Journal of Oceanography, Vol. 55, pp. 505 to 513. 1999
Environmental Factors Affecting the Occurrence and
Production of the Spring Phytoplankton Bloom in Funka
ISAO KUDO and KATSUHIKO MATSUNAGA
Department of Fisheries Oceanography and Marine Sciences, Faculty of Fisheries, Hokkaido University,
Hakodate 041-8611, Japan
(Received 15 June 1998; in revised form 3 February 1999; accepted 22 February 1999)
The concentration of nutrients was measured during the spring phytoplankton bloom in Keywords:
Funka Bay over a 5-year period (1988–92). During the winter mixing period, nutrient ⋅ Spring bloom,
concentrations were similar in every year except in 1990 when a high concentration of ⋅ Funka Bay,
silicate was observed. There was interannual variation in the onset of the bloom, ⋅ nutrient,
presumably depending on the stability of the water column. The bloom developed in early ⋅ primary production,
March when the Oyashio water (OW), which has a lower density than the existing winter
⋅ new production.
water, flowed into the bay and the pycnocline formed near the bottom of the euphotic
zone. In this case, high chl a was found only in the euphotic zone and nutrient utilization
was limited to this zone. In the year when the inflow of OW was not observed by April, the
bloom took place at the end of March without strong stratification and high chl a was found
in the whole water column, accompanied by a decrease in nutrients. Interannual differences
were found not only at the beginning of the decrease, but also in the thickness of the layer
which showed a decrease in nutrients. Primary production from the beginning to the end
of the spring bloom was estimated from the nutrient budget before and after the spring
bloom. The integrated production over the spring bloom period ranged from 25 to 73 g
C m–2, which accounts for 19–56% of the annual production in this bay. We found that
the timing of the bloom was strongly dependent on the inflow of OW, but the amount of
production was not clearly related to this timing.
1. Introduction spring. This renewal sometimes occurred at the time of
Explosive increase in biomass, which is close to their spring bloom formation, so it is interesting to know the
maximal growth, is observed in spring due to the relaxation influence of the hydrography on the primary production.
of a limiting factor for growth, called bloom. A pronounced Primary production was estimated from the nutrient budget
spring bloom is generally observed in coastal subarctic or and the elemental composition (C, N, P and Si) of the
temperate regions. During this period much of the annual phytoplankton. This study is of particular importance to
“new” production occurs because nitrate is the main N understand factors controlling a spring bloom formation and
source and is usually depleted after the bloom until the next the magnitude of primary production, and the energy flow
replenishment by vertical mixing in the fall. The response of from primary production to higher trophic levels in the food
the zooplankton to the rapidly increasing primary produc- web in the subarctic ecosystems.
tion in spring is generally slow, and the grazing impact on
the spring bloom is usually limited (Odate and Maita, 1988; 2. Material and Methods
Waite et al., 1992a, b). A large part of the accumulated Weekly to monthly sampling was carried out on the
phytoplankton biomass therefore sinks out of the euphotic R/V Ushio Maru during 1988–92 (Table 1). The sampling
zone in the late senescent stage of the bloom. Thus, the stations were Stn 17 (42°9.5′ N, 140°48.0′ E) during 1988–
spring bloom dynamics has an impact on the pelagic and 90 and Stn 30 (42°16.2′ N, 140°36.0′ E) during 1991 and
benthic ecology, fisheries and carbon fluxes in coastal areas. 1992 (Fig. 1). Salinity and temperature were monitored with
In this paper, we focused on the relationship between a Neil Brown CTD sensor. Water samples were taken using
hydrography and the development of the phytoplankton a Go-flo sampler attached to a nylon wire which was
bloom in Funka Bay. Funka Bay has a characteristic hy- connected to the end of the hydrowire to prevent samples
drography with subarctic Oyashio water renewal in early from becoming contaminated. Subsamples were transferred
Copyright The Oceanographic Society of Japan.
Table 1. List of sampling date and stations in 1988–1992.
Year Station Date
1988 Stn 17 Feb. 15, 22, 29, Mar. 7, 14, 28, Apr. 11, 18
1989 Stn 17 Feb. 2, Mar. 9, 24, 29
1990 Stn 17 Feb. 20, Mar. 5, 12, 22
1991 Stn 30 Feb. 5, 12, 28, Mar. 5, 12, 22, 27, Apr. 3, 15
1992 Stn 30 Mar. 2, 9, 16, 23, Apr. 20
Fig. 2. Temporal change of salinity along with chl a in 1988–92
Fig. 1. Map showing the sampling stations. Stn 17 was occupied at Stns 17 and 30. Solid lines depict the contour lines of
during 1988–90 and Stn 30 during 1991 and 1992. salinity and dashed lines depict that of chl a. Bold numbers
indicate salinity and italic numbers indicate the concentration
of chl a.
to conventional polyethylene bottles which were previously
soaked in 6N hydrochloric acid. Samples for nutrients were h and measured as reactive silicate by the method of DeMaster
filtered through a Millipore HA filter (0.45 µm) and kept (1981). To obtain dry weight phytoplankton samples on a
frozen until analysis. Nutrients were analyzed with a preweighed GF/F filter were dried at 60°C for 6 h and placed
Technicon Autoanalyzer II and chlorophyll-a was extracted in a desiccator until a constant weight was obtained.
in 90% acetone and measured with a spectrofluorometer
(Parsons et al., 1984). Phytoplankton samples were collected 3. Results
by vertically towing a plankton net (30 µm opening) be-
tween 0 and 30 m depth following passage through a 300 µm 3.1 Hydrography
net to remove large zooplankton. Elemental composition of Ohtani (1971) defined the salinity of typical Funka Bay
phytoplankton was measured with a Yanaco CHN analyzer Winter water (FW) as more than 33.6 and that of Oyashio
MT-2 for C and N and by the method of Menzel and Corwin current water (OW) as less than 33.3. Thus, a contour line of
(1965) for total phosphorus. The silicon content in the salinity 33.3 could be considered as a water mass under the
phytoplankton was digested with 1% Na2CO3 at 85°C for 3 influence of OW. The timing of OW inflow in 1989 was at
506 I. Kudo and K. Matsunaga
the end of February, which was the earliest during the 5-year (Fig. 3). In other years, the maximum chl a was 350–400 mg
observation period (Fig. 2). On the contrary, the contour line m–2. Chl a increased in early March of 1989 and 1990 and
of 33.3 was not observed until the middle of April in 1991. showed a maximum in the beginning of March, while the
This line was observed at the middle or end of March in 1988 maximum concentration in 1988 and 1991 was observed at
and 1992. A sign of OW was observed at the end of February the end of March to early April (Fig. 3).
in 1990, showing a salinity of 33.3, but a salinity of 33.0 was Judging from the temporal change in chl a over the 5-
not found at the end of March. In 1992, a contour line of 33.5 year period, the spring bloom was early in 1989 and 1990,
was observed in early March and just after a lower salinity and occurred at the end of February to the beginning of
of 33.0 appeared, indicating that the replacement of seawater March (referred to as Type I year hereafter). On the contrary,
occurred relatively quickly. Miyake et al. (1988) reported it the bloom started developing at the end of March in 1988
usually took more than a month for OW to replace the whole and 1991 (Type II year), which was a month later than in
area of Funka Bay. During our 5-year study, the physical 1989 and 1990. The occurrence of the bloom in 1992 was in
conditions of Funda Bay were different and thus the timing the middle of March, which was an intermediate year.
of the inflow of OW into the bay showed interannual
variation. 3.3 Nutrients
The nutrient distribution from winter to late spring
3.2 Bloom occurrence showed the same sequence during our 5-year study in this
The standing stock of chl a in the euphotic zone (0–40 bay, despite the difference in the maximum and minimum
m) showed interannual variation with the maximum values concentration (Fig. 3). Before the bloom the maximum
in 1988, 550 mg chl a m–2, followed 500 mg chl a m–2 in 1990 standing stocks of nitrate and silicate were around 500 and
Fig. 3. Temporal change in the standing stock (0–40 m) of nitrate plus nitrite, silicate, phosphate and chl a in 1988–92. Arrows indicate
the peak of the bloom.
5-Year Study of Bloom 507
1000 mmol m–2, respectively except in 1990 when the
maximum standing stock of silicate was 1700 mmol m–2 and
higher than the other years. Nitrate and silicate started
decreasing just before the peak of chl a was observed and
were depleted by the middle of April in every year. Phosphate
showed almost the same sequence as silicate and nitrate in
the bloom period, but a considerable amount of phosphate
remained after the bloom, despite the fact that both nitrate
and silicate were almost depleted. During the mixing period,
nitrate was uniformly distributed vertically with more than
10 µM (Fig. 4). After the beginning of spring bloom, nitrate
decreased in the water column, but the onset of the decrease
was obviously different in every year. The onset was at the
end of February in 1989 and 1990 (Type I year), but in the
middle of March in 1988 and 1991 (Type II year). Interannual
differences were found not only at the beginning of the
decrease, but also in the thickness of the layer which showed
a nutrient decrease. This nutrient decrease was only ob-
served in the euphotic zone in 1989, 1990 and 1992 due to
the presence of strong stratification below the euphotic
zone. Nutrients decreased in the whole water column in
1988 and 1991.
3.4 Elemental composition of phytoplankton
The elemental composition of actively growing phyto-
plankton was estimated from the phytoplankton samples
taken at the peak of the bloom (n = 9): carbon; 18.9 ± 3.85,
nitrogen; 2.8 ± 0.65, phosphorus 0.12 ± 0.04 and silicon;
8.3 ± 1.1 mmol g dry wt–1. The contribution of zooplankton
to this result seems minimal because zooplankton biomass
is relatively small compared to phytoplankton in spring
(Odate and Maita, 1988) and large copepods were removed
by sieving before analysis.
Before the bloom, the N:P atomic ratio for standing
stocks of nutrients in the euphotic zone was 10.1 which was
Fig. 4. Temporal change of nitrate plus nitrite along with chl a in lower than the Redfield ratio of N:P = 16 and decreased at
1988–92 at Stns 17 and 30. Solid lines depict the contour lines the end of the bloom due to the ratio of 23.6 in the plankton,
of nitrate plus nitrite and dashed lines depict that of chl a. Bold which was two times higher than the ratio in seawater (Table
numbers indicate the concentration of nitrate plus nitrite and 2). The Si:P atomic ratio for standing stocks of nutrients also
italic numbers indicate the concentration of chl a.
decreased during the phytoplankton bloom. The Si:N ratio
Table 2. Mean atomic ratios for standing stocks of nutrients during the pre-bloom and post-bloom period, calculated production and
plankton in Funka Bay. Mean values were obtained from 1988–1992 for nutrients and 1988–1991 for plankton. N content for
nutrients means sum of nitrate and nitrite.
N:P Si:P Si:N C:N
pre-bloom 10.1 ± 2.0 21.1 ± 3.0 2.1 ± 0.2
post-bloom 3.4 ± 0.6 11.1 ± 7.5 3.5 ± 1.5
production* 15.6 ± 3.6 30.5 ± 4.3 1.9 ± 0.4
plankton 23.6 ± 2.8 64.3 ± 15.7 2.7 ± 0.4 6.7 ± 0.5
*Calculated production from a nutrient budget (see text).
508 I. Kudo and K. Matsunaga
in seawater, however, showed no change throughout the nium had a minimal effect on this calculation because
bloom because the ratio in the plankton was almost identical ammonium was around 1 µM and showed little change dur-
to the ratio in seawater. This ratio during the post bloom was ing the period (data not shown). Thus, the apparent primary
slightly high, but this increase was not significant because at production was calculated from the difference in the nitrate
the end of the bloom both silicate and DIN were depleted in standing stock before and after the bloom (Fig. 6). This
the euphotic zone. The C:N ratio in plankton, 6.7 ± 0.5 was calculated production was 25–78 g C m–2. The calculation in
similar to the Redfield ratio. 1988 could not be made due to the lack of nutrient data in the
post-bloom period below 60 m. The highest value was 78 g
3.5 Maximum production from nutrient standing stocks C m–2 in 1991 and the lowest was 25 g C m–2 in 1990. A
The maximum primary production was calculated from comparison between this calculated production and the
the standing stock of individual nutrients in the water potential production in each year revealed that the former
column (0–80 m) in the pre-bloom and the elemental com- value accounted for 70–80% of the latter value, except in
position of phytoplankton, assuming that nutrients were the 1990 when it was only 40%.
sole limiting factor and this nutrient was used up during the
bloom. Based on the change in phosphate standing stock the
primary production during the spring bloom period was
150–175 g C m–2; this was several times higher than that
from other nutrients (Fig. 5). The primary production based
on nitrate and silicate ranged from 50 to 100 g C m–2. These
estimates should represent ideal production (i.e. potential
production) because the nutrient would not be utilized in the
whole water column due to limited irradiance (photic zone
was ca. 30–40 m depth).
3.6 Estimation of primary production based on the nutrient
Since silicate is incorporated only into the frustle of
diatom cells and cellular silicon content changes with the
Fig. 6. Comparison of potential production to calculated produc-
silicate concentration in the seawater while cellular carbon
tion in 1989–92. The potential production was calculated from
and nitrogen remains constant with various nitrate concen- the maximum standing stock of nitrate before the bloom and
tration (Kudo, in preparation), we thought the calculation the calculated production was obtained from the difference
based on nitrate content would best represent primary pro- between the maximum and minimum standing stock of nitrate
duction in Funka Bay. Nitrite was negligibly low and ammo- for each year. All production values were expressed as carbon.
Fig. 5. Interannual difference in primary productivity expressed Fig. 7. Change in solar radiation in Muroran in 1988–92. Data are
as carbon based on the standing stock of nitrate, phosphate and plotted as an average over 10 days for upper, middle and end
silicate in 0–85 m. of each month.
5-Year Study of Bloom 509
4. Discussion radiation was a reflection of the increase in the angle of solar
radiation and extending daytime. However, daily radiation
4.1 Triggering factors for the occurrence of the bloom in changed drastically due to cloud cover by a factor of ten
Funka Bay (Fig. 8). The peak of the bloom in each year observed after
During winter there is only a very low production of more than three days of continuous high radiation over 15
phytoplankton due to low light irradiance. Deeper mixed MJ m–2. This high radiation coincided with sunny condition.
layer also reduces the light availability for phytoplankton. In Type II years, even if these favorable conditions contin-
Solar radiation in Murorann, the closest observation station ued for more than three days in the beginning of March, the
to Funka Bay, was a minimum in December and January (5 peak of the bloom was not observed because a mixed layer
MJ m–2) and increased linearly after this period and reached was deeper than that in Type II years. More qualitative
a maximum in summer (Japan Meteorological Agency, discussion including the relationship between critical depth
1988, 1989, 1990, 1991, 1992) (Fig. 7). In the first 10 days and mixed layer depth in Funka Bay will be reported
of March, average solar radiation was about 10 MJ m–2 in a elsewhere.
5 year period. This value increased to about 15 MJ m–2 at the Nutrient concentration is generally high and above the
end of March. This linear increase in the average solar saturation for growth before the bloom because of replen-
Fig. 8. Change in daily solar radiation (line) in Muroran and in standing stock (0–40 m) of chl a (bar) at the station in 1988–92.
510 I. Kudo and K. Matsunaga
ishment of nutrients from deeper layers. Sea (Ohtani, 1985). Funka Bay Winter water (FW) is formed
The sigma-t distribution showed a mirror image of the in the previous fall by mixing of the water present in summer
chl a distribution at the peak of the bloom (Fig. 9). When a with intruding subtropical water (Tsugaru warm current).
strong picnocline presented around 50 m, chl a distributed The temperature of FW in the mixing period is about 4–5°C,
only above the picnocline (Type I: 1989 and 1992). In the while that of OW is less than 3°C. However, the density of
years with a weak picnocline, chl a tended to distribute FW is higher than OW because salinity of FW is more than
homogeneously (Type II: 1988 and 1991). These picnoclines 33.5 compared to that of OW, less than 33.0. Thus, coastal
coincided with haloclines in Funka Bay. In Type I years, OW flows into Funka Bay covering the FW. The nutrient
OW presented over FW, more saline but warmer than OW. concentrations between FW and OW before the bloom were
Although the density of water column is a function of almost same, so nutrient concentration is not crucial to the
salinity and temperature, the density is mainly affected by bloom formation in Funka Bay.
salinity during the spring bloom in Funka Bay. During the 5 years, the spring bloom was early in 1989
and 1990, at the end of February to the beginning of March
4.2 Influence of the inflow of OW on the occurrence of a (Type I year). As was mentioned earlier, the inflow of OW
phytoplankton bloom was early in those two years because salinity was less than
Oyashio water (OW) flows into Funka Bay from the 33.5 at the end of February. This indicates that the bloom
marginal area south east of Hokkaido. This water altered its occurred under the strong influence of OW. The main
composition from the original OW, flowing along the Kurile characteristic of these years is that the distribution of chl a
Islands as a result of mixing with coastal water which is was concentrated at a shallower depth, with the 5 µg l–1
affected by melting water from drifting ice from the Okhotsk contour line occurring above 50 m and nutrients decreasing
Fig. 9. Vertical distribution of chl a (– –) and sigma-t (—) at the peak of the bloom in 1988–92.
5-Year Study of Bloom 511
over the same depth range. This is due to the presence of a factor terminating the spring bloom in Funka Bay based on
strong pycnocline around 50 m. In contrast, in 1988 and the fact that silicate was depleted first and diatoms require
1991, when the spring bloom occurred later, the inflow of silicate for their growth. They compared the utilization rate
OW was late (Type II year). The initial stage of the bloom between silicate and phosphate, and did not consider nitrate.
occurred at salinity 33.5, which meant that the bay was less Although the ratio of silicate to nitrate in Funka Bay was
affected by OW with an absence of a strong pycnocline. As almost similar to the composition ratio for diatoms, we
a result of a deeper mixed layer in 1988 and 1991, the observed that nitrate was depleted earlier than silicate dur-
occurrence of the bloom was delayed until irradiance became ing the spring bloom in this bay. Further evidence in terms
high enough to exceed the production depth (i.e. euphotic of the limiting nutrient terminating the spring bloom in
zone) over a mixed layer depth at the end of March. The Funka Bay will be reported elsewhere (Kudo et al., sub-
euphotic zone in this bay during the bloom was shallower mitted).
than 30 m because the Secchi depth during the bloom was 5–
10 m. High chl a was found below the euphotic zone, with 4.4 Factors affecting primary production during bloom
the 5 µg l–1 contour line extending to 85 m in 1991. Nutrients Primary production is generally divided into “new”
also decreased with depth, even deeper than the euphotic production and “regenerated” production (Dugdale and
zone which corresponded to chl a distribution. Low vertical Goering, 1967). The annual production estimated by the 14C
stability of the water column allowed vertical mixing, re- method would represent the sum of new and regenerated
sulting in a deeper chl a distribution and nutrient decrease. production because nitrate was almost depleted in the eu-
In 1992 the spring bloom was initiated within FW and photic zone after the spring bloom. Thus, the production in
shortly after this, OW flowed into the bay, so the spring summer would run on regenerated production which is
bloom proceeded mainly in the OW. Due to the presence of fueled by recycling ammonium in the euphotic zone. Sub-
the pycnocline, high chl a was observed only in the euphotic stantial export production which equalled new production,
zone and utilization of nutrients was limited to this layer. occurred after the spring bloom using nitrate as a N source.
The presence of OW contributed physically to the stabili- Productivity estimated from the nutrient budget during the
zation of the water column because the pycnocline formed spring bloom ranged from 25 to 73 g C m–2 in this study.
between OW and FW. In another case, when the inflow of Diffusive supply of nutrients from the deeper layer to the
OW was not observed, water stability was weaker than the euphotic zone could also be expected because a vertical
former case. It is generally believed that the spring bloom in gradient of nutrients occurred after the development of the
Funka Bay occurred in the OW (Type I year) because the bloom. Tsunogai and Watanabe (1983) calculated the verti-
inflow of this water stabilized the water column. This study cal diffusive flux of nutrients in order to estimate the
showed that this is not always the case and although the nutrient assimilation rate during the spring bloom in Funka
development of the bloom was delayed until the end of Bay. Their result indicated that the diffusive flux was
March, a bloom also occurred within FW in the year when important before the peak of the bloom, but this flux de-
the inflow of OW was late (Type II year). Townsend et al. creased an order of magnitude after the peak. The calcula-
(1992) reported the development of a spring bloom without tion of production in the present study was based on the
a vertically stratified water column. They explained that difference in the standing stock of nutrients before and after
deep penetration of light in clear late winter waters allows the bloom, so the diffusive flux from the deeper layer would
cells to grow and overcome the vertical excursion rates be included in our estimate. Maita and Yanada (1985)
leading to a bloom when wind-driven vertical mixing was reported the annual production and the production during
weak or absent. the spring bloom using the 14C method in this bay were 130
and 30–40 g C m–2, respectively. New production in this
4.3 Nutrient status and limiting factors study accounts for 19–56% of the annual production by
In a Norwegian fjord, the ratio of SiO4 :NO3 :PO4 was Maita and Yanada (1985). Their production during the
7:12:0.7 and silicate depletion was observed at the end of the spring bloom falls within the range of the values reported in
spring bloom (Skjoldal and Wassmann, 1986). On the other this study.
hand, nitrate exhaustion is primarily responsible for the Odate and Maita (1988) reported that the food re-
senescence of the bloom in Auke Bay, Alaska because quirement (i.e. grazing pressure) was especially small at 3%
silicate was in excess over other nutrients (SiO4:NO3 :PO4 of the primary production during the spring bloom in Funka
ratio was 60:28:2, Waite et al., 1992a). Phosphate limitation Bay. Thus, the effect of grazing by zooplankton is minimal
has been reported in Norwegian waters (Sakshaug and due to low temperatures at the spring bloom.
Olsen, 1986). Diatoms are predominant in the spring bloom The lowest production occurred in 1990 when nutrients
and their nutrient ratios should control which nutrient would below the pycnocline were not utilized due to the presence
become limiting for phytoplankton growth. Tsunogai and of OW, resulting in a low consumption compared to the
Watanabe (1983) reported that silicate was the limiting potential production. On the contrary, this production in
512 I. Kudo and K. Matsunaga
1991 (Type II year) was three times higher than in 1990 and Maita, Y. and M. Yanada (1985): Chemical environment in Funka
similar to the potential production value. If the calculated Bay. p. 113–125. In Coastal Oceanography of Japanese Is-
production from the nutrient budget only depended on the land, Coastal Oceanography Research Committee, The
water column stability which affected the depth for nutrient Oceanographical Society of Japan (in Japanese).
Menzel, D. W. and N. Corwin (1965): The measurement of total
consumption, this production in 1989 and 1990 should be
phosphorus in seawater based on the liberation of organically
lower than that in other years. This study indicates this is not
bound fractions by persulfate oxidation. Limnol. Oceanogr., 10,
the only factor which determines the primary production 280–282.
during the bloom. As shown in Fig. 6, the available nutrient Miyake, H., I. Tanaka and T. Murakami (1988): Outflow of water
inventories before the bloom were different from each year. from Funka Bay, Hokkaido during early spring. J. Oceanogr.
This difference is derived from the proportion of deep water Soc. Japan, 44, 163–170.
remaining when the Tsugaru warm current flowed into the Odate, T. and Y. Maita (1988): Seasonal changes in the biomass
bay and replaced the existing water in the middle to deeper of zooplankton and their food requirement in Funka Bay. J.
layers in the previous fall. Thus, the primary production Oceanogr. Soc. Japan, 44, 228–234.
during a bloom which is defined as the utilized nitrate was Ohtani, K. (1971): Studies on the change of the hydrographic
closely associated with the water column stability and the conditions in the Funka Bay. II. Characteristics of the water
occupying the Funka Bay. Bull. Fac. Fish., Hokkaido Univ., 22,
initial available nutrient inventories.
58–66 (in Japanese).
Ohtani, K. (1985): Physical environment in Funka Bay. p. 102–
Acknowledgements 112. In Coastal Oceanography of Japanese Island, Coastal
We are grateful to the members the Marine Chemistry Oceanography Research Committee, The Oceanographical
laboratory for help with sampling and analyses. The authors Society of Japan (in Japanese).
also express special thanks to the captain and crew of the Parsons, T. R., Y. Maita and C. Lalli (1984): A Manual of
R/V Ushio Maru for assistance with sampling. We with to Chemical and Biological Methods for Seawater Analysis.
thank Dr. P. J. Harrison for reading an early manuscript and Pergamon Press, New York, 173 pp.
helpful comments. We thank Dr. S. Noriki, an editor and Sakshaug, E. and Y. Olsen (1986): Nutrient status of phytoplankton
anonymous referees for helping to improve the manuscript. blooms in Norwegian waters and algal strategies for nutrient
competition. Can. J. Fish. Aquat. Sci., 43, 389–396.
References Skjoldal, H. R. and P. Wassmann (1986): Sedimentation of par-
DeMaster, D. J. (1981): The supply and accumulation of silica in ticulate organic matter and silicium during spring and summer
the marine environment. Geochim. Cosmochim. Acta, 45, in Lindaspollene, western Norway. Mar. Ecol. Prog. Ser., 30,
Dugdale, R. C. and J. J. Goering (1967): Uptake of new and Townsend, D. W., M. D. Keller, M. E. Sieracki and S. G. Ackleson
regenerated forms of nitrogen in primary productivity. Limnol. (1992): Spring phytoplankton blooms in the absence of ver-
Oceanogr., 12, 196–206. tical water column stratification. Nature, 360, 59–62.
Japan Meteorological Agency (1988): Monthly Report of the Tsunogai, S. and Y. Watanabe (1983): Role of dissolved silicate
Japan Meteorological Agency. in the occurrence of a phytoplankton bloom. J. Oceanogr. Soc.
Japan Meteorological Agency (1989): Monthly Report of the Japan, 39, 231–239.
Japan Meteorological Agency. Waite, A., P. K. Bienfang and P. J. Harrison (1992a): Spring
Japan Meteorological Agency (1990): Monthly Report of the bloom sedimentation in a subarctic ecosystem. I. Nutrient
Japan Meteorological Agency. sensitivity. Mar. Biol., 114, 119–129.
Japan Meteorological Agency (1991): Monthly Report of the Waite, A., P. K. Bienfang and P. J. Harrison (1992b): Spring
Japan Meteorological Agency. bloom sedimentation in a subarctic ecosystem. II. Succession
Japan Meteorological Agency (1992): Monthly Report of the and sedimentation. Mar. Biol., 114, 131–138.
Japan Meteorological Agency.
5-Year Study of Bloom 513