ic sys ems (Mucci e al., 2010; Cai e
al., 2010).
5 Conclusions
We epo unique obse a ions o ikai e in unmel ed ice
and e ical p o iles o ikai e abundance and concen a ion
in sea ice o he c ucial season o win e . Ikai e c ys-
als, anging in size om a ew μm o 700μm, we e ob-
se ed o concen a e in he in e s ices be ween he ice
pla ele s in bo h g anula and columna sea ice. Thei con-
cen a ion dec eased wi h dep h om su ace-ice alues o
700–900μmolkg−1ice (∼25×106c ys alskg−1) o alues
o 100–200μmolkg−1ice (1–7×106c ys alskg−1)nea he
ice–wa e in e ace, all o which a e much highe (4–10
imes) han hose epo ed in he ew p e ious s udies. Di ec
measu emen s o TA in su ace laye s ell wi hin he same
ange as ikai e concen a ion, whe eas TA concen a ions in
The C yosphe e, 7, 707–718, 2013 www. he-c yosphe e.ne /7/707/2013/
78 PhD hesis by Do e Haubje g Søgaa d
Ou objec i es a e o in es iga e he physical, adia i e and
he mal en i onmen o os l owe s on new sea ice g own
in a win e polynya o he NE coas o G eenland. We seek
o unde s and he clima e o cing on he o ma ion o hese
l owe s and in pa icula o examine he p ocesses ha c ea e
and d i e hei g ow h and de e io a ion. We hen examine he
po en ial ole o os l owe s in geochemical exchange ac oss
he OSA in e ace and as a mic obial habi a . Mo e speci i -
cally we add ess he ollowing in e ela ed esea ch ques ions:
1. Wha a e he clima ic and geophysical o cing condi ions
associa ed wi h os l owe o ma ion and de elopmen ?
(Clima e Fo cing)
2. How did he abo e condi ions a ec apo di usion p o-
cesses and empe a u es ac oss he young ice su ace and
can we de e mine i a mosphe ic deposi ion o sea ice
sublima ion/e apo a ion o b ine wicking domina ed os
l owe o ma ion? (F os Flowe Fo ma ion)
3. How does he concen a ion o sal in he ice su ace and
os l owe s a ec he ca bona e chemis y o he sea ice
and he exchange o CO2 ac oss he OSA in e ace? (Gas
Exchange)
4. Can bac e ial measu emen s ac oss he OSA in e ace o e
ime help o in o m he os l owe o ma ion p ocess and
he sui abili y o hese unique physical and biogeochemical
en i onmen s as habi a s o mic obes? (Mic obial Habi a )
Me hods
A mul ina ional, mul idisciplina y i eld p og am was coo -
dina ed h ough he Uni e si y o Mani oba’s Canada Excel-
lence Resea ch Chai (CERC) a he Danish/G eenlandic i eld
s a ion, Danebo g. The s a ion is loca ed in he Young Sound
jo d on he NE coas o G eenland (Figu e 1). Gene al de-
sc ip ion o he Danebo g i eld si e and he win e ca bona e
chemis y cycle in he jo d a e ound elsewhe e [Rysgaa d
and Glud, 2007; Rysgaa d e al., 2013].
The hin-ice s a ion, POLY I (74°13.905’N 20°07.701’W, 29-
30 cm hick on 22 Ma ch, snow-co e ed wi h a ying hick-
ness), was si ua ed in a ecu en win e polynya egion abou
3 km o he land as ice edge. An a ea o ~5 × 7 m was opened
nea POLY I a 16:00 GMT on 22 Ma ch o expose he ocean
o he a mosphe e (he eina e e e ed o as he ‘pond’ si e).
The opening o he pond was done using a hand-held ice saw
by cu ing smalle segmen s ha hen we e pushed o he side
unde nea h he ice co e . A ime-lapse came a was ins alled
a he pond si e o documen he de elopmen o os l owe s
as he ice o med in si u. Hal o he pond was eopened on
24 Ma ch a 15:00 GMT; i.e., a e abou 47 h o he ini ial
pond opening. A his ime, he ini ial ice was ~12 cm hick.
The ecu en polynya a his loca ion will occu as open wa e
(as e idenced by sa elli e image y jus p io o ou a i al) o
wi h a young ice co e (like we expe ienced); os l owe s
a e known o occu egula ly on his polynya ice (Rysgaa d
and Glud, 2007).
Componen s o he su ace ene gy balance we e acqui ed us-
ing pu pose-buil me eo ological s a ions loca ed in p oxim-
i y o he pond si e o e he pe iod spanning his expe imen .
Wind speed and di ec ion we e measu ed a he heigh s o 3.8
m and 3.1 m using sonic anemome e s (Gill® Windmas e P o
and Me ek®, model uSonic-3). Incoming and e l ec ed sho -
wa e, and incoming and emi ed longwa e adia ion we e
measu ed a he s a ion (o e a snow su ace) using a ou
componen ne adiome e (Kipp & Zonen®, model CNR4)
a a heigh o 1.5 m. Ai empe a u e and ela i e humidi y
we e measu ed using a HMP45C p obe (Vaisala®) ins alled
in adia ion shields a a heigh o 1.6 m.
Figu e 1. Danebo g i eld s a ion a he
NE G eenland coas (A) showing he lo-
ca ion o he Young Sound jo d (B) and
he newly o med ice in he polynya a
he mou h o he jo d whe e he os
l owe ‘pond’ was cons uc ed (C).

79
PhD hesis by Do e Haubje g Søgaa d
Downwelling inciden scala i adiance and ansmi ed scala
i adiance o e 400–700 nm wa eleng h ange was measu ed
a 1-min in e als a he pond si e using MDS-L pho osyn-
he ically a ailable adia ion (PAR) senso s (Alec Elec onics
Co. L d). The su ace e e ence PAR senso was shielded om
he upwelling componen o he adia ion and ins alled nex
o he pond. Two o he PAR senso s we e ins alled a he ai -
wa e in e ace and a 30-cm dep h in he pond by a aching
hem o a s ing hanging om a ho izon al aluminum pole ha
ex ended ac oss he no he n pond co ne (Figu e 1C).
To i ll gaps be ween measu ed ice hicknesses (H) and su ace
empe a u es (Tsu ace), we used he hou ly a e aged ai em-
pe a u es (Tai ) om he me eo ological s a ion as inpu in an
analy ical ice g ow h model [Mayku 1986]:
–ρice L = kice
dH
d
Tsu ace–Tsw
H+ Fw , (1)
whe e ρice = 920 kg m–3 is he assumed densi y o sea ice
which was kep cons an , L = 333.5 J g–1 he la en hea o
usion, kice = 1.8465 J m–1 s–1 K–1 he mal conduc i i y o sea
ice wi h a salini y o 8 and empe a u e o –5°C, Tsw = –1.7°C
he empe a u e o he unde lying seawa e , Fw = 0 W m–2 is
ocean hea l ux, and
Tsu ace =kice Tsw + c Tai H
kice + c H (2)
The calcula ion o Tsu ace elied on a su ace- o-ai hea ans e
coe i cien C , which was ound by ma ching o obse ed ice
hicknesses and su ace empe a u es. The bes ma ch o obse -
a ions we e ob ained by keeping C a 140 J cm–1 day–1 K–1 o
he i s 7 cm o ice g ow h and hen educing i o 110 J cm–1
day–1 K–1 o he emainde . These alues a e no ably smalle
han he C = 209 J cm–1 day–1 K–1 de i ed by Mayku [1986],
e l ec ing he la ge empe a u e di e ence be ween he wa m
ice su ace and he a mosphe e in his hin ice en i onmen .
The upwelling longwa e adia ion LWu was no measu ed
abo e he pond ice bu calcula ed om Tsu ace ( o a ba e os
l owe ee su ace) using he S e an-Bol zmann law wi h an
emissi i y o 0.985. Su ace and in e io b ine salini ies we e
hen es ima ed assuming a linea empe a u e p o i le om
Tsu ace o Tsw and using he equa ion o he eezing poin o
seawa e ollowing Fo ono and Milla d [UNESCO 1983].
Da a on he In a ed Tempe a u e (IR) o he ice and os
l owe su aces we e collec ed wi h a FLIR sys ems SC660
he mal came a (Wilson ille, OR, USA). The IR came a uses
an uncooled mic obolome e o e he spec al ange o 7.5
o 13.0 μm. The mal images we e s o ed as calib a ed 32-bi
l oa ing poin da a o e a 640 by 480 image plane. Calib a ion
o he FLIR was conduc ed wi h an ex e nal black body and
using an in e nal came a calib a ion sys em. Sys em speci i ca-
ions we e e i i ed h ough a se ies o calib a ion es s con-
duc ed daily. The sys em was capable o an a e age ±0.1°C
p ecision o e he ange o he mal condi ions encoun e ed;
absolu e calib a ion was no es ed bu FLIR indica es i o
be < 1%. Addi ional measu emen s o he ba e ice su ace
empe a u e we e aken wi h a handheld digi al he mome e
(T aceable, model 4000, Con ol Company, USA).
Th oughou his manusc ip we discuss he e ec s o he b ine
skim bu sampled he ‘su ace slush laye ’, which con ains he
b ine and he uppe mos ice g ains o he young ice, o he low-
e mos g ains o he os l owe base, o po en ially some snow
ha migh ha e se led on op o he b ine skim. B ine skim is a
liquid wi h bo h i s salini y and empe a u e con olled by he ac
ha i has o emain abo e i s eezing poin o exis as a liquid,
while he su ace slush addi ionally con ains (sal - ee) ice c ys-
als and possibly p ecipi a ed sal s. Ou esul s show a signi i can
di e ence be ween he salini y o he su ace b ine skim (calcu-
la ed) e sus he bulk salini y o he slush laye . F os l owe and
su ace slush laye samples o salini y and δ18O analysis we e
collec ed wi h a sha p me al spa ula, pu in o plas ic bags, and
mel ed a oom empe a u e p io o analysis. Samples o sea ice
b ine o salini y and δ18O we e collec ed om he op pa o he
ice by sc aping an app oxima e 2-cm deep hole in he su ace
a e he emo al o os l owe s and he su ace slush laye , al-
lowing he in e io sea ice b ine o accumula e o a ew minu es,
hen collec ing i wi h a plas ic sy inge. Sea ice was sampled wi h
a 9-cm diame e ice co e (Ma k II co ing sys em, Ko acs En e -
p ises, Lebanon, USA). A e ex ac ion, ice co es we e imaged
he mally using he a o emen ioned FLIR came a, cu in o laye s,
placed in igh ly closed con aine s, and e u ned o he labo a o y
o mel a oom empe a u e. The salini y o he mel ed os l ow-
e s, su ace slush laye , sea ice, and b ine samples was calcula ed
om conduc i i y and empe a u e using a HACH SENSION5
po able conduc i i y me e (Hach, Lo eland, US A (± 0.01))
calib a ed agains a 15N KCl solu ion a 20°C. Samples o oxy-
gen iso ope composi ion we e ans e ed in o glass ials igh ly
capped wi h polyseal closu es and pa a i lm. Analysis was pe -
o med on a Pica o Iso opic Wa e Analyze , L2120-I (Pica o,
Sunny ale, USA) equipped wi h a PAL au osample (Leap Tech-
nologies, Ca bo o, USA). De ails o he me hod can be ound
elsewhe e [e.g., Ve s eegh e al., 2012]. Resul s a e exp essed in
s anda d δ18O no a ion wi h he V-SMOW s anda d as a e e -
ence alue. Ag eemen be ween iple consecu i e injec ions o
he same sample was wi hin ± 0.1 ‰. The δ18O signal o os
l owe s o a mosphe ic o igin was assumed o be ha o eshly
allen snow; he δ18O signa u e o os l owe s o b ine o igin
was calcula ed om δ18O measu ed in b ine using a ac iona ion
alue o +2.6 [Macdonald e al., 1995].
In he cold labo a o y (–25 o –20°C), co esponding sea ice
co es we e cu in o e ical sec ions (10 × 6 × 1 cm) and ho -
izon al sec ions (6 × 6 × 1 cm), which hen we e moun ed
on o sligh ly wa med glass pla es, cooled un il ozen on o he
pla es, and hinned o 1–2 mm hickness using a mic o ome
(Leica® SM 2010R) and a wood plane [e.g., Ehn e al., 2007].
Each o hese, so called hin sec ions, was hen pho og aphed
80 PhD hesis by Do e Haubje g Søgaa d
(Nikon® D70) be ween c oss-pola ized i l e s o documen
ice ex u e. Subsequen ly, each hin sec ion was inspec ed (in
he i eld) unde a s e eomic oscope (Leica® M125 equipped
wi h a Leica® DFC 295 came a and Leica® Applica ion Sui e
e . 4.0.0. so wa e) o documen he e ical and ho izon al
posi ion o ikai e c ys als in sea ice.
Samples o ikai e om he pond si e we e kep a –20°C o
h ee weeks and b ough o he x- ay labo a o y a he Depa -
men o Geological Sciences a he Uni e si y o Mani oba,
Canada. The e, subsamples o sea ice and os l owe s we e
moun ed on o a cold glass slide es ing on a chilled aluminum
block con aining a 1-cm cen al iewing hole. The c ys als we e
i s examined wi h a pola ized ligh mic oscope o assess hei
op ical p ope ies and hen moun ed o x- ay s udy using a s e-
eo binocula mic oscope. The x- ay di ac ion ins umen con-
sis ed o a B uke D8 h ee-ci cle di ac ome e equipped wi h
a o a ing anode gene a o (MoKα X- adia ion), mul i-laye op-
ics, APEX-II CCD de ec o , and an Ox o d 700 Se ies liquid-N
C yos eam. The in ensi ies o mo e han 100 e l ec ions we e
ha es ed om six ame se ies (each spanning 15° in ei he ω
o φ) collec ed o 60° 2θ using 0.6s pe 1° ame wi h a c ys al-
o-de ec o dis ance o 5 cm. Fu he de ails on he ikai e sam-
pling a e p o ided elsewhe e [Rysgaa d e al., 2013].
The su ace di usi e l ux o CO2 associa ed wi h he e- eez-
ing pond was sampled bo h in he p esence and absence o
os l owe s using an au oma ed chambe l ux sys em (LI-
COR®, model LI-8100) equipped wi h a 20-cm (diame e )
su ey chambe . As pa o he pos -p ocessing p ocedu e, he
CO2 concen a ion ime se ies associa ed wi h each sample
was sc eened, a e which he l uxes we e calcula ed using he
LI-8100A da a analysis so wa e.
F os l owe s o mic obial analyses we e emo ed om he
pond si e in o s e ile 1-L plas ic bags using an e hanol- insed
spa ula. A second sc aping o e he same su ace a ea yielded
he co esponding, ope a ionally de i ned b ine skim; i.e., he
su ace slush laye . Samples o sea ice we e also collec ed, as
desc ibed abo e, along wi h samples o seawa e and snow
om he su ounding a ea. Samples o os l owe s, he un-
de lying su ace slush laye , and snow we e mel ed di ec ly
o e he sho es pe iod possible (always < 12 h, wi h sam-
ple empe a u e emaining a ≤ 0°C), while sea ice samples
we e mel ed in o s e ile 0.2-μm i l e ed b ine acco ding o he
isohaline app oach desc ibed by Ewe e al. [2013]. Immedi-
a ely upon mel ing, samples we e i xed wi h 0.2 μm- i l e ed
o maldehyde o a i nal concen a ion o 2% and s o ed in he
cold and da k un il o al bac e ial ( o al p oka yo ic) abun-
dance was de e mined using epi l uo escence mic oscopy, as
in Bowman and Deming [2010]; i us-like pa icles (VLP)
we e enume a ed on a subse o samples, acco ding o Wells
and Deming [2006]. Pa icula e and dissol ed ex acellula
polysaccha ide subs ances (pEPS and dEPS) we e quan i i ed
using he phenol sul u ic acid assay, as in K embs e al. [2011]
and Ewe e al. [2013]. Salini ies o samples used o hese
bac e ial and EPS analyses we e de e mined by e ac ome e .
To de e mine he dominan membe s o he bac e ial commu-
ni y, DNA was ex ac ed om he di e en sample ypes o
ampli i ca ion and sequencing o he 16S RNA gene using he
phenol-chlo o o m me hod (as in Bowman e al. [2013]); one
pa ch o os l owe s was sampled o ob ain he uppe cen im-
e e po ions sepa a ely om he basal po ions. The V3-V5 e-
gion o he 16S RNA gene was ampli i ed using p ime s 357F
and 926R o 30 cycles. An aliquo o he ampli i ed ma e ial,
along wi h posi i e and nega i e con ols, was isualized on a
gel o insu e p ope agmen leng h. Amplicons we e pu i i ed
using he GeneJe Pu i i ca ion Ki (Fe men as) and submi ed o
he Tu s Uni e si y Sequencing Cen e , whe e amplicons un-
de wen a second ound ampli i ca ion o 10 cycles using ba -
coded p ime s 517F and 967R. Second ound amplicons we e
gel-pu i i ed p io o lib a y cons uc ion. Sequencing was con-
duc ed on he 454 FLX pla o m (Roche) using Ti anium chem-
is y. Read p ocessing and classi i ca ion using Mo hu [Schloss
e al., 2009] ollowed Bowman e al. [2012] excep ha he
G eengenes e e ence axonomy was used o classi i ca ion
(a ailable a he Mo hu websi e, h p://www.mo hu .o g/wiki/
G eengenes- o ma ed_da abases); sequence da a a e a ailable
unde NCBI accession numbe SRP038953). B ine and Milli-Q
blanks we e p ocessed along wi h he en i onmen al samples.
Ope a ional axonomic uni s (OTUs, de i ned a he le el o 97
% simila i y) ound in he blanks we e emo ed om he en-
i onmen al samples p io o downs eam analysis. OTUs ha
classi i ed as chlo oplas s we e also emo ed, assuming hese
eads o be de i ed p ima ily om chlo oplas s [al hough see
Diez e al., 2001; Wale on e al., 2007; Co ell and Ki chman,
2009; and Bowman e al., 2012]. Simila i y in communi y com-
posi ion be ween samples was assessed by andomly sampling
o he dep h o he shallowes sample (1,367 eads), calcula -
ing he ela i e abundance o he 25 mos abundan OTUs, and
applying hie a chal clus e ing o samples using he comple e
linkage me hod h ough he egdis and hclus unc ions as im-
plemen ed by he hea map unc ion in R [Oksanen e al., 2013].
Resul s and Discussion
Clima e Fo cing
The clima e a he Danebo g si e was ypical o condi ions
expe ienced in mid-win e in he Young Sound egion o NE
G eenland [Rysgaa d and Glud, 2007]. Ai masses a e in l u-
enced by he p oximal G eenland ice shee . S o ms a e gen-
e a ed h ough he in e ace o wa m, mois ai masses, as-
socia ed wi h he open wa e a eas o he ma ginal ice zones
su ounding G eenland, in e ac ing wi h he cold d y ai mass-
es associa ed wi h he ice cap.
81
PhD hesis by Do e Haubje g Søgaa d
Figu e 2 summa izes he salien da a om he ‘pond’ me
s a ion. A e a b ie pe iod o >10 m s–1 winds p io o he
opening o he os l owe pool on 22 Ma ch, wind speeds
emained be ween 1 and 3 m s–1 du ing he i s 2 days o he
ice-g ow h expe imen . This pe iod was ollowed by 3 days o
a iable winds gene ally be ween 1 m s–1 and 5 m s–1, eaching
abou 8 m s–1 b ie l y a ound noon on 27 Ma ch (Figu e 2A).
These ini ial low wind speeds a e hough necessa y o os
l owe o ma ion as u bulence du ing highe wind speeds (>5
m s–1) des oys he nea -su ace supe sa u a ed laye [S yle
and Wo s e , 2009].
To allow a di ec compa ison wi h esul s om S yle and Wo -
s e [2009] he measu ed ela i e humidi y epo ed he e is
wi h espec o wa e su aces (as ou pu om senso ) wi h no
co ec ion o ice su aces [see And eas e al., 2002]. Du ing
he g ow h expe imen , ela i e humidi y emained be ween
70 and 85% (Figu e 2B). We no e h ee apid d ops in he
ela i e humidi y on Ma ch 20, 21 and 26 below a smalle
unning mean la gely domina ed by diu nal l uc ua ions due
o changing ai empe a u e.
Ai empe a u e a he s a o he expe imen was a ound –26°C
and d opped o –32°C, he lowes du ing he expe imen , in he
ea ly hou s o 23 Ma ch (Figu e 2C). Ai empe a u es showed
an inc easing end du ing he expe imen , eaching as high as
–10°C on 27 Ma ch, and unde wen diu nal cycling which was
in e up ed o he nigh be ween 25 and 26 Ma ch due o he
p esence o cloud co e (no e pe iod o inc eased downwelling
longwa e adia ion du ing 26 Ma ch in Figu e 2D). The co -
esponding ice su ace empe a u e, calcula ed using Eq. (2),
apidly dec eased o –10°C abou 8 hou s a e he ini ial open-
ing o he pond (Tsu ace1 in Figu e 2C) in esponse o he cold ai
empe a u e and ice hickness g ow h o 4 cm hickness (Hsu -
ace1). The dec ease o –10°C ook 17 hou s o he pond po ion
eopened a ound 15:00 on 24 Ma ch (Tsu ace2 in Figu e 2C), e-
l ec ing he e ec o p e ailing wa me ai empe a u es and ice
hickness g ow h o 5 cm (Hsu ace2). A e he ini ial dec eases,
he su ace empe a u es emained be ween –15 and –6°C o
bo h su aces un il he end o he expe imen when (calcula ed)
ice hicknesses had eached 20 cm o Hsu ace1 and 13 cm o
Hsu ace2 (Figu e 2C).
Scala i adiance o e PAR wa eleng hs was measu ed on he
side o he pond ha was eopened (Figu e 1C; Figu e 2E).
Thus, PAR was measu ed ini ially in ice ha g ew apidly
o abou 12 cm and, o he la e pa o he expe imen , in
ice ha g ew mo e slowly o 13 cm (Figu e 2C). Fo bo h
ice-g ow h phases, scala PAR a abou 30 cm below su ace
ollowed closely he alues eco ded by he cosine-co ec ed
downwelling PAR a he me eo ological owe . The nea in e -
ace PAR, howe e , di e ed be ween he wo cases due bo h
o how he su aces e ol ed ( ull os l owe co e age s.
pa chy co e age) and o di e ences in whe e he senso was
loca ed ela i e o he ice su ace ( op o senso loca ed abou
5 mm below ice su ace s. op o senso a su ace). These
di e ences esul ed in no ably highe in e ace PAR alues
du ing he second g ow h phase and illus a e he impo ance
o nea -su ace p ope ies and os l owe s on ligh ansmis-
sion h ough hin ice. The obse ed PAR le els ha eached
alues > 500 μmol quan a s–1 m–2 a e mo e han su i cien o
suppo p ima y p oduc ion [e.g., Kühl e al., 2001]. This e-
sul may ha e ecosys em implica ions, conside ing he much
la ge spa ial expanse o young, os l owe -co e ed sea ice
in Ma ch in he A c ic (e.g., housands o squa e kilome e s
in he Sou he n Beau o Sea in Feb ua y–Ap il o 2013;
h p://www.you ube.com/wa ch? =KXjb6MRj_5U).
0
500
1000
1500
2000
2500
PAR (ƫmol quan a s m )
-2-1
E (inciden )
d
E (inciden )
0d
E (-0 cm)
0
E (-30 cm)
0
3/19 3/20 3/21 3/22 3/23 3/24 3/25 3/26 3/27
0
50
150
250
350
Radia i e lux (W m )
-2
-50
SWĻ
SWĹ
LWĻ
LWĹQ
*
snow snow snow
LWĹ
LWĹ
su ace2
su ace1
65
75
85
Rela i e
humidi y (%)
-10
0
10
20
Wind (m/s)
A
B
C
D
E
-35
-30
-25
-20
-15
-10
-5
0
Tempe a u e (°C)
T
ai
T
su ace1
T
su ace2
H
su ace1
H
su ace2
H
obse a ion
20
15
10
5
0
25
Ice hickness (cm)
Figu e 2. Su ace ene gy balance condi ions du ing he os l ow-
e g owing expe imen a he pond si e: (A) wind ( he leng h o a -
ows co esponds o wind speed; angle o he a ows co esponds o
wind di ec ion wi h no hwa d wind poin ing upwa d); (B) ela i e
humidi y; (C) ai empe a u e (Tai ) measu ed a he me eo ological
s a ion, ice hickness obse a ions (Hobse a ion), and calcula ed ice
hicknesses (Hsu ace) and su ace ice empe a u es (Tsu ace) whe e
subsc ip s wi h 1 and 2 deno e alues o he ini ially opened and
eopened po ions o he pond, espec i ely; (D) sho wa e (SW),
longwa e (LW) and ne (Q*) adia i e l ux whe e he subsc ip s de-
no e he su ace ype (1, 2, o snow), a ows ei he downwelling o
upwelling; and (E) PAR (E) whe e subsc ip 0 is o scala i adi-
ance and d is o downwelling.
82 PhD hesis by Do e Haubje g Søgaa d
A pa ch o ma u e os l owe s was imaged wi h a FLIR cam-
e a o asce ain he he mal i eld condi ions o he os l owe -
co e ed and ba e ice su ace (Figu e 3A and B, espec i ely).
A pa ch o os l owe s adia ed ou om he ini ial node o
os l owe o ma ion (Figu e 3A. The pa ch consis ed o ma-
u e os l owe s ex ending > 1 cm ( e ically) in o he bound-
a y laye . Top po ions o he os l owe s we e a a empe a-
u e o ~ –7°C, ice su ace empe a u e was ~ –4.5°C, and ai






 


Tempe a u e (°C)
Rela i e pixel loca ion along he p o iles in images A and B.
Figu e 3. The mal IR empe a u e o
young ice showing he empe a u es
o a clus e o os l owe s (A) and
a e his clus e was emo ed (B). P o-
i les in he colo FLIR images a e plo -
ed om le o igh acco ding o he
p o i le line loca ion a A and B (23
Ma ch 2013). La ge and mo e ma u e
l owe s in he clus e ha e a lowe su -
ace empe a u e. Ice su ace empe a-
u es d opped by abou 0.25°C due o
he emo al o he os l owe s.
Figu e 4. The mal IR empe a u e o
young ice showing he empe a u es o
he ice su ace ela i e o ha o h ee
os l owe s. P o i les in he colo FLIR
images a e plo ed om le o igh ac-
co ding o he p o i le line loca ion a
A (23 Ma ch, 19:15 GMT, ~27 hou s
om eeze-up) and B (23 Ma ch,
20:15 GMT, ~28 hou s om eeze-
up). The di e en lines illus a e he
change in ice su ace empe a u e ha
occu ed du ing 1 hou ela i e o he
change in he size and empe a u e o
h ee os l owe s.






 



 
Tempe a u e (°C)
Rela i e pixel loca ion along he p o iles in images A and B.
83
PhD hesis by Do e Haubje g Søgaa d
empe a u e was ~ –27°C. A e os l owe emo al, he ice
su ace empe a u e dec eased by ~ 0.25°C due o di ec con-
ac wi h he cold ai jus abo e he g owing ice (Figu e 3B).
On a e age we ound a di e ence be ween l owe ops and
b ine skim o abou 5°C wi h a ange o ± 1°C based on da a
shown in Figu es 3 and 4 and da a no shown.
The empe a u e i elds associa ed wi h he g ow h o indi-
idual os l owe s we e eco ded wi h he FLIR came a by
cu ing and emo ing a small sec ion o he ice, hen allowing
he ice o e- eeze. A one-hou ime lapse be ween images o
h ee g owing os l owe s (compa e Figu e 4A and B) shows
he he mal e ec o he os l owe o ma ion on he su ace
ice i eld and he e ec o hickening sea ice on he su ace
i eld empe a u e (Figu e 4C). The a e age ela i e di e ence
in he p o i le lines in Figu e 4A and B illus a es he di e ence
in ice su ace empe a u e (A-B) o abou 2.5°C c ea ed by
ice g ow h alone. The h ee l owe s in he he mal IR image
ame g ew 0.5 cm o e 1 hou , which esul ed in a empe a-
u e di e ence o ~ 2.5°C (absolu e di e ence in empe a u e
be ween 19:15 and 20:15 GMT on 23 Ma ch). The g ea es
empe a u e di e ences co espond o he la ge os l owe s
p ojec ing he u hes in o he a mosphe ic bounda y laye .
The lowe su ace empe a u es o he os l owe s dampen
longwa e adia ion loss by be ween ~ 10 Wm–2 (~ 3.6%) and
~ 20 Wm–2 (~ 8%) ela i e o backg ound sea ice and b ine
skim (obse ed ΔTsu ace = –25 and –5°C, espec i ely). To
place he e ec in con ex , he change in longwa e emission
associa ed wi h os l owe o ma ion ela i e o backg ound
b ine-we ed sea ice is be ween 50% and 60% o epo ed win-
e ime sensible hea loss om newly o med g ease ice (Table
1 in Else e al. [2011]), and hence a signi i can p opo ion o
he hea budge o he younges o ms o sea ice.
F os l owe o ma ion
The i s os l owe s we e eco ded a 16:15 GMT on 22
Ma ch ( e y small whi e nodules, Figu e 5A), a e only 15
minu es om when he pond was ini ially opened. Ini ial os
l owe s o med as dis inc nodules in he e y young ice, and
hen expanded om hei ini ial nuclea ion si es adially ou -
wa ds in all di ec ions. F os l owe s g ew signi i can ly o e -
nigh ; by he nex mo ning hey exhibi ed a pa chy dis ibu ion
wi h 9/10 o he pond su ace co e ed. F os l owe s we e well
o med by abou 12 hou s in o he expe imen and exhibi ed
a ypical b anched s uc u e a maximum g ow h (Figu e 5C).
δ18O ± SD [‰] S ± SD T [ºC] υb [%]
B ine –4.9 ± 0.7 88.8
F esh snow –24.6 ± 0.4 0.0
Ice 19h
0.0-2.5 cm –1.6 ± 0.1 10.4 –8.7 8.0
2.5-5.0 cm –1.9 ± 0.5 11.2 –5.4 14.1
5.0-8.0 cm –1.2 ± 0.3 8.7 –2.1 23.9
Ice 26h
0.0-2.0 cm
2.0-5.0 cm 7.6 –5.3 6.6
5.0-8.0 cm 8.6 –2.2 17.7
8.0-11.0 cm 9.2 –1.8 23.6
11.0-14.0 cm 6.6 –1.8 16.8
14.0-17.0 cm 9.5 –1.7 25.9
F os l owe s
19h –0.8 ± 0.3 125.1 –12.7 43.8
26h –1.6 ± 0.3 94.9 ± 13.8 –7.8 56.6
46h –2.1 ± 0.3 90.9 ± 10.1 –7.1 58.9
46h ( ops) –4.0 ± 0.2 92.2 ± 16.5
46h (bo oms) –0.1 ± 0.4 89.6 ± 3.8 –7.1 58.9
70h –4.4 ± 1.2 87.0 ± 9.7
90h –4.4 ± 0.3 64.4
90h ( ops) –7.7 50.2
90h (middles) –4.6 ± 0.4 78.8
90h (bo oms) –0.9 ± 0.3 64.1
114h –4.3 ± 0.1 77.4
Su ace slush laye
19h –0.1 ± 0.5 48.2 –10.3 22.7
26h –0.9 ± 0.4 66.2 ± 7.6
46h 2.1 ± 0.9 77.4 ± 19.1 –8.7 41.7
70h –0.4 ± 0.5 72.7 ± 11.0
90h –0.4 ± 0.0 52.9
114h –1.3 ± 0.5 72.3
Table 1. Measu emen s o δ18O, S, T, and υb in samples o sea ice, b ine, esh snow, os l owe and he su ace slush laye h oughou he
du a ion o he os l owe g ow h expe imen .

84 PhD hesis by Do e Haubje g Søgaa d
O e he cou se o he second nigh , he ini ially pa chy co e -
age e ol ed in o a ela i ely homogeneous co e (lowe pa
o Figu e 5B). A b ine skim o med e y ea ly in he g ow h
cycle o he new ice. This b ine skim is o en seen on young
g owing ice and is he esul o b ine being ejec ed upwa ds
du ing ice g ow h. When he os l owe s o m, his b ine
skim can p o ide bo h a apo sou ce ( o os l owe g ow h)
and b ine o sal mig a ion in o he os l owe s. Wo k on
chemical composi ion o os l owe s co obo a es his ba-
sic p ocess o mic os uc u e de elopmen [Al a ez-A iles e
al., 2008]. In wha ollows we discuss he e ec s o he b ine
skim, ecognizing along he way ha wha we sampled was
he ‘su ace slush laye ’, as de i ned abo e.
The en i onmen al condi ions o os l owe o ma ion du ing
he expe imen we e placed in he con ex o he six egimes
hypo hesized in he SW model based on ela i e humidi y o he
a mosphe e and empe a u e di e ence be ween he a mosphe e
and he ice su ace (Figu e 6; he numbe ing I-VI ollows ha
o S yle and Wo s e [2009]). E apo a ion in o s. condensa ion
om a supe sa u a ed a mosphe e ( ela i e humidi y > 100%)
is cha ac e ized by egimes I and II, espec i ely. Regimes III
o VI a e o exchanges wi h an unde sa u a ed a mosphe e.
Condensa ion om he ai occu s a he su ace in III-IV, while
e apo a ion om he su ace occu s in V-VI. Regimes V and
VI dis inguish be ween unde sa u a ed s. supe sa u a ed wa e
apo condi ions in he nea -su ace laye , as de e mined by he
Clausius–Clapey on equa ion. In bo h V and VI, he sou ce o
wa e apo in he laye di ec ly abo e he sea ice comes om
e apo a ion o sublima ion om he sea ice su ace a he han
om deposi ion om he cold a mosphe e. Fo mos o he ex-
pe imen , pond su ace condi ions emained wi hin egime VI
wi h a b ie ansi ion in o egime V a e abou 84 hou s o
he ini ial pond opening (Figu e 6). S yle and Wo s e [2009]
obse ed ha la ge alues o supe sa u a ion, depic ed by con-
di ions o he le o he dash-do ed cu e in Figu e 6, we e
equi ed o os l owe o ma ion o occu . Ou obse a ions
closely co obo a e his i nding.


Figu e 5. Examples o os l ow-
e s g own a he pond. (A) The
i s e idence o os l owe o -
ma ion (see small whi e do s wi h-
in he black ec angle) a ~ 3 hou s
in o he expe imen (ice < 1 cm
hick; 22 Ma ch 19:00 GMT); (B)
a pa chy os l owe co e age
on he eopened pond hal a e ~
24 hou s o g ow h (ice < 10 cm
hick). Fo eg ound shows he ma-
u e os l owe i eld on he ini-
ially g own ice (~ 17 cm hick);
and (C) a close-up image o a ypi-
cal os l owe s uc u e a ~ 48
hou s in o he expe imen .
Figu e 6. The ela i e di e ences in humidi y and empe a u e du -
ing he pond expe imen se in he con ex o en i onmen al condi-
ions unde which di e en condensa ion, e apo a ion and sublima-
ion p ocesses occu , ollowing S yle and Wo s e [2009]. Red ( i s
opening) and blue (second opening) colo ed numbe s indica e he
ime (hou s) elapsed in 4-hou in e als om he s a o eeze-up
(i.e., om 22 Ma ch a 16:00 GMT and 24 Ma ch a 15:00 GMT,
espec i ely); hei loca ions show he p e ailing condi ions a he
ime. Roman nume als I h ough VI a e he ca ego ical condi ions
unde which a ious apo - empe a u e di e en ials c ea e a ious
ea u es: os l owe , ime, og, dew, e c. The dashed and dash-do -
ed lines a e ed awn om S yle and Wo s e [2009, Figu es 3-4],
while he adjacen solid lines a e ob ained using he same equa ions
bu he empe a u es obse ed du ing he pond expe imen .
85
PhD hesis by Do e Haubje g Søgaa d
In he li e a u e, he e is s ill a deba e as o whe he os l ow-
e s o m due o deposi ion o a mosphe ic wa e apo on o
he ice su ace o sublima ion o e apo a ion o he wa m ice
su ace in o he a mosphe ic bounda y laye . Measu ed δ18O
alues o os l owe s om 19 o 114 hou s a e eeze-up
sugges ha he os l owe s we e composed p ima ily o
b ine (Figu e 7A). F esh (19-h) unsec ioned os l owe s had
a δ18O alue o –0.8 ± 0.3‰ which was simila o he signa-
u e o he su ace slush laye (–0.1 ± 0.5‰) and dec eased
g adually wi h ime o –4.3 ± 0.1‰ a e 114 hou s om
eeze-up (Table 1), poin ing o he in l uence o seconda y a -
mosphe ic deposi ion ei he om apo o om blowing snow
(δ18O alue o –24.6 ± 0.4‰) du ing la e s ages o he expe i-
men . In l uence o ha seconda y a mosphe ic deposi ion is
pa icula ly p onounced in he op sec ions o os l owe s,
while bo om sec ions emained composed almos exclusi ely
o b ine (Figu e 7A). Resul s p esen ed he e a e gene ally in
ag eemen wi h simila s udies on A c ic leads [Douglas e al.,
2005] and o e la ge open wa e expanse o ma ions o young
os l owe co e ed sea ice [Douglas e al., 2012].
Ini ial salini ies o os l owe s we e much highe han hose
measu ed in he su ace slush laye (Figu e 7B), s ongly sup-
po ing he b ine o igin o os l owe s e ealed by he δ18O
signa u e analysis (and u he co obo a ed by he bac e ial
esul s; see below). The low empe a u es o he os l owe s
(Figu e 3) and, pe haps mo e impo an ly, he di ec con ac o
b ine wi h he a mosphe e allowing o high e apo a ion a es,
sugges ha p ecipi a ed sal s mus be p esen [Assu , 1960],
possibly explaining he high salini ies o mel ed os l owe s.
Only a e 46 hou s o he expe imen , when a δ18O a mos-
phe ic signa u e began o appea in he os l owe s and he
collec ed b ine skim had a highe salini y, did he os l owe
salini ies app oach hose measu ed o he b ine skim. P e i-
ous wo k has shown ha b ine can be wicked up in o os
l owe s om a high-densi y b ine skim, bo h in si u [Pe o ich
and Rich e -Menge, 1994] and in labo a o y expe imen s
[Ma in e al., 1995; Roscoe e al., 2011] and sea ice meso-
cosm expe imen s [Islei son e al., 2012]. Lack o da a p io o
19 hou s a e eeze-up in ou expe imen p e en s a de i ni-
i e de e mina ion o ini ial ice-su ace o igin, bu o ma ion
o nuclea ing nodules om ozen b ine and/o p ecipi a ed
sal s seems plausible as: 1) a mosphe ic empe a u es we e
below he empe a u e o eezing o b ine wi h he maximum
modeled salini y o ~ 200 (~ –10 ºC; Figu e 7B), po en ially
allowing o he p ecipi a ion o sal s o occu a e cooling
o b ine o < –5ºC [Ligh e al., 2003]; and 2) ini ial salini ies
o os l owe s we e much highe han hose measu ed in he
su ace slush laye wi hin he ange o modeled salini ies o
he liquid b ine in he op 5 cm o ice (Figu e 8).
B ine skim empe a u es (Figu e 2C) calcula ed using Eq. (2)
ag eed well wi h measu ed su ace slush laye empe a u es
(Table 1). The salini y o he b ine in he su ace slush laye
could hen be es ima ed using seawa e eezing-poin ela-
ionships [Fo on and Milla d, 1983]. The di e ences in sa-
lini y among he samples hen allowed es ima es o he b ine
olume wi hin he su ace slush laye , calcula ed as: 31, 49,
54, 41, 37, and 59% a 19, 26, 46, 70, 90 and 114 hou s, e-
spec i ely, in o he ice-g ow h expe imen . Fu he mo e, as-
suming ha he b ine om he b ine skim wicks in o he os
l owe and o ms pa o i s s uc u e, we es ima ed he ela-
       
í
í
í
í
í
í
í
í
í
í
í
í


!"
!"#&!&
!"#&!# 
!"#&!
!"#&!"
!!
#" !!
#" !!
!
!
#" !% !!##
!
"$!
!"$!
!"$!
$#
$#

į

)

       










 

#'
!!($ $!"
Figu e 7. Tempo al e olu ion o δ18O (A)
and salini y (B) in os l owe s and su -
ace slush laye o e he du a ion o he
os l owe g owing expe imen ha s a ed
16:00 GMT on 22 Ma ch.
86 PhD hesis by Do e Haubje g Søgaa d
i e p opo ions o he sampled os l owe s ha we e o med
h ough b ine wicking e sus apo deposi ion: o he abo e
six sampling occasions, wicked b ine composed abou 81, 70,
64, 49, 45, and 63%, espec i ely, o he mass o he os
l owe s. This exe cise illus a es he impo an ole o b ine
wicking, no only in ans e ing sal (and o ganic ma e , in-
cluding bac e ia), bu also in b inging wa e o he os l ow-
e s. O e he cou se o he ice-g ow h expe imen he e was a
gene al dec ease in he b ine-wicked ac ion and consequen
inc ease in deposi ion om apo , a pa e n en i ely consis en
wi h he δ18O signa u e analysis ha also sugges ed a empo-
al inc ease in he a mosphe ic ac ion (see abo e). Despi e
in e nal consis ency, his app oach o es ima ing he composi-
ion o os l owe s could no accoun o possible changes in
b ine salini ies o e he cou se o he os l owe o ma ion
p ocess, o example due o e apo a ion o sublima ion.
Gas Exchange
The clima e o cing and esul ing os l owe o ma ion con-
cen a e b ine in he uppe laye o he ice and wi hin he os
l owe s. This b ine concen a ion a ec s he ca bona e chem-
is y o he sea ice and hus gas exchanges ac oss he OSA in-
e ace, speci i cally h ough he in l uence o ikai e o ma ion.
Ikai e c ys al concen a ions in os l owe s and b ine skim
we e cha ac e ized and quan i i ed by sampling a di e en imes
and dis ances (cen ime e -scale) in he newly ozen pond (Fig-
u e 9A). Wi hin 1 hou ikai e c ys als we e obse ed o o m in
bo h he hin ice and he os l owe s, indica ing he sho - e m
dynamic condi ions o ikai e o ma ion. Ikai e c ys als become
easily isible when sea ice is mel ed. Fo example, 39 c ys als
we e obse ed in a 3 μg subsample a ew seconds a e mel -





C)
Rela i e pixel loca ion

Figu e 8. Tempe a u e, salini y, b ine olume ac ion (υb) and δ18O
p o i les in sea ice du ing he os l owe g owing expe imen a e
19 (A) and 26 (B) hou s om eeze-up. The mal IR image o small
scale empe a u e o a sea ice co e ex ac om he pond si e a 26
hou s in o he expe imen (coinciden wi h empe a u e p obe meas-
u emen s in B).
Figu e 9. F os l owe s a he pond du ing sampling o ikai e c ys-
als o cha ac e iza ion and quan i i ca ion. (A) F os l owe i eld
wi h ma ks a e sampling a di e en cen ime e dis ances in he
lowe pa o he image abo e he ule . (B) Example o ikai e c ys-
als jus a e ice c ys als in he sample had mel ed. (C) Ikai e con-
cen a ions in os l owe s ( i lled ci cles) and su ace slush laye
(emp y ci cles).
87
PhD hesis by Do e Haubje g Søgaa d
ing 58 mg o new os l owe s (Figu e 9B). This image ep e-
sen s a e y small ac ion (3 μg) o he sample. In his case, he
c ys als co e ed 1.5% o he coun ing a ea, co esponding o an
ikai e concen a ion o 2150 μmol kg–1. The c ys als obse ed
in he sea ice anged in size (maximum dimension) om a ew
mic ome e s o 200 μm, we e highly anspa en wi h a hom-
bic mo phology, and showed uni o m ex inc ion unde c oss-
pola ized ligh , sugges ing ha hey we e simple single c ys-
als. All x- ay e l ec ions i well o a monoclinic C-cen e ed
cell wi h a(Å) = 8.816(3), b(Å) = 8.333(2), c(Å) = 11.043(3),
β(°) = 110.68(2), and V(Å3) = 759.1(6). F om he gene al shape,
op ical p ope ies and uni -cell de e mina ion, he c ys als ex-
amined we e ikai e [Hesse and Küppe s, 1983]. A e age ikai e
concen a ions in he os l owe s we e 1013 μmol kg–1 (± 632,
s anda d de ia ion) and in he su ace slush laye , 1061 μmol
kg–1 (± 736, s anda d de ia ion) (Figu e 9C). The highes con-
cen a ion o ikai e was obse ed in he sea ice column [Rys-
gaa d e al., 2013]. The su ace slush laye and os l owe s
ep esen ed a hin low-densi y laye on op o he ice wi h a
s ikingly high abundance o ikai e c ys als (e.g., compa ed o
Geil us e al. [2013]), ye accoun ed o only 5% o he o al
ikai e in he ice column [Rysgaa d e al., 2013]. Fo ma ion o
ikai e in su ace ice laye s, howe e , can lead o a CO2 l ux
om he ice in o he a mosphe e.
Due o logis ical cons ain s we we e only able o sample he
su ace CO2 l ux a he pond on Ma ch 23 and 24. Sampling
was pe o med on ice hickness o app oxima ely 10 and 13
cm, espec i ely (Figu e 2C), wi h ice su ace empe a u es be-
ween –10 and –13°C. Ai empe a u e anged be ween –21.5
and –19°C du ing sampling. Chambe l ux measu emen s con-
i med an expec ed e l ux o CO2 a he b ine-we ed sea ice
su ace (3.35 ± 0.86 mmol m–2 day–1, n = 8), bu he e was no
disce nible di e ence in his e l ux i os l owe s we e wi hin
he chambe oo p in (3.02 ± 0.76 mmol m–2 day–1, n = 4) o
no (3.67 ± 1.99 mmol m–2 day–1, n = 4). These esul s may
no be unexpec ed gi en he small di e ence in ikai e concen-
a ion be ween he os l owe s and su ounding b ine skim.
Few s udies exis agains which o compa e ou esul s. Fluxes
a e consis en in sign, bu gene ally smalle han chambe l ux
measu emen s epo ed by Geil us e al. [2013] o e young,
land as sea ice nea Ba ow, AK. They obse ed an a e age
e l ux o 6.7 mmol m–2 day–1 (4.2 o 9.6 mmol m–2 day–1) based
on ou measu emen s o e ice ha was hicke (20 cm), co -
e ed wi h olde os l owe s, and ma ginally colde (ice su ace
empe a u e o –14.2 °C) ela i e o he newe ice in ou pond
expe imen . Igno ing possible di e ences in sea ice o ganic
ca bon composi ion be ween si es, he la ge e l ux o CO2 a
he Ba ow si e is possibly he esul o colde nea -su ace
empe a u e. On cooling, pCO2 in he b ine sys em o sea ice
ends o inc ease, gi en he likelihood o ikai e p oduc ion, and
because gases a e gene ally less soluble in s onge elec oly e
solu ions (e.g., Thomas e al., 2010). Gi en ha he b ine ol-
ume was > 5% h oughou hei sea ice p o i le, ou gassing could
occu along a b ine- o-ai pCO2 g adien .
Mic obial Habi a
In gene al, he highe he salini y o he os l owe (o b ine
skim) sample, he g ea e i s bac e ial con en acco ding o
Bowman and Deming [2010], whose da a suppo he iew ha
bac e ia in sea ice b ine a e anspo ed upwa ds as he b ine
wicks in o os l owe s. In ou pond expe imen , he highes
concen a ions o bac e ia we e again obse ed in he sal ies
ea u es (in ank o de , on a mel - olume basis): 1.56–4.46 ×
105 cells ml–1 (n = 32) in os l owe s and he su ace slush
laye , compa ed o 1.10–2.11 × 105 ml–1 (n = 7) in seawa e ,
6.58–7.85 × 104 ml–1 (n = 2) in young sea ice, and only 133–353
ml–1 (n = 2) in eshly deposi ed snow (which had no de ec able
salini y by e ac ome e ). An a mosphe ic sou ce o bac e ia
( ep esen ed by esh snow samples) o he os l owe s and
b ine skim a he pond was hus negligible, including e en o
aging os l owe s showing signs by ∂18O analysis o a mos-
phe ic ice-c ys al deposi ion (Figu e 7). The con ibu ion o
apo -de i ed ice o os l owe s, howe e , ep esen s an una-
oidable dilu ion (upon sample mel ing) o he ue bac e ial
densi y in he habi able b ine olume ac ion o hese s uc-
u es. Al hough es ima es o his dilu ion ac o could be made
in a ew cases, compa ing bac e ial abundance ac oss he da a
se equi ed scaling numbe s o mel ed sample olume.
Ea lie sho - e m labo a o y and mesocosm expe imen s o
e alua e bac e ial con en as a unc ion o os l owe age ha e
no been de i ni i e [Bowman and Deming, 2010; Aslam e al.,
2012], bu his i eld expe imen clea ly e ealed highes con-
cen a ions o bac e ia ea ly in he g ow h phase o hese b iny
ice s uc u es (Figu e 10) when salini y was also highes . F os
Figu e 10. Bac e ial abundance in os l owe s (blue squa es) and su -
ace slush laye ( ed ci cles) o e ime since opening he expe imen-
al pond, wi h abundance in seawa e (g een diamonds) ep esen ing
ime ze o. The i s h ee alues o os l owe s de i e om he second
opening o he pond. Second-o de polynomials we e i o he da a p i-
ma ily o help isualize he di e en appa en ime-cou se ajec o ies
o os l owe s (R2 = 0.609) and su ace slush laye (R = 0.237).

95
PhD hesis by Do e Haubje g Søgaa d
PAPER IV
Ma ine Ecology P og ess Se ies 419:31 – 45, doi: 10.3354/meps08845
D.H. Søgaa d • M. K is ensen • S. Rysgaa d • R.N. Glud •
P.J. Hansen • K.M Hilligsøe
Au o ophic and he e o ophic ac i i y in A c ic
i s -yea sea ice: seasonal s udy om Malene
Bigh , SW G eenland
Pho o: Michael S. Sch øde .
96 PhD hesis by Do e Haubje g Søgaa d
3
Vol. 419: 31–45, 2010
doi: 10.3354/meps08845
MARINE ECOLOGY PROGRESS SERIES
Ma Ecol P og Se
Published No embe 30
Au o ophic and he e o ophic ac i i y in A c ic
i s -yea sea ice: seasonal s udy om Malene Bigh ,
SW G eenland
Do e Haubje g Søgaa d
1, 2,
*, Mo en K is ensen
1, 2
, Sø en Rysgaa d
1
,
Ronnie Nøh Glud
1, 4
, Pe Juel Hansen
2
, Ka en Ma ie Hilligsøe
3
1
G eenland Clima e Resea ch Cen e (c/o G eenland Ins i u e o Na u al Resou ces), Ki ioq 2, Box 570, 3900 Nuuk, G eenland
2
Uni e si y o Copenhagen, Ma ine Biological Labo a o y, S andp omenaden 5, 3000 Helsingø , Denma k
3
Aa hus Uni e si y, Depa men o Biological Sciences, Ny Munkegade 114-116, 8000 Å hus C, Denma k
4
P esen add ess:
The Sco ish Associa ion o Ma ine Science, Oban, A gyll PA37 1QA, Sco land, UK
ABSTRACT: We p esen a s udy o au o ophic and he e o ophic ac i i ies o A c ic sea ice (Malene
Bigh , SW G eenland) as measu ed by 2 di e en app oaches: (1) s anda d incuba ion echniques
(H14CO – and [3H] hymidine incuba ion) on sea ice co es b ough o he labo a o y and (2) co es incu-
ba ed in si u in plas ic bags wi h subsequen mel ing and measu emen s o changes in o al O2 con-
cen a ions. The s anda d incuba ions showed ha he annual succession ollowed a dis inc i e pa -
e n, wi h a low, almos balancing he e o ophic and au o ophic ac i i y du ing Feb ua y and Ma ch.
This pe iod was ollowed by an algal bloom in la e Ma ch and Ap il, leading o a ne au o ophic com-
muni y. Du ing Feb ua y and Ma ch, he oxygen le el in he bag incuba ions emained cons an ,
alida ing he low balanced he e o ophic and au o ophic ac i i y. As he au o ophic ac i i y
exceeded he he e o ophic ac i i y in la e Ma ch and Ap il, i esul ed in a signi ican ne oxygen
accumula ion in he bag incuba ions. In eg a ed o e he en i e season, he sea ice o Malene Bigh
was ne au o ophic wi h an annual ne ca bon ixa ion o 220 mg C m–2, e lec ing he ne esul o a
sea ice- ela ed g oss p ima y p oduc ion o 350 mg C m–2 and concu en bac e ial ca bon demand o
130 mg C m–2. Con e ing he O2 ne exchange o he bag incuba ions in o ca bon u no e es ima ed
an annual ne ca bon ixa ion o 1700 ± 760 mg C m–2 (mean ± SD), which was highe han he annual
ne ca bon ixa ion quan i ied in he s anda d incuba ions.
KEY WORDS: Sea ice · P ima y p oduc ion · Bac e ial ca bon demand · Ne au o ophic ac i i y ·
Ne he e o ophic ac i i y · A enua ion coe icien s
Resale o epublica ion no pe mi ed wi hou w i en consen o he publishe
INTRODUCTION
Sea ice p o ides a low- empe a u e habi a o di e se
communi ies o mic oo ganisms including i uses,
bac e ia and he e o ophic (e.g. lagella es and cilia es)
and au o ophic p o is s (e.g. dia oms) (Kaa okallio e al.
2006). Du ing sea ice o ma ion, mic oo ganisms, in-
o ganic solu es and solids can be inco po a ed in o he
ice and accumula e o concen a ions highe han ha
in he unde lying seawa e (Reimni z e al. 1992, G oss-
mann & Glei z 1993, G adinge & Ikä alko 1998).
O ganisms inco po a ed in o sea ice a e challenged wi h
*Email: [email p o ec ed]l
changes in space, ligh a ailabili y, salini y, nu ien s,
dissol ed ino ganic ca bon (DIC) and O2 concen a ion,
empe a u e and pH (G adinge & Ikä alko 1998).
Especially, ligh a ailabili y wi hin he sea ice has a
majo in luence on he sea ice algal biomass and p o-
duc ion (Co a & Ho ne 1989, Rysgaa d e al. 2001).
Sea ice algae cons i u e an impo an componen o
sympagic communi ies and ha e been ex ensi ely
s udied in A c ic sea ice (e.g. Ho ne & Sch ade 1982,
Gosselin e al. 1997, Glud e al. 2007, Mikkelsen e al.
2008). The sea ice algae ep esen an impo an ood
sou ce o me azoan g aze s, and pho osyn he ic p od-
© In e -Resea ch 2010 · www.in - es.com
O
PENN
ACCCEESSSS
97
PhD hesis by Do e Haubje g Søgaa d
32 Ma Ecol P o
g
Se 419: 31–45, 2010

3
uc s o en ained o ganic ma e ial can lead o ele a ed
bac e ia abundance and p oduc ion wi hin he sea ice
(Smi h e al. 1989, K embs e al. 2002, Meine s e al.
2003, Riedel e al. 2007).
Se e al p e ious s udies ha e shown ha he e o-
ophic bac e ia a e ac i e and abundan in A c ic sea
ice (Bunch & Ha land 1990, Smi h & Clemen 1990,
B inkmeye e al. 2003, Lizo e 2003, Kaa okallio
2004), bu s udies on spa ial and seasonal a ia ions
a e ew (Smi h e al. 1989, G adinge and Zhang 1997).
Sea ice ep esen s a pa ially in e connec ed ne -
wo k o b ine- illed channels comp ising 1 o 30% o
he sea ice olume (e.g. Weeks & Ackley 1986). The
deg ee o in e connec ion o he b ine enclosu es is
gene ally enhanced wi h inc easing empe a u e, and
he po en ial accumula ion o sea ice algae he e o e
inc eases owa ds he pola sp ing. The signi icance o
he e o ophic p ocesses ypically inc eases du ing la e
bloom and pos bloom si ua ions close o sp ing haw
(Vézina e al. 1997, Kaa okallio 2004). In addi ion o
biologically media ed dynamics, haw and eezing
p ocesses induce ex ensi e dynamics in solu e and gas
dis ibu ion (Glud e al. 2002). Consequen ly, he sea
ice ma ix is highly he e ogeneous and dynamic, and
quan i ica ion o in si u algae and bac e ial p oduc i -
i y ep esen a ue challenge.
The o e all objec i e o his in es iga ion was o
desc ibe he dynamics o au o ophic and he e o o-
phic ac i i y in i s -yea sea ice. We measu ed
au o ophic and he e o ophic ac i i y o in ac sea ice
co es unde in si u condi ions in bag incuba ions, and
measu emen s we e compa ed wi h p ima y p oduc-
ion and bac e ial ca bon demand as measu ed in he
concen a ions, sea ice algal p oduc i i y and bac e ial
ca bon demand du ing each sampling campaign, he
la e 2 being e e ed o as ‘s anda d incuba ions’.
On 15 Feb ua y and 19 Ma ch, 10 sea ice co es we e
collec ed using a MARK II co ing sys em (Ko acs En e -
p ises). Co es we e cu in o 2 sec ions o equal leng h
(ca. 22 cm), i.e. op hal and bo om hal , which we e
b ough back o he labo a o y in black plas ic bags
wi hin 1 h o sampling. In a labo a o y cold oom (3 ±
1°C), co e sec ions we e placed in lamina ed anspa -
en NEN/PE plas ic bags (Hansen e al. 2000) i ed
wi h a gas- igh Tygon ube and a al e o sampling.
A i icial seawa e (salini y o 33) wi h a known O2
concen a ion was added (10 o 30 ml) o he NEN/PE
bags (Hansen e al. 2000). The bags we e closed, and
excess ai quickly ex ac ed h ough he al e. The op
and bo om hal es o a single ice co e (i.e. he con en s
o one pai o bags) we e mel ed wi hin 48 h in da k-
ness (3 ± 1°C). Gas bubbles eleased om he mel ing
sea ice we e subsequen ly ans e ed o 12 ml Exe-
aine ubes (Labco) con aining 20 Ǎl HgCl2 (5% w/ ,
sa u a ed solu ion). The gas bubbles we e analysed o
gaseous O2 by gas ch oma og aphy in a lame ioniza-
ion de ec o and he mal conduc i i y de ec o
(FID/TCD, SRI model 8610C). The mel ed wa e was
also ans e ed o Exe aine ubes o O2 measu e-
men s. Dissol ed O2 in he mel ed sea ice was mea-
su ed by Winkle i a ion (G assho e al. 1983). The
emaining 9 sea ice co es in plas ic bags we e ans-
e ed o he d illed holes a he sea ice sampling si e
wi hin 2 h, and he in si u snow co e abo e he co es
was e-es ablished. The sea ice co es, i.e. op and bo -
om hal es, we e sampled a 1 o 2 wk in e als o
labo a o y by s anda d H14CO – and [3H] hymidine de e mine he O2 bulk concen a ion o he sea ice.
incuba ions unde well-de ined condi ions.
MATERIALS AND METHODS
Measu emen s we e conduc ed on i s -yea land-
as sea ice in Malene Bigh in he icini y o Nuuk,
SW G eenland (64° 82’ N, 51° 42’ W). Sampling was
pe o med a 1 o 2 wk in e als om 1 Feb ua y o 14
Ap il 2008. Du ing ha pe iod, sea ice hickness a ied
be ween 0 and 62 cm, while snow co e anged be-
ween 0 and 28 cm. The ne ae obic ac i i y o an en-
closed sea ice communi y was ollowed in si u by de e -
mining he O2 concen a ions in sea ice co es sealed in
plas ic bags and placed unde na u al snow co e ,
he ea e e e ed o as ‘bag incuba ions’. These mea-
su emen s we e supplemen ed wi h s anda d measu e-
men s o empe a u e, salini y, i adiance a enua ion in
snow and ice, nu ien concen a ions (phospha e,
Pe meabili y o he NEN/PE bags was quan i ied
by adding a i icial seawa e (salini y o 33) o 9 plas-
ic bags con aining a gas- igh al e o sampling
(Hansen e al. 2000). The bags we e spiked wi h HgCl2
(200 Ǎl o a 5% solu ion pe li e seawa e ) o p e en
biological ac i i y du ing he incuba ion. The wa e
was hen lushed wi h N2 gas un il he O2 concen a-
ion in he bags eached 50% o a mosphe ic sa u a-
ion. The al es in he bags we e hen closed ensu ing
ha no gas phase was le inside he 1 l bag. Th ee
bags we e incuba ed a –20°C, 3 bags a 3°C and 3
bags a 20°C o 7 d. Ini ially he O2 concen a ion was
measu ed in all bags. A e he incuba ion, O2 samples
we e collec ed and measu ed as desc ibed p e iously.
Ice o ma ion in he bags (–20°C expe imen ) led o
bubble o ma ion and he O2 concen a ion was he e-
o e calcula ed as he sum o O2 in gas and wa e . The
maximum O2 lux ac oss he plas ic bags was calcu-
la ed om he O2 concen a ion change be ween ini ial
3–
– – –1 –1
PO
4
; ni a e and ni i e, NO3
+
+ NO2 ; silicic acid, and inal eadings o 0.48 ± 0.20 Ǎmol O2 l d (mean
Si(OH)4; and ammonium, NH4 ), chlo ophyll a (chl a) ± SD). This pu a lowe limi on he ac i i y ha can be
98 PhD hesis by Do e Haubje g Søgaa d
Søgaa d e al.: Au o ophic and he e o ophic ac i i
y
in A c ic sea ice 33

4
+
–
–
esol ed by his p ocedu e and may comp omise o e - (Tu ne Designs). The emaining il e ed sea ice was
3–
– –
all ne ac i i y a low biomass. In eali y he pe meabil- ozen (–18°C) o la e analysis o PO4
+ , NO3
– + NO2 ,
–
i y a he incuba ion empe a u e (–5 o 0°C) was sig-
ni ican ly less han ou maximum alue.
O2 bulk concen a ion in he sea ice (Ci) was calcu-
la ed as:
C W C W
Si(OH)4 and NH4 . Concen a ions o NO3 + NO2 we e
measu ed by anadium chlo ide educ ion (B aman &
Hend ix 1989). Concen a ions o Si(OH)4 and PO 3–
we e de e mined by spec opho ome ic analysis
(S ickland & Pa sons 1972, G assho e al. 1983). Con-
+
C
i
m m a
a
(1)
cen a ions o NH4 we e de e mined by a luo ome ic
W
i
whe e Cm is he O2 concen a ion in he mel ed sea ice
me hod (Holmes e al. 1999). The lowe de ec ion limi
o he nu ien measu emen s we e 0.002 Ǎmol l–1 o
3– –1
– – –1
(gas bubble + mel ed sea ice), Wm, is he weigh o he
PO
4
, 0.5 Ǎmol l o NO3 + NO2 , 0.002 Ǎmol l o
mel ed sea ice, Ca, is he O2 concen a ion in he a i icial Si(OH)4 and 0.10 Ǎmol l–1 o NH4 (lowe de ec ion
sea wa e , Wa is he weigh o he a i icial sea wa e , and
Wi is he weigh o he sea ice (Rysgaa d & Glud 2004).
The season was di ided in o Se ies 1 and Se ies 2.
Se ies 1 consis ed o 10 co es, which we e incuba ed
on 15 Feb ua y and we e indi idually collec ed o
analysis om 15 Feb ua y o 25 Ma ch. Se ies 2 con-
sis ed o ano he 10 co es incuba ed on 19 Ma ch and
we e indi idually collec ed o analysis om 19 Ma ch
o 11 Ap il. Linea eg ession was pe o med on each
o he 4 se ies (i.e. op hal and bo om hal o co es in
Se ies 1 and Se ies 2). The slope o he eg ession line
was es ed in all 4 se ies by means o S uden ’s - es .
On each o he 7 sampling occasions, iplica e ice
co es we e collec ed using a MARK II co ing sys em.
The 3 sea ice co es we e cu in o sec ions using a s ain-
less s eel handsaw, and e ical empe a u e p o iles
limi is calcula ed using he - alue o 2.99 co espond-
ing o a 99% con idence in e al wi h d = 7). Conduc-
i i y o he mel ed sea ice sec ions was measu ed using
a conduc i i y cell (The mo O ion 3-s a wi h an O ion
013610MD conduc i i y cell) and con e ed o bulk
salini y (G assho e al. 1983). Sea ice b ine salini y
was calcula ed as a unc ion o empe a u e (Cox &
Weeks 1983) and he b ine olume as a unc ion o bulk
salini y, densi y and empe a u e. B ine olume was
calcula ed acco ding o Leppä an a & Manninen (1988)
o empe a u es g ea e han –2°C and acco ding o
Cox & Weeks (1983) o empe a u es less han –2°C.
P ima y p oduc ion was de e mined in mel ed
(mel ed wi hin 48 h in da k condi ions a 3 ± 1°C) sea
ice wa e a 3 ligh in ensi ies (42, 21 and 9 Ǎmol pho-
ons m–2
s–1) and co ec ed wi h one da k incuba ion us-
we e measu ed in d illed holes o he cen e o each ing he H14CO3 incuba ion echnique (S eeman-
sec ion using a he mome e (Tes o he mome e ).
Downwelling i adiance was measu ed di ec ly abo e
and below he snow wi h a da a logge (LI-1400, Li-Co
Biosciences). In addi ion, ai empe a u e was mea-
Nielsen 1952). The sea ice samples we e mel ed and
subsequen ly acclima ised a 20 Ǎmol pho ons m–2 s–1
o a ew hou s (> 2 h). Then he sea ice samples we e
pou ed in o 120 ml gas igh glass bo les and 4 ǍCi o
su ed 2 m abo e he snow, and snow and sea ice hick- H14CO3 we e added o each bo le. The bo les we e
nesses we e de e mined using a measu ing s ick. The
sea ice sec ions we e placed in plas ic con aine s in a
da k insula ed anspo box (The mobox) and b ough
back o he labo a o y. Ligh a enua ion o he sea ice
samples was measu ed wi h a LI-1400 da a logge in a
da k he mally egula ed oom using a ib e lamp wi h
a spec um close o na u al sunligh (15 V, 150 W, ib e-
op ic ungs en–halogen bulb). The LI-1400 da a logge
was placed unde he sea ice sec ion and he ib e lamp
was placed abo e. Ligh a enua ion was measu ed
his way o each sea ice sec ion. All sea ice samples
we e mel ed wi hin 48 h o analysis o nu ien s, salin-
i y, chl a, p ima y p oduc ion and bac e ial ca bon
demand analysis a 3 ± 1°C in da kness.
The mel ed sea ice was il e ed on o 25 mm GF/F
il e s (Wha man) o chl a analysis. The il e s we e ex-
ac ed o 18 h in 96% e hanol (Jespe sen & Ch is o -
e sen 1987) and analysed luo ome ically (Tu ne TD-
700 luo ome e , Tu ne Designs) be o e and a e
incuba ed on a plank on wheel a 1 pm o 5 h a 3 ±
1°C a he di e en ligh in ensi ies. Illumina ion was
p o ided by cool whi e luo escen lamps wi h a spec-
um close o na u al sunligh (Mas e TL-D 36W/840,
Phillips) and i adiance was measu ed using a LI-1400
da a logge . Incuba ions we e e mina ed by adding
200 Ǎl o 5% ZnCl2 and subsequen ly il e ed on o
25 mm GF/F il e s (Wha man). The il e s we e placed
in scin illa ion ials con aining 200 Ǎl 1 M HCl o e-
mo e labelled, un ixed ino ganic ca bon, and hen ex-
ac ed in scin illa ion liquid (Pe kinElme Ul ima Gold)
o 22 h and coun ed using a liquid scin illa ion analyse
(T icCa b 2800 TR, Pe kinElme ). Dissol ed ino ganic
ca bon (DIC) concen a ions in mel ed sea ice we e
measu ed on a CO2 coulome e as desc ibed by Rys-
gaa d & Glud (2004). A e liquid scin illa ion coun ing,
coun s we e con e ed o po en ial p ima y p oduc ion
(unde pho oinhibi ion) (PPi, Ǎg C l–1 h–1) as:
addi ion o 200 Ǎl o a 1 M HCl solu ion. The luo o-
me e was calib a ed agains a pu e chl a s anda d PPi DPMac i i y DICsea ice Fdisc Mc
DPMadded inc
(2)
99
PhD hesis by Do e Haubje g Søgaa d
34 Ma Ecol P o
g
Se 419: 31–45, 2010

whe e DPMac i i y is he 14C assimila ed ca bon co -
ec ed o assimila ed ca bon in he da k (disin e-
g a ions pe minu e [dpm] on il e ), DICsea ice is dis-
sol ed ino ganic ca bon in mel ed sea ice (~450 Ǎmol
l–1), Fdisc = 1.05 is he disc imina ion ac o o algae
assimila ion o 12CO2 and 14CO2, Mc is he mola mass
o ca bon (12.01 g mol–1), DPMadded is he speci ic ac i -
i y o he 14C labelled medium (dpm ml–1) in which
cells we e labelled, and inc is he incuba ion ime (5 h).
The po en ial p ima y p oduc ion (unde no pho oin-
a ed h ough 25 mm mixed cellulose es e il e s
(po e size 0.2 Ǎm, Ad an ec MSF) and he il e s we e
placed in scin illa ion ials. Scin illa ion ials we e
insed wi h 5 ml o 5% cold TCA. Subsequen ly, il e s
we e insed 7 imes wi h 1 ml o cold 5% TCA and hen
ex ac ed in scin illa ion liquid (Ul ima Gold, Pe kin-
Elme ) o 22 h and coun ed using a liquid scin illa ion
analyse (T icCa b 2800, Pe kinElme ).
Bac e ia p oduc ion (BP, Ǎg C l–1 h–1) was calcula ed
as:
hibi ion) (PP, Ǎg C l–1 h–1) measu ed in he labo a o y DPM
N
N
M
sample cells c c
a di e en sea ice dep hs was plo ed agains he 3
labo a o y ligh in ensi ies, 42, 21 and 9 Ǎmol pho ons
BP SA
inc V il
(4)
m–2 s–1, and i ed o he ollowing unc ion desc ibed
by Pla e al. (1980):
whe e DPMsample is he a e age dpm o he li e ea -
men sub ac ed om he a e age dpm o he TCA-
PP Pm
1
exp
E
PAR
(3)
killed con ols, Ncells is he con e sion ac o (2.09
1018 cells mol–1 3H, acco ding o Smi h & Clemen
P
m
1990), inc is he incuba ion pe iod ( inc = 6 h), V il is he
whe e Pm (Ǎg C l–1 h–1) is he maximum pho osyn he ic
a e a ligh sa u a ion, (Ǎg C m2 s Ǎmol pho ons–1 l–1
h–1) is he ini ial slope o he ligh cu e and EPAR (Ǎmol
pho ons m–2
s–1) is he labo a o y i adiance. The pho o-
adap a ion index, Ek (Ǎmol pho ons m–2 s–1) was cal-
olume o he subsamples (V il = 0.01 l), Mc is he mol-
ecula mass o ca bon and SA is he speci ic ac i i y o
he hymidine solu ion (2.24 1016 dpm mol–1).
Nc = 5.7 10–8 Ǎg C cell–1 is calcula ed as:
cula ed as
P
m
/
.
In si u downwelling i adiance was measu ed a
N
c
Cellsize C ac o
1000000 (5)
g ound le el wi h a py ome e (Kipp & Zonen, model
CM21, spec um ange o 305 o 2800 nm) once e e y 5
min, and hou ly a e ages we e p o ided by Asiaq
(G eenland Su ey). Hou ly downwelling i adiance
was con e ed in o hou ly pho osyn he ically ac i e
adia ion (PAR) (ligh spec um, 300 o 700 nm) a e
in e calib a ion ( 2 = 0.99, p < 0.001, n = 133) wi h a LI-
whe e Cellsize is he a e age bac e ia cell size (0.473 Ǎm3)
(Smi h & Clemen 1990), and C ac o is he ac o used
o con e cell olume o ca bon (0.12 pg C Ǎm–3) acco d-
ing o Smi h & Clemen (1990).
Bac e ial ca bon demand (BCD, Ǎg C l–1 h–1) was
calcula ed as:
1400 da a logge (Li-Co ). The in si u hou ly PAR i a-
diance was calcula ed a di e en dep hs, depending BCD BP
BGE
(6)
on sea ice and snow hickness, using he a enua ion
coe icien s measu ed du ing he sea ice season.
In si u p ima y p oduc ion was calcula ed o each
hou a di e en sea ice dep hs using hou ly in si u
PAR i adiance (see Eq. 3). To al daily (24 h) in si u p i-
ma y p oduc ion was calcula ed as he sum o hou ly in
si u p ima y p oduc ion o each dep h. The dep h-
in eg a ed ne p ima y p oduc ion was calcula ed
using apezoid in eg a ion.
Bac e ia p oduc ion in mel ed sea ice samples was
de e mined by measu ing he inco po a ion o
[3H] hymidine in o DNA. T iplica e subsamples ( ol-
ume, V il = 0.01 l) we e incuba ed in da kness a 3 ±
1°C wi h 10 nM o labelled [3H] hymidine (New Eng-
land Nuclea , speci ic ac i i y, 10.1 Ci mmol–1). T i-
chlo oace ic acid (TCA)-killed con ols we e made o
measu e he abio ic adso p ion. A he end o he incu-
ba ion pe iod ( inc = 6 h), 1 ml o 50% cold TCA was
added o all he subsamples (Fuh man & Azam 1982).
Subsamples o [3H] hymidine we e s o ed a 3 ± 1°C in
scin illa ions ials un il il a ion. Subsamples we e il-
whe e BP is he bac e ia p oduc ion and BGE is he
bac e ial g ow h e iciency o 0.5 acco ding o Ri kin
& Legend e (2001). To ex apola e bac e ial ca bon
demand o daily in si u ca bon demand we assumed
ha he espi a ion was ligh -independen (mul iply by
24). The dep h-in eg a ed bac e ial ca bon demand
was calcula ed using apezoid in eg a ion.
On 23 Feb ua y, 10 sea ice co es we e collec ed along
a 10 m long sec ion o in es iga e he spa ial (ho izon al
and e ical) a iabili y o he bio ic condi ions (chl a,
p ima y p oduc ion and bac e ial ca bon demand) and
he abio ic condi ions (sea ice empe a u e, bulk salin-
i y and b ine olume) in he sea ice. The 10 sea ice co es
we e collec ed close o he plas ic bag incuba ions a
1 m in e als. The co es we e cu in o 2 sec ions, i.e. op
and bo om hal es, and b ough back o he labo a o y
in a da k The mobox o u he analysis as desc ibed.
To ex end he e alua ion o spa ial a iabili y, a
la ge-scale in es iga ion o ho izon al a iabili y was
conduc ed on 29 Feb ua y. Fi y- wo sea ice co es we e
collec ed along a 367 m long ansec o in es iga e he

100 PhD hesis by Do e Haubje g Søgaa d
Søgaa d e al.: Au o ophic and he e o ophic ac i i
y
in A c ic sea ice 35

he e ogenei y o chl a, sea ice empe a u e, b ine salin-
i y and b ine olume in he bo om o he sea ice. The
snow and sea ice hickness was measu ed as desc ibed
p e iously. Fou sea ice co es we e sampled a 20 cm
in e als a posi ion 1 m. The i s co e was cu e i-
cally in o 2 pieces a e e y sampling posi ion. Sea ice
co es we e collec ed a dis ances o 1, 3, 5, 7, 9, 20, 31,
42, 53, 64, 165, 266 and ca. 367 m. All he ice co es
we e b ough back o he labo a o y in a da k The mo
box o de e mina ion o b ine salini y, olume and
chl
a
concen a ion as desc ibed.
Spa ial au oco ela ion (Legend e & Legend e 1998)
was used o analyse he ho izon al dis ibu ion o chl a,
sea ice empe a u e, bulk salini y, snow hickness and
sea ice hickness. We assume ha he a iabili y along
one line is he same as along a pe pendicula line. The
au oco ela ion was es ima ed by Mo an’s I coe i-
cien s (Mo an 1950, Legend e & Legend e 1998). This
coe icien was calcula ed o each o he ollowing
in e als along he ansec (dis ance classes): 0 o
0.50 m, 0.50 o 1.5 m, 1.5 o 2.5 m, 2.5 o 5.5 m, 5.5 o
10.5 m, 10.5 o 20.5 m, 20.5 o 50.5 m, 50.5 o 100.5 m
and >100.5 m. The au oco ela ion coe icien s es i-
ma ed by Mo an’s I coe icien we e es ed o signi i-
cance acco ding o he me hod desc ibed in Legend e
& Legend e (1998). A 2- ailed es o signi icance was
used. The null hypo hesis o andom spa ial dis ibu-
ion was ejec ed a he speci ied le el o signi icance
when an indi idual au oco ela ion coe icien ex-
ceeded a c i ical alue (posi i e o nega i e). A signi i-
cance le el o p < 0.05 was used.
RESULTS
Abio ic pa ame e s
Tempe a u es wi hin he snow co e a ied om –13
o 0°C and he sea ice empe a u e om –5 o 0°C
(Fig. 1a). Minimum empe a u es o he sea ice we e
a)
Sea ice and snow empe a u e (°C)
40
–13
25 –10
0
d)
Bulk salini y
40
25
0 9
–3
–20
–35
–50
–65
–6 –5 –1
–4
–2
–20
–35
–50
–65
7 8 1
5
1
40
25
0
–20
–35
–50
–65
b)
Ligh a enua ion coe icien (m
–1
)
20 32
13
10
8
6
c)
I adiance (mol pho ons m
–2
s
–1
)
40
25
0
–20
–35
–50
–65
e)
Rela i e b ine olume (%)
0
–20
–35
–50
–65
7 7 76
60
40
7
500
250
0
Fig. 1. Sea ice de elopmen in Malene Bigh , NW
G eenland, du ing he 2008 season: (a) sea ice and
snow empe a u e (°C), (b) ligh a enua ion coe i-
cien (m–1), (c) a e age daily pho on i adiance in
sea ice and a su ace (PAR, Ǎmol pho ons m–2 s–1),
(d) bulk salini y, (e) ela i e b ine olume ac ion
(%). The sea ice in la e Ma ch om 0 o 16 cm
is a laye o g anula snow–ice. The black do s
ep esen iplica e measu emen s
Feb
Ma Ap
80
10
10 60
20
Feb Ma
A
p
Sea
ice
and
snow
dep h
(cm)
Sea
ice
and
snow
dep h
(cm)
Sea
ice
and
snow
dep h
(cm)
P
AR
(ȝ
mol
pho ons
m
–2
s
–1
)
101
PhD hesis by Do e Haubje g Søgaa d
36 Ma Ecol P o
g
Se 419: 31–45, 2010

32
4
–
–
obse ed in Feb ua y, a e which he empe a u e
g adually inc eased o maximum alues jus be o e sea
ice b eak-up in mid-Ap il. The high ligh e lec ance
and sca e om he snow co e caused s ong ligh
a enua ion h oughou he sea ice season, wi h a e -
age a enua ion coe icien s o K
snow = 23 m–1 and Kice =
8 m–1 (Fig. 1b). Ea ly in he sea ice season, i adiance a
he bo om o he sea ice was low (0.03 o 0.15 Ǎmol
pho ons m–2 s–1) (Fig. 1c). Ligh a ailabili y inc eased
he sea ice (Fig. 1e). In la e Ma ch mel ing om he op
o he sea ice was ini ia ed, which esul ed in high el-
a i e b ine olumes and low bulk salini ies in he
uppe mos pa o he sea ice. Du ing his pe iod ai
empe a u es a ied be ween 0 and 7°C.
Nu ien pa ame e s
3–
along wi h inc easing day leng h and declining snow Bulk PO4 concen a ion was ini ially 0.05 o
co e , eaching a maximum downwelling i adiance o
76 Ǎmol pho ons m–2 s–1 in he uppe mos sea ice
sec ion and 7 Ǎmol pho ons m–2 s–1 in he bo om on
Ap il 4.
The bulk salini y dec eased o e ime wi h a maxi-
mum salini y o 9 in ea ly Ma ch and a minimum salin-
i y o 1 om la e Ma ch and onwa d (Fig. 1d). Bulk
salini y a ied e ically wi hin sea ice co es du ing
0.20 Ǎmol l–1 (Fig. 2a), and inc eased o a maximum o
0.70 Ǎmol l–1 in Ma ch bu subsequen ly dec eased o
0.05 Ǎmol l–1 in Ap il. Bulk Si(OH)4 concen a ion
emained cons an be ween 1.8 and 2.3 Ǎmol l–1 in he
sea ice om Feb ua y o la e Ma ch wi h lowes alues
encoun e ed a he bo om o he sea ice. In la e Ma ch,
Si(OH)4 concen a ion dec eased apidly o below
1.4
Ǎmol l–1 (Fig. 2b). The ini ial bulk concen a ion o
win e , wi h lowe alues encoun e ed in he bo om
NO
–
+ NO
–
was 3.2 Ǎmol l–1 a he op and 2.0 Ǎmol l–1
sea ice om mid-Feb ua y un il la e Ma ch. F om
25 Ma ch o 4 Ap il, bulk salini y was lowes in he
uppe mos pa o he sea ice. The ela i e b ine ol-
a he bo om o he sea ice (Fig. 2c) and dec eased
h oughou he season, eaching a minimum o
1.5
Ǎmol l–1 in Ap il. The ini ial bulk concen a ion o
ume inc eased h oughou he sea ice season, wi h a
NH
+
was 4.0 Ǎmol l–1 a he op and bo om o he sea
+
maximum in la e Ma ch in he uppe mos sec ion o ice. Du ing he sea ice season he NH4 concen a ion
Fig. 2. Nu ien concen a ion (Ǎmol l–1) in bulk sea ice: (a) phospha e, (b) silicic acid, (c) NO3 + NO2 , (d) ammonium. LOD is
lowe limi o de ec ion, which is calcula ed using he - alue o 2.99 co esponding o a 99% con idence in e al wi h d = 7. The
black do s ep esen iplica e measu emen s
102 PhD hesis by Do e Haubje g Søgaa d
Søgaa d e al.: Au o ophic and he e o ophic ac i i
y
in A c ic sea ice 37

15 Feb
3 Ma
10 Ma
19 Ma
25 Ma
4 Ap
c) NO
–
+ NO
–
3 2
a)
Phospha e b) Silicic acid
8 8
6 6
4 4
2 2
0 0
d) Ammonium
8 8
6 6
4 4
2 2
0 0 2
4 6
8 0
10 0 2 4
6
8 10
Bulk salini y
– –
Fig. 3. Concen a ions o (a) phospha e, (b) silicic acid, (c) NO3 + NO2 and (d) ammonium e sus bulk salini y in sea ice. The solid
line indica es he expec ed dilu ion line p edic ed om salini y and nu ien concen a ions in seawa e (0 o 10 m dep h, salini y
o 33). See ‘Resul s: Nu ien pa ame e s’ o explana ion
inc eased, eaching alues o 6 o 12 Ǎmol l–1 a he
bo om o he sea ice and 5.0 o 6.0 Ǎmol l–1 a he op o
he sea ice (Fig. 2d).
Bulk nu ien concen a ions o each sampling da e
we e plo ed as a unc ion o bulk salini y and com-
pa ed wi h he expec ed dilu ion line (acco ding o
Cla ke & Ackley 1984). I alues we e below he line,
deple ion o nu ien s occu ed in he sea ice. I al-
ues we e abo e he dilu ion line, p oduc ion o ne
deposi ion o he solu e ook place. Plo s o salini y–
ion (bulk) o 40 Ǎg C l–1 d–1 was encoun e ed in he
middle pa o he sea ice p o ile in Ap il.
The highes sea ice bac e ial ca bon demand o
27 Ǎg C l–1 d–1 was encoun e ed in he cen al ice a he
onse o he mel ing pe iod (Fig. 4c). Howe e , a single
peak o 9.00 Ǎg C l–1 d–1 was obse ed on 10 Ma ch in
he uppe 10 cm sec ion o he sea ice.
Dep h in eg a ion o he ac i i y e lec ed low
au o ophic and he e o ophic p oduc i i y du ing
win e , ollowed by a sligh ly ne he e o ophic pe iod
3–
– –
PO
4
, salini y–Si(OH)4, salini y–NO3
+ + NO2
and
in la e Feb ua y and Ma ch. Finally, a ne au o ophic
salini y–NH
4
in sea ice we e gene ally all abo e he
+
pe iod was obse ed om la e Ma ch un il he end o
dilu ion line (Fig. 3), bu mos explici ly o NH4 ,
which clea ly accumula ed wi hin he sea ice (Fig. 3d).
Bio ic pa ame e s
Sea ice p o iles o he algal biomass, exp essed as chl
a, showed ha he highes bulk concen a ion in he
lowe 15 cm o he sea ice co es was 2.80 Ǎg l–1 on
25 Ma ch (Fig. 4a). Subsequen ly, he algal biomass in
he lowe 15 cm o he sea ice co es dec eased apidly
eaching a alue o 1.50 Ǎg l–1 in Ap il.
Sea ice p ima y p oduc ion in eg a ed o he sea ice
p o ile inc eased h oughou win e om 0.09 mg C
m–2 d–1 (15 Feb ua y) o 12.60 mg C m–2 d–1 (4 Ap il)
(Fig. 4b). The highes olume-speci ic p ima y p oduc-
he s udy pe iod (Fig. 5). In eg a ed o e he en i e
measu ing season (i.e. om 15 Feb ua y o 14 Ap il)
he sea ice o Malene Bigh was ne au o ophic. An
annual ne ca bon ixa ion o 220 mg C m–2 was calcu-
la ed by sub ac ing he ne esul o a sea ice- ela ed
g oss p ima y p oduc ion o 350 mg C m–2 om he
bac e ial ca bon demand o 130 mg C m–2.
Bag incuba ions
Oxygen le els du ing Feb ua y and Ma ch indica ed a
low ne oxygen accumula ion in he op sea ice co es o
0.50 ± 3.00 Ǎmol O2 l–1 d–1 (mean ± SD) and in he bo om
sea ice co es o 1.30 ± 5.00 Ǎmol O2 l–1 d–1. Howe e ,
none o hese alues we e signi ican ly di e en om
Nu ien
concen a ion
(mol
l
–1
)
103
PhD hesis by Do e Haubje g Søgaa d
38 Ma Ecol P o
g
Se 419: 31–45, 2010

40
25
0
–20
–35
–50
–65
40
25
a)
Chlo ophyll (g l–1 mel ed sea ice)
b)
P ima y p oduc ion (g C l–1 mel ed sea ice d–1)
14
12
10
8
6
4
2
0
Fig. 5. P ima y p oduc ion (PP, mg C m–2 d–1) and bac e ial
ca bon demand (BCD, mg C m–2 d–1) in bulk sea ice
0
–20
–35
–50
–65
c)
Bac e ial ca bon demand (g C l–1 mel ed sea ice d–1)
40
25
350
300
250
200
150
100
50
a)
Top sea ice co es
Se ies 1
Se ies 2
PP
BCD
100
80
60
40
20
0
0
–20
–35
–50
–65
5
9 2
27
2
13
350
300
250
200
b)
Bo om sea ice co es 100
80
60
40
Feb
Ma Ap
Fig. 4. (a) Chl a concen a ions (Ǎg chl a l–1) in bulk sea ice.
(b) P ima y p oduc ion (Ǎg C l–1 mel ed sea ice d–1). (c) Bac e-
ial ca bon demand (Ǎg C l–1 mel ed sea ice d–1) calcula ed
acco ding o Ri kin & Legend e (2001). The black do s
ep esen iplica e measu emen s
150
100
50
20
0
Feb
Ma Ap
ze o (p > 0.05) (Fig. 6). Au o ophic ac i i y exceeded
he e o ophic ac i i y in la e Ma ch and Ap il, esul ing
in a signi ican ly high ne oxygen accumula ion in
he bo om sea ice co es o 6.30 ± 2.30 Ǎmol O2 l–1 d–1 (p <
0.01) whe eas no signi ican oxygen accumula ion
(0.80 ± 3.50 Ǎmol O2 l–1 d–1) was obse ed in he op sea
ice co es.
Assuming a pho osyn he ic quo ien o 1.00 CO2
e ol ed pe O2 consumed, a ne annual ca bon ixa-
ion o 1700 ± 760 g C m–2 was calcula ed o he bag
incuba ions.
Fig. 6. Measu emen s o O
2
concen a ion (
d
,
z
: Ǎmol O
2
l
–1
mel ed sea ice), p ima y p oduc ion (PP)
(
j
:
Ǎg C l
–1
mel ed
sea ice d
–1
) and bac e ial ca bon demand (BCD)
(
h
j
:
Ǎg C l
–1
mel ed sea ice d
–1
) in (a) op hal and (b) bo om hal o sea ice
co es. Dashed lines ep esen eg ession lines
He e ogenei y
On 23 Feb ua y he abio ic and bio ic condi ions, i.e.
bulk salini y, b ine olume, empe a u e, chl a, p i-
ma y p oduc ion and bac e ial ca bon demand, we e
measu ed in op and bo om sec ions o he sea ice
co es (n = 10) along a 10 m ansec (Table 1). The e
0.2
0.5
1 0.3
1.5
2
22..88
2
7
4 23
7
4 3
3355
4
4400
22
277
PP
BCD
Feb
Ma
Ap
Sea
ice
and
snow
dep h
(cm)
O
2
conc.
(ȝ
mol
O
2
l
–1
mel ed
sea
ice)
PP
,
BCD
(mg
C
m
–2
d
–1
)
PP,
BCD
(ȝ
g
C
l
–1
d
–1
)
110 PhD hesis by Do e Haubje g Søgaa d
Søgaa d e al.: Au o ophic and he e o ophic ac i i
y
in A c ic sea ice 45

ߨ
Thomas DN, La a RJ, Eicken H, Ka ne G, Skoog A (1995)
Dissol ed o ganic ma e in A c ic mul i-yea sea ice du -
ing win e : majo componen s and ela ionships o ice
cha ac e is ics. Pola Biol 15:477–483
T odahl HJ, Buckley RG (1990) Enhanced ul a iole ans-
mission o An a c ic sea ice du ing he aus al sp ing.
J Geophys Res 17:2177–2179
ߨ
Vézina AF, Deme s S, Lau ion I, Sime-Ngando T, Junipe SK,
De ine L (1997) Ca bon lows h ough he mic obial ood
web o i s -yea ice in Resolu e Passage (Canadian High
A c ic). J Ma Sys 11:173–189
Weeks WF, Ackley SF (1986) The g ow h, s uc u e and p op-
e ies o sea ice. In: Un e s eine N (ed) The geophysics o
sea ice. NATO ASI Se ies Se B Physics. Plenum P ess,
New Yo k, NY, p 9–164
Welle G, Schwe d ege P (1967) Radia ion pene a ion in
An a c ic pla eau and sea ice. In: Pola me eo ology.
Wo ld Me eo ol O g Tech No e 87:120–141
Edi o ial esponsibili y: G aham Sa idge,
Po a e y, UK
Submi ed: Janua y 8, 2010; Accep ed: Sep embe 27, 2010
P oo s ecei ed om au ho (s): No embe 26, 2010

111
PhD hesis by Do e Haubje g Søgaa d
PAPER V
Pola Biology 34:1157 – 1165, doi: 10.1007/s00300-011-0976-3
D.H Søgaa d • P.J. Hansen • S. Rysgaa d • R.N. Glud
G ow h limi a ion o h ee A c ic sea ice algal
species: e ec s o salini y, pH, and ino ganic
ca bon a ailabili y
Sea ice dia om (F agila iopsis sp.) Pho o: Diana W. K awczyk.
112 PhD hesis by Do e Haubje g Søgaa d
ORIGINAL PAPER
G ow h limi a ion o h ee A c ic sea ice algal species: e ec s
o salini y, pH, and ino ganic ca bon a ailabili y
Do e Haubje g Søgaa d •Pe Juel Hansen •
Sø en Rysgaa d •Ronnie Nøh Glud
Recei ed: 14 Oc obe 2010 / Re ised: 18 Janua y 2011 / Accep ed: 19 Janua y 2011 / Published online: 3 Ma ch 2011
ÓSp inge -Ve lag 2011
Abs ac The e ec o salini y, pH, and dissol ed ino -
ganic ca bon (TCO
2
) on g ow h and su i al o h ee
A c ic sea ice algal species, wo dia oms (F agila iopsis
nana and F agila iopsis sp.), and one species o chlo o-
phy e (Chlamydomonas sp.) was assessed in con olled
labo a o y expe imen s. Ou esul s sugges ha he chlo-
ophy e and he wo dia oms ha e di e en ole ance o
fluc ua ions in salini y and pH. The wo species o dia oms
exhibi ed maximum g ow h a es a a salini y o 33, and
g ow h a es a a salini y o 100 we e educed by 50%
compa ed o a a salini y o 33. G ow h ceased a a salini y
o 150. The chlo ophy e species was mo e sensi i e o high
salini ies han he wo dia om species. G ow h a e o he
chlo ophy e was g ea ly educed al eady a a salini y o 50
and i could no g ow a salini ies abo e 100. A salini y 33
and cons an TCO
2
concen a ion, all species exhibi ed
maximal g ow h a e a pH 8.0 and/o 8.5. The wo dia om
species s opped g owing a pH [9.5, while he chlo o-
phy e species s ill was able o g ow a a a e which was 1/3
o i s maximum g ow h a e a pH 10. Thus, Chlamydo-
monas sp. was able o g ow a high pH le els in he suc-
cession expe imen and he e o e ou compe ed he wo
dia om species. Complemen a y expe imen s indica ed ha
g ow h was mainly limi ed by pH, while ino ganic ca bon
limi a ion only played an impo an ole a e y high pH
le els and low TCO
2
concen a ions.
Keywo ds A c ic Sea ice algae Salini y pH TCO
2

In si u succession pa e ns
In oduc ion
Sea ice is pe mea ed wi h po es and b ine channels, which
hos a unique mic obial communi y. The o al b ine chan-
nel olume o sea ice ypically anges be ween 1 and 30%
depending on salini y, empe a u e, and ionic composi ion
o he b ine fluid (Weeks and Ackley 1986). When he
empe a u e dec eases, he he modynamic phase equilib-
ium d i es he sea ice owa d a lowe b ine olume wi h
highe salini ies (Cox and Weeks 1983). Thus, he em-
pe a u e and b ine o sea ice a e in e ela ed. A b ine
empe a u e be ween -1.9 and -6.7°C, he b ine salini y
may ange om 34 o 108 (Glei z e al. 1995). Howe e ,
when sea ice is exposed o empe a u es below -20°C, he
b ine salini y can be well abo e 200 (Cox and Weeks
1983). In he summe when sea ice mel s, he salini y o he
b ine can be as low as one- hi d o no mal sea wa e , e.g.,
salini y 10 (Ryan e al. 2004). B ine salini y also fluc u-
a es e ically wi hin he sea ice, wi h he lowes salini ies
usually encoun e ed in he bo om sea ice laye s (G adinge
1999; Lizo e 2003; Ryan e al. 2004). Thus, sea ice algae
mus cope wi h se e e physicochemical s ess ac o s
caused by na u al a ia ions in salini y. Only a ew s udies
D. H. Søgaa d (&)S. Rysgaa d R. N. Glud
G eenland Clima e Resea ch Cen e
(C/O G eenland Ins i u e o Na u al Resou ces),
Ki ioq 2, Box 570, 3900 Nuuk, G eenland
e-mail: [email p o ec ed]
P. J. Hansen
Ma ine Biological Labo a o y, Uni e si y o Copenhagen,
S andp omenaden 5, 3000 Helsingø , Denma k
D. H. Søgaa d R. N. Glud
Uni e si y o Sou he n Denma k, Campus ej 55,
5230 Odense M, Denma k
R. N. Glud
Dums a nage Ma ine Labo a o y,
Sco ish Associa ion o Ma ine Science,
PA37 1QA Dunbeg, Sco land, UK
123
Pola Biol (2011) 34:1157–1165
DOI 10.1007/s00300-011-0976-3
113
PhD hesis by Do e Haubje g Søgaa d
include in es iga ions in o he e ec o salini y s ess on
g ow h a es o sea ice algae (G an and Ho ne 1976;
A igo and Sulli an 1992; Thiel e al. 1996; Ryan e al.
2004). A s udy on he sea ice dia oms (Amphip o a ku -
e a hii, Ni zschia, and Thalassiosi a An a c ica) isola ed
om ice co es om Weddell Sea e ealed g ow h a
salini ies up o 90 a -5.5°C, and he dia oms we e ound
o su i e o 20 days a salini ies up o 145 (Thiel e al.
1996). O he s udies ha e shown ha he g ow h o di -
e en sea ice dia oms om he Weddell Sea ceased a
salini ies abo e 50 (G an and Ho ne 1976). Fu he mo e,
s udies ha e shown ha mos flagella e species o algae
e.g., chlo ophy es, dinoflagella es, and ch ysophy es ha e
been epo ed in he sea ice (A igo e al. 2010). These
flagella e species is especially ound in he op sea ice
laye , whe e he highes salini y and lowes empe a u es is
encoun e ed.
The di e ences in ole ance be ween sea ice algal
species ha e been asc ibed o a ious abili ies o osmo ic
acclima ions, e.g., p oduc ion o osmoly es (such as p o-
line), which balances he ionic p essu e du ing changes in
salini y (Glei z and Thomas 1992). Mo eo e , changes in
sea ice salini y and associa e ac o s may be he key
d i e s o mic obial succession in sea ice communi ies
(Mikkelsen e al. 2008). The sea ice algal species ha a e
capable o cope wi h a b oad ange o salini y may ha e
an ad an age and become dominan in he sea ice com-
muni y. An unde s anding o he e ec o fluc ua ing
salini ies in sea ice b ine indica es which ac o s d i e he
dis ibu ion and succession o sea ice algae and migh
gi e impo an in o ma ion ha can be used o modeling
sea ice species succession and ca bon dynamics wi hin he
b ine.
Va ia ion in seawa e pH le els can also ha e a ma ked
e ec on he g ow h and su i al o sea ice algae. In sea
ice, a numbe o biological and physical p ocesses influ-
ence pH. S udies o sea ice ha e shown ha in egions
cha ac e ized by high p ima y p oduc ion, he sea ice
b ine has conside ably educed concen a ions o dis-
sol ed ino ganic ca bon (TCO
2
) and ele a ed pH le els as
high as 10.0 (Glei z e al. 1995; Thomas e al. 2001).
Fu he mo e, changes in ca bon chemis y alone can
esul in significan changes in pH o he sea ice b ine.
One mechanism behind his is CaCO
3
p ecipi a ion ha
can occu a low empe a u es (Rysgaa d e al. 2007;
Dieckmann e al. 2008). Ca bona e p ecipi a ion will
ini ially lead o a buildup o CO
2
in he b ine sys em
leading o a dec ease in pH. Wi h ime, CO
2
can be
anspo ed o he wa e column h ough b ine d ainage.
This ne expo o CO
2
ou o he sea ice b ine will lead
o inc eased pH in he b ine, especially when sea ice
s a s o mel . In sea wa e , changes in pH influence he
equilib ium o he ca bona e sys em and he e o e he
in e -specia ion o TCO
2
, i.e., CO
2
(aq), HCO
3
-
,CO
3
-2
,
which may influence mic oalgae species succession and
dis ibu ion (Hansen 2002; Ros e al. 2003; T imbo n
e al. 2008). Limi a ion in he supply o CO
2
due o
ele a ed pH le els may es ic pho osyn hesis and g ow h
o some algal species (Hansen 2002; Ros e al. 2003;
Hansen e al. 2007) and a o species ha u ilize HCO
3
-
as an ino ganic ca bon sou ce (Ko b e al. 1997; Hue as
e al. 2000; Hansen 2002). Dia oms ha e been ound o
ac i ely ake up HCO
3
-
and con e i in o in acellula
CO
2
by ex acellula enzymes (e.g., Ko b e al. 1997;
To ell e al. 1997). Addi ionally, dia oms can u ilize
HCO
3
-
di ec ly o ca bon fixa ion h ough C
4
pho o-
syn hesis (To ell e al. 1997; Rein elde e al. 2000).
Howe e , a p e ious s udy has shown ha he abili y o
ole a e high pH is no ela ed o pa icula e algal g oups,
bu a he is species specific (Hansen 2002). In his s udy,
we in es iga ed he uppe limi o g ow h wi h espec o
salini y, pH, and TCO
2
o h ee common A c ic sea ice
algal species; he dia oms F agila iopsis nana,F agila -
iopsis sp. and he chlo ophy e Chlamydomonas sp. The
physiological esponse owa d hese s ess ac o s a e
e alua ed and discussed in ela ion o in si u succession
pa e ns. This s udy is impo an o unde s and he ac o s
con olling he g ow h, su i al, composi ion, and dis i-
bu ion o he sea ice algal communi ies wi hin he b ine
and can be used o accu a ely model he species succes-
sion and he p oduc i i y o his complex sys em.
Ma e ials and me hods
Algae species and main enance o sea ice algae cul u es
Th ee sea ice algae we e selec ed o he s udy (A igo
e al. 2010). The dia om F agila iopsis sp. (CCMP2297)
and he chlo ophy e Chlamydomonas sp. (CCMP2294)
o igina ed om sea ice om Ba fin Bay and we e p o ided
by Guilla d Na ional Cen e o Cul u e o Ma ine Phy o-
plank on (CCMP), and he dia om F agila iopsis nana
(SCCAP K-0637) was isola ed om he Lab ado Sea and
p o ided by he Scandina ian Cul u e Collec ion o Algae
and P o ozoa, Depa men o Phycology, Uni e si y o
Copenhagen. The wo species o dia oms we e selec ed as
ep esen a i es o penna e dia oms, which a e e y com-
mon in sea ice (A igo e al. 2010). We delibe a ely chose
F agila iopsis nana because i is a ela i ely small species
(leng h 8.0–9.4 lm; wid h 1.9–2.0 lm) and F agila iopsis
sp. because i is somewha la ge dia om species (leng h
12–16 lm; wid h 6–10 lm). The chlo ophy e Chlamydo-
monas sp. (leng h 8–10 lm; wid h 4.0–6.0 lm) was
selec ed because chlo ophy es a e common in sea ice as
well (A igo e al. 2010).
1158 Pola Biol (2011) 34:1157–1165
123
114 PhD hesis by Do e Haubje g Søgaa d
Algal cul u es we e g own in L1 g ow h medium
(Guilla d and Ha g a es 1993) based on au ocla ed sea-
wa e wi h a salini y o 33. The s ock cul u es we e
main ained a 3 ±1°C and 50 lEm
-2
s
-1
ollowing a
ligh :da k cycle o 16:8 h. Illumina ion was p o ided by
cool fluo escen lamps, and i adiance was measu ed using
a LiCo 1400 (Li-Co , NE, USA).
Expe imen al condi ions
All expe imen s we e ca ied ou a 3 ±1°C and a an
i adiance o 50 lEm
-2
s
-1
ollowing a ligh :da k cycle
o 16:8 h. Only cells om exponen ially g owing cul u es
we e used o inocula ion o he expe imen s. Howe e ,
he fi s 6–10 days we e conside ed as an acclima ion
pe iod; he e o e, cell coun s om hese samplings we e
no included in he calcula ions o g ow h a es. All
expe imen s we e ca ied ou in 62-ml polys ylene bo les,
excep o he pH-d i expe imen s ha we e ca ied ou
in gas- igh lamina ed NEN/PE plas ic bag (Hansen e al.
2000) fi ed wi h a gas- igh Tygon ube and al e o
sampling. All expe imen s we e ca ied ou in iplica es,
i.e., each expe imen was ca ied ou in h ee sepa a e
bo les.
Cul u es we e kep suspended h ough he use o a
plank on wheel, and an ex e nal cooling sys em was used o
p e en hea ing associa ed wi h adia ion abso p ion. The
L1 g ow h medium was selec ed o make su e ha algal
cul u es we e no nu ien limi ed a any ime du ing he
expe imen .
Enume a ion o cells was ca ied ou using subsamples
fixed in acidic Lugol’s iodine (2.5% final concen a ion),
and cells we e coun ed in a Sedgewick-Ra e chambe .
Each coun was based on a leas 400 cells.
G ow h a es (l) we e measu ed as inc ease in cell
numbe and we e calcula ed assuming exponen ial g ow h:
lðd1Þ¼ðln N1ln N0Þ
ð 1 0Þð1Þ
whe e N
0
and N
1
a e numbe o cells a ime
0
and
1
, and
is he di e ence in ime (d) be ween
0
and
1
samples
(Hansen 2002). We de e mined he exponen ial phase o
g ow h (s aigh line). Two poin s, N
0
and N
1
, a he
ex emes o his linea phase was aken and subs i u ed
in o he equa ion (same app oach was used o de e -
mining he wo poin s,
0
and
1
). All expe imen s we e
ca ied ou in iplica es; hus, his was done o each
eplica e and he mean o he h ee maximum g ow h
a es was de e mined. The calcula ions o he g ow h a es
we e co ec ed o any dilu ions. pH alues we e mea-
su ed using a Sen on
Ò
2001 pH-me e equipped wi h a
Red Line elec ode, which is an ISFET
Ò
senso (Semi-
conduc o Ion Field E ec T ansis o ) wi h de ec ion limi
o 0.01. The pH senso was calib a ed (2 poin ) using
Sen on bu e s o pH 7.0 and 10.0. The concen a ion o
dissol ed ino ganic ca bon (TCO
2
) was measu ed in he
g ow h medium by ans e ing samples (12 ml) o Exe-
aine ubes (12 ml Exe aine
Ò
, Labco High Wycombe,
UK) spiked wi h 20 ll HgCl
2
(sa u a ed solu ion, 5%
w/ ) and was measu ed using a CO
2
analyze (CM5012
CO
2
Coulome e ).
Expe imen al se up
G ow h a eo sea ice algae a di e en salini ies
In he fi s se o expe imen s, g ow h a es o he h ee
sea ice algae F agila iopsis nana, F agila iopsis sp, and
Chlamydomonas sp. we e measu ed a di e en salini ies
anging om 5 o 150 (i.e., salini y o 5, 20, 33, 50, 75,
100, and 150). The salini y was adjus ed om a salini y
o 33 by addi ion o a ificial seawa e based on Red Sea
sal wi h known TCO
2
concen a ions o he L1 medium.
The pH alue was kep cons an a 8.0 h oughou he
expe imen . I he pH di e ed by mo e han 0.03 om he
se poin , i was adjus ed by he aliquo addi ion o 0.1 M
NaOH o HCl. The expe imen was ini ia ed wi h an
inocula ion o 1,000 cells ml
-1
and was allowed o un
o minimum 18 d and maximum 20 d. E e y second day,
pH was measu ed, and subsamples (1 ml) we e aken o
enume a ion o algae cells. A e subsampling, he bo les
we e efilled o capaci y wi h L1 g ow h medium (1 ml).
The L1 g ow h medium was a each e en adjus ed o he
co ec salini y o p e en salini y in he expe imen al
bo les o d i . Salini y and TCO
2
concen a ions we e
measu ed ini ially and a he e mina ion o he
expe imen .
To es he e ec o lowe ed salini y on he g ow h o
he h ee species o sea ice algae, a second se o expe i-
men s was conduc ed. The algal cul u es we e g own in L1
g ow h medium (Guilla d and Ha g a es 1993) based on
au ocla ed seawa e wi h a salini y o 75 and a known
TCO
2
concen a ion o a mon h. The salini y was adjus ed
om a salini y o 75 o di e en salini ies o 5, 20, and 33
o mimic he ansi ion om cold o mel ing sea ice. The
pH alue was kep cons an a 8.0 h oughou he expe i-
men . I he pH di e ed by mo e han 0.03 om he se
poin , i was adjus ed by he aliquo addi ion o 0.1 M
NaOH o HCl. The expe imen was ini ia ed wi h an
inocula ion o 1,000 cells ml
-1
and was allowed o un o
a minimum o 18 d and a maximum o 20 d.
G ow h a eo he sea ice algae a di e en pHand TCO
2
In he fi s se o pH expe imen s, g ow h a es o he h ee
species o sea ice algae: F agila iopsis nana,
Pola Biol (2011) 34:1157–1165 1159
123
115
PhD hesis by Do e Haubje g Søgaa d
F agila iopsis sp., and Chlamydomonas sp. we e measu ed
a di e en pH alues anging om 8.0 o 10.0 (i.e., pH 8.0,
8.5, 9.0, 9.5, and 10.0). The salini y was 33 h oughou he
expe imen . The pH was adjus ed by addi ion o 0.1 M HCl
o NaOH o he medium. The expe imen was ini ia ed by
inocula ing 1,000 cells ml
-1
and was allowed o un o 20
d. The TCO
2
concen a ion was, in all ins ances, 2.4 mM.
E e y second day, pH o he cul u e media was measu ed,
and subsamples (1 ml) we e aken o enume a ion o algae
cells. A e subsampling, he bo les we e efilled o
capaci y wi h pH adjus ed-L1 g ow h medium (i.e., pH 8.0,
8.5, 9.0, 9.5, 10.0), and he bo les we e emoun ed on he
plank on wheel. I he pH di e ed by mo e han 0.03 om
he se poin , i was adjus ed by addi ion o aliquo s o
0.1 M HCl o NaOH.
In he pH-d i expe imen , F agila iopsis sp. and
Chlamydomonas sp. we e inocula ed (1,000 cells ml
-1
)in
media wi h a pH o 8.0 and ini ial TCO
2
concen a ions o
c. 1.2 o 2.4 mM and we e allowed o g ow in o s a iona y
g ow h phase (up o 26 d). The 1.2 mM TCO
2
concen a-
ion medium was ob ained by mixing he 2.4 mM L1
g ow h medium wi h a e y low TCO
2
concen a ion
medium ( 0.5 mM). The e y low TCO
2
medium was
p epa ed by acidi ying he g ow h medium ( o pH 3),
ollowed by hea ing o 110°C o 30 min and ae a ing he
medium. The pH was hen adjus ed o 8.0 by he addi ion
o 0.1 o 1.0 M NaOH o HCl (Hansen e al. 2007). The
expe imen was ca ied ou in gas- igh lamina ed NEN/PE
plas ic bags (Hansen e al. 2000). E e y second day, he pH
o he cul u e medium was measu ed, and subsamples we e
wi hd awn o enume a ion o algae cell concen a ion
(3 ml) and o measu emen s o he TCO
2
concen a ion
(36 ml). The NEN/PE plas ic bags we e no efilled a e
each sampling. The [CO
2
?HCO
3
-
] concen a ions we e
calcula ed om measu emen s o TCO
2
, empe a u e,
salini y, and pH (Lewis and Wallace 1998).
Succession expe imen
The h ee sea ice algal species we e inocula ed in a mixed
cul u e (i.e., 1,000 cells ml
-1
) a a pH o 8.0 and a
salini y o 33 and an ini ial TCO
2
concen a ion o
2.4 mM. The h ee sea ice algal species we e allowed o
g ow well in o s a iona y g ow h phase (up o 22 d).
E e y second day, pH o he cul u e media was measu ed,
and subsamples (1 ml) we e aken o enume a ion o he
mixed algae cells. A e subsampling, he bo les we e
efilled o capaci y wi h pH adjus ed-L1 g ow h medium
and he bo les we e emoun ed on he plank on wheel.
TCO
2
concen a ion was measu ed a he ini ia ion and
he e mina ion o he expe imen o ensu e ha he
concen a ion was su ficien o algae g ow h du ing he
expe imen s.
Resul s
E ec o salini y on he g ow h a es o h ee A c ic sea
ice algae
The wo sea ice dia oms exhibi ed simila g ow h a es as a
unc ion o salini y, and no significan di e ences we e
obse ed be ween acclima ion salini ies (33 o 75) used in
he salini y expe imen s (S uden ’s - es , P[0.05).
Maximum g ow h a es we e ob ained a a salini y o 33
(Fig. 1a). A salini ies abo e 33, g ow h a es g adually
dec eased wi h salini y. Howe e , g ow h a es a a salini y
o 100 we e educed by 50%, and none o he dia oms
could g ow a a salini y o 150. A salini ies below 33,
g ow h a es o he wo dia oms dec eased only sligh ly and
hey we e s ill qui e high a a salini y o 5 (Fig. 1a, b). The
wo dia om species showed significan ly mo e educed
g ow h a es a high salini ies han a low salini ies
(F agila iopsis nana OLS, P=0.005, one-sided and
-0.4
-0.2
0.0
0.2
0.4
0.6
52033 50 75 100 150
150
10075
5033205
0.6
0.4
0.2
0.0
-0.2
-0.4
F agila iopsis nana
F agila iopsis sp.
Chlamydomonas sp.
Salini y
g ow h a es (d-1)
g ow h a es (d-1)
F agila iopsis nana
F agila iopsis sp.
Chlamydomonas sp.
A
B
Fig. 1 F agila iopsis nana, F agila iopsis sp., and Chlamydomonas
sp. G ow h a es o he h ee sea ice algae as a unc ion o salini y.
aSalini y adjus ed om 33 o he expe imen al salini y. bSalini y
adjus ed om 75 o he expe imen al salini y. Da a poin s ep esen
ea men means SE ±(n=3)
1160 Pola Biol (2011) 34:1157–1165
123

116 PhD hesis by Do e Haubje g Søgaa d
F agila iopsis sp., P=0.005). The g ow h esponse o he
chlo ophy e as a unc ion o salini y was simila o ha o
he wo dia oms showing mo e educed g ow h a es a
high salini ies (Chlamydomonas sp. OLS, P=0.009, one-
sided) (Fig. 1a). G ow h a es o he chlo ophy e was
g ea ly educed al eady a a salini y o 50 and i could no
g ow a salini ies abo e 100 (Fig. 1a, b). E ec o pHand
TCO
2
limi a ion on he g ow h a eo he h ee sea ice
algae.
A e y p o ound e ec o he pH was obse ed on he
g ow h a es o all h ee species (Fig. 2). All species
exhibi ed maximum g ow h a es a a pH o 8.0–8.5
(Fig. 2). Abo e a pH o 8.5, a nega i e e ec o inc easing
pH was obse ed on he g ow h a e o all species. How-
e e , all species we e s ill able o g ow a hal he maxi-
mum g ow h a e a pH 9.5. The dia om species could no
g ow a pH 10, while he chlo ophy e species demons a ed
a g ow h a e o one- hi d i s maximum. The g ow h a es
o he wo dia oms we e significan ly educed a pH [9.0
(S uden ’s - es F agila iopsis nana, P =0.0066; F agi-
la iopsis sp., P=0.0061).
In he pH-d i expe imen s o F agila iopsis sp., he pH
eached a maximum o 9.5 and 9.7 in he expe imen s
ini ia ed a a TCO
2
concen a ion o 1.4 and 2.4 mM,
espec i ely (Fig. 3). Final TCO
2
concen a ions in hese
expe imen s we e 1.0 and 1.5 mM, espec i ely.
Fo he pH- ole an species, Chlamydomonas sp., he pH
eached 9.8, when g own a ini ially high and low TCO
2
concen a ions (Fig. 3). The final TCO
2
concen a ion was
1.0 mM in he expe imen s ini ia ed a a high TCO
2
con-
cen a ion, whe eas he concen a ions dec eased om 1.4
o 1.0 mM o his species in expe imen s ini ia ed a a low
TCO
2
concen a ion (Fig. 3).
Succession expe imen
The impo ance o pH in succession o sea ice algae species
was s udied using mixed cul u es o h ee sea ice algal
species (F agila iopsis nana,F agila iopsis sp., and
Chlamydomonas sp.) wi h an ini ial pH o 8.0 (Fig. 4). All
h ee species g ew un il pH eached 9.4 o 9.5 on Day 18.
A Day 20, he pH had inc eased o abo e 9.6, and he wo
dia oms s opped g owing, while he chlo ophy e species
main ained a posi i e g ow h a e.
Discussion
G ow h o sea ice algae a di e en salini ies
The abili y o sea ice algae o g ow wi hin he physio-
chemical g adien ound in he sea ice sugges s ha he
algae a e well adap ed o cope wi h fluc ua ions in ligh ,
empe a u e, salini y, pH, and TCO
2
concen a ions.
Howe e , salini y has a p onounced e ec on g ow h,
pho osyn he ic e ficiency, and me abolism (Mis a e al.
2001). Some mic oalgae a e conside ed eu yhaline, since
hey can adap o a ying ex e nal salini ies (Hellebus
1985). Howe e , he salini y ange o e which ac i e
g ow h akes place di e s g ea ly among species, and he
physiochemical condi ions in he sea ice will p o ide a
selec ion p essu e ha influences he final communi y
composi ion (Ryan e al. 2004). A p e ious s udy has
indica ed ha mos sea ice algae a e mo e ole an o
educed, a he han ele a ed salini ies (Ba es and Co a
1986). The p esen s udy suppo s he esul s o Ba es and
Co a (1986), as he h ee sea ice algae showed mo e
educed g ow h a es a high salini y le els han a low
salini y le els. The expe imen s sugges ha he wo dia-
oms ha e a compe i i e ad an age in sea ice, whe e b ine
salini y is g ea e han 50. These salini y condi ions a e
ypically encoun e ed whe e he sea ice empe a u e is
be ween -1.9 and -6.7°C (Glei z e al. 1995).
When sea ice mel s, he algae a e exposed o al e ed
salini ies and subsequen ly he algal g ow h may be influ-
enced. A p e ious s udy showed ha dia om species a e
only sligh ly a ec ed by dec easing salini ies, whe eas
dec easing salini ies may esul in subs an ial losses o
cilia es and flagella e species (Ga ison and Buck 1986;
Ryan e al. 2004; Mikkelsen and Wi kowski 2010). In he
p esen s udy, he h ee sea ice algal species we e exposed
o changes in salini y condi ions wi h di e en ini ial
salini ies o 33 o 75 o es he e ec o apid shi s in
salini y om high o low on he g ow h a es. The salini y
s ess had he smalles e ec on he g ow h a e o he wo
dia oms compa ed o he e ec on he chlo ophy e. This
sugges s ha sea ice dia oms a e less a ec ed by
-0.4
-0.2
0.0
0.2
0.4
0.6
8.0 8.5 9.0 9.5 10.0
F agila iopsis nana
F agila iopsis sp.
Chlamydomonas sp.
pH
g ow h a es (d
-1
)
Fig. 2 F agila iopsis nana, F agila iopsis sp., and Chlamydomonas
sp. G ow h a es o he h ee sea ice algae as a unc ion o di e en
fixed pH le els. Dissol ed ino ganic ca bon (TCO
2
) concen a ion
was ini ially be ween 2.2 and 2.4 mM in he expe imen flasks. Da a
poin s ep esen ea men means SE ±(n=3)
Pola Biol (2011) 34:1157–1165 1161
123
117
PhD hesis by Do e Haubje g Søgaa d
dec easing salini ies and hus may ha e a compe i i e
ad an age du ing summe and sp ing haw when sea ice
salini y becomes low. This esul compa es wi h p e ious
s udies showing ha sea ice dia oms domina es du ing sea
ice summe and sp ing haw (Palmisano and Ga ison
1993; Ika
¨ alko and Thomsen 1997; Mikkelsen e al. 2008).
Fu he mo e, he p esen s udy shows ha some sea ice
algal species a e be e adap ed o changes in salini y han
o he algal species, and hus, he changes in sea ice salini y
may d i e species succession o sea ice algae.
All he expe imen s we e conduc ed a highe empe a-
u es (3 ±1°C) han he in si u empe a u es obse ed in
sea ice (Søgaa d e al. 2010). P e ious s udies ha e shown
ha pho osyn he ic a es in sea ice algae a e influenced by
empe a u e (Palmisano e al. 1987; Ralph e al. 2005).
This sugges s ha he g ow h a es o he h ee sea ice algae
migh be o e es ima ed compa ed o g ow h a es a in si u
empe a u es. Howe e , i is a non i ial ask o incuba e
samples a low bulk salini y a 0 o sub-ze o empe a u e
wi hou in oducing eezing and haw a i ac s.
Tole ance o sea ice algae o ele a ed pH
The e ec o high pH on he g ow h a es o ma ine
plank onic algae is well es ablished. Some species a e e y
sensi i e o ele a ed pH and canno g ow when pH exceeds
8.8, while o he s s ill g ow a pH abo e 10 (e.g., Hansen
2002; Lundholm e al. 2004). Se e al s udies ha e also
shown ha he ole ance o high pH is species specific, and
la ge di e ences exis wi hin impo an ma ine algal
g oups, such as dia oms and dinoflagella es (Hansen 2002;
Lundholm e al. 2004; Søde be g and Hansen 2007).
The knowledge o he e ec o high pH on g ow h a es
o sea ice algae is howe e , e y limi ed. High pH is
obse ed in sea ice wi h high p ima y p oduc ion (Glei z
e al. 1995; Thomas e al. 2001) and hus p e ails du ing
sp ing when i adiance inc ease (Co a and Ho ne 1989;
Ku
¨hl e al. 2001). In he p esen s udy, influence o high pH
le els on he g ow h a e o he h ee species o sea ice
algae was s udied a pH le els anging om pH 8.0 o 10.0
in nu ien - ich g ow h media. We did no measu e he
nu ien concen a ions in he nu ien ich media du ing all
he expe imen s, bu he amoun o nu ien s le a he
e mina ion o he expe imen s assuming Redfield s oichi-
ome y documen s ha nu ien s we e no limi ing a any
poin du ing he expe imen s (see Table 1).
In he p esen s udy, ou esul s clea ly demons a e ha
all species we e es ic ed by high pH e en a a high ini ial
TCO
2
concen a ion. G ow h a es we e significan ly
educed o bo h dia om species a pH [9.0 (Fig. 2).
Abo e pH 9.5, he wo sea ice dia oms s opped g owing
i espec i e o TCO
2
, showing ha pH had a di ec e ec
on algal g ow h. Only a limi ed numbe o dia oms ha e
been s udied wi h espec o e ec o pH on g ow h;
howe e , among hose s udied he e ec o pH on g ow h
a ied (Lundholm e al. 2004). Lundholm e al. (2004)
ound ha smalle dia oms ha e a highe uppe pH limi o
g ow h han la ge dia om. In p esen s udy, we
0
1
2
3
4
pH
7.5
8.0
8.5
9.0
9.5
10.0
pH
TCO2[mM] and CO2+HCO3-[mM]
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
7.5
8.0
8.5
9.0
9.5
10.0
0
1
2
3
4B
02468101214161820222426
0
1
2
3
4D
7.5
8.0
8.5
9.0
9.5
10.0
pH
pH
TCO2[mM] and CO2+HCO3-[mM]
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
F agila iopsis sp. Chlamydomonas sp.
Cell concen a ion in 105 [cells ml-1]
7.5
8.0
8.5
9.0
9.5
10.0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
Time(d) Time(d)
0
1
2
3
4
A
0 2 4 6 8 10121416182022
C
Cell concen a ion
pH
TCO2
[CO2+HCO3-]
Cell concen a ion
pH
TCO2
[CO2+HCO3-]
Cell concen a ion
pH
TCO2
[CO2+HCO3-]
Cell concen a ion
pH
TCO2
[CO2+HCO3-]
Fig. 3 F agila iopsis sp. and Chlamydomonas sp. Cell concen a ion
in 10
5
, pH, dissol ed ino ganic ca bon (TCO
2
) and a ailable
ino ganic ca bon [CO
2
?HCO
3
-
] as a unc ion o ime o pH-d i
expe imen s a ini ial TCO
2
concen a ion o he wo sea ice algae.
Ini ial TCO
2
:(a,b) 2.4 mM and (c–d) 1.4 mM. Da a poin s ep esen
ea men means ±(n=3)
1162 Pola Biol (2011) 34:1157–1165
123
118 PhD hesis by Do e Haubje g Søgaa d
delibe a ely chose F agila iopsis nana because i is a ela-
i ely small dia om species and F agila iopsis sp. because i
is somewha la ge dia om species. Despi e he di e ence in
cell olume, he wo dia om species showed he same uppe
pH limi . The sea ice chlo ophy e showed an ex eme pH
ole ance as only a small educ ion in g ow h a e was
obse ed abo e his pH le el. The esul s sugges ha hese
sea ice algal species a e no limi ed by ino ganic ca bon a pH
8.0–9.0 (Fig. 3), a pH le el close o ha ound in sea ice b ine
(Papadimi iou e al. 2007). Howe e , pH inc eases in col-
onized sea ice because o a decline in TCO
2
as a esul o
pho osyn he ic ca bon assimila ion (Thomas e al. 2010).
This may a ec species such as F agila iopsis nana and
F agila iopsis sp (Figs. 2,3and 4). O he species such as
Chlamydomonas sp. can ole a e much highe pH le els and
hus will ha e a compe i i e ad an age in sea ice wi h high
pH le els (Figs. 2,3and 4). Howe e , his species was
limi ed by low TCO
2
concen a ions and hus may be ou -
compe ed by algal species in he sea ice, which a e able o
g ow a high pH le els and e y low TCO
2
concen a ions.
Algae can only u ilize CO
2
and HCO
3
-
o pho osyn hesis
(e.g., S umm and Mo gan 1996; Ko b e al. 1997), and i is
well known ha he specia ion o ino ganic ca bon species
depends upon pH. Fo ins ance, a pH 9.3, only hal o he
TCO
2
is a ailable in he o m o [CO
2
and HCO
3
-
]. In he
pH-d i expe imen s ini ia ed a a low TCO
2
, he algae we e
able o deple e he a ailable ino ganic ca bon [CO
2
and
HCO
3
-
] o a lowe limi o 0.45 mM o F agila iopsis sp.
and 0.25 mM o Chlamydomonas sp., assuming equilib ium
in he ca bona e sys em (Fig. 3). A hose low concen a ions
o [CO
2
and HCO
3
-
], g ow h a es o he algal species may
ha e become es ic ed by ca bon, as has been shown p e-
iously o dinoflagella es (Hansen e al. 2007).
Fo plank on communi ies, pH changes ha e been
shown o d i e species succession, because many plank-
onic algae appea o be qui e sensi i e o high pH (Hansen
2002; Pede sen and Hansen 2003). Howe e , possible ole
o ele a ed pH in he succession o A c ic sea ice algae has
ecei ed li le a en ion. The succession expe imen ca ied
ou in he p esen s udy sugges ed ha ele a ed pH may
well d i e species succession, as he pH- ole an species
(Chlamydomonas sp.) ou -g ew he wo sea ice dia oms
(Fig. 4). Howe e , how can we be su e ha he obse ed
succession pa e n in he s udy is due o pH changes and
no due o o ins ance p oduc ion o oxic subs ances
(allelochemicals) ha a ec he g ow h o he wo dia oms?
Well, we canno comple ely ou ule ha he chlo ophy e
exudes allelochemicals, as we did no es his specifically.
Howe e , no ma ine chlo ophy es ha e ye been
0 2 4 6 8 10121416182022
0
0.5
1.5
2.0
2.5
3.0
3.5
1.0
Time(d)
Cell concen a ion in 105 [cells ml-1]
pH
0 2 4 6 8 10 12 14 16 18 20 22
7.5
8.0
8.5
9.0
9.5
10.0
B
A
F agila iopsis nana
F agila iopsis sp.
Chlamydomonas sp.
Fig. 4 Succession expe imen . aChange in cell concen a ion in 10
5
o he h ee sea ice algae species. F agila iopsis nana, F agila iopsis
sp., and Chlamydomonas sp. as a unc ion o ime (d) om
inocula ion a pH 8.0. bpH as a unc ion o ime om inocula ion.
Da a poin s ep esen ea men means SE ±(n=3)
Table 1 Es ima ed upda e o C, N, and P in pH-d i expe imen s a dissol ed ino ganic ca bon concen a ion (TCO
2
) o 2.4 and 1.4 mM in
F agila iopsis nana,F agila iopsis sp., and Chlamydomonas sp. cul u es ha ha e eached maximum cell concen a ion in L1 medium
Algae species Maximum cell concen a ion
(cell ml
-1
)
TCO
2
ini ial 2.4 mM
C up ake
(lm)
N up ake
(lm)
P up ake
(lm)
C up ake
(lm)
N up ake
(lm)
P up ake
(lm)
F agila iopsis nana 4.1 910
5
919 139 8.7 329 50 3.1
F agila iopsis sp. 2.4 910
5
887 139 8.4 405 61 3.8
Chlamydomonas sp. 4.0 910
5
1,384 209 13.0 398 60 3.8
Addi ion o N and P o he seawa e in L1 medium was 1,111 and 47 lm, espec i ely. Es ima ion was based on a Redfield a io o 106C:16N:1P
Pola Biol (2011) 34:1157–1165 1163
123
119
PhD hesis by Do e Haubje g Søgaa d
con incingly shown o p oduce allelochemicals (see
e iew by G ane
´li and Hansen 2006). Secondly, he g ow h
o he wo dia om species in he mixed cul u e expe imen
can be explained by pH changes alone, and he e a e no
indica ions in ou da a se which sugges ha allelochem-
icals we e p oduced. Ou s udy has only deal wi h a ew
species o ice algae. Thus, much mo e a en ion is equi ed
in his opic. I would be pa icula ly in e es ing o s udy
how ele a ed pH and he e o e dec easing TCO
2
a ec s in
si u succession pa e n and he g ow h a es o he algal
species in he sea ice.
Conclusions
Ou esul s sugges ha he h ee sea ice algal species ha e
di e en ole ance o fluc ua ions in salini y and pH. The
esul s sugges ha he h ee sea ice algal species we e
mainly limi ed by pH, whe eas TCO
2
concen a ions only
played a ole a high pH le els and low TCO
2
concen a-
ions. The salini y s ess had he smalles e ec on he
g ow h a e o he wo dia oms compa ed o he e ec on
he chlo ophy e. This sugges s ha sea ice dia oms a e less
a ec ed by salini ies changes and hus may ha e a com-
pe i i e ad an age compa ed o he chlo ophy e in sea ice
wi h apid fluc ua ions in salini y. Finally, he fluc ua ions
in pH le els may d i e species succession o sea ice algae.
The chlo ophy e was able o ole a e much highe pH le els
han he wo dia om species. Thus, Chlamydomonas sp.
was able o g ow e en a high pH le els in he succession
expe imen and he e o e ou compe ed he wo dia om
species.
Consequen ly, sea ice algal species, which a e able o
g ow a fluc ua ing pH and salini y condi ions, may ha e an
ad an age in su i ing in he ha sh en i onmen o o ming
and mel ing sea ice.
Acknowledgmen s We hank Anna Haxen and Michael R. Sch ø-
de o assis ance in field and labo a o y and Thomas Juul-Pede sen,
K is ine A end and Paul Ba y o aluable commen s. Fu he mo e,
we wan o hank Nina Lundholm om he Scandina ian Cul u e
Collec ion o Algae and P o ozoa, Depa men o phycology, Uni-
e si y o Copenhagen o p o iding he dia om F agila iopsis nana.
The s udy ecei ed financial suppo om he Danish Agency o
Science, Technology and Inno a ion, KVUK Commission o Sci-
en ific Resea ch in G eenland and is a pa o he G eenland Clima e
Resea ch Cen e (GCRC 6507) and a con ibu ion o he Nuuk Basic
and Zackenbe g Basic p og ammes.
Re e ences
A igo KR, Sulli an CW (1992) The influence o salini y and
empe a u e co a ia ion on he pho ophysiological cha ac e is-
ics o An a c ic sea ice mic oalgae. J Phycol 28:746–756
A igo KR, Mock T, Lizo e MP (2010) P ima y p oduce s and sea
ice. In: Thomas DN, Dieckmann GS (eds) Sea ice, 2nd edn.
Blackwell Science, Ox o d, pp 283–325
Ba es SS, Co a GF (1986) Flu o escence induc ion and pho osyn he ic
esponses o a c ic ice algae o sample ea men and salini y.
J Phycol 22:421–429
Co a GF, Ho ne EPW (1989) Physical con ol o a c ic ice algal
p oduc ion. Ma Ecol P og Se 52:111–121
Cox GFN, Weeks WF (1983) Equa ions o de e mining he gas and
b ine olume in sea ice samples. J Glaciol 29:306–316
Dieckmann GS. Neh ke G, Papadimi iou S, Go
¨ liche J, S eininge
R, Kennedy H, Wol -Glad ow D, Thomas DN (2008) Calcium
ca bona e as ikai e c ys als in An a c ic sea ice. Geophys Res
Le 35:L08501. doi:10.1029/2008GL033540
Ga ison DL, Buck KR (1986) O ganisms losses du ing mel ing: a
se ious bias in sea ice communi y s udies. Pola Biol 6:237–239
Glei z M, Thomas DN (1992) Physiological esponses o a small
An a c ic dia om (Chae oce os sp.) o simula ed en i onmen al
cons ain s associa ed wi h sea ice o ma ion. Ma Ecol P og Se
88:271–278
Glei z G, Ru ge s VD, Thomas DN, Dieckmann GS, Mille o FJ
(1995) Compa ison o summe and win e ino ganic ca bon,
oxygen and nu ien concen a ions in An a c ic sea ice b ine.
Ma Chem 51:81–91
G adinge R (1999) Ve ical fine s uc u e o he biomass and
composi ion o algal communi ies in A c ic pack ice. Ma Biol
133:745–754
G ane
´li E, Hansen PJ (2006) Allelopa hy in ha m ul algae: a
mechanism o compe e o esou ces? In: G ane
´li E, Tu ne JT
(eds) Ecology o Ha m ul Algae, Ecological S udies se ies o
Sp inge -Ve lag, ol 189. Sp inge , Chap 15 pp 189–201
G an WS, Ho ne RA (1976) G ow h esponses o salini y a ia ion
in ou A c ic ice dia oms. J Phycol 12:180–185
Guilla d RRL, Ha g a es PE (1993) S ichoch ysis immobilis is a
dia om, no a ch ysophy e. Phycologia 32:234–236
Hansen PJ (2002) The e ec o high pH on he g ow h and su i al o
ma ine phy oplank on: implica ions o species succession.
Aqua Mic ob Ecol 28:279–288
Hansen JW, Thamd up B, Jø gensen BB (2000) Anoxic incuba ion o
sedimen in gas- igh plas ic bags: a me hod o biogeochemical
p ocess s udies. Ma Ecol P og Se 208:273–282
Hansen PJ, Lundholm N, Ros B (2007) G ow h limi a ion in ma ine
ed- ide dinoflagella es: e ec s o pH e sus ino ganic ca bon
a ailabili y. Ma Ecol P og Se 334:63–71
Hellebus JA (1985) Mechanisms o esponse o salini y in halo ol-
e an mic oalgae. Plan Soil 89:69–81
Hue as IE, Colman B, Espie GS, Lubian LM (2000) Ac i e anspo
o CO
2
by he h ee species o ma ine mic oalgae. J Phycol
36:314–320
Ika
¨ alko J, Thomsen HA (1997) The Bal ic Sea ice bio a (Ma ch
1994): a s udy o he p o is an communi y. Eu J P o is ol
33:229–243
Ko b RE, Sa ille PJ, Johns on AM, Ra en J (1997) Sou ces o
ino ganic ca bon o pho osyn hesis by h ee species o ma ine
dia oms. J Phycol 33:433–440
Ku
¨hl M, Glud RN, Bo um J, Robe s R, Rygaa d S (2001)
Pho osyn he ic pe o mance o su ace-associa ed algae below
sea ice as measu ed wi h a pulse-ampli ude-modula ed (PAM)
fluo ome e and O
2
mic osenso s. Ma Ecol P og Se 223:1–14
Lewis E, Wallace D (1998) P og am de eloped o CO2 sys em
calcula ions. En i onmen al Science Di ision. No. 4735
Lizo e MP (2003) The mic obiology o sea ice. In: Thomas DN,
Dieckmann GS (eds) Sea ice an in oduc ion o i s physics,
chemis y, biology and geology. Blackwell Science, Ox o d,
pp 184–210
1164 Pola Biol (2011) 34:1157–1165
123
126 PhD hesis by Do e Haubje g Søgaa d
Kaa okallio e al.: Bac e ial p oduc ion and cell p
ope
ies in sub-A
ic sea ice 61

ï
3
Ice
dep h
(cm)
and 50 nmol lï1 TdR o all sample ypes, 200, 700,
1500, 2000 and 2500 nmol lï1 Leu o ice and b ine
samples, and 100, 350, 750, 1000 and 1250 nmol lï1
Leu o wa e samples.
Expe imen al da a we e i ed wi h he single
kine ic Michaelis-Men en equa ion V= Vmax × S/[(K
+ Sn)+S], (whe e V= inco po a ion a e, Vmax = maxi-
mum inco po a ion a e, K + Sn = appa en hal -
sa u a ion cons an , i.e. hal -sa u a ion cons an +
na u al leucine/ hymidine concen a ion), S= added
leucine/ hymidine concen a ion), using SigmaPlo
10 so wa e (SPSS). Sa u a ing concen a ions we e
de ined as subs a e concen a ions whe e 90% o
Vmax was eached and calcula ed as Vmax90 = 9 × K +
Sn (Ayo e al. 2001). Iso ope dilu ion was assessed as
a a io o Vmax and Vmeasu ed a he sa u a ing concen-
a ion ( an Looij & Riemann 1993, Buesing & Gess-
ne 2003). Inco po a ion kine ics expe imen s we e
conduc ed wi h a 16 h incuba ion a 4 h in e als o
all 3 sample ypes on Day 3 using wa e om imme-
dia ely unde he ice, and b ine and bo om ice sam-
ples om Day 3 sampling.
S a is ical analyses
S a is ical analyses we e done using base, Psych,
mos 24 cm o ice), ‘bo om ice’, he lowe mos 4 cm
ice sec ion a ice-wa e in e ace, and ‘middle ice’,
he ice sec ions be ween he o he 2 classes. B ine
samples om bo h b ine sampling ho izons as well
as wa e samples om 2 sampling dep hs we e
pooled o o m sample classes ‘b ine’ and ‘wa e ’,
espec i ely.
Physical and chemical pa ame e s
Ice empe a u e, salini y and b ine olumes a e
p esen ed in Fig. 1 and Table 1. Di e ences in he 3
pa ame e s be ween sampling days we e no s a is i-
cally signi ican as e i ied by he KW es s. The ice
empe a u e a ied om ï3.8 o ï0.8°C, displaying
ypical nea -linea p o iles inc easing wi h ice dep h
(Fig. 1). Ice bulk salini y was signi ican ly highe in
bo om ice compa ed o uppe and middle ice (KW 2
= 17.64, p < 0.001, Wp-h bo h p < 0.008). Calcula ed
b ine olumes we e gene ally below 10% excep in
he bo om ice. The unde lying wa e had median
salini y o 33.1 wi h small empo al a iabili y.
Ino ganic and o ganic nu ien concen a ions a e
p esen ed in Table 1. Bo h DOC and DON concen a-
ions we e highe in bo om ice compa ed o middle
and uppe ice (KW, DOC: 2 = 8.39, p = 0.015, DON:
Vegan and MASS packages o R so wa e (R De el- 2 = 11.89, p < 0.003). NO3 concen a ions we e sig-
opmen Co e Team 2011). Di e ences be ween mo e ni ican ly highe in bo om ice compa ed o middle
han 2 sample ypes we e e i ied using he K uskal-
ice (NO
ï
:
2
= 10.61, p < 0.005) bu no uppe ice
ï
Wallis ank (KW) sum es wi h he pai wise
(Wp-h o DOC, DON and NO
3
all p < 0.05). Su p is-
Wilcoxon ank sum pos hoc es (Wp-h) and Bon e -
oni adjus men o , when compa ing 2 sample ypes
using Wilcoxon ank sum es s (W). Fo non-me ic
mul idimensional scaling (NMDS),
Spea man’s ank-o de co ela ions 0
be ween bac e ial and en i onmen al
pa ame e s we e calcula ed. NMDS
was pe o med on a B ay-Cu is dis-
simila i y ma ix o bac e ial pa ame- 20
e s wi h me aMDS w appe ou ine
included in he Vegan package
(Oksanen e al. 2012). En i onmen al 40
pa ame e s we e i ed on he NMDS
plo using he unc ion, e , included in
he Vegan package.
60
RESULTS
ingly, apa om silica e, b ine and ice concen a-
ions o dissol ed ino ganic nu ien s we e simila ,
despi e he salini y con as be ween bulk ice and
T
Uppe ice
Middle ice
Bo om ice
–4 –3 –2 –1
0 0 2 4 6 8 10 0 10 20 30 40
Fo analyses, ice co e sec ions we e
Tempe a u e (°C) Salini y
B ine olume (%)
di ided in o 3 classes: ‘uppe ice’,
he 2 uppe mos ice sec ions (uppe -
Fig. 1. Tempe a u e (T), salini y (S) and b ine olume (Vb) p o iles in ice on Days 2,
4 and 6. The 3 ice laye s used a e indica ed by di e en shading
V
b
S
Da
y
2
Day 4
Day 6

[Document text truncated for crawler view.]