1
Fe nández, A., Ma añón, E., Bode, A., 2014. La ge-scale me idional and zonal a iabili y 1 in he ni ogen iso opic composi ion o plank on in he A lan ic Ocean. J. Plank on Res. 2 (2014) 36(4): 1060-1073 doi:10.1093/plank / bu041 3 4
P e-p in e sion 5
6
7
La ge-scale me idional and zonal a iabili y in he ni ogen iso opic composi ion o plank on in he 8
A lan ic Ocean 9
Ana Fe nández
a, *
, Emilio Ma añón
a
, and An onio Bode
b
10
11
a
Dp o Ecoloxía e Bioloxía Animal, Uni e sidade de Vigo, E-36310 Vigo, Spain. 12
b
Ins i u o Español de Oceanog a ía, Cen o Oceanog á ico de A Co uña, E-15080 A Co uña, Spain. 13
14
Co esponding au ho : * Tel: +34 986 814087 Fax: +34 986 812556 e-mail: [email p o ec ed] 15
16
E-mail add esses: a ca e a@u igo.es (A. Fe nández), [email p o ec ed] (A. Bode), and 17
[email p o ec ed] (E. Ma añón). 18
19
Keywo ds: ni ogen iso opes, δ
15
N, phy oplank on, zooplank on, diazo ophy, A lan ic Ocean. 20
21
2
ABSTRACT 22
23
The zonal (ca. 15º-40ºW along 26-29ºN) and me idional (ca. 30ºN-30ºS along 28º-29ºW) 24
a iabili y o δ
15
N o suspended pa icles and zooplank on (>40 µm) was s udied o assess he 25
in luence o ni ogen ixa ion in he iso opic budge o he opical and sub opical A lan ic ocean. 26
Two c uises we e conduc ed in Oc obe -No embe 2007 and Ap il-May 2008 comp ising a zonal 27
and me idional ansec each. In he egion be ween 30º-15ºN, he concu en ly measu ed ni ogen 28
ixa ion was insu icien o explain he consis en pa ch o suspended pa icles wi h δ
15
N < 2‰ and 29
poin s o a signi ican con ibu ion o a mosphe ic deposi ion o ligh ni ogen o he iso opic 30
budge . The equa o ial egion (15ºN-10ºS) is subjec o in ense ni ogen ixa ion, which, acco ding 31
o a wo-end-membe mixing model, may explain 40-60% o he obse ed δ
15
N in suspended 32
pa icles and 3-30% in zooplank on. In he Sou h egion be ween 10ºS-30ºS, low alues (<4‰) 33
we e measu ed in suspended pa icles and zooplank on du ing 2008. The alues o δ
15
N o 34
suspended pa icles sugges ha ni ogen ixa ion, which is usually low (<10 µmol N m
−2
d
−1
), may 35
ep esen 50-60% o phy oplank on ni ogen in his egion. Hence, diazo ophy in he Sou h 36
A lan ic may be mo e impo an han p e iously hough . 37
38
39
40
3
INTRODUCTION 41
In many ma ine ecosys ems, p ima y p oduc ion is limi ed by he a ailabili y o ni ogen (Vi ousek 42
and Howa h, 1991; Ka l e al., 1992; Moo e e al., 2013). Reac i e ni ogen is supplied o he 43
eupho ic zone by di e en physical, chemical and biological p ocesses such as ad ec i e di usion, 44
a mosphe ic deposi ion and biological ni ogen ixa ion. The la e is media ed by o ganisms and, in 45
he oligo ophic egions o he oceans, is a ele an sou ce o new ni ogen (Pae l and Zeh , 2000). 46
The a io o s able iso opes in phy oplank on (
15
N:
14
N exp essed as δ
15
N in ‰) is a iable, due o 47
he con as ing p e e ences o he o ganisms o each iso ope. The me abolic pa hways usually 48
disc imina e agains he hea y iso ope (
15
N), a disc imina ion ha is measu ed by he iso opic 49
ac iona ion ac o (Mon oya, 2008). Besides, he di e en o ms o ino ganic ni ogen ha e 50
dis inc signa u es o δ
15
N. Deep-ni a e ypically anges be ween 3-6‰ (Mon oya, 2008), 51
a mosphe ic dini ogen is, by de ini ion, 0‰, and deep-ammonium lies be ween 6-8‰ (Miyake and 52
Wada, 1967). Hence, a e y di e en δ
15
N o o ganic ma e is expec ed, acco ding o he sou ce o 53
ni ogen, i his is comple ely consumed. The iso opic signa u e o phy oplank on will depend hen 54
on he signa u e o he sou ce o ni ogen and he deg ee o ac iona ion du ing up ake. Ye , he 55
in e p e a ion o δ
15
N is no so s aigh o wa d. In he case o animals (i.e. uppe ophic le els), a 56
ophic e ec is also obse ed whe eby he issues o he consume a e usually 2-4‰ hea ie han 57
he ood, whe eas he animal’s exc e a, mainly in he o m o ammonium, can be 2-4‰ ligh e han 58
he ood (Mon oya, 2008; and e e ences he ein). In addi ion, cul u ed cyanobac e ia g owing on 59
excess ni a e showed a s ong ac iona ion ac o , yielding δ
15
N alues simila o hose p oduced 60
by g ow h on dini ogen (Baue sachs e al., 2009). 61
In he A lan ic Ocean, expe imen al da a e ie ed du ing la ge-scale su eys show ha 62
T ichodesmium, he mos well-s udied diazo oph, is dis ibu ed p e e en ially be ween 0-20ºN 63
(Ty ell e al., 2003; Moo e e al, 2009; Fe nández e al., 2010; Luo e al., 2012). In addi ion, 64
ni ogen ixa ion, mos ly measu ed wi h he me hod o Mon oya e al. (1996), is mo e signi ican 65
be ween 0º-15ºN (Moo e e al., 2009; Fe nández e al., 2010; Luo e al., 2012). The δ
15
N o 66
diazo ophs usually anges be ween −1‰ and −2‰ (Mon oya e al., 2002). Howe e , he measu ed 67
iso opic signa u e o ni ogen in suspended pa icles and he biogeochemical es ima es o excess 68
ni ogen a ailable in he li e a u e sugges ha ni ogen ixa ion is mo e ele an in a egion u he 69
no h, be ween 15º-30ºN (G ube and Sa mien o, 1997; Maha ey e al., 2003; Maha ey e al., 70
2004; Reynolds e al., 2007; Hansell e al., 2004). The ime scales e lec ed by hese measu emen s 71
a e di e en : in si u ni ogen ixa ion a es gene ally ep esen ins an aneous a es o e a ew hou s 72
o 1 day, while δ
15
N and excess ni ogen a e indica o s o he diazo ophic ac i i y o e longe 73
pe iods o days o mon hs. Howe e , he de e minan s o his disag eemen emain unde ined. Duce 74
4
e al. (2008) a gued ha he a mosphe ic deposi ion o eac i e ni ogen in he oceans has inc eased 75
due o human ac i i ies and is as app oaching he ma ine N
2
ixa ion budge . O he s udies ha e 76
also shown an inc ease o he a mosphe ic deposi ion o
15
N-deple ed ni ogen in high and 77
empe a e la i udes (Has ings e al., 2009; Ma a e al., 2009; Mo in e al., 2009; Hol g ie e e al., 78
2011), as a esul o he inc easing an h opogenic p oduc ion o eac i e ni ogen and/o na u al 79
specia ion p ocesses. In addi ion, Bake e al. (2007) and Knapp e al. (2010) epo ed deposi ional 80
luxes o low δ
15
N simila o measu ed N
2
ixa ion a es in he A lan ic Ocean. 81
As pa o a wide p ojec , we ha e p e iously desc ibed he la i udinal and longi udinal dis ibu ion 82
o measu ed communi y ni ogen ixa ion in he opical and sub opical A lan ic Ocean (Fe nández 83
e al., 2010; Fe nández e al., 2013) and he ela i e con ibu ion o ni ogen ixa ion and ni a e 84
eddy di usion in supplying new ni ogen o he eupho ic laye (Mou iño-Ca ballido e al., 2011). 85
He e we epo on he dis ibu ion o δ
15
N in suspended pa icles and wo size- ac ions o plank on, 86
wi h he aim o desc ibing he la ge-scale la i udinal a iabili y o ni ogen iso opic signa u es in he 87
A lan ic Ocean and compa ing hese in e ed pa e ns o diazo ophy wi h concu en , di ec 88
measu emen s o in si u N
2
ixa ion a es. 89
90
METHODS 91
Sampling, hyd og aphy and chlo ophyll a 92
Two esea ch c uises we e conduc ed in he opical and sub opical A lan ic Ocean du ing 17 93
No embe -8 Decembe 2007 and 13 Ap il-2 May 2008 on boa d he BIO ‘Hespé ides’. The acks 94
ollowed by he essel comp ised a zonal and a me idional ansec in each season (Fig. 1). 95
The e ical dis ibu ion o empe a u e, salini y, dissol ed oxygen and luo escence was measu ed 96
by a SBE 911plus CTD a ached o a ose e equipped wi h 12-L Niskin bo les which was i ed o 97
300m dep h, always be o e dawn. The e ical p o iles o luo escence and oxygen a each s a ion 98
we e used o choose he sampling dep hs o he de e mina ion o ino ganic nu ien s concen a ion, 99
chlo ophyll a concen a ion, communi y
15
N
2
ixa ion and na u al abundance o ni ogen iso opes in 100
suspended pa icles. 101
The concen a ion o chlo ophyll a was measu ed a 6-7 dep hs dis ibu ed h ough he eupho ic 102
laye . A each dep h a 250-mL sample was il e ed, using low acuum p essu e, h ough 0.2 µm 103
po e-size polyca bona e il e s. The pigmen s we e ex ac ed o e nigh in 90% ace one a -4ºC. 104
Fluo escence was subsequen ly measu ed on boa d wi h a Tu ne Designs 700 luo ome e , 105
calib a ed wi h pu e chlo ophyll a (Fluka). 106
107
5
Ra es o N
2
ixa ion by he whole plank onic communi y in a 24-hou incuba ion pe iod we e 108
de e mined in each s a ion a he su ace (5m), an in e media e dep h (30-80m) and he dep h o he 109
deep chlo ophyll maximum (DCM), and a e al eady desc ibed in Fe nández e al. (2010) and 110
Fe nández e al. (2013). B ie ly, we incuba ed iplica e samples ollowing he Mon oya e al. 111
(1996) p o ocol o he
15
N
2
-up ake echnique wi h he modi ica ions o Rees e al. (2009). The 112
equa ions o Weiss (1970) and Mon oya e al. (1996) we e used o calcula e he ini ial N
2
113
concen a ion (assuming equilib ium wi h a mosphe e) and N
2
ixa ion a es, espec i ely. The limi 114
o de ec ion, es ima ed ollowing Mon oya e al. (1996), was 0.001 µmol N m
−3
d
−1
. 115
Na u al abundance o ni ogen iso opes in suspended pa icles 116
Fo he de e mina ion o δ
15
N signa u e in suspended pa icles (δ
15
N
sp
), 2-L samples we e aken a 6 117
dep hs h ough he eupho ic laye in each p e-dawn s a ion and il e ed h ough a 25-mm diame e 118
GF/F il e (Wha man). All il e s we e d ied a 40ºC du ing 24 h and hen s o ed un il pelle iza ion 119
in in capsules. The measu emen o pa icula e o ganic ni ogen (PON) and
15
N a om% was ca ied 120
ou wi h an elemen al analyze combined wi h a con inuous- low s able iso ope mass-spec ome e 121
(FlashEA112 + Del aplus, The moFinnigan) and using an ace anilide s anda d as e e ence. The 122
limi o de ec ion o he equipmen was 0.20 µg N. 123
The iso opic signa u e obse ed in he suspended pa icles may be a ec ed by he p esence o o he 124
ypes o ma e ial in addi ion o phy oplank on (i.e. bac e ia, de i us, zooplank on). The exis ence o 125
a ela ionship be ween he pa icula e o ganic ni ogen (PON) o chlo ophyll a (chl-a) a io and he 126
δ
15
N o suspended pa icles is an indica o o such a ophic e ec (Wase e al., 2000). The Pea son 127
p oduc -momen co ela ion coe icien o PON:chl-a and δ
15
N
sp
was calcula ed o es his 128
possibili y. 129
The weigh ed mean o δ
15
N o suspended pa icles in he eupho ic laye was used as an in eg al o 130
he signa u e o phy oplank on in he eupho ic zone o simpli y he compa ison wi h he δ
15
N o he 131
wo size- ac ions o zooplank on (40-200 µm and >200 µm). I was calcula ed, ollowing Land um 132
e al. (2011), as: 133
Weigh edmeanδ N


=∑PN

×∆z

×δ N




∑PN

×∆z



Whe e [PN]
i
is he concen a ion (µM) o pa icula e ni ogen, δ
15
N
sp
is he ni ogen iso opic 134
composi ion o suspended pa icles (
15
N:
14
N, ‰), and ∆z
i
is he dep h in e al (m). 135
The ac ion con ibu ion o diazo oph ni ogen o he bulk suspended pa icles de ined by 136
Mon oya e al. (2002) was also calcula ed as: 137

6
% !"#$%&$'ℎ)=100×, δ N


−δ NO
/
0

δ N
12345647

−δ NO
/
0

8
Whe e δ
15
N
diazo oph
is he ni ogen iso opic composi ion o diazo ophs (
15
N:
14
N, ‰) and δ
15
NO
3‒
is 138
he ni ogen iso opic composi ion o deep-ni a e (
15
N:
14
N, ‰). As poin ed by hese au ho s, his 139
wo-end-membe mixing model is sensi i e o he alues o he end membe s chosen (δ
15
N
diazo oph
140
and δ
15
NO
3‒
). In o de o ep esen only he ni a e in he uppe he mocline, and a oid he e ec o 141
ecen ly ixed ni ogen ecycled be ween he uppe wa e column and he he mocline in he 142
calcula ions, he δ
15
NO
3−
used was 4.5‰, which is he global a e age o deep-ni a e (Liu and 143
Kaplan, 1989; Sigman e al., 1997). Due o he ac ha mos o ou s a ions a e oligo ophic, no 144
addi ional ac iona ion ac o du ing ni a e up ake was added. As a conse a i e choice 145
ep esen ing he leas con ibu ion o ni ogen ixe s, and conside ing he ac ha li le ac iona ion 146
occu s du ing N
2
ixa ion (Mon oya, 2007), he δ
15
N
diazo oph
used was −2‰ (Mon oya e al., 2002). 147
148
Na u al abundance o ni ogen iso opes in plank on 149
A each p e-dawn s a ion, zooplank on we e collec ed by e ical ows o a 40µm ne o 30cm in 150
diame e h ough he uppe 200m o he wa e column a a cons an owing speed o 60 m min
−1
. 151
The con en o he collec o was suspended in 500 mL o 20 µm- il e ed seawa e . Two 60 mL sub-152
samples we e p ese ed, one in Lugol's solu ion and he o he in o maldehyde, o he 153
de e mina ion o abundance o T ichodesmium and o he plank on by mic oscopical examina ion. 154
T ichodesmium ichomes we e mo e abundan in he ac ion 40-200µm while colonies we e 155
p esen in he >200µm ac ion. The es o he sample was sepa a ed in o wo size ac ions by 156
passage h ough nylon sie es o 40 and 200µm. Each ac ion was hen e-suspended in 200 mL o 157
20µm- il e ed seawa e and subsequen ly il e ed in p e-weigh ed 45-mm diame e GF/F il e s by 158
low acuum p essu e. All il e s we e d ied o 24 h a 40ºC and s o ed un il measu emen o 159
pa icula e o ganic ni ogen and
15
N a om% as p e iously desc ibed. 160
The ela i e con ibu ion o diazo oph N o zooplank on biomass was es ima ed ollowing Mon oya 161
e al. (2002) wo-end-membe mixing model o zooplank on: 162
% !"#$%&$'ℎ)=100×,δ N
:2;<54;

−δ N
6=>=6=;?=:

δ N
12345647

−δ N
6=>=6=;?=:

8
Whe e δ
15
N
plank on
s ands o he ni ogen iso opic composi ion o he plank on size- ac ion 163
(
15
N:
14
N, ‰), and δ
15
N
e e ence pl
is he δ
15
N o e e ence zooplank on. Again, a conse a i e alue o 164 ‒2‰ was used o δ
15
N
diazo oph
. The δ
15
N o he e e ence plank on was calcula ed as he mean o 165
he δ
15
N
40
o δ
15
N
200
measu ed in he s a ions whe e he lowes abundance o T ichodesmium and 166
7
ni ogen ixa ion we e ound, i.e., he s a ions be ween 0-20ºS in he la i udinal leg o 2007 c uise, 167
whe e no in luence o ni ogen ixa ion in he samples is expec ed. The alues used we e: 4.6‰ o 168
he δ
15
N
e e ence 40
, and 5.9‰ o δ
15
N
e e ence 200.
This model is based on he use o e e ence 169
plank on o accoun o he ophic e ec , i.e., he e e ence plank on se es as a p oxy in bo h 170
e ms o he calcula ion; he e o e, no addi ional ac iona ion e m o he ophic e ec was 171
needed. The assump ions ha a e implied a e: i) he size dis ibu ion o g aze s in he sample and 172
he e e ence plank on a e simila , ii) he ophic ac iona ion in he sample and he e e ence is 173
simila , and iii) in bo h loca ions he iso opic composi ion o he ni a e suppo ing he ood web is 174
he same (J. P. Mon oya, A lan a, pe sonal communica ion). 175
176
RESULTS 177
Hyd og aphy and luo escence 178
In he la i udinal ansec s, he Equa o ial upwelling was clea ly de ined by he ising o he isolines 179
o empe a u e (T) and salini y (S) in bo h c uises (Fig. 2a, b, c, d). In o de o simpli y he analysis 180
o da a, we use he changes in he dep h o he 16ºC iso he m, abo e and below 150m, o iden i y 181
he a ea a ec ed by he Equa o ial upwelling and delimi h ee main egions along he ansec s, 182
i.e., No h gy e (29º-15ºN), equa o ial egion (15ºN-10ºS) and Sou h gy e (10º-33ºS). 183
The hyd og aphic se ings ound in hese egions we e simila in bo h legs. Su ace wa e s in he 184
equa o ial egion we e always wa me (> 24ºC) and less saline (< 35 psu) han in he gy es in bo h 185
seasons. In u n, he s abili y o he wa e column in he gy es was weake han ha ound in he 186
equa o ial egion, whe e he a e age B un -Väisäla equency in he uppe 125m was highe 187
(Fe nández e al., 2010). The luo escence p o iles showed a well-de ined deep chlo ophyll 188
maximum (DCM) associa ed wi h he he mocline in bo h ansec s (Fig. 2e, ). This DCM was 189
shallowe and be e de ined in he equa o ial egion han in he gy es. By con as , in he 190
longi udinal sec ions, wa e s we e wa me and sligh ly mo e saline in au umn 2007 han in sp ing 191
2008 (Fig. 2e), leading o a s onge s abili y o he wa e column, as indica ed by he highe B un -192
Väisäla equency measu ed in his c uise (Fe nández e al., 2013). The DCM was loca ed a ca. 193
100m in bo h zonal legs and no appa en end in dep h was obse ed (Fig. 2e, ). 194
195
S able ni ogen iso opes in suspended pa icles (δ
15
N
sp
) and pa icula e o ganic ni ogen (PON) 196
The Pea son p oduc -momen co ela ion coe icien o PON:chl-a and δ
15
N
sp
showed no signi ican 197
ela ionship, nei he posi i e no nega i e, in ou wo c uises (Fig. 3). The PON:chl-a a io 198
ep esen s he con ibu ion o o he componen s o he ood web han phy oplank on. I any 199
8
ela ionship be ween PON:chl-a a io and δ
15
N
sp
is ound, his will sugges a signi ican e ec o 200
de i al ma e and/o o he non-phy oplank onic o ganisms on he obse ed signal o suspended 201
pa icles. Thus, we can assume ha he δ
15
N o suspended pa icles in ou da a mainly e lec s he 202
composi ion o phy oplank on (Wase e al., 2000). 203
The zonal dis ibu ion o PON showed no appa en end and he measu ed concen a ions we e 204
simila in magni ude in bo h c uises (Fig. 4b, d), wi hou any signi ican di e ences be ween c uises 205
( - es , n=60). The mean PON concen a ions o he zonal ansec s we e 0.22±0.12 µM in 2007 206
and 0.24±0.06 µM in 2008. In he la i udinal sec ions, he measu ed concen a ions in he Sou h 207
gy e we e simila in bo h c uises, showing simila alues and e ical a iabili y (Fig. 4a, c). In he 208
No h gy e and equa o ial egions, measu ed PON di e ed signi ican ly be ween c uises ( - es , p < 209
0.01, n=54 and p < 0.05, n=84, espec i ely). In he No h gy e egion, he concen a ions o PON 210
in 2007 c uise we e highe han ha o 2008 c uise. In he equa o ial egion, PON depic ed a 211
dec easing pa e n in 2007 c uise, bu no linea end was obse ed in 2008. 212
In he zonal ansec o 2007 c uise, he δ
15
N
sp
inc eased sha ply by 4-6‰ om 80-100m o he base 213
o he eupho ic laye , p obably e lec ing he in luence o he A ican upwelling in he eas e nmos 214
s a ions (Fig. 5a). Howe e , he δ
15
N
sp
dis ibu ed uni o mly in he eupho ic laye in 2008 c uise 215
(Fig. 5b). In he la i udinal ansec s, he δ
15
N
sp
was lowe in 2007 han in 2008, bo h in he No h 216
gy e (Fig 5c, d) and equa o ial egion (Fig 5e, ). By con as , in he Sou h gy e egion, he e ical 217
dis ibu ion o δ
15
N
sp
di e ed be ween s a ions in each c uises (Fig 5g, h), wi h a wide ange o 218
alues be ween −4 and 4‰. The di e ence be ween c uises we e signi ican in he zonal ansec , 219
he No h gy e and he equa o ial egion ( - es , p < 0.01, n=72), bu no in he Sou h gy e egion. 220
The co ela ions o δ
15
N o suspended pa icles wi h ammonium concen a ion and wi h ni a e 221
concen a ion a e shown in Table 1. Conside ing all he s a ions in each c uise, δ
15
N
sp
co ela ed 222
wi h ammonium concen a ion in 2007 (p < 0.05, n=128) and wi h ni a e in 2008 (p < 0.05, 223
n=119). 224
225
Ni ogen iso opic signa u e in he eupho ic laye 226
In o de o compa e he δ
15
N o suspended pa icles and plank on ne ows, we calcula ed he 227
weigh ed mean o δ
15
N o suspended pa icles o ob ain an in eg a i e δ
15
N
sp
signa u e o he whole 228
eupho ic laye . The pa e ns desc ibed by he δ
15
N o he plank onic 40-200 µm (δ
15
N
40
) and >200 229
µm (δ
15
N
200
) size- ac ions we e e y simila in all s a ions, wi h a ew excep ions in he equa o ial 230
egion and he zonal legs, and closely ma ched ha o δ
15
N
sp
(Fig. 6). The a e age di e ences 231
be ween δ
15
N
sp
and he wo plank on size ac ions we e in he ange p e iously desc ibed 232
9
(Minagawa and Wada, 1984): be ween he δ
15
N
sp
and he δ
15
N
40
ha di e ence was 3.2‰ in 2007 233
and 2.6‰ in 2008; be ween he δ
15
N
sp
and he δ
15
N
200
was 4.3‰ in 2007 and 1.9‰ in 2008. 234
In he au umn 2007 me idional ansec , he iso opic signa u e o suspended pa icles showed wo 235
minima (< ‒2‰) in he No h gy e and Sou h gy e egions. In he equa o ial egion, δ
15
N
sp
236
oscilla ed a ound 0‰ (Fig. 6). The δ
15
N
40
and δ
15
N
200
oughly ollowed hese pa e ns. By con as , 237
he dis ibu ions we e dome-shaped in sp ing 2008, eaching peak alues in he equa o ial egion. In 238
bo h c uises, he gy es p esen ed low δ
15
N alues in mos o he s a ions. Besides, a posi i e 239
s a is ical co ela ion be ween δ
15
N in he h ee ac ions sugges s a egula impac o ligh ni ogen 240
ac oss ophic le els (Table 2). 241
A wo-way ac o ial ANOVA indica ed signi ican di e ences be ween egions and c uises, and o 242
δ
15
N
40
and δ
15
N
200
, a signi ican in e ac ion egion-c uise, which enhances he di e ence (Table 3). 243
The di e ences be ween egions appea ed o be signi ican only o he No h gy e-equa o ial 244
egion (pos -hoc Tukey HSD es ), as can be also seen in igu e 6. 245
We ied o es ima e i T ichodesmium could be he majo in luence on he pa e ns obse ed bu no 246
signi ican co ela ion (Pea son’s ) was ound be ween he measu ed ilamen abundance 247
(Fe nández e al., 2010; Fe nández e al., 2013) and he δ
15
N o suspended pa icles (p=n.s., n=42), 248
he 40-200 µm (p=n.s., n=41) o he >200 µm plank on size- ac ions (p=n.s., n=42). 249
Ni ogen ixa ion a es (Fig. 6) we e p e iously epo ed in Fe nández e al. (2010) and Fe nández 250
e al. (2013). B ie ly, in he longi udinal ansec s no appa en end was depic ed in 2007, while a 251
clea inc easing pa e n o he Eas appea ed in 2008 (Fig. 6b, d). In he sp ing 2008 zonal leg, he 252
a e age e ically in eg a ed N
2
ixa ion was 7- old highe han ha o au umn 2007 (8.3±3.3 µmol 253
N m
−2
d
−1
s. 1.2±0.5 µmol N m
−2
d
−1
). In bo h me idional ansec s, he highes in eg a ed a es (ca. 254
250 and 150 µmol N m
−2
d
−1
in 2007 and 2008, espec i ely) we e measu ed a s a ions loca ed 255
wi hin he equa o ial egion (Fig. 6a, c). Besides, he No h gy e showed highe diazo ophic 256
ac i i ies han he Sou h gy e. Bu , while N
2
ixa ion sou h o he Equa o was almos unde ec able 257
du ing he 2007 c uise, subs an ial a es we e measu ed in he Sou he n Hemisphe e in 2008 (Fig. 258
6a, c). 259
260
Diazo oph ni ogen con ibu ion o δ
15
N in he eupho ic laye 261
The con ibu ion o diazo ophs o he obse ed δ
15
N o suspended pa icles, 40-200µm and >200 262
µm plank on size- ac ions, es ima ed by he wo-end-membe mixing models, dec eased o he 263
Sou h in 2007 c uise (Table 4). In 2008, he minimum was obse ed in he equa o ial egion (Table 264
4). The impo ance o his con ibu ion is highe in 2007, excep in he Sou h gy e, whe e he 265
16
ha e add essed he deposi ional and diazo ophic luxes oge he (Bake e al., 2007; Knapp e al., 469
2011) and u he s udies a e needed o accu a ely de ine he s eng h, equency and iso opic 470
composi ion o he a mosphe ic deposi ional lux agains he lux o ni ogen ixa ion in he No h 471
A lan ic. The equa o ial egion (15ºN-10ºS) is subjec o ela i ely in ense ni ogen ixa ion 472
h oughou he yea (Moo e e al., 2009; Fe nández e al., 2010; G osskop e al., 2012) which may 473
explain 40 o 60% o he obse ed δ
15
N
sp
signal. Howe e , his ni ogen o diazo ophic o igin 474
seems o be ine icien ly ans e ed o uppe ophic le els. In he Sou h gy e, he low δ
15
N
sp
and 475
he daily es ima ed con ibu ion o ni ogen ixa ion o he supply o new ni ogen (Mou iño-476
Ca ballido e al., 2011) sugges ha diazo ophs can con ibu e up o hal o he ni ogen in 477
phy oplank on a di e en ime scales (Fig. 4). E en hough he measu ed ni ogen ixa ion a es a e 478
low (Moo e e al., 2009; Fe nández e al., 2010; G osskop e al., 2012), hei impac in he ni ogen 479
iso opic budge o his egion may be la ge. Hence, a e-e alua ion o he impo ance o diazo ophy 480
in he Sou h A lan ic Ocean is needed h ough new s udies ha should add ess he annual a iabili y 481
in ni ogen ixa ion a es as well as he dis ibu ion and ela i e impo ance o he di e en g oups 482
o diazo ophs. 483
484
ACKNOWLEDGEMENTS 485
486
The au ho s app ecia e he ca e ul e ision and hou ough commen s o wo anonymous e iewe s 487
who helped o imp o e he quali y o he manusc ip . We hank N. Lluch and P. Chouciño o 488
echnical assis ance. S able iso opes we e analyzed a SXAI-Uni e sidade da Co uña. We also 489
hank he o ice s and c ew o he BIO Hespé ides and he s a o he Ma ine Technology Uni 490
(UTM), o hei suppo du ing he wo k a sea. A. F. was suppo ed by g an 491
PGIDIT05PXIC31201PN o he Xun a de Galicia. This is a con ibu ion o he p ojec 492
TRYNITROP (T ichodesmium and N
2
ixa ion in he opical A lan ic Ocean) unded by he 493
Spanish Minis y o Science and Technology h ough g an s CTM2004-05174-C01 and CTM2004-494
05174-C02 o A. B. and E. M., espec i ely. 495
496
497

17
REFERENCES 498
Bake , A. R., Wes on, K., Kelly, S. D e al. (2007) D y and we deposi ion o nu ien s om he 499 opical A lan ic a mosphe e: Links o p ima y p oduc i i y and ni ogen ixa ion. Deep-Sea 500 Resea ch Pa I-Oceanog aphic Resea ch Pape s, 54, 1704-1720. 501 Baue sachs, T., Schou en, S., Compao e, J. e al. (2009) Ni ogen iso opic ac iona ion associa ed 502 wi h g ow h on dini ogen gas and ni a e by cyanobac e ia. Limnology and Oceanog aphy, 503 54, 1403-1411. 504 Checkley, D. M. and Mille , C. A. (1989) Ni ogen iso ope ac iona ion by oceanic zooplank on. 505 Deep-Sea Resea ch Pa a-Oceanog aphic Resea ch Pape s, 36, 1449-1456. 506 Duce, R. A., La oche, J., Al ie i, K. e al. (2008) Impac s o a mosphe ic an h opogenic ni ogen on 507 he open ocean. Science, 320, 893-897. 508 Fang, Y. T., Koba, K., Wang, X. M., e al. (2011) An h opogenic imp in s on ni ogen and oxygen 509 iso opic composi ion o p ecipi a ion ni a e in a ni ogen-pollu ed ci y in sou he n China. 510 A mosphe ic Chemis y and Physics, 11, 1313-1325. 511 Fe nandez, A., G ana, R., Mou ino-Ca ballido, B. e al. (2013) Communi y N-2 ixa ion and 512 T ichodesmium spp. abundance along longi udinal g adien s in he eas e n sub opical No h 513 A lan ic. Ices Jou nal o Ma ine Science, 70, 223-231. 514 Fe nandez, A., Mou ino-Ca ballido, B., Bode, A., e al. (2010) La i udinal dis ibu ion o 515 T ichodesmium spp. and N-2 ixa ion in he A lan ic Ocean. Biogeosciences, 7, 3167-3176. 516 Gao, Y., Kau man, Y. J., Tan e, D. e al. (2001) Seasonal dis ibu ions o aeolian i on luxes o he 517 global ocean. Geophysical Resea ch Le e s, 28, 29-32. 518 Glibe , P. M. and B onk, D. A. (1994) Release o dissol ed o ganic ni ogen by ma ine 519 diazo ophic cyanobac e ia, T ichodesmium spp. Applied and En i onmen al Mic obiology, 520 60, 3996-4000. 521 G osskop , T., Moh , W., Baus ian, T. e al. (2012) Doubling o ma ine dini ogen- ixa ion a es 522 based on di ec measu emen s. Na u e, 488, 361-364. 523 G ube , N. and Sa mien o, J. L. (1997) Global pa e ns o ma ine ni ogen ixa ion and 524 deni i ica ion. Global Biogeochemical Cycles, 11, 235-266. 525 Hansell, D. A., Ba es, N. R. and Olson, D. B. (2004) Excess ni a e and ni ogen ixa ion in he 526 No h A lan ic Ocean. Ma ine Chemis y, 84, 243-265. 527 Has ings, M. G., Ja is, J. C. and S eig, E. J. (2009) An h opogenic Impac s on Ni ogen Iso opes o 528 Ice-Co e Ni a e. Science, 324, 1288-1288. 529 Hawse , S. P., Oneil, J. M., Roman, M. R.e al. (1992) Toxici y o blooms o he cyanobac e ium – 530 T ichodesmkium o zooplank on. Jou nal o Applied Phycology, 4, 79-86. 531 Hea on, T. H. E. (1990) 15N/14N a ios o NOx om ehicle engines and coal- i ed powe s a ions. 532 Tellus B, 42, 304-307. 533 Hol g ie e, G. W., Schindle , D. E., Hobbs e al. (2011) A Cohe en Signa u e o An h opogenic 534 Ni ogen Deposi ion o Remo e Wa e sheds o he No he n Hemisphe e. Science, 334, 535 1545-1548. 536 Ka l, D., Michaels, A., Be gman, B. e al. (2002) Dini ogen ixa ion in he wo ld's oceans. 537 Biogeochemis y, 57, 47-+. 538 Ka l, D. M., Le elie , R., Hebel, D. V. e al. (1992) T ichodesmium blooms and new ni ogen in he 539 No h Paci ic gy e. Vol. 362. 540 Kiga, T., Yoshikawa, K., Tsunogai, U. e al. (2000) E alua ion o NOx o ma ion in pul e ized coal 541 i ing by use o ni ogen iso ope a ios, pape p esen ed a In e na ional Join Powe 542 Gene a ion Con e ence, Am. Soc. o Mech. Eng., Miami Beach, Fl, USA, 23–26 July. 543 Knapp, A. N., Has ings, M. G., Sigman, D. M. e al. (2010) The lux and iso opic composi ion o 544 educed and o al ni ogen in Be muda ain. Ma ine Chemis y, 120, 83-89. 545 Knapp, A. N., Sigman, D. M., Lipschul z, F. e al. (2011) In e basin iso opic co espondence 546 be ween uppe -ocean bulk DON and subsu ace ni a e and i s implica ions o ma ine 547 ni ogen cycling. Global Biogeochemical Cycles, 25. 548
18
Land um, J. P., Al abe , M. A. and Mon oya, J. P. (2011) Basin-scale dis ibu ions o s able 549 ni ogen iso opes in he sub opical No h A lan ic Ocean: Con ibu ion o diazo oph 550 ni ogen o pa icula e o ganic ma e and mesozooplank on. Deep-Sea Resea ch Pa I-551 Oceanog aphic Resea ch Pape s, 58, 615-625. 552 Liu, K. K. and Kaplan, I. R. (1989) The eas e n opical Paci ic as a sou ce o N-15-en iched ni a e 553 in seawa e o sou he n Cali o nia. Limnology and Oceanog aphy, 34, 820-830. 554 Luo, Y. W., Doney, S. C., Ande son, L. A. e al. (2012) Da abase o diazo ophs in global ocean: 555 abundances, biomass and ni ogen ixa ion a es. Ea h Sys . Sci. Da a, 4, 47-73. 556 Maha ey, C., Williams, R. G., Wol , G. A. e al. (2004) Physical supply o ni ogen o 557 phy oplank on in he A lan ic Ocean. Global Biogeochemical Cycles, 18. 558 Maha ey, C., Williams, R. G., Wol , G. A. e al. (2003) Biogeochemical signa u es o ni ogen 559 ixa ion in he eas e n No h A lan ic. Geophysical Resea ch Le e s, 30. 560 Ma a, P., Mihalopoulos, N., Gogou, A. e al. (2009) Iso opic composi ion o ni a e in we and d y 561 a mosphe ic deposi ion on C e e in he eas e n Medi e anean Sea. Global Biogeochemical 562 Cycles, 23. 563 Minagawa, M. and Wada, E. (1984) S epwise en ichmen o N-15 along ood chains - u he 564 e idence and he ela ion be ween del a-N-15 and animal age. Geochimica E Cosmochimica 565 Ac a, 48, 1135-1140. 566 Mino, Y., Saino, T., Suzuki, K. e al. (2002) Iso opic composi ion o suspended pa icula e ni ogen 567 (del a N-15(sus)) in su ace wa e s o he A lan ic Ocean om 50 deg ees N o 50 deg ees S. 568 Global Biogeochemical Cycles, 16. 569 Miyake, Y. and Wada, E. (1967) The abundance a io o 15N/14N in ma ine en i onmen s. Rec. 570 Oceanog . Wo ks Jpn., 9, 37–53. 571 Mompean, C., Bode, A., Beni ez-Ba ios, V. M. e al. (2013) Spa ial pa e ns o plank on biomass 572 and s able iso opes e lec he in luence o he ni ogen- ixe T ichodesmium along he 573 sub opical No h A lan ic. Jou nal o Plank on Resea ch, 35, 513-525. 574 Mon oya, J.P. (2007) Na u al abundance o 15N in ma ine plank onic ecosys ems. In Michene , R. 575 and Laj ha, K. (Eds.) S able Iso opes in Ecology and En i onmen al Science. Blackwell 576 Publishing L d, Malden, pp. 176-201. 577 Mon oya, J. P. (2008) Ni ogen S able Iso opes in Ma ine En i onmen s. In Capone, D. G., B onk, 578 D. A., Mulholland, M. R. and Ca pen e , E. J. (Eds) Ni ogen in he Ma ine En i onmen . 579 Else ie Inc., London. pp 1277-1302. 580 Mon oya, J. P., Ca pen e , E. J. and Capone, D. G. (2002) Ni ogen ixa ion and ni ogen iso ope 581 abundances in zooplank on o he oligo ophic No h A lan ic. Limnology and 582 Oceanog aphy, 47, 1617-1628. 583 Mon oya, J. P., Voss, M., Kahle , P. e al. (1996) A simple, high-p ecision, high-sensi i i y ace 584 assay o N-2 ixa ion. Applied and En i onmen al Mic obiology, 62, 986-993. 585 Moo e, C. M., Mills, M. M., Ach e be g, E. P. e al. (2009) La ge-scale dis ibu ion o A lan ic 586 ni ogen ixa ion con olled by i on a ailabili y. Na u e Geoscience, 2, 867-871. 587 Moo e, C. M., Mills, M. M., A igo, K. R. e al. (2013) P ocesses and pa e ns o oceanic nu ien 588 limi a ion. Na u e Geoscience, 6, 701-710. 589 Mo in, S., Sa a ino, J., F ey, M. M. e al. (2009) Comp ehensi e iso opic composi ion o 590 a mosphe ic ni a e in he A lan ic Ocean bounda y laye om 65 deg ees S o 79 deg ees N. 591 Jou nal o Geophysical Resea ch-A mosphe es, 114. 592 Mulholland, M. R. (2007) The a e o ni ogen ixed by diazo ophs in he ocean. Biogeosciences, 4, 593 37-51. 594 Mou ino-Ca ballido, B., G ana, R., Fe nandez, A. e al. (2011) Impo ance o N-2 ixa ion s. 595 ni a e eddy di usion along a la i udinal ansec in he A lan ic Ocean. Limnology and 596 Oceanog aphy, 56, 999-1007. 597 Mulholland, M. R. (2007) The a e o ni ogen ixed by diazo ophs in he ocean. Biogeosciences, 4, 598 37-51. 599
19
Oneil, J. M., Me zle , P. M. and Glibe , P. M. (1996) Inges ion o N-15(2)-labelled T ichodesmium 600 spp and ammonium egene a ion by he ha pac icoid copepod Mac ose ella g acilis. Ma ine 601 Biology, 125, 89-96. 602 Oneil, J. M. and Roman, M. R. (1994) Inges ion o he cyanobac e ium T ichodesmium spp. by 603 pelagic ha pac icoid copepods Maco se ella, Mi acia and Oculos ella. Hyd obiologia, 293, 604 235-240. 605 Pae l, H. W. and Zeh , J. P. (2000) Ma ine ni ogen ixa ion. In Ki chman DL (Ed) Mic obial 606 Ecology o he Oceans, Wiley-Liss Inc., Wilming on, pp 387−426. 607 Rau, G. H., Teyssie, J. L., Rassoulzadegan, F. e al. (1990) C-13/C-12 and N-15/N-14 a ia ions 608 among size- ac iona ed ma ine pa icles – implica ions o hei o igin and ophic 609 ela ionships. Ma ine Ecology P og ess Se ies, 59, 33-38. 610 Rees, A. P., Gilbe , J. A. and Kelly-Ge eyn, B. A. (2009) Ni ogen ixa ion in he wes e n English 611 Channel (NE A lan ic Ocean). Ma ine Ecology P og ess Se ies, 374, 7-12. 612 Reynolds, S. E., Ma he , R. L., Wol , G. A. e al. (2007) How widesp ead and impo an is N-2 613 ixa ion in he no h A lan ic ocean? Global Biogeochemical Cycles, 21. 614 Sigman, D. M., Al abe , M. A., Michene , R. e al. (1997) Na u al abundance-le el measu emen o 615 he ni ogen iso opic composi ion o oceanic ni a e: an adap a ion o he ammonia di usion 616 me hod. Ma ine Chemis y, 57, 227-242. 617 Ty ell, T., Ma anon, E., Poul on, A. J. e al. (2003) La ge-scale la i udinal dis ibu ion o 618 T ichodesmium spp. in he A lan ic Ocean. Jou nal o Plank on Resea ch, 25, 405-416. 619 Vi ousek, P. M. and Howa h, R. W. (1991) Ni ogen limi a ion on land an in he sea – How can i 620 occu . Biogeochemis y, 13, 87-115. 621 Wannicke, N., Liskow, I. and Voss, M. (2010) Impac o diazo ophy on N s able iso ope signa u es 622 o ni a e and pa icula e o ganic ni ogen: case s udies in he no h-eas e n opical A lan ic 623 Ocean. Iso opes in En i onmen al and Heal h S udies, 46, 337-354. 624 Wase , N. A., Ha ison, W. G., Head, E. J. H. e al. (2000) Geog aphic a ia ions in he ni ogen 625 iso ope composi ion o su ace pa icula e ni ogen and new p oduc ion ac oss he No h 626 A lan ic Ocean. Deep-Sea Resea ch Pa I-Oceanog aphic Resea ch Pape s, 47, 1207-627 1226. 628
629
630
20
Table and igu e legends 631
Table 1. Pea son p oduc -momen co ela ion coe icien be ween δ
15
N o suspended pa icles 632
(δ
15
N
sp
) and nu ien concen a ion: ammonium (NH
4
) and ni a e (NO
3
), in 2007 and 2008 c uises. 633
Numbe s in b acke s ep esen he o al numbe o samples used o he analysis. 634
Table 2. Pea son p oduc -momen co ela ion coe icien be ween δ
15
N o suspended pa icles 635
(δ
15
N
sp
), δ
15
N o 40-200µm plank on size- ac ion (δ
15
N
40
), and δ
15
N o >200µm plank on size-636
ac ion (δ
15
N
200
) in he la i udinal ansec s o 2007 and 2008 c uises. ** p< 0.01, n=17. 637
Table 3. Two-way ac o ial ANOVA ( egion, c uise) o δ
15
N o suspended pa icles (δ
15
N
sp
), δ
15
N 638
o 40-200µm plank on size- ac ion (δ
15
N
40
), and δ
15
N o >200µm plank on size- ac ion (δ
15
N
200
) 639
in he 2007 and 2008 c uises. DF, deg ees o eedom; SS, sums o squa es; MS, mean o squa es; 640
F, F s a is ic; p, p obabili y. 641
Table 4. Mean±S anda d de ia ion o he con ibu ion o diazo oph ni ogen o δ
15
N o suspended 642
pa icles, 40-200µm plank on size- ac ion (δ
15
N
40
) and >200µm plank on size- ac ion (δ
15
N
200
) 643
acco ding o he wo-end-membe mixing model p oposed by Mon oya e al. (2002). The e e ence 644
zooplank on used in each ac ion co esponded o he a e age o he s a ions sampled in he Sou h 645
gy e du ing 2007, whe e T ichodesmium abundance was < 1 ichome L
−1
, δ
15
N
40
= 4.6‰, δ
15
N
200
= 646
5.9‰. The alues o % o diazo oph N abo e 100 and below 0 we e disca ded in he calcula ion o 647
he egions a e age. Numbe o samples is indica ed in pa en heses. 648
Figu e 1. Sampling s a ions du ing he TRYNITROP c uises on boa d he BIO ‘Hespé ides’. Whi e 649
ci cles ep esen he au umn 2007 c uise (17 No embe - 8 Decembe 2007), and g ey iangles he 650
sp ing 2008 c uise (13 Ap il - 2 May 2008). 651
Figu e 2. Zonal and me idional e ical dis ibu ion o empe a u e (ºC), salini y and luo escence in 652
au umn 2007 and sp ing 2008 c uises. Dashed lines in he empe a u e panels de ine he limi s o 653
he h ee majo egions iden i ied by he dep h o 16ºC iso he m: No h gy e, equa o ial egion and 654
Sou h gy e. 655
Figu e 3. Rela ionship be ween he δ
15
N o suspended pa icles (δ
15
N
sp
) and he pa icula e o ganic 656
ni ogen (PON) o chlo ophyll a (chl-a) a io du ing he au umn 2007 (a) and he sp ing 2008 (b) 657
c uises. 658
Figu e 4. Zonal and me idional dis ibu ions o pa icula e o ganic ni ogen o suspended pa icles 659
(PON) du ing he au umn 2007 (a, b) and sp ing 2008 (c, d) c uises. Dashed lines de ine he limi s 660
o he h ee majo egions iden i ied by he dep h o 16ºC iso he m in he me idional ansec : No h 661
gy e, equa o ial egion and Sou h gy e. In he legend z1 o z6 ep esen he sampled dep hs om 662
deepe dep h, z1 (DCM), o shallowe dep h, z6 (5m). 663
21
Figu e 5. Ve ical dis ibu ion o δ
15
N o suspended pa icles (δ
15
N
sp
) in au umn 2007 and sp ing 664
2008 c uises g ouped by egion: a, b) zonal ansec (15º-38º W), c, d) No h gy e egion (30º-665
15ºN), e, ) equa o ial egion (15ºN-10ºS), and g, h) Sou h gy e egion (10º-30ºS). 666
Figu e 6. Zonal and me idional dis ibu ions o he weigh ed mean o δ
15
N o suspended pa icles 667
(δ
15
N
sp
), he δ
15
N o 40-200µm plank on size- ac ion (δ
15
N
40
), he δ
15
N o >200µm plank on size-668
ac ion (δ
15
N
200
), and concu en measu ed N
2
ixa ion (Fe nández e al., 2010; Fe nández e al., 669
2013) in au umn 2007 (a, b) and sp ing 2008 (c, d).Dashed lines de ine he limi s o he h ee majo 670
egions iden i ied by he dep h o 16ºC iso he m in he me idional ansec : No h gy e, equa o ial 671
egion and Sou h gy e. 672
673

22
Table 1. 674
2007 c uise 2008 c uise
All
s a ions No h
gy e
a
Equa o ial
egion Sou h
gy e All
s a ions No h
gy e
a
Equa o ial
egion Sou h
gy e
NH
4
0.22*
[128] 0.55*
[54] 0.43**
[38] n.s. n.s. n.s. 0.47**
[42] −0.38*
[36]
NO
3
n.s. 0.30*
[54] n.s. n.s. 0.21*
[120] n.s. n.s. 0.51**
[36]
** p < 0.01, * p < 0.05, n.s. no signi icance
675
a
includes he zonal and me idional legs in he No h gy e.
676
677
23
Table 2. 678
δ
15
N
sp
δ
15
N
40
2007 2008 2007 2008
δ
15
N
40
0.66 ** 0.82 **
− −
δ
15
N
200
0.66 ** 0.74 **
0.90 ** 0.78 **
679
24
Table 3. 680
681
δ
15
N
sp
δ
15
N
40
δ
15
N
200
DF
SS MS F p SS MS F p SS MS F p
egion 2 23.6 11.8 6.3 0.004
26.6 13.3 11.5 0.000
25.7 12.8 15.4 0.000
c uise 1 16.0 16.0 8.6 0.006
4.8 4.8 4.2 0.048
15.9 15.9 19.2 0.000
egion
s.c uise 2 6.0 3.0 1.6 0.213
20.9 10.5 9.0 0.000
21.7 10.9 13.1 0.000
682
25
Table 4. 683
684
Region Suspended pa icles 40-200µm
zooplank on >200µm zooplank on
2007 2008 2007 2008
2007 2008
Longi udinal ansec
81
(29) 59
(3) 52
(14) 16
(9) 43
(4) 31
(6)
No h gy e egion 85
(14) 61
(10) 48
(22) 21
(11) 41
(12) 36
(8)
Equa o ial egion 62
(27) 39
(8) 25
(18) 3
(2) 11
(3) 29
(12)
Sou h gy e egion 49
(12) 58
(18) 15
(20) 36
(16) 13
(15) 40
(17)
685
686