Real-time monitoring of nutrient concentrations and red-tide outbreaks in the southern sea of Korea



[1] In order to determine the physical and chemical factors controlling the outbreak of red tides, we monitored nutrients and other environmental parameters using a novel real-time monitoring buoy system during the summer of 2003 in the southern sea of Korea, where red-tide outbreaks occur every year. The real-time monitoring data on bioluminescence, which may indicate the presence of bioluminescent dinoflagellate species, showed a sudden increase under the lowest concentrations of dissolved inorganic nutrients. Our monitoring system for the first time provides real-time variations in nutrients and associated outbreaks of dinoflagellate red tides. This result supports the previous hypothesis by others that the outbreak of dinoflagellate red tides is associated with the limited growth of diatoms under depleted DIP or DIN conditions. We suggest that this real-time monitoring system can be utilized as a powerful tool for studying and predicting harmful dinoflagellate red tides in the coastal ocean.

1. Introduction

[2] Dinoflagellate blooms have occurred every summer in the coastal areas of the southern sea of Korea since their first occurrence in 1982 [Cho et al., 2001; Cho and Costas, 2004]. Although Cochlodinium polykrikoides is known to be the dominant red-tide species in this region, Gyrodinium impudicum often predominates over other species [Jeong et al., 2000]. The red tides persist from late summer to fall, leading to serious consequences for fin-fish farms, which cover large areas of the coastal ocean. The estimated economic loss caused by C.polykrikoides blooms was about $95.5 million in 1995 [Kim et al., 1997].

[3] Many studies have been conducted to determine the factors that control the outbreak of dinoflagellate red tides. Dinoflagellates can adapt well to low nutrient levels since, although their half-saturation constant (Ks) is higher than that for diatoms [Smayda, 1997] and many dinoflagellate species undergo vertical migration, descending at night, which is an efficient means of taking up nutrients in the subsurface layer [MacKenzie, 1991; Kamykowski, 1995; Koizumi et al., 1996; Park et al., 2001]. Some dinoflagellate species are capable of taking up dissolved organic nitrogen (DON) and/or phosphorus (DOP) using oxidase and alkaline phosphatase enzymes, respectively [Palenik and Morel, 1990; Carlsson et al., 1998; Glibert et al., 2001; Oh et al., 2002; Berg et al., 2003], and some phototrophic dinoflagellates have turned out to be mixotrophic [Smalley et al., 1999; Skovgaard, 2000; Jeong et al., 2004]. Thus, dinoflagellate red tides have been suggested to be associated with the limited growth of diatoms under depleted conditions of inorganic nutrients [Cembella et al., 1984; Egge, 1998; Oh et al., 2002; Lunven et al., 2005].

[4] In order to elucidate real-world controls of red-tide outbreaks, we need to obtain high-resolution time-series information on physical and biogeochemical variations that may be associated with climate and ocean variability. For this purpose, we deployed a monitoring buoy system which houses various environmental sensors for measuring temperature, salinity, currents, waves, winds, nutrients, chlorophyll a, bioluminescence, and methane. The details of buoy-system instrumentation have been reported by Nam et al. [2005]. We thought that these real-time data would provide the best information on the actual environmental conditions for red-tide outbreaks in collaboration with shipboard observations or laboratory experiments.

2. Materials and Methods

[5] The study area is bordered by the Goheung-Oenarodo and Yeosu-Kuemhodo peninsulas along the southern coast of Korea (Figure 1). The water depth is shallower than 20 m, but the oligotrophic Kuroshio water passes through the open areas outside Yeoja Bay (Figure 1). The average water depth and spring tidal range were about 9.6 m and 3.2 m, respectively, at the buoy station. A real-time monitoring buoy (Figure 2) was deployed from August 13 to September 4, 2003 (Figure 1).

Figure 1.

A location map of the monitoring buoy (semi-filled circle) and water sampling sites (solid triangles) in the southern sea of Korea.

Figure 2.

A real-time monitoring system for measuring nutrients, chlorophyll a, bioluminescence, CTD (temperature and salinity), and currents using an ocean buoy [from Nam et al., 2005].

[6] The chlorophyll a and bioluminescence were measured using a Fluorometer (Seapoint Co.) and Glowtracka (Chelsea Instruments Ltd.). Glowtracka measures bioluminescence as an index of dinoflagellate abundance since some dinoflagellates themselves exhibit a large range of luminescent output. The concentrations of nutrients (NO3, NO2, PO43−, and Si(OH)4) were measured using a submersible multi-channel analyzer (Eco-LAB). The commercially available Eco-LAB uses conventional colorimetric methods, which are a well-established wet chemistry technique.

[7] This buoy system utilizes Code Division Multiple Access (CDMA) as a communication platform for two-way communication between the sensors on the buoy and a database computer and/or personal cellular phone. The system was programmed to produce data every 10 minutes for temperature, chlorophyll a, and bioluminescence, and every 3 hours for nutrients at 1.5 m below the surface. Although we attempted to monitor all inorganic nutrient species and hydrographic parameters, the system did not work for NH4+ and salinity because of mechanical problems.

[8] In order to validate the buoy data, we collected surface seawater samples from station 1 (5 samples from nearby areas) on August 12 and stations 1–5 on August 19 and 22, 2003, for the analyses of nutrients and pigments (Figure 1). For the grab samples, nutrients (NO3, NO2, NH4+, PO43−, and Si(OH4)) were measured using an auto-analyzer (TRAACS 2000, Bran+Luebbe Co.). We define DIN as the sum of NO3, NO2, and NH4+ and DIP as PO43−. Photosynthetic pigments in seawater were measured using reverse-phase high-performance liquid chromatography (HPLC) (Waters Co. System) with a Rexchrom-S5-100-ODS column (Regis, USA, 250 × 4.6 mm, particle size: 5 μm) after a slight modification of the method by Wright et al. [1991], changing the linear gradient of solvents A (methanol:ammonium acetate, 80:20), B (acetonitrile:H2O, 87.5:12.5), and C (ethyl acetate). The dominant phytoplankton communities were calculated using the CHEMTAX program for pigments [Mackey et al., 1996]. An initial pigment ratio to chlorophyll a for each marker used in the CHEMTAX calculation was from the Southern Ocean data [Mackey et al., 1996].

3. Results and Discussion

[9] During the monitoring period, the water temperature increased from 23.3°C to 27.3°C in August 2003 (Figure 3). The concentrations of Si(OH)4 were higher than 10 μM, ample to stimulate diatoms [Kudo, 2003; Fujiki et al., 2004]. However, the concentrations of PO43− were lower than 0.5 μM, except on 14 August (2.96 μM). The DIN showed a large fluctuation ranging from 0.03–9 μM for NO3 and from 0.15–2.62 μM for NO2 (Figure 3). In general, the ratios of DIN/DIP were lower than 16 (the Redfield Ratio, the average ratio of DIN/DIP utilized by phytoplankton) in seawater samples taken from the study sites (Table 1), indicating a lack of bio-available DIN relative to DIP in the study season.

Figure 3.

The variations of temperature, PO4, Si(OH)4, NO3, NO2, chlorophyll a, and bioluminescence at the Buoy station, the southern sea of Korea, from 13 August to 4 September 2003. The real-time data were obtained automatically using chemical sensors.

Table 1. Concentrations of Dissolved Inorganic Nutrients, Chlorophyll a, and the Composition of Dinoflagellates and Diatoms in Areas Outside Yeoja Bay in 2003a
 August 12August 19August 22
  • a

    The values in parentheses are the average and the standard deviation of 5 surface-water samples, and the unit of nutrients is μM.

Temperature24.4–24.6 (24.5 ± 0.1)24.9–26.0 (25.4 ± 0.4)24.0–25.2 (24.3 ± 0.5)
Salinity29.8–29.9 (29.9 ± 0.0)29.8–30.3 (30.1 ± 0.2)27.8–29.6 (28.9 ± 0.7)
NH41.02–2.39 (1.54 ± 0.62)0.32–2.07 (1.13 ± 0.69)0–0.33 (0.13 ± 0.15)
NO20–0.32 (0.11 ± 0.12)0.38–0.74 (0.56 ± 0.14)0.07–0.14 (0.09 ± 0.04)
NO30.64–2.39 (1.47 ± 0.66)0.44–1.29 (0.79 ± 0.32)0.13–0.45 (0.28 ± 0.11)
DIN1.85–4.47 (3.12 ± 1.18)1.35–3.99 (2.48 ± 1.01)0.32–0.83 (0.50 ± 0.22)
DIP0.19–0.27 (0.23 ± 0.03)0.17–0.27 (0.22 ± 0.05)0.02–0.12 (0.07 ± 0.05)
Si(OH)413.6–19.2. (15.9 ± 2.2)14.0–17.9 (15.8 ± 1.4)13.4–16.7 (15.0 ± 1.5)
DIN/DIP7.0–23.6 (14.3 ± 7.0)7.7–14.6 (11.2 ± 2.5)2.7–41.5 (13.7 ± 15.9)
Chlorophyll a (μg L−1)0.16–0.74 (0.36 ± 0.24)0.55–1.58 (0.86 ± 0.41)1.00–11.3 (3.47 ± 4.40)
Community Composition of Dinoflagellates and Diatoms, %
Dinoflagellates0.06–14.9 (4.10 ± 6.16)1.62–8.25 (4.47 ± 2.72)3.24–59.2 (17.8 ± 23.5)
Diatoms16.0–70.8 (47.2 ± 23.8)45.7–51.8 (48.6 ± 2.6)36.4–80.6 (53.2 ± 16.5)

[10] The concentration of chlorophyll a exhibited a sudden increase from 25 August 2003, following a gradual increase during the period 15–24 August 2003. This trend is similar to that of bioluminescence indicating the presence of bioluminescent dinoflagellates. The increase of dinoflagellates from 19 August was confirmed by the analysis of photosynthetic pigments in grab samples determined by HPLC (Table 1). Diatoms and dinoflagellates were the dominant species in the study area, as indexed from the marker pigments, fucoxanthin and peridinin, respectively, with the highest proportion of dinoflagellates and visible red-tide outbreaks on 22 August 2003 (Table 1).

[11] Such a trend was also found in the cell-density of C. polykrikoides, which is one of the major red-tide species in this region [National Fisheries Research and Development Institute, 2003]. The cell density was below 300 (20–300) cells mL−1 before 15 August 2003, but increased to 300–1500 cells mL−1 between 15–23 August 2003, and then up to 6000 cells mL−1 after 24 August 2003 until red tides disappeared (around mid-October, 2003). So far, physical aggregation effect has been suggested to be the most likely mechanism of such a sudden increase of dinoflagellates in the red tide patch zones [Jeong et al., 2000; Litaker et al., 2002; Lee and Kang, 2003; Park et al., 2005]. The change in cell density in this region is associated with physical and biogeochemical conditions of inshore waters rather than the transport of water with red tide already present. This was confirmed by SF-6 [Park et al., 2005] and 224Ra [Hwang et al., 2005] tracers, which are chemically soluble artificial and natural tracers, respectively, in seawater.

[12] Although the reproduction rates of diatoms are about 3∼5 fold higher than those of C. polykrikoides [Smayda, 1997; Kim et al., 2001], the intensity of fucoxanthin increased only about 4 times during an about 50-fold increase of peridinin from August 19 to 22, 2003. Thus, the much larger increase of dinoflagellates appears to be due to the limited growth of diatoms [Jeong et al., 2000; Litaker et al., 2002; Lee and Kang, 2003; Park et al., 2005]. Indeed, the observed outbreak of chlorophyll a occurred under minimum DIN and DIP concentrations assuming that the trend of NH4+ concentrations was similar to that of NO3 (as shown from grab samples measured in the laboratory, Table 1). Thus, this real-time monitoring result supports the hypothesis by others [Cembella et al., 1984; Egge, 1998; Oh et al., 2002; Lunven et al., 2005] that outbreaks of red tides are associated with the limited growth of diatoms under conditions of minimum dissolved inorganic nutrient concentrations.

4. Conclusions

[13] Our real-time monitoring results show that the outbreak of bioluminescence (caused by bioluminescent dinoflagellates) coincides with the increased total abundance of dinoflagellates including non-bioluminescent dinoflagellates such as Cochlodinium, which is one of the major red-tide species in this region. This suggests that this type of real-time monitoring system can be used as a powerful tool for studying and predicting the outbreak of harmful dinoflagellate red tides in the coastal ocean. The monitoring may be more powerful if all other biological, physical, and chemical observations are collated. For example, one should resolve the question of how dinoflagellates reproduce or accumulate so quickly that they are visible within a few days, and how red tides are maintained for a few months, and the mechanisms terminating red tides.


[14] This research was fully supported by Korea Research Foundation Grant (KRF, R08-2003-000-10328-0). We thank D. Hutchins (at the University of Delaware) for valuable comments on an earlier version of this manuscript.