Sequential Nutrient Uptake as a Potential Mechanism for Phytoplankton to Maintain High 1 Primary Productivity and Balanced Nutrient Stoichiometry 2 3

25 We hypothesize that phytoplankton have the sequential nutrient uptake strategy to 26 maintain nutrient stoichiometry and high primary productivity in the water column. 27 According to this hypothesis, phytoplankton take up the most limiting nutrient first until 28 depletion, continue to drawdown non-limiting nutrients and then take up the most limiting 29 nutrient rapidly when it is available. These processes would result in the variation of ambient 30 nutrient ratios in the water column around the Redfield ratio. We used high resolution 31 continuous vertical profiles of nutrients, nutrient ratios and on-board ship incubation 32 experiments to test this hypothesis in the Strait of Georgia. At the surface in summer, ambient 33 NO3 was depleted with excess PO4 3and SiO4 remaining, and as a result, both N:P and N:Si 34 ratios were low. The two ratios increased to about 10:1 and 0.45:1, respectively, at 20 m. 35 Time series of vertical profiles showed that the leftover PO4 3continued to be removed, 36 resulting in additional phosphorus storage by phytoplankton. The N:P ratios at the nutricline 37 in vertical profiles responded differently to mixing events. Field incubation of seawater 38 samples also demonstrated the sequential uptake of NO3 (the most limiting nutrient) and then 39 PO4 3and SiO4 (the non-limiting nutrients). This sequential uptake strategy allows 40 phytoplankton to acquire additional cellular phosphorus and silicon when they are available 41 and wait for nitrogen to become available through frequent mixing of NO3 (or pulsed 42 regenerated NH4). Thus, phytoplankton are able to maintain high productivity and balance 43 nutrient stoichiometry by taking advantage of vigorous mixing regimes with the capacity of 44 the stoichiometric plasticity. To our knowledge, this is the first study to show the in situ 45 dynamics of continuous vertical profiles of N:P and N:Si ratios, which can provide insight 46 into the in situ dynamics of nutrient stoichiometry in the water column and the inference of 47 the transient status of phytoplankton nutrient stoichiometry in the coastal ocean. 48


Introduction
The stoichiometry of the C : N : P Redfield ratio (Redfield, 1958) remains a central tenet in oceanography as it couples ecosystem processes with ocean biogeochemistry, which is driven by physical processes in oceans.The Redfield ratio of C : N : P varies widely across a wide range of environmental conditions.Laboratory cultures of phytoplankton that are in the steady state usually display variable cellular N : P ratios with the nutrient N : P supply ratios (Geider and La Roche, 2002).Recently, Martiny et al. (2013) found strong latitudinal patterns of the elemental ratios, which are closely related to ambient levels of nutrients, by making comparative analysis of elemental ratios of organic matter between different latitudes.Even at a fixed site (the Bermuda Atlantic Time-series Study station in the North Atlantic Ocean), the C : N : P ratio is quite variable (Singh et al., 2015).Four mechanisms have been proposed to explain the variability in C : N : P ratios in marine plankton, as summarized by Weber and Deutsch (2010).The first mechanism emphasizes the relationship between cellular elemental stoichiometry of phytoplankton and ambient nutrient ratios, i.e. the stoichiometry of nutrients in the water column.Based on the average Redfield ratio, this mechanism has been used to infer the most limiting nutrient for phytoplankton and to debate which nutrient (nitrogen or phosphorus) should be managed to control eutrophication effects.The second mechanism suggests that the elemental stoichiometry is taxonomy specific.Diatoms were reported to draw down nutrients with low C : P and N : P ratios (Geider and La Roche, 2002;Elser et al., 2003;Price, 2005), while marine cyanobacteria have higher C : P and N : P ratios (Karl et al., 2001;Bertilsson et al., 2003).Such different uptake ratios of N : P by phytoplankton can influence the magnitude of ocean N fixation (Mills and Arrigo, 2010).Based on the resource allocation theory, the third proposed mechanism is the "growth rate hypothesis", which states that the elemental stoichiometry within a cell is controlled by the biochemical allocation of resources to different growth strategies (Falkowski, 2000;Elser et al., 2003;Klausmeier et al., 2004).Fast-growing cells may have a lower N : P ratio due to a larger allocation to P-rich assembly machinery of ribosomes (Loladze and Elser, 2011), whereas competitive equilibrium favours a greater allocation to P-poor resource acquisition machinery and therefore higher N : P ratios.The fourth mechanism is related to the interference from dead plankton or organic detritus with the measurement of elemental composition of organic matter, and such interference is difficulty to assess due to lack of the measurements of non-living organic matter in oceans and coastal waters.However, the X-ray microanalysis (XRMA) technique was recently used to produce simultaneous quotas of C, N, O, Mg, Si, P and S in single-cell organisms (Segura-Noguera et al., 2016), which will not only help to understand the fourth mechanism but also understand the variability of stoichiometry of phytoplankton in the oceans.
In culture experiments, continuous uptake of non-limiting nutrients has been demonstrated for diatoms under N and Si limitation (Conway et al., 1976;Conway and Harrison, 1977;Harrison et al., 1989).Surge uptake of the limiting nutrient occurs when it is added to the nutrient-starved phytoplankton culture, while the uptake of the non-limiting nutrient is slowed or stopped until the diatom has overcome its nutrient debt.Hence, the sequence of which nutrient is taken up first is directly related to the nutrient status of the phytoplankton.It is difficult to assess the nutritional status of phytoplankton in the field, but the application of laboratory results to the interpretation of vertical nutrient profiles can provide information on their nutritional status.To date, there have been no studies of sequential uptake of nutrients in the field using a series of high-resolution vertical profiles of nutrients and their application to nutritional status of the phytoplankton.
In this study, we used high-resolution continuous vertical profiles of N : P and N : Si ratios to examine how N : P and N : Si ratios respond to the mixing in a highly dynamic coastal water column as well as the uptake of nutrients.Onboard ship incubation experiments were conducted to support the observations of changes in vertical profiles of N : P and N : Si ratios.We constructed seven conceptual profiles to illustrate how a vertical profile of N : P ratios changes with mixing and uptake of nitrogen and phosphorus, and how they could indicate the nutritional status of the phytoplankton assemblage.The conceptual model also explains how N : P ratios respond to mixing, particularly at the nutriclines (nitracline for NO − 3 , phosphacline for PO 3− 4 and silicacline for SiO − 4 ), and indicates which nutrient (NO − 3 or PO 3− 4 ) is taken up first in the water column.To our knowledge, this is the first study to show the dynamics of continuous vertical profiles of N : P and N : Si ratios and to examine responses of phytoplankton to the supply of nutrients from water column mixing.We believe that our approach can add a new dimension to examining the in situ dynamics of nutrients in the water column and illustrate the ecological role of phytoplankton stoichiometry in phytoplankton competition for nutrients.

Conceptual model of variability in vertical N : P ratios
The Strait of Georgia (hereafter the Strait) is an inland sea that lies between Vancouver Island and the mainland of British Columbia (LeBlond, 1983).It is an ideal area for studying the interactions between mixing, nutrient vertical profiles and phytoplankton nutrient uptake because of its relatively high biomass, frequent wind mixing and shallow (15 m) photic zone.The Strait is biologically productive, reaching a daily production of up to 5 g C m −2 d −1 and annual production of up to about > 300 g C m −2 yr −1 (Harrison et al., 1983(Harrison et al., , 1991)), but inorganic nitrogen is often undetectable in productive seasons in the surface layer.The nutricline sitting within the euphotic zone is often associated with the pycnocline.In the Strait, the ambient N : P ratio of nutrients is ∼ 10 : 1, similar to other coastal areas (Hecky and Kilham, 1988).
We illustrate the conceptual model of variability in vertical profiles of N : P ratios based on seven (C0 to C6) vertical profiles that we encountered in our field studies and suggest events that likely occurred to produce these nutrient profiles (Fig. 1).C0: in winter or after a strong wind speed event, the water column is homogeneously mixed, and NO − 3 and PO 3− time, the N : P ratio is near 0 and some PO 3− 4 remains in the water column.
C3: the remaining PO 3− 4 is completely taken up and stored as extra/surplus intracellular PO 3− 4 .
C4: after cross-pycnocline mixing occurs, the ambient N : P ratio in the newly mixed water should be the same as the ratio in the deep water.As a result, the vertical profile of the N : P ratio will form a right angle on the top part of the nutricline.
C5: depending on how long the phytoplankton are nutrient limited, their response to the mixed limiting nutrient can be different.When N-deficient phytoplankton take up N only, the curve of the N : P ratio parallels the NO − 3 distribution curve, and PO 3− 4 is left behind in the water column.
C6: on the other hand, if phytoplankton take up PO 3− 4 before NO − 3 (e.g. if phytoplankton were severely N starved, and there is a lag in NO − 3 uptake), the N : P ratio would be higher at the nutricline than below (Fig. 1).
Similarly, this conceptual model can be applied to N, SiO − 4 and N : Si ratios.The ambient (N : Si) ratio is about 0.5 : 1 at 20 m in the Strait, with 20 µM NO − 3 and 40 µM SiO − 4 .As the average uptake ratio of N : Si is about 0.7-1 : 1 (equivalent to Si : N = 1.5-1 : 1) (Brzezinski, 1985), the N : Si ratio decreases with depth.SiO − 4 is rarely depleted, and therefore the N : Si ratio is mainly determined by the distribution of NO − 3 .The continuous uptake of SiO − 4 without the uptake of NO − 3 can be inferred based on the comparison between the gradient of N : Si and the silicacline.For example, a sharper gradient of the N : Si ratio than the silicacline would indicate the continuous uptake of SiO − 4 without the uptake of NO − 3 as in C5 (Fig. 1) 2 Materials and methods

Station locations
The transect started from station S2, 8 km beyond the Fraser River mouth and under the influence of the river plume, and extended 108 km northwest to S1 (well beyond the plume) in the Strait of Georgia (Fig. 2).The station numbers are consistent with previous studies (Yin et al., 1997a).

Sampling and data processing
The sampling was designed to investigate the distribution of nutrients (NO − 3 , PO 3− 4 and SiO − 4 ) and N : P and N : Si ratios associated with mixing processes during 6-14 August 1991.Data, either at an anchored station for 24 h or along a transect of a few stations within 10 h, were used.At each station, a vertical profile (0-25 m) of temperature, salinity, in vivo fluorescence and selected nutrients (NO − 3 + NO − 2 , PO 3− 4 and SiO − 4 ) was obtained.Only vertical profiles of nutrients are presented in this study.Other data (salinity, temperature and fluorescence) are published elsewhere (Yin et al., 1997a).The vertical profiling system has been described in detail by Jones et al. (1991) and Yin et al. (1995a).Basically, a hose connected to a water pump on deck was attached to the conductivity-temperature-depth (CTD) probe or S4 (InterOcean ® ) which has the dual function of a CTD probe and a current meter.Seawater from the pump was connected into the sampling tubing of an AutoAnalyzer ® on-board the ship for in situ nutrient measurements, while the CTD probe was lowered slowly into the water at 1 m min −1 .Each sampling produced a high-resolution continuous vertical profile of physical and biological parameters, and thus the relationship between these parameters in the water column can be easily recognized.Data from a vertical profile (a datum point every 3 s) were smoothed over 15 s intervals.This smoothing reduced the fluctuations caused by the ship's motion.

Analysis of nutrients
All nutrients were determined using a Technicon AutoAnalyzer II.Salinity effects on nutrient analyses were tested on-board the ship and were found to be small.Therefore, no correction was made for salinity effects.NO − 3 + NO − 2 and  1967) and ammonium analysis followed Parsons et al. (1984).A water sample for particulate organic carbon and nitrogen (POC and PON) was filtered onto a GF/F filter, and POC/PON on the filter were analysed with a Carlo Erba model NA 1500 NCS elemental analyser, using the dry combustion method described by Sharp (1974).

Field incubation experiments
Incubation experiments of water samples were conducted to examine the disappearance and change in ambient nutrient ratios in the water samples from different cruises in 1989, 1990and 1991. Niskin bottles (5 L) . Niskin bottles (5 L) were used to take seawater samples and the samples were transferred to acid-cleaned carboys (10 L).Subsamples of seawater were transferred to transparent polycarbonate flasks (1 L) and placed in Plexiglas tanks.The tanks were kept at the same temperature as the surface water by pumping seawater (from the ship's intake at 3 m) through the tank.At Station S3, on 8 June 1989, water samples taken at 1 and 16 m were incubated in the tanks and wrapped with one and four layers of neutral density screening, respectively, which reduced the light intensity to the corresponding sampling depths (50 and 6 % of the surface light).At Station P3, on 11 August 1991, a water sample was incubated under one layer of neutral density screening (about 50 % of the surface light).In this incubation experiment, NO − 3 , PO 3− 4 and SiO Vertical profiles and seawater samples for in situ incubation which were used in this study were collected at different stations and different sampling times.Water column conditions such as salinity, temperature and fluorescence have been described in the listed publications, as shown in Table 1.

Vertical profiles of nutrients and nutrient ratios
At S3, near the edge of the Fraser River plume, the profiles documented changes before (T1) and after wind mixing (T7).At T1, both NO − 3 and PO 3− 4 were low in the surface layer and N : P ratios were low (< 2 : 1) and increased to ∼ 8 : 1 at 20 m  (Fig. 3).At T7, higher N : P ratios of 16-20 : 1 occurred due to an increase in NO − 3 in the deep water.SiO − 4 was ∼ 30 µM at the surface due to input from the Fraser River, and increased to 37 µM at 20 m (Fig. 3).The N : P ratio curve nearly formed a right angle at the top of the nutriclines at T7 when the gradient of the nitracline was larger than that of the phosphacline.At T1, the N : Si ratio was near zero because NO − 3 was near the detection limit, but started to increase along the nitracline at the depth of the SiO − 4 minimum.At T7, N : Si increased more rapidly with the nitracline.
A strong wind speed event occurred on 7 August and the water column was mixed (Yin et al., 1997b).We followed the change in the nutrient profiles and nutrient ratios from S3 near the Fraser River plume to P4 and P6 and well beyond the plume to S1.At S3, N : P ratios in the water column were greater than 7 : 1 when both NO − 3 and PO 3− 4 were high after wind mixing, with N : Si ratios being less than 0.5 : 1 (Fig. 4).As the post-wind bloom of phytoplankton developed along P4-P6 due to the newly supplied nutrients (Yin et al., 1997b), the N : P ratio followed the distribution of NO − 3 at P4, and decreased to zero as NO − 3 was depleted at the surface at P6 (Fig. 4).It was clear that little PO 3− 4 was consumed while NO − 3 was taken up.At the same time, the silicacline deepened and paralleled the nitracline.At S1, N : P and N : Si ratios formed almost a vertical line.N : P and N : Si ratios were ∼ 8 : 1 and 0.5 : 1, respectively, in the deep water (Fig. 4).
The time series (T1, T3, T8 and T11) of 8-9 August captured changes over 1 or 2 days after the wind mixing event at S1 that were well beyond the river plume (Fig. 5).At T1, N : P and N : Si ratios were ∼ 9 : 1 and 0.45 : 1, respectively, with NO − 3 and PO 3− 4 being 15 and 1.7 µM, respectively, at the surface.At T3, the N : P ratio remained constant at ∼ 9 : 1, while NO − 3 and PO 3− 4 decreased by 10 and 1.0 µM, respectively, indicating an uptake N : P ratio of 10 : 1.In comparison, the N : Si ratio decreased from T1 to T3 when SiO − 4 was 35 µM at T1 and decreased by > 10 µM at T3, producing an uptake N : Si ratio of ∼ 1 : 1.At T8, the N : P ratio followed the NO − 3 distribution as NO − 3 decreased to ∼ 0 µM at the surface while PO 3− 4 was still ∼ 0.5 µM.This indicated that NO − 3 uptake was more rapid than PO 3− 4 uptake, and hence NO − 3 mainly determined the ambient N : P ratios.The N : Si uptake ratio of ∼ 1 : 1 continued until T8.However, at T11, the N : P ratio spiked higher in the top 5-10 m of the nutricline, suggesting a more rapid uptake of PO 3− 4 relative to NO − 3 in the upper portion of the phosphacline (Fig. 5).
Changes in the profiles after the wind event on 7 August were followed over 5 days (10-14 August) at P5, which was still within the influence of the river plume, as evidenced by the higher surface SiO − 4 at the surface (Fig. 6).On 10-11 August, N : P ratios were higher at the surface where the post-wind induced bloom occurred 2 days earlier, suggesting that uptake of PO 3− 4 had caught up with uptake of NO − 3 .The right-angle shape of the N : P ratio on 12 August occurred as the nutriclines became sharper due to entrainment of nutrients.By 13 August, more NO − 3 was taken up at depth and the N : P ratio followed the deepening of the nitracline, and PO 3− 4 was left behind.On 14 August, PO 3− 4 started to decrease.During 10-14 August, a minimum in SiO − 4 was present at an intermediate depth (5-10 m), coinciding with the top of the nitracline, and the silicacline followed the nitracline below 10 m.

Changes in nutrient ratios during field incubations
On-deck incubation experiments were used to examine changes in uptake ratios by eliminating any effects due to mixing.Ambient N : P and N : Si ratios were lower at the surface than at depth, indicating higher uptake of NO − 3 at the surface.The indication of a higher uptake ratio of N : P and N : Si was supported by field incubation experiments.During nutrient addition (NO − 3 , PO 3− 4 and SiO − 4 ) bioassays on a sample from 1 m at P3, all nutrients decreased as fluorescence increased (Fig. 7).Ambient N : P and N : Si ratios decreased to almost 0.0 after 96 h, indicating more rapid uptake of NO − 3 than uptake of PO 3− 4 and SiO − 4 .The temporal decline in the N : P and N : Si ratios resembled the temporal progression during a bloom, as illustrated in C0-C3 of the conceptual profiles (Fig. 1) and in the water column (S3, P4, P6) on 8 August (Fig. 4) and during the time se- time, and the accumulative uptake ratio of N : Si exceeded 3 : 1 in the presence of PO 3− 4 (Fig. 9c).

Discussion
The Strait is highly productive, reaching up to 2700 mg C m −2 d −1 in August (Yin et al., 1997b).This is due to pulsed nutrient supplies and multiple phytoplankton blooms in the shallow photic zone interacting with wind events (Yin et al., 1997b) and fluctuations in river discharge (Yin et al., 1997a(Yin et al., , 1995c)).Our results revealed sequential nutrient uptake as a potential mechanism to optimize nutrient uptake efficiency and generate high primary productivity of phytoplankton by taking advantage of pulsed nutrients in this highly dynamic, relatively shallow photic zone.

Responses of N : P and N : Si ratios to vertical mixing and uptake of nutrients
A vertical profile of N : P and N : Si ratios represents a snapshot of the mixing and the uptake of N, P and Si by phytoplankton in the water column.The depletion zone of the most limiting nutrient in the euphotic zone ends at a depth where the uptake of nutrients just balances the upward flux of nutrients through the nutricline, as indicated in C3 in the conceptual profiles (Fig. 1).Different responses of nutrient uptake to pulsed nutrients by mixing appeared to depend on the previous stability of the water column, the depth of the euphotic zone and nutritional status of phytoplankton.Our observations spanned all seven conceptual profiles (Fig. 1) and indicated the dynamic processes influencing the sequence of nutrient uptake.The change in the profiles of the N : P ratio from S3 to P6 (Fig. 4) displayed the spring bloom-like progression, as illustrated in conceptual profiles of C0-C3 (Fig. 1) after the wind mixing event.Various responses illustrated in the conceptual profiles C4, C5 and C6 (Fig. 1) were observed in the observations, including the right angle in the N : P ratio (T7 -Fig.3; P5, 12 August -Fig.6), parallel lines between the nitracline and the N : P ratio curve on 12 August, (Fig. 6), and a spike in the N : P ratio curve at T11 at S1 due to continued uptake of PO 3− 4 , with NO − 3 being depleted during the time period from T1 to T8 (Fig. 5), which was frequently observed on 10 August at P5 (Fig. 6).The recycling of nutrients in different preferences, such as faster P regeneration than N, which in turn is recycled faster than Si, also contributes to the variability of nutrient ratios as shown above.

Sequential nutrient uptake for balanced stoichiometry and nutritional optimization
Phytoplankton can take advantage of the dynamic mixing regimes and optimize their growth rates by taking up nutrients sequentially.The disappearance of nutrients during the incubation resembled the temporal progression of a bloom as illustrated in C0-C3 of the conceptual profiles (Fig. 1) and in the water column (S3, P4, P6; Fig. 4), or during the time series at S1 (Fig. 5).
Nutrient deficiency results from a decrease in the cellular content of the limiting nutrient and continuous uptake of other non-limiting nutrients.Earlier studies found that N limitation results in excess cellular content of P and Si (Conway and Harrison, 1977;Healey, 1985;Berdalet et al., 1996).Some phytoplankton develop enhanced uptake of the limiting nutrients, such as NH 4 and PO 3− 4 , upon its addition after a period of nutrient limitation or starvation, and there is an accompanying shutdown of the non-limiting nutrient (Conway et al., 1976;Conway and Harrison, 1977;McCarthy and Goldman, 1979).A few hours of enhanced N uptake quickly overcome the N debt since the enhanced uptake rate is many times faster than the growth rate (Conway et al., 1976).For example, enhanced uptake of phosphorus could double internal P within 5 min to 4 h depending on the degree of P limitation and the pulsed PO 3− 4 (Healey, 1973).After the nutrient debt has been overcome by enhanced uptake, the uptake of non-limiting nutrients returns to normal after the cell quota of the limiting nutrient is maximal (Collos, 1986).The sequential uptake of a limiting nutrient and then the uptake of both the non-limiting and limiting nutrients are advantageous to allow phytoplankton to maintain maximum growth rates over several cell generations.

Significance of sequential uptake of nutrients
There are two essential strategies used by phytoplankton to cope with the limiting nutrient (Collos, 1986).One strategy is the "growth" response, where phytoplankton uptake of the limiting nutrient, and cellular growth are coupled when the limiting nutrient is available.The other strategy is the "storage" response, where phytoplankton have the capability of accumulating large internal nutrient pools, resulting in extensive uncoupling between uptake and growth, and a lag in cell division of up to 24 h following a single addition of the limiting nutrient.The former strategy would have the competitive advantage under frequent pulses of the limiting nutrient, whereas the latter strategy presents an ecological advantage when the nutrient pulsing frequency is lower than cell division rate.A phytoplankton assemblage can be assumed to contain both strategies in the water column.Phytoplankton species composition in subsurface waters was more or less similar at three stations (S1, S2 and S3), considering a span of 100 km across a large salinity gradient (Clifford et al., 1992).Cryptomonads and Chrysochromulina spp.and Micromonas pusilla were dominant at S2, S3 and S1 in cell density (Clifford et al., 1992).The common diatom species included Chaetoceros spp.and Thalassiosira spp.(Clifford et al., 1992), which are said to use the growth and storage strategies, respectively (Collos, 1986).At Station S2, the chlorophyll maximum at 7 m on 7 August contained 4 times more phytoplankton cells than at the surface (Clifford et al., 1992) and was frequently observed at or associated with the nutricline (Cochlan et al., 1990;Yin et al., 1997a).Phytoplankton there could use either the growth or storage strategy by different species.The storage strategy of non-limiting nutrients would allow phytoplankton to utilize the limiting nutrient when it is available and thus maximize phytoplankton growth by saving the energy expenditure associated with taking up non-limiting nutrients under limiting irradiance.This may explain why there were various modes or patterns of the N : P ratio at the nutricline, which indicates the different strategies of taking up nutrients sequentially based on the nutritional status of phytoplankton.The sequential uptake strategy allows some phytoplankton species to use the storage www.biogeosciences.net/14/2469/2017/Biogeosciences, 14, 2469-2480, 2017 capacity for non-limiting nutrients and other phytoplankton species to use the growth response for the most limiting nutrient when it becomes available by mixing processes.Sequential uptake of nutrients by phytoplankton can be a fundamental mechanism in maintaining high productivity in the water column where there are frequent mixing events in coastal waters.The sequential uptake strategy largely occurs at the nutriclines near or at the bottom of the photic zone.There is a consistent association between the nutriclines and the chlorophyll maximum in various aquatic environments (Cullen, 2015) and it is also common in the Strait (Harrison et al., 1991).There is a frequent upward flux of nutrients through the nutricline due to entrainment in the Strait (Yin et al., 1995a, b, c) and by internal waves in the open ocean (Pomar et al., 2012).Phytoplankton in the chlorophyll maximum are generally exposed to nutrients, and when these cells are brought up to the surface during entrainment or wind mix-ing (Yin et al., 1995a) they can quickly photosynthesize (Yin et al., 1995c).When phytoplankton exhaust the most limiting nutrient, their internal nutrient pool decreases and they sink down to the nutriclines, possibly due to the formation of clumps, and take up the abundant nutrients there.Thus, the cycle of sequential uptake of the limiting and then the nonlimiting nutrients may reduce nutrient deficiency in phytoplankton.
Sequential uptake of nutrients can be an important process to maintain the phytoplankton nutrient stoichiometry.Carbon fixation continues after a nutrient becomes deficient (Elrifi and Turpin, 1985;Goldman and Dennett, 1985) and the storage of organic carbon of a higher POC : N ratio is common in phytoplankton (Healey, 1973).When phytoplankton cells with excessive organic carbon, due to limitation of a nutrient, sink from the upper euphotic zone to the nutricline where light becomes limiting, uptake of other nutrients occurs by utilizing stored organic carbon, leading to an increase in the cellular N and P quotas.Thus, the ratios of carbon to other nutrients approach optimum stoichiometry.POC : N ratios at Stations S2 and S3 were observed to be between 6 : 1 and 7 : 1 in the water column, even though both ambient NO − 3 and PO 3− 4 were near detection limits (Fig. 10).In addition, the POC : N ratio was slightly higher than 7 : 1 (Fig. 10) at Station S1 where nitrogen was more frequently under the detection limit than at Stations S2 and S3.This might suggest the lack of ambient nitrogen limitation on the cellular nutrient stoichiometry.However, using the C : N ratio in particular matter to infer the nutrient limitation has its limitation as particular C : N ratios do not necessarily reflect phytoplankton elemental composition alone, especially in estuarine-influenced waters.

Conclusions
The use of in situ continuous vertical profiles in this study shows a high variability of ambient N : P and N : Si ratios in the water column, suggesting the dynamics of nutrient uptake ratios, as illustrated in the conceptual model of Fig. 1.The incubation experiments demonstrated the sequential uptake of nutrients by phytoplankton, which suggests that deficiency of a nutrient that is based on the ambient nutrient ratio could be transient and overcome by the sequential uptake of the most limiting and non-limiting nutrients.The capacity of sequential uptake of nutrients is an important strategy for phytoplankton to maintain high primary productivity and near-optimum cellular nutrient stoichiometry in the water column.The sequential nutrient uptake strategy also offers another mechanism for the explanation of the variability in the nutrient stoichiometry of phytoplankton in the euphotic zone.

Figure 1 .
Figure1.Conceptual model for changes in vertical profiles of N, P and N : P ratio due to sequential nutrient uptake.C0 to C3 represent a time series of nutrient uptake during bloom development, and C4 to C6 indicate subsequent vertical mixing of nutrients and subsequent uptake.The short horizontal line near the middle of the depth axis indicates the euphotic zone depth.N disappears first at C2, and P is left and continues to be taken up at C3. C4 represents mixing of nutrients into the bottom of the photic zone and phytoplankton have not taken up these nutrients yet.At C5, N is taken up first before P, while at C6, P is taken up first before N.

Figure 2 .
Figure 2. Map of the Strait of Georgia showing the study area and the sampling stations.Note that the Fraser River is located to the right, with two river channels flowing into the Strait of Georgia.

Figure 8 .
Figure 8.Time course of NO − 3 , PO 3− 4 and SiO − 4 during the field incubation of water samples taken at Station S3 on 8 June 1989.(a) Sample taken at 1 m and incubated under one layer of screening; (b) sample taken at 16 m and incubated under four layers of screening.

Figure 9 .
Figure 9.Time course of ambient NO − 3 , PO 3− 4 and SiO − 4 during the field incubation of a water sample taken at Station S1 on 4 June 1990; (a) pre-incubation: no nutrients were added to the sample during the first 28 h; (b) after pre-incubation, nutrients were added in eight treatments: +N, +P, +Si, +N+P, +N+Si, +P+Si and +N+P+Si, as described in the materials and methods section; (c) ambient nutrient ratios were calculated from measured ambient nutrients during the time course of incubation in panel (b) .Ambient +N / +P indicates the ratio of N in +N (the added N alone) over P in +P (the added P alone).Please note that this N : P ratio is not for N and P in the same bottle.Similarly, ambient +N / +Si indicates N in +N over Si in +Si (the added Si alone).The accumulative uptake ratio was directly calculated from the decreasing concentrations over time in panel (b).

Figure 10 .
Figure 10.Vertical profiles of particulate organic C : N ratios at Stations S2, S3 and S1 along the increasing distance from the river during 20-23 August 1990.

Table 1 .
Water column conditions, such as vertical profiles of salinity, temperature and fluorescence, that can be found in the listed publications for the sampled stations in the Strait of Georgia.