Journal of Marine Research, 50, 267-296, 1992
The dynamic changes of stable isotopic ratios of carbon and
nitrogen in suspended and sedimented particulate organic
matter during a phytoplankton bloom
by Takeshi Nakatsukal, Nobuhiko Handa', Eitaro Wada2•3 and Chi Shing Wong"
ABSTRACT
The dynamic changes of carbon and nitrogen stable isotopic ratios in suspended
and
sedimented particulate matter were observed together with many other chemical and biological properties during a phytoplankton
bloom induced by nutrient addition in a controlled
ecosystem enclosure (CEE, about 70 m3) in Saanich Inlet, British Columbia, Canada.
Both of the stable isotopic ratios of carbon (&I3C) and nitrogen (&15N) in suspended
particulate organic matter showed characteristic patterns of variations in surface water during
the bloom. The &I3Cof suspended particulate matter increased with the growth of phytoplankton population and decreased gradually after the depletion of N03- and NOz-. The &15Nof
suspended particulate matter was very low soon after the beginning of phytoplankton
bloom,
but the value increased rapidly with the decrease in N03 - and NOz -, and reached maximal
value following nutrient depletion, after which the &15Nremained high until the end of the
experiment.
In order to understand such variations of &I3C and &15N,we made the mass and isotopic
balance models of carbon and nitrogen for the upper layer of the CEE, and simulated the
temporal variations of &13Cand &15Nof particulate organic matter using them in connection
with several hypotheses on the isotope fractionations
associated with the uptake of inorganic
substrates by phytoplankton.
While neither change in the dissolved inorganic carbon (i.e., its
isotope ratio and/or molecular CO2 concentration)
nor the phytoplankton
species compositions can well explain the variation of &I3C, this variation can be well simulated considering the
effect of change in the specific production rate of particulate organic carbon. On the other
hand, the variations of &ISN can be clearly understood by a first-order isotope fractionation
model under the assumption of large isotopic fractionation
during the assimilation of N03and NOz - by phytoplankton.
The particulate organic matter produced in the nutrient controlled phytoplankton
bloom
can be classified into three phases from an isotopic viewpoint: (I) the early stage of the
phytoplankton
bloom when N03- plus NOz- were still in excess in sea water (high &13Cbut
low &15N),(II) the late stage of the bloom when N03 - plus N02 - had just been depicted (high
&13Cand high 8J5N) and (III) the steady state phase, a few days after the depletion of N031. Water Research Institute, Nagoya University, Chikusa-ku, Nagoya 464-01, Japan.
2. Mitsubishi Kasei Institute of Life Sciences, 11 Minamiooya, Machida, Tokyo 194, Japan.
3. Present address: Center for Ecological Research, Kyoto University, Shimosakamoto, Otsu, Shiga
520-01, Japan.
4. Centre for Ocean Climatic Chemistry, Institute of Ocean Sciences, 9860 W. Saanich Road, Sidney,
BC, V8L 4B2, Canada.
267
268
Journal of Marine Research
[50,2
plus N02- (low Sl3Cbut high S15N).The cooperative variation of Sl3C and S15Nin the
suspended and sedimcnted particulate organicmatter was also demonstrated.
1. Introduction
The isotopic ratios of carbon and nitrogen (813Cand 815N)in marine environment
have been studied frequently to date. In particular, the variations in isotopic ratios of
organic carbon and nitrogen in oceanic plankton have been discussed and explained
in several ways by many researchers.
For the carbon isotope ratio, Sackett et ai. (1965) firstly suggested the correlation
between the 813Ccomposition of plankton and the ambient temperature, which was
also confirmed by Degens et ai. (1968) and Fontugne and Duplessy (1981). Rau et ai.
(1989; 1991) suggested the relationship between oceanic plankton 813Ccompositions
and molecular CO2 concentration as the mechanism for explaining field 813Cchanges
due to temperature. Wong and Sackett (1978), however, showed that species changes
greatly affect the ol3C compositions by culture experiments. In addition, Takahashi et
ai. (1991) also proposed changes of the phytoplankton growth rate as the primary
reason for the changes in 013C of plankton with continuous culture experiments
undrr identical conditions of the CO2 concentration which equilibrated with atmosphere CO2• Many problems remain before we may obtain a complete understanding
of isotope fractionation associated with the fixation of inorganic carbon by phytoplankton.
On the other hand, the 815Ncompositions of oceanic plankton have always been
explained by the nitrogen dynamics in oceanic surface waters. Wada and Hattori
(1976) suggested that the low 815Nvalues of plankton are associated with N?, fixation
in low latitudinal ocean and large isotope fractionation, during the uptake of N03by phytoplankton, in high latitudinal ocean. Minagawa and Wad a (1986) and Saino
and Hattori (1987) also attributed the low 815Nof plankton in low latitudinal ocean
to N2 fixation. Checkley and Miller (1989), however, showed that large isotope
fractionation occurred during the ammonium excretion by zooplankton, and explained the low 815Nvalue of plankton in oligotrophic waters by ordinary isotope
mass balance without N2 fixation. Altabet (1988) also explained the 815Ncomposition
of particulate organic matter in surface water by an isotope mass balance model
containing upward N03 - flux and downward particle sinking associated with isotope
fractionations along the food chain.
Although many researchers have been dealing with the 8l3C and 815Ncompositions in quasi-steady state conditions, there have been few studies on the l3 and
815Nin the non-steady state marine environment such as the spring phytoplankton
bloom. Cifuentes et at. (1988) observed the seasonal changes of 813C and 815N
compositions of particulate organic matter in an inland bay and reported that the
8l3C compositions correlated with the primary productivity of waters and the 815N
compositions were controlled by the nutrient dynamics like N03 - uptake by phyto-
ac
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
bloom
269
plankton and nitrification of NH4 + by bacteria, which directly change the B15N
compositions in N03 - and N~ + due to large isotope fractionation associated with
uptake of these substrates by micro-organisms (Horrigan et ai., 1990). Goering et al.
(1990) reported the changes in B13Cand B15Nof marine organisms during a spring
bloom in a high latitudinal inland bay and showed the B13Cand B15Ncompositions
correlated with the water temperature and the N03- concentration, respectively.
Recently, Altabet et al. (1991) clearly demonstrated that the BI5N of oceanic
plankton increases together with the decrease in ambient N03- concentration
during a spring bloom.
In this study, we observed the temporal variations of B13Cand B15Nin the
suspended and sinking particulate organic matter together with the changes in
particulate organic carbon, nitrogen and nutrient concentrations, dissolved organic
carbon concentration, phytoplankton species and carbon assimilation rate by phytoplankton during the course of a phytoplankton bloom, under nutrient limitation,
which took place within a well-controlled ecosystem enclosure (CEE). Continuous
observation of the same water mass in this CEE must provide inherent information
on the chemical and biological variables during phytoplankton bloom. The nutrientcontrolled phytoplankton bloom is often encountered in high latitude ocean, local
upwelling areas and coastal region as one of the most exaggerated biological events
in the sea (Parsons et ai., 1984b) and thought to be the main source of sinking particle
and sedimented organic matter in the marine environment (Wefer, 1989).
Our study has two objectives. Firstly, we discuss which factors, as mentioned
above, actually control the BI3Cand BI5Nof suspended particulate organic matter.
Secondly, we generalize a typical pattern of change in the B13Cand BI5Nof suspended
and sedimented particulate matter in a nutrient-controlled phytoplankton bloom.
2. Materials and methods
The isotopic dynamic study using polyethylene controlled ecosystem enclosure
(CEE, 2.5 m diameter x 16 m depth with conical bottom) was carried out in Saanich
Inlet, British Columbia, Canada, in the summer of 1986. A column of water was
enclosed in the CEE, by SCUBA diving deployment (Menzel and Case, 1977) on 15
July 1986 (day 0). Immediately, nutrients (nitrate: 1800 mg at.; silicate: 3000 mg at.;
and phosphate: 180 mg at.) were added with gentle bubbling to obtain uniform
vertical distribution. In order to prevent contamination of nutrients, (i.e., bird feces),
the CEE was covered with a white woven plastic sheet. The chemical and biological
parameters of the ecosystem in the CEE were monitored for 23 days.
Water samples were collected from each of three depth intervals (0-4 m, 4-8 m,
8-12 m) at 0900 on days 1, 2, 4, 6, 8, 10, 13, 16,20 and 23 by a peristaltic pump and
were filtrated through precombusted (450°C, 4 h) glass fiber filters (Whatman
GF IC). The suspended particles collected on the filters were analyzed for particulate
organic carbon (POC), nitrogen (PON) and their BI3Cand B15Ncompositions. The
270
Journal of Marine Research
[50,2
filtrates were used for the measurement of the concentrations of dissolved organic
carbon. The changes in inorganic nutrients, chl.a concentration, cell numbers of
major phytoplankton and photosynthetic production rates of total organic carbon
and various organic compounds have already been reported and discussed in detail
by Hama et al. (1988). Phytoplankton cell counting was performed with the water
mixture of three depth intervals (i.e. average cell numbers throughout the water
column 0-12 m).
Sedimented particles deposited during sampling interval were collected from the
conical bottom of the bag by pumping through a vinyl hose on the sampling days 4, 8,
10, 13, 16, 20 and 23. The sedimented material was collected onto GF /C filters for
analysis of pac, paN content and their 013Cand ol5N compositions. The filtrates
were also used for the measurement of regenerated ammonium and dissolved
organic carbon.
For the measurement of pac and paN, the filters with the suspended or
sedimented particle were exposed to vapor of Hel in an enclosed glass aquarium for
a few hours to remove carbonate materials. Portions of the filters were analyzed for
pac and PN by CHN corder (Yanaco MS-1). The concentrations of dissolved
organic carbon in filtrates were measured by the wet oxidation method (Menzel and
Vaccaro, 1964). The residues of pac and paN analysis were analyzed for stable
isotopic ratios of carbon and nitrogen by the combustion method described in
Minagawa et at. (1984). The compositions of 13Cwere determined by mass spectrometer (Varian Mat 250) and those of 15Nby Hitachi RMU-6R. Stable isotopic ratios
were calculated in terms of oX as follows:
oX = [RsamPle - 1] X 103 (%0)
Rstd
where X is I3C or 15N,and R is 13C/12Cor 15N/14N,respectively. The standards for
carbon and nitrogen are Peedee Belemnite and N2 in air, and the standard deviation
of the measurement was less than 0.1 and 0.3%0for a13c and a15N,respectively. In the
case of a small amount of sample such as suspended particulate nitrogen less than 50
"",gN, the standard deviation sometimes increased to 1.0%0' However, the 015Nof
those samples varied far more than the analytical uncertainty discussed in this paper.
3. Results
a. Chemical and biological variables. N03 - plus N02 - usually nearly equal to N03 -,
were distributed uniformly at 26 "",gat.N 1-1 throughout the water column on day 1
(Fig. 1a). It decreased rapidly after day 2 to zero on day 6 at 0-4 m depths. Almost
the same trend was found in the 4-8 m depths with depletion occurring on day 8.
N03 - plus N02 - were never exhausted at 8-12 m depth until the end of the
experiment. Ammonium was usually found between 0.5 and 1.5 "",gat.N 1-1 through-
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
bloom
271
30
_
25
~
..•
20
:
15
'"
10
5
0
--
2.5
"-z
2.0
-;l t5
01
to
"-
0.5
0
ChLa
25
(c)
_20
-
"-
01
15
'"
10
5
o
o
2
4
6
8
10 ~
W
ffi
m
20 ~
~
Days
a 3
Phytoplankton
Centric dlatoms
C
numbers
cell
( 0-12m
(
d)
)
i:J.
100
o
"
u
; 40
....
_20
o
...
•
12
:: 8
"-z
..
..
:. 4
o
02466
10
12
14
16
18
20
22
24
Days
Figure 1. Temporal changes of chemical and biological variables in the CEE (a) N03- plus
N02- concentration; (b) NH4+ concentration; (c) ChI. a concentration; (d) Phytoplankton
cell numbers; (e,f) Concentration and sedimented flux of particulate organic carbon and
nitrogen; (g) Concentration of dissolved organic carbon; (h) Production rate of particulate
organic carbon; (i) Production rate of particulate organic carbon per unit concentration of
particulate organic carbon.
Journal of Marine Research
272
[50,2
200
..
160
" 120
u
:
""
-
8-12m
80
40
(9)
DOC
°
Production
rate 01 POC
(h)
-.c 4
"
::: 3
u
:;:2
..
';1
°
0.06
.c
;::0.04
0.02
°°
2
4
6
8
ro
12 W
16 16 W ~ ~
Days
Figure 1. (Continued)
out the period of the experiment (Fig. 1b). However relatively higher concentrations
were observed at 0-4 m on days 1, 2 and at deeper layers after day 16.
Other nutrients, phosphate and silicate, also tended to decrease rapidly at the
beginning of the experiment, but they were never depleted throughout the experiment in contrast to nitrate and nitrite.
High variabilities of chl.a with maxima of 18, 25 and 29 f.LgI-Ion day 4, 6 and 8
were observed at depth intervals of 0-4, 4-8 and 8-12 m, respectively (Fig. lc),
indicating that phytoplankton cells following rapid growth in the CEE were promptly
transferred to the water layers below by sinking.
The phytoplankton bloom was dominated by centric diatoms of Chaetoceros spp.
and Thalassiosira spp., and by small (mainly 2-5 f.Lm)nanoflagellates (Fig. 1d). The
maximum cell numbers of the phytoplankton were observed on day 8 in the whole
CEE. While centric diatoms were predominant at the maximum of phytoplankton
bloom, small nanoflagellates became dominant after day 13 and centric diatoms
completely disappeared after day 20. Other phytoplankton species were < 100 cells'
ml-1 throughout the experiment.
Suspended particulate organic carbon and nitrogen (PaC and paN) shown in
Figure Ie and If, increased with time to maxima between days 4 and 8 in the three
depth intervals and then decreased toward the end of the experiment as observed in
chl.a (Fig. 1c). This would indicate that changes in the suspended pac and paN are
1992]
Nakatsuka et al.: Nutrient controlled phytoplankton bloom
273
associated with the phytoplankton bloom developed after the spike of nutrients to
the CEE. However, the concentrations of POC and PON decreased much slower
than that of chl.a after the maximum period. Vertical fluxes of sedimented POC and
PON at the bottom of the CEE varied with time with maxima for days 8 to 13 (Fig. Ie
and If). Thus, phytoplankton particles settled from the 0-4 m depth interval to the
bottom of the bag (16 m depth) in 4-9 days.
Dissolved organic carbon (DOC) increased gradually throughout the experiment
(Fig. Ig). However, large variations in the concentration were found in the early and
final phases of the experiment.
Primary productivity measured by l3C-tracer incubation method (Hama et aI.,
1988) increased rapidly early in the experiment to reach the maxima between days 4
and 8 at the three depth intervals (Fig. Ih). Although it decreased greatly at 0-4 m
after the maximum, second maxima were also found on day 16 at 4-8 and 8-12 m
depth intervals. The major part of the primary production of the whole water column
occurred at 0-4 m throughout the experiment.
b. Budget of nutrients and organic matter in whole CEE. Budgets of organic carbon
and total nitrogen in the whole CEE throughout the experiment are illustrated in
Figure 2. They contain the changes in the standing stocks of POC, DOC, PON, N03and NH4 + in the water column, the primary production of POC and the fluxes of
sedimented POC and PN, including the regenerated NH4 + and DOC associated with
the sedimented matter, during the experiment. Because the water chemistry below
12 m depth to the bottom of this CEE was not identified, the concentrations of POC,
DOC, PON, N03 - and NH4 + at that depth interval are assumed to be equal to those
at 8-12 m for the calculation ofthe changes in the standing stocks in the whole CEE.
Because of uncertainty in the calculation of total primary production in the CEE
throughout the experiment using data of primary productivity per hour, the presented organic carbon budget is not complete. However, it can be at least concluded
that about half of the primary production of POC settled to the bottom of the CEE as
sinking particles (Fig. 2a), which must correspond to the sinking rate of POC below
the euphotic zone in oceanic surface water during the phytoplankton bloom.
Total budget of nitrogen indicates that there is missing nitrogen (570 mg at.N /
Bag) after the experiment, which is equivalent to a quarter of the initial standing
stock of total nitrogen (2140 mg at.N/Bag). It can, however, be explained partly by
the change of dissolved organic nitrogen (DON) because DOC increased in the
water (870 mg at.C/Bag); it also accumulated on the bottom of CEE together with
the sedimented particles (2270 mg at.C). If the C/N atomic ratio of dissolved organic
matter (DOM) in this CEE is about 7, the missing nitrogen can be compensated
completely by the change in concentration and accumulation of DON. The C/N ratio
of DOM is, however, thought to be higher than 7 (Williams, 1975), and thus some
part of the missing nitrogen must be explained by some other components in the
274
Journal of Marine Research
[50,2
mg at. C / Bag
Bottom
Sediment
+8,110
( a)
mg at. N / Bag
Bottom
Sediment
+1,070
(b)
Figure 2. Budgets of organic carbon (a) and total nitrogen (b) in whole CEE throughout the
experiment. In Figure a, "a" means a factor with which production rates of POC per hour
measured in the daytime are converted to those per day. Bottom DOC (a) and NIL+ (b)
were measured in waters sampled together with bottom sediment.
1992]
Nakatsuka et al.: Nutrient controlled phytoplankton bloom
gJ
-15
-16
Suspended [ ~::
6.
poe
•
sed~~e~ted
275
~
a-12m
16
m
EO
~ -17
•
9
B
~
..•
7
- 6
Z
Q. 5
'0 4
Z
~
3
2
o
2
4
6
B
W
U
M ffi m W~
M
Days
Figure 3. Temporal changes in carbon (a) and nitrogen (b) isotopic ratios of suspended and
sedimented particulate organic matter in the CEE.
CEE. The nitrogen in the water depth below 12 m is the other factor responsible for
the missing nitrogen, because the vertical distributions of N03 - and NH4 + in the
water column on day 23 suggests the existence of a higher N03- and NH4+
concentration near the bottom of CEE. Moreover, there must have been much PON
deposited
C.
on the bottom of CEE which had not been sampled yet.
fj/3C and fj/5N in suspended and sedimented POM. At interval 0-4 m, fj13C of
suspended POC increased rapidly from - 20.3%0 (day 1) to a maximum of -17.6%0
(day 6), then gradually decreased to -20.5%0 on day 23 (Fig. 3a). Almost the same
trend of fj13C was observed for depth intervals 4-8 and 8-12 m with fj13C maxima on
day 6. In later period of this experiment, fj13C of the suspended POC in deeper layers
became higher than that in the 0-4 m layer. The fj13C of sedimented POC collected at
the bottom of the CEE changed following those of suspended POC in the water
column, although rather higher values of a13c were found after day 14 (Fig. 3a).
fjl5N of suspended PON decreased rapidly in the beginning of this experiment till
day 4, returned to about 7%0 throughout the water column in the CEE on days 6 and
8, and then remained relatively constant until the end of this experiment (Fig. 3b).
The fJ15N of the sedimented PON showed almost identical trends as the suspended
PON but the former's variation in amplitude was smaller (Fig. 3b).
Journal o[ Marine Research
276
Table 1. Equations
[50,2
in the mass balance model of carbon and nitrogen.
1. Conservative
equations of carbon
d[POC]/dt = P(poc) - D(poc) - S(poc)
d[DOC]/dt = P(doc) - D(doc)
d[DIC]/dt
= D(poc) + D(doc) - P(poc) - P(doc)
P(doc) = 0.2' P(poc)
2. Conservative equations of nitrogen
d[PON]/dt
= P(pon) - D(pon)
- S(pon)
d[DON]/dt
= P(don) - D(don)
P(pon) + P(don) = U(N03) + U(N~)
d[N03]/dt
= - U(N03)
d[NH4]1dt = D(pon) + D(don) - U(NH4)
3. Assumptions on the C/N ratio of organic compounds
S(poc)/S(pon)
= C/N(pom)
D (poc)/D (pon) = C/N(pom)
P(doc)/P(don)
= D(doc)/D(don)
= 15
(1-1)
(1-2)
(1-3)
(1-4)
(1-5)
(1-6)
(1-7)
(1-8)
(1-9)
(1-10)
(1-11)
(1-12) (1-13)
The terms are defined as follows: [POC], [PON] concentration
of suspended particulate
organic carbon and nitrogen (I1mole/I); [DOC], [DON] concentration
of dissolved organic
carbon and nitrogen (I1molelI); [DIC], [N03], [NH4] concentration
of dissolved inorganic
carbon, nitrate and ammonium (I1molelI); P(poe), P(pon), P(doe), P(don) production rate of
particulate organic carbon, nitrogen, dissolved organic carbon and nitrogen (I1mole/l/day);
U (N03), U(NH4) uptake rate of N03and NH4+ by phytoplankton
for photosynthesis
(I1molell/day);
D(poe), D(pon), D(doe), D(don) decomposition
rate of particulate organic
carbon, nitrogen, dissolved organic carbon and nitrogen (I1mole/l/day);
S(poe), S(pon) sinking
. rate of particulate organic carbon and nitrogen, defined per unit volume of the concerned
surface water layer (I1mole/l/day);
C/N(pom) C/N ratio of particulate organic matter.
4. Models and discussions
a. Mass balance model o[ carbon and nitrogen in CEE. In order to understand what
factors actually govern the temporal variations in I)I3Cand 1)15N values of suspended
paM, we make mass and isotope balance models of carbon and nitrogen. Firstly, we
propose the mass balance model of carbon and nitrogen which can completely
simulate the actual changes of carbon and nitrogen budgets in the CEE during this
experiment, using the data of temporal changes in pac, paN, DOC and production
rate of pac, under several assumptions regarding the carbon and nitrogen cycling in
the CEE. On the basis of the reconstructed variations in the carbon and nitrogen
budgets, the variations in I)I3Cand 1)15N of each organic and inorganic component in
the CEE will be discussed with the other models which represent isotope mass
balance of carbon and nitrogen.
The mathematical notations and the schematic illustration of the mass balance
model are presented in Table 1 and Figure 4, respectively. The following seven
components of carbon and nitrogen are considered: pac, paN including mainly
phytoplankton and additionally other organisms and detritus; DOC, DON; DIC
including CO2, HC03-, and COl-; N03- and NH4+.
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
#
S(poc)
#
bloom
277
S(pon)
Figure 4. Schematic illustration of mass balance model for carbon and nitrogen cycling in the
CEE. Closed stars indicate budgets and flux whose magnitudes can be calculated explicitly
using the data in Figure 1. Symbols of # mean POM and DOM fluxes whose C/N ratios are
prescribed for mass balance calculation. All names of fluxes in this figure correspond to the
notation in Table 1.
The particulate organic matter (POM) is assumed to be produced only by the
photosynthesis of phytoplankton, and the production of POM through the microbial
loop is neglected here (Eqs. 1-1 and 1-5), because the photosynthetic activity of
phytoplankton was very high in the phytoplankton bloom in this experiment (Hama
et al., 1988). It is also necessary to simplify the condition of calculation under the
limitation of data on the carbon and nitrogen budgets in this study. The production
of DOM is also assumed to occur only as the exudation of phytoplankton accompanied with the photosynthetic production of POM, and the contribution of heterotrophers was neglected (Eq. 1-4). The phytoplankton exudation fraction of organic
carbon during photosynthesis is assumed to be constant at 20% although it may be
affected by change in environmental factors (Ignatiades, 1973) (Eq. 1-4). All vertical
flux of carbon and nitrogen outside the concerned layer in the CEE is attributed to
only the sinking of POM, and the diffusive transport of dissolved matter is neglected
here (Eqs. 1-1, 1-2, 1-3, 1-5, 1-6, 1-8 and 1-9), except for some DOM flux in the later
period of the experiment mentioned below, because the water column was stratified
by temperature gradient after day 1 and there are usually only small gradients in the
concentrations of dissolved components between 0-4 m, 4-8 m and 8-12 m layers
(Fig. 1). PON and DON are produced using substrates N03- and NH4+ (Eq. 1-7).
The decomposition of POM and DOM directly contributes to the increases of DIe
and NH4 + (Eqs. 1-3 and 1-9), and it is assumed that nitrification cannot occur in CEE
(Eqs. 1-8 and 1-9). The production of calcium carbonate is also neglected (Eq. 1-3)
because CaC03 producing plankton like coccolithus were not identified in the CEE.
The transfer of CO2 through the water surface is neglected (Eq. 1-3) because the
CEE was covered with the plastic sheet as mentioned above, and almost no exchange
278
Journal of Marine Research
[50,2
of gases can be expected through the water surface under completely windless
conditions as in this study (Etcheto et al., 1988). Even if it is assumed that the wind
speed on the water surface is 1 m/ s, the CO2 transfer velocity k must be calculated as
0.17 cm/h (about 4 em/day) at 20°C (Wong and Chan, 1991), which cannot affect the
DIC concentration in the CEE significantly. This is because, under the condition k =
4 em/day, the CO2 flux through the water surface cannot exceed 1000 !-Lmole/m2/day,
which is approximately equivalent to only 1/20000 of the DIC pool of 10 m water
column in the CEE, even when CO2[aq] has been consumed greatly by phytoplankton and LlPCO2reaches - 300 ppm.
In order to perform the mass balance calculation with the limited data, we define
the C/N ratios of the following four fractions of organic matter as follows. The C/N
ratios of both the sinking POM and the POM fraction, which is to be decomposed,
are assumed to be equal to the C/N ratio of POM standing stock at that time (Eqs.
1-10 and 1-11). The C/N ratios of both the produced and the decomposed DOM
(and as a result, DOM standing stock itself) are put at 15 (Eqs. 1-12 and 1-13),
because the C/N ratio of DOM is usually much higher than that of POM (Williams,
1975). The constancy in C/N ratio of DOM does not appear realistic for strict
calculation of nitrogen mass balance, because we have no information on DON in the
CEE and DON concentration must always be determined using the data of DOC in
Fig. Ig and this C/N ratio. The variation in DOM concentration is, however, much
slower than that in POM (Fig. 1), therefore, this assumption on DOM does not affect
budget of POM seriously, even if the C/N ratio of DOM is, in fact, not constant but
variable.
Using these thirteen independent equations in Table 1, we can simulate the
temporal variations of carbon and nitrogen budgets in the CEE explicitly, because
Ll[POC] (Ll[POC] means that d[POC]/dt . Llt), Ll[DOC], Ll[PON], Ll[N03], Ll[NH4],
C/N(pom) and P(poc) in the equations can be calculated directly from the observed
data in Figure 1 and there remain only the following thirteen unknown variables
(whose number is equal to that of the equations), D(poc), S(poc), P(doc), D(doc),
Ll[DIC], P(pon), D(pon), S(pon), Ll[DON], P(don), D(don), U(N03) and U(NH4).
The initial concentration of [DIC] is put at 1830 !-Lmole/las an approximate estimate
with the temperature (15°C) and salinity (29) data by the method of Persons et al.
(1984a). The photosynthetic production of POC (P(poc)) per day is set as 22 times (a
in Fig. 2a; higher than 12 but lower than 24) that observed per hour in Fig. Ih,
because the photosynthetic activities per hour were measured during day time from
10:00 to 15:00 and the highest photosynthetic activities are usually observed not at
noon but early in the morning (Taguchi, 1976).
The simulation of the temporal variations in carbon and nitrogen budgets is
carried out on the data at the 0-4 m depth interval of the CEE. To simplify the
conditions of calculation, the rates of transformations of carbon and nitrogen
components (e.g. P(poc), D(poc), S(poc)) in the concerned layer are assumed to be
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
bloom
279
constant during each sampling interval (2 or 3 days). The P(poc) during each
sampling interval is defined as the average of those at both the beginning and the end
of that interval which can be calculated with the data in Figure Ih. Although these
assumptions on the conditions of calculation would not appear highly realistic, the
simulation will present useful results in sufficient detail for the assessment of isotopic
dynamics on the basis of mass balance because the sampling intervals are very short
in this study.
The results of simulation at 0-4 m depth interval are presented in Figure 5, in
which all units of fluxes and budgets are unified to j.Lmole/l/day and j.Lmole/l,
respectively. The sinking rate and decomposition rate of poe per unit concentration
of poe varied from 0 (-0.1; see below) to 0.9 and from 0.1 to 1.1 (lday), respectively
(Fig. 5b, 5c and Fig. Ie). This specific sinking rate at 0-4 m depth interval is
equivalent to the sinking rate 0 to 4 m per day, which is reasonable for phytoplankton
and detritus (Walsby and Reynolds, 1980). The relatively high specific decomposition rates are mainly governed by the high specific production rate of poe during the
phytoplankton bloom, which varied from 0.25 to 1.5 in this experiment (22 times
those in Fig. Ih). The maximum sinking poe flux outside 0-4 m is observed at the
sampling interval of days 4-6, which precedes the maximum in the sedimented poe
flux identified at 16 m depth for about 5 days. This result is consistent with the
observation of the temporal changes in poe concentration at three depth intervals
in the eEE.
Because the temporal variations in the concentrations of carbon and nitrogen
components in the eEE (Fig. 1) were measured independently by the different
methods which have different accuracy and there are some bold assumptions in the
equations in Table 1, the results of simulation on the basis of those data may have
some contradiction. In fact, there are two discrepancies in this simulation. One is the
occurrence of "upward sinking" flux of POM at three sampling intervals (Fig. 5c and
5i); and the other is the fact that the decomposition rate of DOM is calculated as the
value under zero during the last interval from day 20 to day 23 (Fig. 5e and 5k).
In relation to the sinking POM flux in the first discrepancy, the following equation
can be derived from Eq. (1-5) to (1-9).
S(pon) . 6.t
= -(6.[PON] + 6.[DON] + 6.[N03] + 6.[NH4]),
(1)
where 6.t means the time length of the sampling interval. Because 6.[PON], 6.[N03]
and 6.[NH4] can be calculated explicitly with the observed data in Figure 1 and
6.[DON] can be guessed from 6.[DOC] (as like Eqs. 1-12 and 1-13), the S(pon) can
be thought to be the most reliable value obtained with the equations in Table 1. The
calculated "upward flux of paM" is, therefore, not the mistake derived from the
model situations but an actual result associated with other phenomena like turbulent
mixing of particles in the CEE and/or diffusive flux of dissolved DON or DIN. As the
occurrence of "upward flux" (day 10-13,13-16,20-23) are, in fact, always associated
280
Journal of Marine Research
_80
poe ( a )
Production rate of
•...•
[50,2
Q-4m
~ 60
-"-
040
<.;l
"'20
:0.
0
_ 80
Decomposition
i;'
rate of
poe ( b )
Q-4m
~ 60
-"-
040
<.;l
"'20
:0.
0
_ 80
.••..
Sinking rate of POC ( c )
::::
60
0-4m
::::
....
040
'"20
:0.
0
o
2
4
6
8
m ~ M
W
m 20 ~ ~
Days
20
Production rate of DOC ( d )
Q-4m
,,
,
,,
,
20
Decomposition
1;' 15
:-- .._-~
"
"-
;::: 10
0-4m
:,
o
<.;l
rate of DOC ( e )
,,
,
5
'"
____ J
';0
20
Upward flux of DOC ( f )
,..----.,
,
,
,
,,
''
,
J
,
,,
,..
':
I
•
o
2
4
6
8
0-4m
,'
I
m ~
M
W
m
20 ~
~
Days
Figure 5. Reconstructed variations in carbon and nitrogen cycling rates (a-n) and concentration of dissolved inorganic carbon (0) at 0-4 m depth interval of the CEE, calculated using
the data in Figure 1 and the equations in Table 1 (See text).
1992]
Nakatsuka
et al.: Nutrient control/ed phytoplankton
Production mte of PON ( g )
-12
B
0-4m
::: 8
"z
..
m
4
III
:s.
o
Decomposition
mte of PON ( h )
0-4m
o
Sinking mte of PON ( i )
0-4m
024
6
8
ro n w w re w
n ~
Days
-1.2
Production mte of DON ( j )
B
0-4m
r-- ~-~
::: 0.8
"-
,:,
~ 0.4
"
III
:l.
0
i1.2
Decomposition
rate of DON ( k)
"'0
::: 0.8
"'
~0.4
m
III
:l.
0
-1.2
Upward flux of DON ( I )
>-
-l!l
",0.8
0-4m
:----1
"z 0.4
.."
,,
,
,,
,
:0
••
024
6
8
t
.1
ro n
M
re re w n ~
Days
Figure 5. (Continued)
,
bloom
281
282
Journal of Marine Research
-..,
Uptake of N03-1
~8
+
1m)
0-4m
"-::::6
z
1;1
N02-)
[50,2
4
'"
';2
o
o
-
18
~1.75
...
to
~t7
~
-o. __
~
165
16
o
2
4
6
8
W U M ~
Days
-~
m w ~ ~
Figure 5. (Continued)
with the relatively large increases
in DOC pool (Fig. Ig), the upward fluxes are here
assumed to be attributed to diffusive fluxes of DaM. The second discrepancy of
"minus decomposition" rate is apparently caused by the extraordinary high DOC
value on day 23 (Fig. Ig). Although some analytical error in the DOC measurement
may affect this value, it is assumed that during the last sampling interval, the
production rate of DOC was much higher than 20% of that of pac to compensate
the "minus decomposition."
The corrected version of the simulation is also presented in Figure 5 (see dashed
lines). Because there was no information on the sinking rate of paM and/or the
decomposition rate of DaM during the corrected periods mentioned above, these
rates are put at zero during those periods. Although this is not realistic, this
reformation can be thought to be trivial compared to the whole of the carbon and
nitrogen cycle in the eEE.
b. Isotope mass balance model on CEE carbon budget. On the basis of the reconstructed temporal variations in the carbon budget at 0-4 m depth interval (Fig. Sa, b,
c, d, e, f and 0), we simulate the variations of l)I3C values of the three components of
carbon (paC, DOC, DIe) at the concerned layer, using several assumptions on the
isotope fractionations during transformation processes of carbon.
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
bloom
283
Table 2. Equations of the isotope mass balance model of carbon.
= (&13C(dic)+ .6.&13C)· P(poc)
- &13C(pOC)
. (S(poc) + D(poc»
d(&13C(doc) . [DOC))/dt
= (&13C(dic) + .6,&13C)· P(doc) - &13C(doc)
. D(doc)( +&13C(UPdoc) . UP(doc»*
d(&13C(dic)'
[DIC))/dt
= -(&I3C(dic) + .6,&13C)·(P(poc) + P(doc»
+ &13C(pOC)'D(poc) + &13C(doc)· D(doc)
d(&13C(pOC)'
[POC))/dt
(2-1 )
(2-2)
(2-3)
The terms are defined as follows: &13C(poc), &13C(doc), &13C(dic) carbon isotope ratios of
suspended particulate organic matter, dissolved organic matter and dissolved inorganic
carbon, defined as %0 (PDB); .6.&/3C isotope discrimination during photosynthesis production
of organic carbon by phytoplankton; UP(doc), &13C(UPdoc) flux (IJ.mole/l/day) and its isotope
ratio of upward diffusive transport of dissolved organic carbon; The rests are the same as those
in Table 1.
*The occurrence of upward diffusive transport of DOC is limited to the three periods (day
10-13,13-16 and 20-23) on the basis of the result of mass balance calculation (see text). Their
fluxes are not so large compared to the whole of the carbon cycle in CEE (Fig. 5).
The mathematical notations of the carbon isotope mass balance model are
presented in Table 2. These equations are, in fact, approximate ones because l)13e is
not the 13C/(I2C + \3C) itself. Within the variations of about 20%0 in l)\3C, however,
they can simulate the actual isotope mass balance almost completely. The l)13Cof
newly produced POC is represented as the sum of the l)13Cof ole (total inorganic
carbon) and 6.l)13C,the isotope discrimination during the photosynthetic production
of POC from DIC (Eq. 2-1). Because there is no information on the l)13e of newly
produced DOC, it is made equal to that of POC (Eq. 2-2) under the assumption that
DOC is always produced by phytoplankton associated with the photosynthesis. The
s\3e values of the sinking poe, decomposed poe and the decomposed DOe are
assumed to be equal to the s13e of bulk POC and DOC, respectively, at that time
(Eqs. 2-1 and 2-2). The 813Cvalues in the "upward"
DOC fluxes during days 10-13,
13-16 and 20-23 (Fig. 5f) are also assumed to be equal to that of DOC standing
stock. The last assumption has no basis, yet does not seem to affect the simulated
result of l)13Cof POC significantly, because it can hardly be assumed that the l)13C
value of DOC changes greatly in vertical profile, and their fluxes are not so large as
other fluxes of carbon (Fig. 5). The initial l)13Cvalues of DOC and DIC on day 1 are
arranged to the values, equal to that of POC, and 0-2%0 (Kroopnick, 1985),
respectively.
While it is believed that the isotope discrimination 6.l)13C is highly variable
reflecting the change in the environmental conditions and/or the phytoplankton
species, we firstly simulate the variation in the l)13e of poe and DIC under the
condition that 6.l)\3C in Eqs. 2-1 to 2-3 is constant (Fig. 6). Because the isotope
discrimination during the photosynthesis by phytoplankton is usually about - 20%0
and the DIC concentration decreased more than 10% during this experiment, the
l)13Cof DIC increased about 2%0by the end of experiment (Fig. 6b). This increase in
284
Journal of Marine Research
[50,2
-15
( a )
o'3e
of
poe
tnIUal6'3C
0-4m
of 2: CO2:
1,
/H''c, -21
~-17
o
~-19
o
o
0
-21
..~
;:
2
~ 0
-2
cl'3e
(c)
01 DOC 0-4m
;; -17
~
-21
o
2
4
6
8
10
W
~
~
~
~
n
M
Days
-15
cl'3e 01 poe 0-4m
( d)
tJ.
613c, -20
/~-~--------------~~-o,,,J'
~-19
9/
0
~.~_••• - .-- --- - •••
//,,/
..g- - -. -
-0·-~-"--.. _:~~--_.
Initial 8'3C of
~__ .-_.'
t CO2;
0
., °/.0
0
-21
;-17
~
~ -19
-21
(I)
..
o'3e
;;; -17
01 poe 0-4m
o
a
[Q'_I~~~~_~:~~_?!
~ c?.~:_~:
tJ. 6'3C, -22
/
0---
~ -19
~i-~~~-""-~-=-~
~o,
.20
o
.21
o
2
4
6
8
ID ~
Days
M
ffi
0
m w ~ ~
Figure 6. Simulated results of temporal variations in l)13Cof POC, IC02 and DOC at 0-4 m
depth interval, using equations in Table 2 and results shown in Figure 5, with assumption
that magnitude of isotope fractionation during uptake of dissolved inorganic carbon by
phytoplankton (~l)\3C) is constant. Figures a, b, c show results of POC, IC02 and DOC,
calculated under same conditions for initial l)\3Cvalue of IC02 (1%0) and MI3C (-21%0).
Figures d, e and f indicate results of POC, calculated under different conditions for ~l)\3C
( - 20, - 21 and - 22), initial l)13Cof ICOz (2, 1 and 0) and both of them, respectively. (Open
circles show the actual change in l)\3Cof suspended POC at 0-4 m.)
~13eof DIe may possibly explain more than the half of the actual increase in the
~13eof poe at 0-4 m until day 6 (Fig. 3a). The most unique characteristic of the
temporal variation of the ~13e
of poe in this experiment is, however, the fact that the
~13eof poe decreased together with the decrease of poe concentration to reach
the
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
bloom
285
the 813C value equal or lower than that of initial POC (Fig. 3a). This change in the
813C value of POC in the late part of phytoplankton bloom in this experiment cannot
be explained by the constant il813C model.
Possible reasons for the great decrease in 813C of POC in the late part of the
phytoplankton bloom are as follows.
(1) Decrease in 813C of DIC
(2) Change in CO2[ aq] concentration
(3) Change in the kind of DIC substrate (C02 or HC03-) used by phytoplankton
for photosynthesis
(4) Change in the pathways of enzymatic reaction to fix the DIC by phytoplankton
(5) Change in the predominant species of phytoplankton
(6) Change in the photosynthetic production rate ofPOC per individual phytoplankton cell
The first hypothesis, however, does not seem applicable because the rapid
penetration of CO2 from the atmosphere into the water could not occur in this CEE
as discussed above, and the 813C of DIC proved to be almost constant after day 6 by
the mass and isotope balance models, which do not consider air-water CO2 exchange
(Fig. 6b). Even if the air-water CO2 exchange occurs to a certain degree, it is difficult
for the 813C of DIC to return to the initial value by the end of this experiment as
observed in the 813C of POC (Fig. 3a).
The second is attributed to the hypothesis that the isotope fractionation during
photosynthesis by phytoplankton is mainly controlled by the ambient CO2 concentration, proposed by Rau et al. (1989, 1991). Also, in the CEE of this study, the
concentration of CO2[ aq] must have changed greatly by the uptake of CO2 by
phytoplankton. This hypothesis, however, cannot explain the decrease in 313C of
poe in the late part of this experiment for the same reason as the first one
mentioned above, because the concentration of CO2 must decrease together with the
phytoplankton bloom and cannot increase much until the end of the experiment.
Because 813C of HC03 - is commonly 7%0more or less higher than that of CO2[ aq]
(Deuser and Degens, 1967), the change in the photosynthetic substrate between
these two may possibly alter the 813C of POe. This third mechanism, however, seems
not to have produced the observed change in the 813C of POC because the
concentration of CO2, which has a lower 813C value and may produce the lower 313C
of POC being assimilated by phytoplankton, must have not increased greatly even in
the late part of experiment.
Descolas-Gros and Fontugne (1985) observed that the higher the contribution of
PEP (phosphoenolpyruvate) carboxylase (the enzyme of C4 pathway) to the whole
photosynthetic CO2 fixation by phytoplankton, the higher the 813C of produced POC,
in both the incubation experiment and the open ocean observation. This fourth
possibility also does not seem to correspond to the situation of this study because the
marine diatoms, which were predominant at the time when the 813C of poe reached
286
[50,2
Journal of Marine Research
the maximum (Fig. Id and 3a), do not have PEP carboxylase activity (Descolas-Gros
and Fontugne, 1985) which make the 013C of produced pac higher.
The temporal changing pattern of phytoplankton species composition (the proportion of centric diatoms in Fig. Id) appeared to correspond well with that of the Ol3C
of pac (Fig. 3a). In fact, Gearing et al. (1984) reported that the ol3e of large
phytoplankton
like diatom is about 2%0 higher than that of nanoplankton
in
Narragansett
Bay. If this correlation is applicable in this study too, the decrease in
the Ol3C of poe in the late part of this experiment may be explained by the change in
the phytoplankton species as suggested by many previous studies (Wong and Sackett,
1978; Fontugne and Dup1essy, 1978). Upon closer examination, however, the change
in the composition of phytoplankton
species proves not to correspond with that in
l3
the
e of poe at all. For example, on day 16, the sub-maximum of l3e of poe can
be observed (Fig. 3a), associated with that of poe concentration (Fig. Ie), although
the ratio of the number of centric diatoms to that of nanoflagellates became very low
at that time, due to the second bloom by nanoflagellate (Fig. Id). This fact indicates
that the Ol3C of pac is not simply governed by the composition of phytoplankton
o
o
specIes.
The photosynthetic production rates of pac in three depth intervals (Fig. Ih) also
appeared to correspond well with the variations in Ol3C of pac (Fig. 3a), expect that
the decreases in the production rates after their maxima are much faster than those
in ol3e. This sixth hypothesis can, moreover, explain the existence of the submaximum in the Ol3C of poe on day 16, because it can be attributed to the peak of
the production rate, irrespective of the phytoplankton species change (Fig. Ih).
The positive relationship between the o13e of poe and the production rate of
pac has been interpreted by the effect of limitation of DIC supply for photosynthe-
et al., 1988) and the supply of DIC is usually thought
through cell membrane (O'Leary, 1981;
Farquhar et al., 1989; Popp et al., 1989). However, if the carbon for photosynthesis of
sis in phytoplankton
(Cifuentes
to be loaded by the diffusion
phytoplankton
of CO2[aq]
is always supplied
through
the diffusion
of CO2,
the change
in
CO2 [ aq] concentration
must affect critically the Ol3C of POC as well. Such relation-
ship between
and Sl3C of POC was not suggested
CO2[aq]
mentioned above. As an alternative
that the active transport of HC03
in this experiment
as
for the diffusion of CO2, Berry (1989) suggested
- is the main pathway bringing carbon into the
phytoplankton
cell for photosynthesis,
due to the fact that many phytoplankton
species have the ability to uptake HC03- (e.g., Lucas and Berry, 1985) and the
CO2[ aq] concentration
in the cell sometimes
exceeds
that outside
the cell. He
showed a simple model of CO2 concentrating system by phytoplankton
(Fig. 7) and
proposed the following equation on the isotope discrimination during photosynthesis, assuming that the fractionation
accompanied
with the active transport
of HC03-
can be neglected.
(2)
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
CO 2
287
bloom
HCO 3 - (In)
(In)~
CO 2 (out)~HCO
3 -(out)
Figure 7. Schematic illustration of phytoplankton HC03- uptake model, presented by Berry
(1989). FI>F2 and F3 mean the rates of active transport of HC03- into phytoplankton cell,
RuBP carboxylation and diffusive leak of CO2 out of cell.
where dS13C* means the isotope fractionation from air CO2 to phytoplankton, and a,
band c are the discrimination factors associated with the processes of dissolution of
CO2, equilibrium conversion of gaseous CO2 to HC03 -, and RuBP carboxylation,
respectively. FI and F2 mean the fluxes of active transport of HC03 - into the cell and
the rate of RuBP carboxylation in the cell, respectively (see Fig. 7).
F2, in Eq. 2, appears proportional to the production rate of pac per unit cell (also
to the specific production rate of paC). Takahashi et al. (1991) thought that F2 is
proportional to the specific growth rate of phytoplankton
and FI is nearly constant
because of the large and relatively consistent amount of HC03 - pool in the water,
and actually found the linear relationship between d/)13C* (from CO2 in feeding air
to paC) and ~ (specific growth rate) with Chlamydomonas Reinhardtii (fresh water
green algae) in the water equilibrated
with air CO2 by the continuous culture
experiment as follows.
dSI3C*
=
-25.3
+ 35.4,
J.L
(r = -0.92)
(3)
where -25.3 and 35.4 correspond to (a + b) + c and -c' (F2/FI)/~ (which become a
proportional
constant when Fl is constant and F2 is proportional
to ~) in Eq. 2,
respectively. If (a + b) in Eq. 2, the isotope fractionation
during equilibrium
conversion between air CO2 and HC03 - in water, is 7%0, Eq. (3) can be rewritten as
follows.
(4)
where dS13Ct is the isotope discrimination
from HC03
-
to
pac. While
Takahashi
et
288
JournaL of Marine Research
[50,2
-15
A :
.•f
-32.3 . B
:
35.4
::: -17
o
0
o
0
o
( a ) 6'3C of POC 0-4m
.•..
c
:.!
-24.
0
-25.
4
-2
( d ) 6'3C of DOC 0-4m
.•f
--11
:-: -19
-21
o
2
468
mUM
$
m
~
~
~
Days
Figure 8. Simulated results of temporal variations in l)J3Cof poe, ~C02 and DOC at 0-4 m
depth interval, using equations in Table 2 and results shown in Figure 5, with assumption
that MJ3C changes according to Eg. 5. Figure a shows result of poe, calculated with values
-32.3 and 35.4 for constants A and B in Eq. 5, which were deduced from Takahashi et at.
(1991). And Figure b, c and d indicate results of pac, ~C02 and DOC, calculated with
three combinations of values for A and B ("-23 and 2," "-24 and 3" and "-25 and 4"). In
each case, initiall)13C of ~C02 is set as 1%0' (Open circles show the actual change in l)J3Cof
suspended poe at 0-4 m.)
al. (1991) confessed that the coefficients in Eq. 3 does not seem applicable directly to
marine phytoplankton, this good correlation suggested that the 813C of POC may be
governed mainly by the specific growth rate of phytoplankton.
In order to confirm the relationship between J.L and 6.813C, following equation on
the 6.813C is arranged to set in the isotope mass balance model (Table 1) for the
simulation of 813C in POc.
6.813C
= A + B .X
(5)
where X is the specific production rate of POC (the production rate of POC per unit
POC concentration per day), which is nearly equivalent to the specific growth rate of
phytoplankton (J.L) in the particles consisting of phytoplankton mainly, and!A and B
are some constants as shown in Figure 8. In this equation, 6,813C is defined as the
magnitude of isotope discrimination from total DIC to newly produced POc.
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
bloom
289
t
However, this .6.&l3Cis approximately equivalent to .6.&l3C(isotope
discrimination
from HC03- to paC) because HC03- is highly dominant in DIC pool and its 8l3C
value is nearly equal to that of total DIC.
The results of simulation using Eq. 5, which bring the growth rate control onto the
.6.8l3C,are presented in Figure 8. When Eq. 4 is used directly instead of Eq. 5, the
simulated result of 8l3C of pac increases extraordinarily together with the development of phytoplankton bloom (Fig. 8a). Even if it is considered that the uncertainty
for the calculation of the specific production rate of pac per day using that per hour
(Fig. Ii), this results of the simulated 8l3C of pac seems too high. As suggested by
Takahashi et at. (1991), the quantitative correlation between J.L and .6.813C,observed
in fresh water algae, cannot be applied directly to the marine phytoplankton.
The observed variation in the &l3Cof pac, however, can be well reconstructed
when suitable values are arranged for A and B in Eq. 5 (Fig. 8b), except for the
following discrepancies. In the initial stage of the experiment, the simulated 8l3C of
pac does not increase as fast as the observed one. On the other hand, the simulated
813Cof pac decreases much faster than the observed one just after day 6. The first
discrepancy, however, can be explained by the fact that the production rate of pac
measured on day 2 was extremely low by chance due to the very cloudy condition at
that time (Hama et al., 1988). The second discrepancy can also be explained by the
possibility that, just after the maximum stage of phytoplankton bloom on day 6, the
specific production rate of pac was underestimated in relation to the specific growth
rate of phytoplankton, due to much detritus in the suspended particles at the time.
The coefficients A and B in Eq. 5 are arbitrarily defined in this case (Fig. 8).
Nevertheless, the well reconstructed variation in the 8l3C of pac using Eq. 5 (Fig.
8b) clearly suggests the impact of specific growth rate of phytoplankton on the &\3Cof
pac. Further studies will be required for the reliable determination of the factors in
Eq. 5, using marine phytoplankton.
c. Isotope mass balance model on eEE nitrogen budget. The variations of 815Nvalues
in paN, DON, N03 - and NH4 + at 0--4 m depth interval are also simulated based on
the results of the reconstructed temporal variations in the nitrogen budget at 0--4 m
depth interval (Fig. 5g, h, i, j, k, 1, m and n). The mathematical notations of the
nitrogen isotope mass balance model are presented in Table 3.
The 815Nof newly produced paN is affected by proportions of N03 - and N~ + for
the substrates utilized by phytoplankton (Eq. 3-3). While it is assumed that there is
no isotope fractionation at NH4 + assimilation, a large isotope fractionation (.6.&15N)
is set to occur during uptake of N03 - by phytoplankton, which is the process
involving the largest isotope fractionation of nitrogen in this model (Eq. 3-3). Just as
in the case of carbon isotope mass balance model, the 815Nof newly produced DON
is assumed to be equal to that of PON (Eq. 3-2), and it is assumed that no isotope
fractionations occur during the sinking and decomposing processes of paN and
[50,2
Journal of Marine Research
290
Table 3. Equations of the isotope mass balance model of nitrogen.
d (&15N(pon). [PON))/dt
d(&15N(don) . [DON))/dt
&15N(p) . P(pon) - &15N(pon) . S(pon)
_(&15N(pon) - 3)' D(pon)
= S15N(p)' P(don) - S15N(don)' D(don)
=
(3-1)
(+&15N(UPdon) . UP(don»*
SI5N(p)
=
«S15N(N03) + M15N) . U(N03)
»
(3-2)
+ &15N(N",,)
. U(NH4»/(U(N03)
+ U(NH4
d(&15N(N03)'
[N03))/dt = _(SI5N(N03) + MI5N)· U(N03)
d(S15N(NH4)'
[NH4))/dt = -S15N(NH4) . U(NH4) + (&15N(pon)- 3)' D(pon)
(3-3)
(3-4)
+ &15N(don)' D(don)
(3-5)
The terms are defined as follows: &J5N(pon), &J5N(don), S15N (N03), S15N(NH4) nitrogen
isotope ratios of suspended particulate organic matter, dissolved organic matter, nitrate and
ammonium, defined as %0 (AIR); &15N(P) nitrogen isotope ratio of newly produced organic
matter by phytoplankton; t:.&15Nisotope discrimination during uptake of nitrate by phytoplankton; UP(don), &lW(UPdon) flux (IJomole/l/day) and its isotope ratio of upward diffusive
transport of dissolved organic nitrogen; The rests are the same as those in Table 1.
*The occurrence of upward diffusive transport of DON is limited to three sampling intervals
(See in Table 2).
DON, except the decomposition of paN in which NH4 +, 3%0 lower than paN
(Check1ey and Miller, 1989), is regenerated (Eqs. 3-1 and 3-2). In addition, the l)J5N
values in the "upward" DON fluxes are made equal to that of DON standing stock.
The initial l)15Nvalues of DON and NH4 + on day 1 are arranged to 3%0lower than
that of paN on the assumption
that the isotopic equilibrium
between
paN, DON
and NH4 + had been established there before day 1. Because there is no information
on the initial l)15Nvalues of N03- which was supplied artificially into the CEE,
severall)15N values of initial l)15Nof N03 - were tried for the simulations.
The results of simulations using the isotope mass balance model in Table 3 are
shown in Figure 9. The actual pattern of temporal variation in the l)15Nof PQN at 0-4
m can be very well simulated by this nitrogen isotope model (Fig. 9a). The l)J5Nvalue
of N03 - increases rapidly together with the development of bloom (Fig. 9b), which
directly results in the change of l)15Nof phytoplankton. This clearly demonstrates
that, during phytoplankton bloom, the temporal variation in l)15N of paN is
principally governed by the isotope fractionation process during uptake of N03 - by
phytoplankton. This result is consistent with previous observations in the natural
waters (Saino and Hattori, 1985; Goering et al., 1990; Altabet et ai., 1991). On the
other hand, the l)15Nof NH4+ follows the change of l)15Nin paN (Fig. 9c) because
the standing stock of NH4 + in surface water is always sustained by the regenerated
NH4+ from paN and/or DON. The NH4+ regeneration process (the other isotope
fractionation process in this model) does not appear to affect the l)J5N of paN
greatly, because the almost same variation in l)15Nof paN can be reconstructed
without the isotopic discrimination during regeneration of NH4 + (Fig. 9h).
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
bloom
291
10
o
o
o
InJtlal615N
of N03-:
3.
6 6,5N (NOJ!,-6
=6
-2
-------------------o
2
.66
m
U
~
$
m ~
n
M
Days
Figure 9. Simulated results oftemporal variations in !)15N of PON, N03-, NlL+ and DON at
0-4 m depth interval, using equations in Table 3 and results shown in Figure 5. Figures a, b,
c and d show the results of PON, N03-, NH4+ and DON, respectively, which were
calculated under same conditions for initial !)I5N of N03 - (3%0) and magnitude of isotope
fractionation during uptake of N03- by phytoplankton, ~!)15NN03_ (-6%0)' Figures e, f and
g indicate results of PON, calculated under different conditions for ~!)I5NN03 (-8, -6 and
-4), initial !)15N of N03- (5,3 and 1) and both of them, respectively. In Figure h, a result is
shown as dashed line, derived from the assumption that no isotope fractionation occur
during regeneration of NH.+ from paN. (Open circles show the actual change in 815N of
suspended PON at Q-4 m.)
From the quantitative viewpoint, the actual variation at 0-4 m depth seems to fit
best the result from combination of 3, as the initial S15N value of NO) -, and - 6, as
the size of isotope fractionation
during uptake of NO) -. The value of -6 as the
isotope fractionation during uptake of NO) - is very similar to those in the previous
reports (Horrigan et aI., 1990; Wada, 1980). On the other hand, the value of 3 as the
initial SI5N of N03- seems to be somewhat low because the weighted mean S15N
value of suspended and sedimented PON collected by day 23 is about 6%0' This
apparent discrepancy may be caused by some other isotope fractionation processes
than NO) - uptake, which are not involved in this model, like DON production and
decomposition.
In this model, the variability
of the magnitude
ation during uptake of NO) - (Wada and Hattori,
of isotope fraction-
1978) is also neglected. However,
Journal of Marine Research
292
[50,2
10
f~---~----------_·tr-~~~_~~~~~~~:~_~:·i··
/ --~_£._-----.Q.
\
//
----_/
/ lJ. 615N(No.l ,-6
615N of
\_//
-2
o
2
,..-'"
----- ,
466
ro u
~
$
PON 0-4m ( I)
m w n
~
Days
Figure 9. (Continued)
we cannot discuss these possibilities more in detail because the data for &15N values
of DON and N03 - are not presently available now.
5. Conclusion
The phytoplankton bloom in the controlled ecosystem enclosure (CEE) revealed
the predominant temporal variations in the 813C and 815N value of suspended paM.
Using the detailed mass balance model, the carbon and nitrogen dynamics in the
CEE were reconstructed, and on the basis of the results, the isotope mass balance
models of carbon and nitrogen were produced and used to understand the temporal
changing pattern of 813C and 815N of POM, not only qualitatively but also quantitatively.
As the result, it was suggested that the main factor which control the variation of
13
8 C of POM during phytoplankton bloom is not the change in carbon dioxide system
nor phytoplankton species change in the water but the change of phytoplankton
growth rate. This may have great implications for the 13C biogeochemistry of organic
matter in the ocean, especially for the paleoceanographic study of 813C in organic
matter. Because most sedimentary organic matter in high productive ocean derives
from the settling particles produced during phytoplankton bloom (Wefer, 1989), this
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
bloom
293
9
lDl
8
20
7
_ 6
a:
;: 5
l-4
...•
3
(I)
2
POM
0-4m
-20
-19
6'3C
-18
( POe)
Figure 10. Distribution pattern of 813C and 815N of suspended particulate organic matter at
0-4 m depth interval. Small numbers in figure indicate the day of sampling. The Roman
numerals indicate "phases" during development of phytoplankton bloom (see text).
correlation between Ol3Cof paM and phytoplankton growth rate during bloom must
provide a firm constraint for the interpretation of sedimentary 013Crecord.
On the contrary, the temporal variation in l)15Nof paM during this phytoplankton
bloom was consistent with the previous reports on the temporal variation in 015Nof
paM in natural waters, both from the qualitative and quantitative viewpoints. It
could be explained principally by only one isotope discrimination process, the N03uptake by phytoplankton.
In this study, however, the Ol3Cand l)15Nvalues of dissolved carbon and nitrogen
were not measured, and the carbon dioxide system in the water was not observed
either. If we use these data, several assumptions in the mass and isotope mass
balance calculations in Table 1, 2, 3 can be cancelled, and more distinct constraints
must be provided for the magnitudes of isotope fractionations at some other carbon
and nitrogen transformation processes than uptake of DIC and N03 - by phytoplankton.
The temporal observation of 013C and 015N of paM during the course of this
nutrient controlled phytoplankton bloom (Fig. 3) led us to classify the paM into
three phases of the bloom from an isotopic viewpoint (Fig. 10): (I) The early stage of
the phytoplankton bloom when N03 - are still in excess in water (high 013Cand low
l)15N);(II) the late stage of the bloom when N03 - has just been depleted (high 013C
and high l)15N);and (III) the quasi-steady state phase, a few days after the depletion
of N03 - (low l)13Cand high l)15N).In phase I, rapid uptake of newly supplied N03by phytoplankton (low l)15N)supports high production rate of pac (high l)l3e). In
phase II, although the N03 - has been depleted (high l)15N),the production rate of
294
Journal of Marine Research
[50,2
POC has not decreased (high /)13C) and organic compounds of high C/N ratios are
being produced (Goldman, 1980; Hama et at., 1988). In phase III, quasi-steady state
balance between NH4 + and PON has been established (high /)15N) and the production rate of POC has become lower (low /)13C).
In addition, it is notable that the changes in /)13Cand /)15Nof the sedimented POM
followed those in suspended POM consistently. This means that the characterization
of POM by /)I3C and /)15Nvalues, in the viewpoint of development of phytoplankton
bloom, is possible not only in suspended POM but also in sinking POM and sediment
(AJtabet and Deuser, 1985; Handa et al., 1992).
Acknowledgments. We thank T. Hama, M. Takahashi, F. Whitney, E. Matsumoto and Y.
Hirano for their help in the field study and valuable discussions. Thanks are also due to N.
Nakai and S. Yoshioka for their technical advise in isotopic analysis. This work was supported
by a grant of the Japan Society for Promotion of Science to the Japan-Canada cooperative
study entitled "Experimental studies on dynamics of chemical substances in marine environment by using MESOCOSM."
REFERENCES
Altat-et, M. A. 1988. Variations in nitrogen isotopic composItIon between sinking and
suspended particles: implications for nitrogen cycling and particle transformation in the
open ocean. Deep-Sea Res., 35, 535-554.
Altabet, M. A. and W. G. Deuser. 1985. Seasonal variation in natural abundance of 15Nin
particles sinking to deep Sargasso Sea. Nature, 315, 218-219.
Altabet, M. A., W. G. Deuser, S. Honjo and C. Stienen. 1991. Seasonal and depth-related
changes in the source of sinking particles in the North Atlantic. Nature, 354, 136-139.
Berry, J. A. 1989. Studies of mechanisms affecting the fractionation of carbon isotopes in
photosynthesis, in Stable Isotopes in Ecological Research, P. W. Rundel, 1. R. Ehleringer
and K. A. Nagy, cds., Springer-Verlag, New York, 82-94.
Checkley, D. M., Jr. and C. A. Miller. 1989. Nitrogen isotope fractionation by oceanic
zooplankton. Deep-Sea Res., 36, 1449-1456.
Cifuentes, L. A., J. H. Sharp and M. L. Fogel. 1988. Stable carbon and nitrogen isotope
biogeochemistry in the Delaware estuary. Limno!. Oceanogr., 33, 1102-1115.
Degens, E. T., R. R. L. Guillard, W. M. Sackett and 1. A. Hellebust. 1968. Metabolic
fractionation of carbon isotopes in marine plankton-I Temperature and respiration experiments. Deep-Sea Res., 15, 1-9.
Descolas-Gros, C. and M. R. Fontugne. 1985. Carbon fixation in marine phytoplankton:
carboxylase activities and stable carbon-isotope ratios; physiological and paleoclimatological aspects. Mar. Bio!., 87, 1-6.
Deuser, W. G. and E. T. Degens. 1967. Carbon isotope fractionation in the system CO2(gas)CO2(aqueous)-HC03(aqueous). Nature, 215, 1033-1035.
Etcheto, J. and L. Merlivat. 1988. Satellite determination of the carbon dioxide exchange
coefficient at the ocean-atmosphere interface: A first step. J. Geophys. Res., 93, 15,66915,678.
Farquhar, G. D., K. T. Hubick, A. G. Condon and R. A. Richards. 1989. Carbon isotope
fractionation and plant water-use efficiency, in Stable Isotopes in Ecological Research, P.
W. Rundel, J. R. Ehleringer and K. A. Nagy, eds., Springer-Verlag, New York, 21-40.
1992]
Nakatsuka
et al.: Nutrient controlled phytoplankton
bloom
295
Fontugne, M. R. and J. C. Duplessy. 1978. Carbon isotope ratio of marine plankton related to
surface water masses. Earth. Planet. Sci. Let., 41, 365-671.
--1981.
Organic carbon isotopic fractionation by marine plankton in the temperature range
-1 to 31°C. Oceanol. Acta, 4, 85-90.
Gearing, J. N., P. J. Gearing, D. T. Rudnick, A. G. Requejo and M. J. Hutchins. 1984. Isotopic
variability of organic carbon in a phytoplankton-based, temperature estuary. Geochim.
Cosmochim. Acta, 48, 1089-1098.
Goering, J., V. Alexander and N. Haubenstock. 1990. Seasonal variability of stable carbon and
nitrogen isotope ratios of organisms in a North Pacific bay. Est. Coast. Shelf Sci., 30,
239-260.
Goldman, J. C. 1980. Physiological processes, nutrient availability, and the concept of relative
growth rate in marine phytoplankton ecology, in Primary Productivity in the Sea, P. G.
Falkowski ed., Plenum, New York, 179-194.
Hama, T., N. Handa, M. Takahashi, F. Whitney and C. S. Wong. 1988. Change in distribution
patterns of photosynthetically incorporated C during phytoplankton bloom in controlled
experimental ecosystem. 1. Exp. Mar. BioI. Ecol., 120, 39-56.
Handa, N., T. Nakatsuka, M. Fukuchi, H. Hattori and T. Hoshiai. 1992. Vertical fluxes and
ecological significance of organic materials during the phytoplankton bloom during austral
summer in Breid Bay, Antarctica. Mar. BioI., 112, 467-478.
Horrigan, S. G., J. P. Montoya, J. L. Nevins and J. J. McCarthy. 1990. Natural isotopic
composition of dissolved inorganic nitrogen in the Chesapeake Bay. Est. Coast. Shelf Sci.,
30, 393-410.
Ignatiades, L. 1973. Studies on the factors affecting the release of organic matter by
Skeletonema costa tum (greville) cleve in field conditions. J. Mar. BioI. Assoc. U.K., 53,
923-935.
Kroopnick, P. M. 1985. The distribution of l3C of IC02 in the world oceans. Deep-Sea Res.,
32,57-84.
Lucas, W. J. and J. A. Berry. 1985. Inorganic carbon transport in aquatic photosynthetic
organisms. Physiol. Plant., 65, 539-543.
Menzel, D. W. and J. Case. 1977. Concept and design: controlled ecosystem pollution
experiment. Bull. Mar. Sci., 27, 166-184.
Menzel, D. W. and R. F. Vaccaro. 1964. The measurement
of dissolved organic and
particulate carbon in seawater. Limnol. Oceanogr., 9, 138-142.
Minagawa, M. and E. Wada. 1986. Nitrogen isotope ratios of red tide organisms in the East
China Sea: a characterization of biological nitrogen fixation. Mar. Chern., 19, 245-259.
Minagawa, M., D. A. Winter and 1. R. Kaplan. 1984. Comparison of Kjeldahl and combustion
methods for measurement of nitrogen isotope ratios in organic matter. Anal. Chern., 56,
1859-1861.
O'Leary, M. H. 1981. Carbon isotope fractionation in plants. Phytochem., 20, 553-567.
Parsons, T. R., Y. Maita and C. M. Lalli. 1984a. A Manual of Chemical and Biological
Methods for Seawater Analysis. Pargamon Press, Oxford, 173 pp.
Parsons, T. R., M. Takahashi and B. Hargrave. 1984b. Biological Oceanographic Processes.
3rd ed. Pergamon Press, Oxford, 330 pp.
Popp, B. N., R. Takigiku, J. M. Hayes, J. W. Louda and E. W. Baker. 1989. The post-paleozoic
chronology and mechanism of l3Cdepletion in primary marine organic matter. Amer. J. Sci.,
289, 436-454.
Rau, G. H., T. Takahashi and D. 1. Des Marais. 1989. Latitudinal variations in plankton Sl3C:
implications for CO2 and productivity in past oceans. Nature, 341, 516-518.
296
Journal of Marine Research
[50,2
Rau, G. H., T. Takahashi, D. 1. Des Marais and C. W. Sullivan. 1991. Particulate organic
matter /)13Cvariations across the Drake Passage. J. Geophys. Res., 96, 15,131-15,135.
Sackett, W. M., W. R. Eckelmann, M. L. Bender and A. W. H. Be. 1965. Temperature
dependence of carbon isotope composition in marine plankton sediments. Science, 148,
235-237.
Saino, T. and A. Hattori. 1987. Geographical variation of the water column distribution of
suspended particulate nitrogen and its 15Nnatural abundance in the Pacific and its marginal
seas. Deep-Sea Res., 34, 807-827.
--1985.
Variation in 15Nnatural abundance of suspended organic matter in shallow oceanic
waters, in Marine and Estuary Geochemistry, A. C. Sigleo and A. Hattori, eds., Lewis
Publishers, Chelsea, 1-13.
Taguchi, S. 1976. Short-term variability of photosynthesis in natural marine phytoplankton
populations. Mar. Biol., 37, 197-207.
Takahashi, K., E. Wada and M. Sakamoto. 1991. Relationship between carbon isotope
discrimination and the specific growth rate of green algae Chlamydomonas reinhardtii. Jpn.
J. Limnol., 52, 105-112.
Wada, E. 1980. Nitrogen isotope fractionation and its significance in biogeochemical processes
occurring in marine environments, in Isotope Marine Chemistry, E. D. Goldberg, Y. Horibe
and K. Saruhashi, eds., Uchida Rokakuho, Tokyo, 375-398.
Wada, E. and A. Hattori. 1976. Natural abundance of 15Nin particulate organic matter in the
North Pacific Ocean. Geochim. Cosmochim. Acta, 40, 249-256.
--1978.
Nitrogen isotope effects in the assimilation of inorganic nitrogenous compounds by
marine diatoms. Geomicrobiol. J., 1,85-101.
Walsby, A. F. and C. S. Reynolds. 1980. Sinking and floating, in The Physiological Ecology of
Phytoplankton, I. Morris ed., Blackwell, Oxford, 371-412.
Wefer, G. 1989. Particle flux in the ocean: effects of episodic production, in Productivity of the
Ocean: Present and Past, W. H. Berger, V. S. Smetacek and G. Wefer, eds., John Wiley &
Sons, Chichester, 139-153.
Williams, P. J. 1975. Biological and chemical aspects of dissolved organic material in sea water,
in Chemical Oceanography v. 2., J. P. Riley and G. Skirrow, eds., Academic Press, London,
301-362.
Wong, C. S. and Y. H. Chan. 1991. Temporal variations in the partial pressure and flux of CO2
at ocean station P in the subarctic northeast Pacific Ocean. Tellus, 43B, 206-223.
Wong, W. W. and W. M. Sackett. 1978. Fractionation of stable carbon isotopes by marine
phytoplankton. Geochim. Cosmochim. Acta, 42, 1809-1815.
Received: 11 December, 1990; revised: 4 February, 1992.