ABSTRACT: We investigated the effects of specific light wavelengths from light-emitting diodes (LEDs) on the growth of the dinoflagellate Heterocapsa circularisquama, which kills bivalves, and the diatom Skeletonema costatum, which is an important food source for bivalves. Growth of H. circularisquama was obviously inhibited at 590 nm and a photon flux density less than 75 μmol quanta/m2/s. However, growth of S. costatum was not suppressed by irradiance from any LEDs tested from near-ultraviolet to near-infrared wavelengths at 75 μmol quanta/m2/s. The growth rate of H. circularisquama in an experimental treatment group with irradiance provided by both cool-white fluorescent lamps (12:12 h L : D cycle) and a 590-nm LED (continuous irradiance) was 0.43/day. In the control group with irradiance provided only by cool-white fluorescent lamps (12:12 h L : D cycle), the growth rate was 0.63/day, indicating that growth of H. circularisquama was suppressed by 590 nm (less than 75 μmol quanta/m2/s) irradiance from the LED and the continuous irradiance. The use of 590-nm LEDs in bivalve culture at irradiance levels less than 75 μmol quanta/m2/s might encourage the growth of the useful diatom S. costatum without stimulating growth of the harmful dinoflagellate H. circularisquama.
Phytoplankton synthesize organic matter from using light as an energy source.1,2 Certain physiological processes of phytoplankton (e.g. carbon metabolism, cell pigment composition and enzymatic activities) are affected by the spectral composition of light.3–9 Moreover, the physiological effects of the spectral composition of light are variable among different algal groups.10–12 Researchers have explored the possibility of obtaining live food for filter feeders by large-scale monocultures and photobioreactors using light sources with varying spectral characteristics.8,13–17 For instance, more amino acids and protein may be found in algal biomass produced under blue light than under equal levels of white light, whereas red light has been shown to favor glycogenesis over proteogenesis.3 However, the physiological effects of light wavelengths on dinoflagellates remain unsolved because reports are rare, and some harmful dinoflagellates have a baneful influence on bivalve cultures.
In coastal waters of Japan, plankton causing outbreaks of red tide include up to 200 species,18 and the dominant species tend to shift between diatoms and dinoflagellates.19 Some dinoflagellates with no previous history of bloom formation have recently caused serious problems.20–22Heterocapsa circularisquama is a dinoflagellate that has caused mass mortalities of bivalves in Japan only since 1988.20,23–25 In 1992, an extensive red tide of H. circularisquama occurred in Ago Bay.26 Thereafter, this species caused heavy damage to cultured shellfish in western Japan. Cultured pearl oysters, in particular, were affected by H. circularisquama at lower cell densities than other harmful dinoflagellates,27,28 whereas harmful effects on wild fish populations or cultured fish have not yet been reported.29,30
Recently, light-emitting diodes (LEDs) were used as a light source in large-scale monocultures and photobioreactors, instead of gas discharge lamps16,31,32 because LED light sources have much lower electrical power consumption and a longer life than gas discharge lamps. Thus, we investigated the effects of specific wavelengths of light from LEDs on the growth of the harmful dinoflagellate H. circularisquama and the diatom Skeletonema costatum, an important food source for bivalves in batch culture. If wavelengths and irradiance of LEDs, which can obviously inhibit the growth of H. circularisquama, can be successfully observed and used, LEDs may be able to control the biomass of H. circularisquama around intensive bivalve culture grounds, such as Ago Bay.
MATERIALS AND METHODS
Phytoplankton strains and culture conditions
Heterocapsa circularisquama was isolated from Nishiura, Mie Prefecture, Japan (HCAS003N7) and S. costatum was obtained from the Microbial Culture Collection of the National Institute for Environmental Studies, Japan (MCC-NIES, strain N324). Aged sea water collected from the Okinoshima area (N34°15′11″, E130°04′51.6″) in the Tsushima Current was enriched with nutrients to the concentrations of f/2 medium33 after sterilization through a membrane filter (Sterivex-GS, 0.22 μm filter unit with a filling bell; Millipore, Billerica, MA, USA). Heterocapsa circularisquama was cultured in medium enriched without silicate. For the culture of S. costatum, silicate was added to a final concentration of 107 μmol/L. All culture media were diluted to 30 psu with ultrapure water and adjusted to a pH of 8.0 by the addition of HCl or NaOH. Stock cultures were maintained at 25°C under cool-white fluorescent lamps (FLR40S. EX-N/M/36-H, Toshiba, Tokyo, Japan) at a photon flux density (PFD) of 100 μmol quanta/m2/s (12:12 h L : D cycle). Experiments were conducted using batch cultures in test tubes (18 mm × 180 mm; borosilicate glass, Duran, Mainz, Germany). Prior to the experiments, all equipment and glassware were washed with 30% HCl, thoroughly rinsed with distilled water and then autoclaved at 202 kPa for 20 min.
Effect of irradiance source and photon flux density on the growth rate of Skeletonema costatum and Heterocapsa circularisquama
Skeletonema costatum and H. circularisquama were incubated at 25°C, 30 psu and 12:12 h L : D cycle. Illumination was provided by cool-white fluorescent lights (FLR40S. EX-N/M/36-H, Toshiba) at PFDs of 10, 50, 75, 100, 200 and 250 μmol quanta/m2/s, and the 590-nm is provided by six different irradiances (10, 30, 50, 75, 100 and 190 μmol quanta/m2/s) for S. costatum and seven different irradiances (30, 55, 75, 100, 120, 160 and 300 μmol quanta/m2/s) for H. circularisquama. The PFDs were measured with a radiometer (model QSL-2101; Biospherical Instruments, San Diego CA, USA). We measured cell densities daily (at 10:00 hours) using a Sedgewick–Rafter chamber under an inverted microscope (Type-210; Nikon, Tokyo, Japan). Specific growth rates (μ; /day) were calculated for cultures in exponential growth phase using a least-square regression on log-transformed cell density data:
where N0 and Nt denote the initial and final cell densities during the exponential phase (cells/mL), respectively, and Δt represents the time interval of the exponential growth phase (days).
As there was no apparent photoinhibition at the PFDs used, a rectangular hyperbolic curve was fitted for the relationship between specific growth rate and PFD:34
where, μmax is the maximum specific growth rate (/day), I is the PFD (μmol quanta/m2/s), I0 is the compensation PFD (μmol quanta/m2/s) and Ks is the PFD at μmax/2 (half-saturation light intensity). We conducted the experiments in triplicate. Data showing an obviously different trend from the other two replicates were excluded from calculations.
Effects of wavelength on the growth of Skeletonema costatum and Heterocapsa circularisquama
To assess the effects of light wavelengths on growth, we used nine wavelengths ranging from near-ultraviolet to near-infrared (405, 470, 505, 525, 568, 590, 623, 644 and 660 nm). Light was supplied by LEDs (Luxeon Star, San Jose, CA, USA) with relative spectral distributions as shown in Figure 1. The LEDs and their power supply were placed in the incubator. When S. costatum and H. circularisquama were incubated under the LED of each wavelength, the inoculum size was adjusted to be approximately 100 cells/mL in 10 mL (f/2 medium) because an initial bloom of H. circularisquama was observed from 92 cells/mL around Tategami of Ago Bay on 6 August 2004. The irradiance of each wavelength was 15 μmol quanta/m2/s and 75 μmol quanta/m2/s, respectively, on a 12:12 h L : D cycle (06.00–18.00 hours light). Fine adjustment of the irradiance was made by varying voltage to the LEDs. The growth rate was calculated from the increase in cell density during the exponential growth phase as described above (eqn 1).
Incubations with LEDs and simulated natural illumination
To investigate the possible use of LEDs for the suppression of H. circularisquama in the natural environment, we set up culture conditions with light by cool-white fluorescent lamps (PFD = 100 μmol quanta/m2/s; 12:12 h L : D cycle) as the multi-wavelengths, although the spectra between cool-white fluorescence lamps and sunlight might not be the same, with additional irradiance from a 590-nm LED (PFD = 70 μmol quanta/m2/s; 24-h illumination). For a control treatment, we used only the cool-white fluorescent lamps. Each species was inoculated to an initial cell density of approximately 100 cells/mL. Cell densities were counted daily (at 10:00 hours) and growth rates were calculated as previously described.
Absorption coefficients of Heterocapsa circularisquama and Skeletonema costatum
To measure the absorption coefficients (aph) of H. circularisquama and S. costatum incubated under a temperature of 25°C, a salinity of 30 psu and a light intensity of 100 μmol quanta/m2/s (12:12 h L : D cycle with fluorescent lamp) we used the quantitative filter technique (QFT35). Cells were collected in the late exponential phase by filtration onto 25-mm GF/F filters under a vacuum of less than 100 mmHg. The absorbances of the material on the filters were measured from 400 to 750 nm using a spectrophotometer (U-2001, Hitachi, Tokyo, Japan). The absorption coefficients were measured at three places on the filter to decrease the possible error from an uneven distribution of cells. The absorbance at 750 nm was subtracted to correct for the effect of back-scattering (reflection) because of suspended particulate material collected on the filter. We also measured the absorption coefficient of detritus, such as dead cells and cell walls, which remained after removing pigments from material on filters using 99.5% ethanol and subtracted these values to get a corrected aph. Chlorophyll-a (chl a) was measured by acetone extraction as suggested by SCOR-UNESCO36 and aph was determined as m2/mg chl a.
Effects of irradiance levels on growth of Skeletonema costatum and Heterocapsa circularisquama
The growth rates of both S. costatum and H. circularisquama increased with increasing irradiance levels under cool-white fluorescent lamps (Fig. 2). Fitting the data to a rectangular hyperbolic curve (eqn 2) yielded μ = 1.58(I − 1.00)/(I − 29.0) (r = 0.90) for S. costatum and μ = 1.36(I − 17.1)/(I − 90.8) (r = 0.98) for H. circularisquama. Maximum growth rate (μmax) and compensation PFD (I0) were 1.58/day and 1.00 μmol quanta/m2/s for S. costatum and 1.36/day and 17.1 μmol quanta/m2/s for H. circularisquama, respectively.
Effects of irradiance wavelength on the growth of Skeletonema costatum and Heterocapsa circularisquama
At an irradiance level of 15 μmol quanta/m2/s, S. costatum grew under wavelengths of 405, 470, 505, 525 and 568 nm, but not at 590, 623, 644 and 660 nm (Fig. 3), whereas H. circularisquama grew at wavelengths of 405, 470, 505 and 525 nm, but not at 568, 590, 623, 644 and 660 nm. At an irradiance level of 75 μmol quanta/m2/s, S. costatum grew at all wavelengths tested (Fig. 3). However, H. circularisquama only grew under wavelengths of 405, 470, 505, 525 and 660 nm and not at 590, 623 and 644 nm. Growth of H. circularisquama was suppressed at 590 nm (μ = −0.42/day). The maximum growth rate for both species was under 405-nm illumination: at 15 μmol quanta/m2/s irradiance, S. costatum and H. circularisquama had maximum growth rates of 0.61/day and 0.37/day, respectively; at 75 μmol quanta/m2/s irradiance, the maximum growth rates were 1.27/day and 0.83/day, respectively.
Effect of 590-nm LED on the growth of Skeletonema costatum and Heterocapsa circularisquama
At 590 nm, S. costatum grew at PFDs greater than 30 μmol quanta/m2/s and H. circularisquama grew at PFDs greater than 100 μmol quanta/m2/s (Fig. 4). Fitting the growth rates during exponential growth as a function of PFD using a rectangular hyperbolic curve (eqn 2) yielded μ = 1.41(I − 17.9)/(I + 13.2) (r = 0.95) for S. costatum and μ = 0.83(I − 77.9)/(I − 53.0) (r = 0.98) for H. circularisquama. The compensation PFD at 590 nm (I0−590) was 17.9 μmol quanta/m2/s for S. costatum and 77.9 μmol quanta/m2/s for H. circularisquama.
Incubations with LEDs and simulated natural illumination
In treatment groups with irradiance provided by cool-white fluorescent lamps (PFD = 100 μmol quanta/m2/s; 12:12 h L : D cycle) and a 590-nm LED (PFD = 70 μmol quanta/m2/s; 24-h illumination), the growth rate of S. costatum was 1.40/day (Fig. 5). This growth rate was similar to that of the control experiment with irradiance provided by cool-white fluorescent lamps only (12:12 h L : D cycle; Fig. 5). Growth rates of H. circularisquama in the treatment and control groups were 0.43/day and 0.63/day, respectively (Fig. 5). This substantial difference (P < 0.05; t-test) in growth rates might indicate that the additional irradiance at 590 nm decreased the growth rate of H. circularisquama.
Absorption coefficients of Heterocapsa circularisquama and Skeletonema costatum
The absorption spectra of S. costatum and H. circularisquama obtained using the QFT method are shown in Figure 6. The absorption coefficient aph of S. costatum was higher than that of H. circularisquama over most of the absorption spectra. The maximum aph was at 663 nm for S. costatum (aph = 0.031 m2/mg chl a) and at 435 nm for H. circularisquama (aph = 0.016 m2/mg chl a). The aph of S. costatum over the wavelengths emitted by the 590-nm LED (approximately 550–650 nm) ranged from 0.008 to 0.017 m2/mg chl a, whereas the aph of H. circularisquama over the same range was lower by a factor of 3, ranging from 0.003 to 0.006 m2/mg chl a.
At irradiance levels of 15 and 75 μmol quanta/m2/s, growth rates of S. costatum and H. circularisquama were highest under the 405-nm LED, which has a violet-blue color, compared with growth rates under LEDs of other wavelengths (Fig. 3). Moreover, although the 15 μmol quanta/m2/s of the 405-nm LED was lower than the I0 (17.1 μmol quanta/m2/s) of H. circularisquama under fluorescent lamps, H. circularisquama grew well under 405-nm. The growth rate under fluorescent light at 15 μmol quanta/m2/s calculated for S. costatum from the hyperbolic equation (Fig. 2) was 0.50/day, which is lower than the value measured at 15 μmol quanta/m2/s under the 405-nm LED (0.63/day).
Few studies have examined the effects of violet-blue light on the growth of algae, although many studies have investigated the effects of blue light (which is near violet-blue light) on the growth of algae. For example, growth of Chlorophyceae, Pheophyceae and Rhodophyceae was stimulated under blue light.1,11,37 Growth of the diatoms Cyclotella nana, Thalassiosira pseudonana15 and Haslea ostrearia9 was also markedly stimulated under blue light. Moreover, the blue light caused major chlorophyll increases in some diatoms compared with white light.5 However, the growth of benthic diatoms and the dinoflagellate Heterocapsa pygmaea was not stimulated by blue light.12,38,39 Nielsen and Sakshaug7 and Tremblin et al.40 also reported that blue and blue-green light did not affect the growth of S. costatum. Our study indicated that the growth rate of S. costatum under blue light (470 nm) was not different from that under cool-white fluorescent lamps, although H. circularisquama showed stimulated growth under blue light. These results suggest that growth rates of S. costatum might be stimulated by near-violet wavelengths. Thus, in large-scale culture of S. costatum, for example, as food for fish or juvenile bivalves, violet-blue light LEDs may be more effective in stimulating growth than sunlight or fluorescent lamps. However, it is difficult to explain why the growth of S. costatum under near-violet wavelengths is stimulated because of no physiological data of S. costatum under near-violet light. Thus, physiological research (i.e. variation in pigment composition and photosynthesis) of S. costatum and H. circularisquama under each wavelength is necessary.
At irradiance levels of 15 μmol quanta/m2/s neither species grew at wavelengths higher than 568 nm. At 75 μmol quanta/m2/s, S. costatum was able to grow at all wavelengths tested, but H. circularisquama grew at wavelengths below 568 nm and at 660 nm. At 590 nm (yellow light) in particular, the growth of S. costatum (μ = 0.96/day) was completely different from that of H. circularisquama (μ = −0.42/day). This result is similar to that obtained for the benthic diatom Haslea ostrearia, which also grew well under yellow light at a PFD of 100 μmol quanta/m2/s, although the growth rate was lower than that under green, red and far-red light.9 The growth rate of H. circularisquama in an experimental treatment group with irradiance provided by both cool-white fluorescent lamps (12:12 h L : D cycle) and a 590-nm LED (continuous irradiance) was lower than that of the control group with irradiance provided only by cool-white fluorescent lamps (12:12 h L : D cycle). The decreased growth rate of H. circularisquama in the 590-nm LED continuous irradiance might have happened because of the effects of the 590 nm finding from our study and the continuous irradiance. As far as continuous irradiance is concerned, Brand and Guillard41 reported that most species from oceanic regions are harmed by continuous irradiance, although species of coastal regions are insensitive to continuous irradiance. Heterocapsa circularisquama may be one of the species that grows poorly under continuous irradiance despite being commonly observed in semiclosed bays, such as Ago Bay, Uranouchi Bay and Hiroshima Bay, Japan.25 In contrast, Gilstad et al.42 observed that growth of S. costatum was higher under continuous irradiance than under a 12-h day length, although growth of S. costatum was not different between the two groups in our study. The different results between Gilstad et al.42 and our study might result from different physiological conditions of the isolated strains.43 However, under a 590-nm LED, growth of H. circularisquama recovered when irradiance levels exceeded 75 μmol quanta/m2/s (Fig. 4). This indicates that irradiance at 590 nm does not suppress the growth of H. circularisquama. The absorption coefficient (aph) of H. circularisquama at approximately 590 nm was low, but not zero (Fig. 6), indicating that H. circularisquama was able to grow under higher irradiance levels at 590 nm.
In the Ago Bay, a rugged ria shoreline on the Pacific coast of Japan, decades of pearl culture have caused a gradual degradation of water quality. Today, mass die-offs of pearl oysters because of toxic blooms of dinoflagellate algae threaten the century-old aquaculture industry in this area.26,27,44,45 In particular, the noxious dinoflagellate H. circularisquama, which specifically damages shellfish, bloomed in 1992 for the first time in Ago Bay killing 30–60% of the pearl oysters in the nurseries.26 This alga has spread in the sea along the coast of western Japan and now forms red tides nearly every year.28,46 Cultivation tests have shown that pearl oysters are killed at lower densities of H. circularisquama cells than other noxious dinoflagellates.27,28 Thus, the use of 590-nm LEDs in pearl culture at irradiance levels less than 75 μmol quanta/m2/s might allow for the growth of the useful diatom S. costatum, which is a dominant diatom in Ago Bay, without stimulating growth of the harmful dinoflagellate H. circularisquama.
This study was supported by the Environmental Restoration Project on Enclosed Coastal Seas under the CREATE (Collaboration of Regional Entities for the Advancement of Technological Excellence) Program of the Mie Industry and Enterprise Support Center (MIESC) in Japan and a Korea Research Foundation Grant funded by the Korean Government (MOEHRD) (KRF-2006-353-C00060).