INTRODUCTION
Dinoflagellates are ubiquitous and a major component of marine ecosystems (Jacobson and Anderson 1996, Glibert et al. 2012, Jeong et al. 2013, 2021, Sellers et al. 2014, Lee et al. 2021). They play diverse roles as primary producers, prey, predators, symbiotic partners, and parasites in marine ecosystems (Coats 1999, Jeong et al. 2010, 2012, Davy et al. 2012, Park et al. 2013, Spilling et al. 2018, Ok et al. 2021). Marine dinoflagellates are known to produce diverse useful materials, such as omega-3, amino acids, pigments, and bioluminescent materials (Wynn and Ratledge 2005, Sugawara et al. 2007, Valiadi and Iglesias-Rodriguez 2013, Onodera et al. 2014, Jang et al. 2017, Lim et al. 2018, 2020, Park et al. 2021b). However, obtaining pure cultures on a large scale is difficult, which may limit research and commercial utilization (Gupta et al. 2015, Jerney and Spilling 2018). Thus, the development of an automatic or semi-automatic system for cultivating useful dinoflagellate species on a large scale is required.
Noctiluca scintillans is a heterotrophic dinoflagellate that often causes red tides (Kiφrboe and Titelman 1998, Miyaguchi et al. 2006, Harrison et al. 2011, Zhang et al. 2016). This dinoflagellate is transparent, but becomes red or orange in color after digesting prey cells in the protoplasm (Turkoglu 2013). This species can feed on diverse prey, including phytoplankton, heterotrophic protists, and eggs of metazoans (Kimor 1979, Kiφrboe and Titelman 1998, Zhang et al. 2016). Cells of N. scintillans are bioluminescent (Sweeney 1971, Tett 1971, Buskey et al. 1992). Thus, N. scintillans often causes a red ocean during the day and a glowing ocean at night. There have been many studies on the red tides and bioluminescence of N. scintillans (Sweeney 1971, Tett 1971, Uhlig and Sahling 1990, Buskey 1995, Valiadi and Iglesias-Rodriguez 2014, Valiadi et al. 2019). Red-tide patches containing N. scintillans provide a large-scale bright bioluminescence field when bioluminescent cells in the patches are hit by waves, swimming fish and mammals, moving boats, ships, and submarines (Tarasov 1956, Bityukov 1971, Hastings 1975, Morin 1983, Williams and Kooyman 1985, Rhor et al. 1998). Thus, many scientists wish to conduct research on large-scale cultures of N. scintillans.
While it is easy to cultivate N. scintillans at scales <1 L, large-scale cultivation (>100 L) is challenging. Cells of N. scintillans are almost spherical and measure 200–1,000 μm in length, making it the largest known dinoflagellate (Nakamura 1998, Löder et al. 2012). However, this dinoflagellate swims slowly and usually stays near the surface (Kiφrboe and Titelman 1998, Johnson and Shanks 2003). In large tanks, some prey (e.g., diatoms and eggs) tend to sink and accumulate at the bottom of the tanks, while others (e.g., nano-, micro-, and dinoflagellates) swim fast enough to escape being eaten by N. scintillans. The highest reported growth rate of N. scintillans is approximately 0.8 d−1 (Tada et al. 2004, Stauffer et al. 2017). However, the spatial separation between N. scintillans and its prey cells in large tanks inhibits the rapid growth of N. scintillans. In large tanks, dense prey should be supplied to surface waters where N. scintillans cells stay, or the waters where N. scintillans cells and prey cells co-exist should be mixed. Thus, it is necessary to develop an automatic system for cultivating N. scintillans on a large scale while overcoming these shortcomings.
In the present study, to obtain mass-cultured N. scintillans cells, we developed an automatic system for cultivating them on a 100-L scale. We harvested 100-L dense culture of N. scintillans every 10 d using this system, and this culture was used for bioluminescence experiments. The results of this study provide insights into scaling up the volume of pure cultures of N. scintillans and other heterotrophic dinoflagellates to tens or hundreds of tons.
MATERIALS AND METHODS
Cultures of Noctiluca scintillans and its prey Dunaliella salina
Cells of N. scintillans NSDJ2010 were isolated from plankton samples collected from surface waters off the coast of Dangjin, Korea, when the water temperature and salinity were 19.5°C and 28.8, respectively. A clonal culture was established using two serial single-cell isolations and was grown on Dunaliella salina DSJH1710 (approximately 80,000 cells mL−1) in six-well plates. Cells of D. salina DSJH1710 were originally isolated from the coastal waters of Jinhae, Korea. As the abundance of N. scintillans increased, the cells were transferred to 50-mL and 250-mL plate flasks and 1-L polycarbonate (PC) bottles containing fresh prey. The 1-L PC bottles were placed on a rotating wheel (0.00017 ×g) and incubated at 20°C under 20 μmol photons m−2 s−1 illumination provided by light-emitting diode (LED) lamps with a 14 : 10 h light : dark cycle. The culture of D. salina was placed on a shelf and maintained in 800-mL culture flasks containing a f/2-Si medium (Guillard and Ryther 1962).
Design of an automatic system of cultivating Noctiluca scintillans on a 100-L scale
To develop an automatic system for cultivating N. scintillans on a 100-L scale, the optimal conditions, such as the optimal prey species supporting high growth rate, time intervals and volumes at which nutrients in the nutrient tank were supplied to the prey in the prey culture tank, time intervals and volumes at which the prey in the prey culture tank were supplied to N. scintillans in the predator culture tank, and light intensities supporting high growth rates of the prey and predators were investigated. Furthermore, the results of our previous studies on the development of a semi-continuous system of cultivating mixotrophic dinoflagellates on a 10-L scale were used to develop an automatic system for cultivating N. scintillans on a 100-L scale (Jeong and Lim 2020, Lim et al. 2020). In the automatic system for cultivating N. scintillans on a 100-L scale, a stirring subsystem was set up for mixing.
To establish an automatic system for cultivating N. scintillans on a 100-L scale, we developed hardware for cultivation and a system controller for controlling the entire system. With respect to hardware, we chose 200-L round PC water tanks; LED lamps for illumination for prey and predators; spectrometers for measuring light intensities; magnetic pumps for supplying large volumes of nutrients and prey as main pumps; diaphragm pumps for supplying small volumes of nutrients and prey as support pumps; sensors for measuring water temperature, salinity, pH, and dissolved oxygen (DO); air pumps; spargers; stirrers; low-speed motors mounted on each stirrer; weighing systems; and polyvinyl chloride or silicone pipes (Fig. 1).
In this study, a software program was developed for use with the system controller (Fig. 2). The program was designed to control and monitor the hardware of the automatic system. The program contained parts controlling temperature, light intensity, and light : dark cycling for prey and predators, as well as opening and closing of valves between the two tanks. Furthermore, it controlled the speed, time interval, and duration of the stirring subsystem; air volume, time interval, and duration of aeration; and time intervals of numerical data saving. Moreover, it monitored temperature, light intensities, light : dark cycle, elapsed time, functions operating in each tank, water quality of prey and predator cultures, volumes in the prey and predator culture tanks, and current system mode (i.e., automatic or manual mode).
Cultivating Noctiluca scintillans on a 100-L scale using the automatic system
The newly developed automatic system for cultivating N. scintillans on a 100-L scale was operated by providing D. salina as prey. After several preliminary tests, we chose 20°C as the optimal temperature, 20 μmol photons m−2 s−1 for the optimal light intensity for N. scintillans and 300 μmol photons m−2 s−1 for D. salina under a 14 : 10 h light : dark cycle.
At the beginning of the experiment, 50 L of f/2-Si medium was added to the nutrient tank, 90 L of D. salina culture (approximately 120,000 cells mL−1) to the prey culture tank, and 30 L of N. scintillans culture (approximately 25 cells mL−1) to the predator culture tank. Every 2 d, the valves between the prey and predator culture tanks were opened to allow the prey to be transported, and subsequently the valves between the nutrient and prey culture tanks were opened to allow the medium to be transported. Every day, 10-mL aliquots were taken from the waters in the prey and predator culture tanks and fixed with 5% acidic Lugol’s solution. In the prey culture tank, subsampling was performed both before and after the f/2-Si medium was added. Subsampling was also performed both before and after the prey culture was transported to the predator culture tank. All or >200 cells of each species were counted in three 1-mL Sedgwick-Rafter chambers using inverted microscopes of 40× or 100× magnification. This procedure was repeated three times.
The specific growth rate of N. scintillans (μ, d−1) was calculated as follows:
, where Ct1 and Ct2 indicate the cell concentrations at time points t1 (t = 0 d) and t2 (t = 1 d), respectively. The growth rates of N. scintillans were measured every day for 6 d during each experimental cultivation period.
The prey removal rate (%) of N. scintillans was calculated as follows:
, where Pt1 represents D. salina abundance (cells mL−1) immediately after the culture was added to the predator culture tank, and Pt2 represents the remaining D. salina abundance (cells mL−1) 2 d after the culture was added to the predator culture tank. The prey removal rates of N. scintillans were measured every 2 d for 6 d in each experimental cultivation period.
During the experimental cultivation periods, salinity, pH, and DO were monitored but not controlled. All cultivation periods were conducted under the same conditions as described above.
The bioluminescence capability of N. scintillans cultivated using the automatic mass cultivation system was tested when the volume of N. scintillans culture was 90 L in the predator culture tank. The N. scintillans culture was mechanically stimulated by placing the predator culture tank in the dark for 3 h. It was then stimulated using three air pumps (10 L min−1) for aeration, and three spargers for even air distribution. During mechanical stimulation, bioluminescence images were captured using a Canon EOS R with a Canon RF 50 mm F1.2L lens (Canon, Tokyo, Japan) at aperture F2.2, shutter speed 8 s, and ISO200.
RESULTS
Developed hardware of the automatic system for cultivating Noctiluca scintillans on a 100-L scale
The hardware components of the automatic system were successfully established based on the design and results of several trials (Fig. 3). The hardware consisted of four 200-L PC water tanks; eight full spectrum LED lamps (maximum 1,000 μmol photons m−2 s−1; Full spectrum LED lamp; Yunlighting, Namyangju, Korea); four spectrometers (UM-2280; OtO Photonics Inc., Hsinchu, Taiwan); four magnetic pumps (NH-100PX; Panworld Co., Ltd., Tokyo, Japan); four diaphragm pumps (KM212; Cheonsei Co., Ltd., Ansan, Korea); four sensors for measuring water temperature (°C), salinity, pH, and DO (Aqua TROLL 500; In-Situ Inc., Fort Collins, CO, USA); four air pumps (maximum 10 L min−1); four spargers; four stirring subsystems each consisting of a stirrer equipped with a low-speed motor (maximum 100 rpm); four weighing systems (HBS-200, CI-600; CAS Co., Ltd., Yangju, Korea); and polyvinyl chloride or silicone pipes (Fig. 3). A temperature-controlled walk-in chamber enclosing all hardware parts could maintain a temperature between 15 and 25°C.
When the valves between the prey and predator culture tanks were opened, 20 L of the prey culture in the prey culture tank was transported to the predator culture tank at a rate of 10 L min−1. Subsequently, when the valves between the nutrient tank and prey culture tank were opened, 20 L of the nutrient medium in the nutrient tank was transported to the prey culture tank at a rate of 10 L min−1. Prior to opening the valves, the culture in the prey culture tank was mixed using a stirrer at a rate of 5 rpm for 2 min to homogenize prey abundance. This minimized the difference in the abundances of the remaining and transported prey.
Cells of D. salina tended to stay near the bottom of the prey culture tank, and thus, mixing was necessary. To mix the culture in the prey culture tank, air was supplied to the bottom of the tank at a rate of 9 L min−1 using an air pump and a sparger for 10 min twice per day. Additionally, the culture was mixed using a stirrer at a rate of 5 rpm for 10 min twice per day. Mixing was also required to prevent the predator cells from being separated from the prey cells in the predator culture tank. To mix the culture in the predator culture tank, air was supplied to the bottom of the tank at a rate of 9 L min−1 using an air pump and a sparger for 10 min twice per day. The culture was mixed using a stirrer at a rate of 5 rpm for 10 min every 2 h.
Developed software of the automatic system for cultivating Noctiluca scintillans on a 100-L scale
The system controller program was designed to control and monitor the hardware functions (Fig. 4). The program can automatically control the temperature in the chamber as well as various functions of the hardware (light source operation, mix operation, pump operation, and aeration operation), and save numerical data (volume, light intensity, water temperature, salinity, pH, and DO in each tank) measured in real time at a set time interval. In addition, it is possible to monitor the current system operation status (automatic or manual mode), incubation period (elapsed time), and environmental changes of cultures in each tank (water temperature, salinity, pH, DO, and light intensity) in real time (Fig. 4).
Operation of the automatic system for cultivating Noctiluca scintillans on a 100-L scale
Before the beginning of the operation, 50 L of f/2-Si medium was added to the nutrient tank, 90 L of D. salina culture (approximately 120,000 cells mL−1) to the prey culture tank, and 30 L of N. scintillans culture (approximately 25 cells mL−1) to the predator culture tank. When the automatic system for cultivating N. scintillans on a 100-L scale was operated in automatic mode, the hardware and software of the system controller worked as follows. A predetermined volume (20 L) of D. salina culture in the prey culture tank was transported to the predator culture tank. The prey culture tank was then refilled with 20 L of nutrient medium in the nutrient tank. The prey culture tank was maintained at a constant volume (90 L). Subsequently, in the predator culture tank, the volume of the N. scintillans culture gradually increased with the addition of D. salina culture. Thus, when N. scintillans culture in the predator culture tank reached a predetermined volume (90 L), some of the volume (70 L) was transported to the predator storage tank. And then, 10 L of 0.2-μm filtered sea water was manually added to the remaining N. scintillans culture (20 L) in the predator culture tank. Thus, the N. scintillans culture in the predator culture tank became 30 L and was semi-continuously cultivated by repeating the above process using this automatic mass cultivation system. Seventy liters of N. scintillans culture in the predator storage tank was maintained with adding 30 L of D. salina culture.
The transport of the nutrient medium, D. salina culture, and N. scintillans culture was conducted through two-stage pumping systems. All transportation was performed automatically at set times. In addition, N. scintillans and D. salina cultures were mixed for a set time (2 min) before being transported and continuously mixed during transport to the next tank using stirrers equipped with a low-speed motor at a predetermined speed (5 rpm).
Meanwhile, to suspend cells of D. salina in the prey culture tank, the culture was aerated with a predetermined air volume (9 L min−1) for a set time (10 min) and at a set time interval (twice every day) using an air pump and a sparger; and mixed with a stirrer at a predetermined speed (5 rpm) for a set time (10 min) and at a set time interval (twice every day). In addition, to prevent spatial separation of D. salina cells and N. scintillans cells by depth in the predator culture tank, the culture was aerated with a predetermined air volume (9 L min−1) for a set time (10 min) and at a set time interval (twice every day) using an air pump and a sparger; and mixed with a stirrer at a predetermined speed (5 rpm) for a set time (10 min) and at a set time interval (once every 2 h).
Each tank was irradiated with predetermined light intensities, 20 μmol photons m−2 s−1 for N. scintillans and 300 μmol photons m−2 s−1 for D. salina, for a set duration (a 14 : 10 h light : dark cycle). The light intensities were measured using spectrometers, and monitored and controlled by the program in real time.
In addition, the elapsed time in automatic mode, water quality of cultures in each tank (water temperature, salinity, pH, and DO), and liquid volumes of each tank were measured in real time.
Cultivating Noctiluca scintillans on a 100-L scale using the automatic system
Among the tested dinoflagellates Lingulodinium polyedra, Scrippsiella acuminata, Prorocentrum cordatum, Amphidinium carterae, Heterocapsa steinii, Alexandrium pacificum, and A. minutum (previously A. lusitanicum), and the chlorophyte D. salina in screening tests using 6-well plates, D. salina was selected as the optimal prey to support the high growth rate of N. scintillans.
The light intensity supporting the high growth rate of D. salina was 300 μmol photons m−2 s−1 in a 14 : 10 h light : dark cycle, whereas that supporting the high growth rate of N. scintillans was 20 μmol photons m−2 s−1 in a 14 : 10 h light : dark cycle. LED lamps were used because they provided the target light intensity for a long time, and the heat from the LED lamp was lower than that from other light sources such as fluorescent lamps and halogen lamps.
To determine the volume of the f/2-Si medium transported from the nutrient tank to the prey culture tank, as well as the volume of prey culture transported from the prey culture tank to the predator culture tank, the growth rate of D. salina and the growth and ingestion rates of N. scintillans feeding on D. salina were measured in our preliminary test. The growth rate of D. salina in a stationary stage was 0.1–0.2 d−1 and the growth and ingestion rates of N. scintillans feeding on D. salina were 0.6 d−1 and 2,725 cells predator−1 d−1, respectively. Furthermore, when N. scintillans was manually cultivated with D. salina, D. salina cultures with an abundance of >100,000 cells mL−1 were required to maintain the abundance of N. scintillans at 10–20 cells mL−1. Thus, in this automatic system, the abundance of D. salina in the prey culture tank was maintained at >100,000 cells mL−1. At the beginning of the 6-day cultivation experiment, 20 L of 90 L D. salina culture (approximately 120,000 cells mL−1) in the prey culture tank was transported to 30 L of N. scintillans culture (approximately 25 cells mL−1) in the predator culture tank. Subsequently, 20 L of D. salina culture in the prey culture tank was transported to the predator culture tank every 2 d until N. scintillans culture in the predator culture tank became 90 L.
This cultivation experiment was repeated three times (periods 1–3). At the beginning of each period (day 0), the abundance of D. salina in the prey culture tank containing 90 L of prey culture was 111,500–122,330 cells mL−1 in periods 1–3 (Fig. 5). Ten minutes later, 20 L of D. salina culture from the prey culture tank was automatically transported to the predator culture tank. Subsequently, 20 L of f/2-Si medium in the nutrient tank was added to the prey culture tank containing 70 L prey culture (on day 0.1); thus, the total volume of the prey culture became 90 L. Thus, due to dilution, the abundance of D. salina in the prey culture tank decreased to 86,660–96,500 cells mL−1 in periods 1–3. Two days later (day 2), D. salina in the prey culture tank grew, and its abundance was 113,660–126,330 cells mL−1 in periods 1–3. This process was repeated on days 2–4 and days 4–6.
At the beginning of each period (day 0), the abundance of D. salina in the predator culture tank containing 30 L predator culture was 100–170 cells mL−1 in periods 1–3 (Fig. 6). Ten minutes later, 20 L of D. salina culture in the prey culture tank was automatically transported to the predator culture tank (on day 0.1). The abundance of D. salina in the predator culture tank immediately after the prey culture was added was 44,000–47,800 cells mL−1 in periods 1–3. Most D. salina cells in the predator culture tank were eaten by N. scintillans cells 2 d later. On day 0, the abundance of N. scintillans in the predator culture tank containing 30 L predator culture was 25–33 cells mL−1 in periods 1–3 (Fig. 6). After 20 L of D. salina culture was transported to the predator culture tank (on day 0.1), the abundance of N. scintillans in the predator culture tank immediately after the prey culture was added was 10–12 cells mL−1 in periods 1–3. Cells of N. scintillans grew well on D. salina and the abundance of N. scintillans was 16–26 cells mL−1 on day 2 in periods 1–3. This process was repeated on days 2–4 and days 4–6. The maximum abundance of N. scintillans in the predator culture tank was 45 cells mL−1 during periods 1–3.
The average and maximum growth rates of N. scintillans in periods 1–3 were 0.29–0.35 and 0.62–0.90 d−1, respectively. The prey removal rate (%) by N. scintillans 2 d after prey cultures were added was 98–99% in periods 1–3 (Fig. 6).
Several single prey cells were observed inside the protoplasm of N. scintillans cells 1 h after the addition of prey cells, while packages of aggregated prey cells were observed 6 h later (Fig. 7). Most N. scintillans cells contained single prey cells or packages of aggregated prey cells 48 h later.
Cells of N. scintillans fed on D. salina in the predator culture tank produced bright bioluminescence after mechanical stimulation (Fig. 8).
DISCUSSION
In the present study, the newly developed automatic system for cultivating N. scintillans could continuously produce 100 L of dense N. scintillans culture every 10 d. Prior to the present study, there were no reports on the development of automatic systems for cultivating N. scintillans. Thus, the newly developed automatic system for cultivating N. scintillans will enable scientists to conduct diverse experiments that require large volumes of N. scintillans. We are currently conducting experiments to test whether the bioluminescence of N. scintillans in several 10-L tanks can be detected remotely using drones or aircraft. For these experiments, several hundred liters of pure N. scintillans culture are needed simultaneously, and the newly developed automatic system provides this.
In the present study, the highest abundance of cultivated N. scintillans using the newly developed automatic system was 45 cells mL−1. In our previous experiments, the highest abundance of N. scintillans when manually cultivated was <20 cells mL−1. When the newly developed automatic system was used, 98–99% of prey cells in the predator culture tank were eliminated by N. scintillans after 2 d incubation. Furthermore, the maximum growth rate of N. scintillans in the predator culture tank in periods 1–3 was 0.62–0.90 d−1, which was similar or higher than that of N. scintillans when manually cultivated (Jeong 1995, Tada et al. 2004, Stauffer et al. 2017). Thus, supplying a designated amount of prey culture at a designated interval may help N. scintillans feed on prey cells effectively and support high maximum growth rates of N. scintillans. This newly developed automatic system for cultivating N. scintillans can significantly reduce labor costs, as only 1–2 people may be needed to produce several tons of N. scintillans culture.
Mixing cultures using air pumps and spargers for air supply and stirrers prevents predator cells from separating from the prey cells in the predator culture tank. However, strong mixing often causes high turbulence, which can inhibit the growth of dinoflagellates (Thomas and Gibson 1990, Berdalet 1992). Therefore, the rates, durations, and time intervals of air supply or stirring at which cultures are well mixed without harming prey and predator cells should be standardized. The high growth rates of N. scintillans on D. salina in the predator culture tank indicates that the rates, durations, and time intervals of air supply using an air pump and a sparger and mix using a stirrer selected in this study were suitable and safe.
The light intensity illuminating the predator culture tank was selected to be 20 μmol photons m−2 s−1, whereas that illuminating the prey culture tank was 300 μmol photons m−2 s−1. High light intensity sometimes causes active swimming of prey cells; thus, it is critical to choose a light intensity at which prey cells survive, but swim slowly, and are easily eaten by predator cells. The light intensity of 20 μmol photons m−2 s−1 was ideal for the survival and weak swimming of D. salina and active ingestion by N. scintillans. A light intensity of 20 μmol photons m−2 s−1 was used in the experiments on feeding by other heterotrophic dinoflagellates, such as Gyrodinium dominans, Oxyrrhis marina, Polykrikos kofoidii, and Stoeckeria changwonensis (Lee et al. 2014, Lim et al. 2014, Jeong et al. 2018, You et al. 2020, Eom et al. 2021, Park et al. 2021a).
This 100-liter-scaled automatic system can be used as a unit, and several units can be combined to obtain a larger volume of N. scintillans cultures. Furthermore, a one-ten- or one-hundred-ton-scaled automatic system can be established based on this unit.
The newly developed automatic system for cultivating N. scintillans on a 100-L scale can be applied to cultivate other heterotrophic dinoflagellates or mixotrophic dinoflagellates. However, prey species supporting a high growth rate of a target heterotrophic or mixotrophic dinoflagellate should be selected. Furthermore, the growth rates of the optimal prey and also the target predator feeding on the prey should be determined. Based on the growth rates of the prey species and target dinoflagellate, the time interval and duration at which the valves between the nutrient tank and the prey culture tank, or between the prey culture tank and the predator culture tank open should be modified. Furthermore, suitable rates, durations, and time intervals of air supply using an air pump and a sparger, and those of mixing using a stirrer supporting the high growth of the predators in the predator culture tank should be chosen. Moreover, light intensities that support high growth of predators should be selected. The conditions for the target predator and its prey can be set up by changing the parameters in the program of the system controller in this study.