1 INTRODUCTION

The first record of a Phaeocystis bloom in China was from Zhelin Bay, South China Sea in the fall of 1997, covering over 3 000 km² and lasting for about 6 months (Chen et al., 1999). The bloom resulted in massive fish mortalities, and severely devastated the net-cage fish aquaculture in the Fujian and Guangdong Provinces. It was estimated that the direct economic loss was more than 180 million RMB (about 26 million US dollars; Wang et al., 2007). A further study confirmed that the causative species could produce hemolytic toxins, which were probably responsible for the massive fish kills (He et al., 1999). The causative organism was identified as Phaeocystis globosa rather than Phaeocystis pouchetii or other species, based on microscopic morphological observations, culturing behavior, physiological characteristics and sequence analysis of its partial 18S rDNA (Shen et al., 2000; Wang et al., 2000; Chen et al., 2002; Xu et al., 2003; Qi et al., 2004). During the following decade, P. globosa blooms were frequently reported in Chinese coastal waters throughout the year, involving extensive areas from the Bohai Sea (the North China Sea) to the coast of the Hainan Province (the South China Sea), causing heavy damage to local fisheries (Wang et al., 2007). Owing to its frequent occurrence, broad spread and severe damage to fisheries, tourism and marine ecosystems in the past decade, P. globosa is thought to be one of the most important harmful algal bloom causative species in China (Qi et al., 2004; Wang et al., 2007; Liu et al., 2010).

In addition to temperate and sub-tropical coastal waters like the southeastern coastal waters of China, Phaeocystis blooms have also been frequently recorded in Polar Regions, the North Sea and the Arabian Sea (Garrison et al., 1987; Lancelot et al., 1987; Al-Hasan et al., 1990; Schoemann et al., 2005). It is considered to be an ecosystem disruptive algal bloom (EDAB) species (Sunda et al., 2006), endangering marine ecosystems and being hazardous to human beings. Furthermore, Phaeocystis is also a major producer of sulphur compounds (dimethyl sulphide) and thus is treated as a special phytoplankton group in Ocean Biogeochemical Climate Models (OBCM), which address the role of marine phytoplankton and the 'biological pump' in the oceanic biogeochemical cycles (Schoemann et al., 2005). As well as the different Phaeocystis species, there is also increasing evidence that the most widespread species, P. globosa, can be subdivided into at least five different ecotypes, which differ in their ecophysiological properties (Riegman and van Boekel, 1996). Most research has been performed on the P. globosa ecotype North European (English Channel/North Sea isolates) (Riegman and van Boekel, 1996). It is noteworthy that the ecological features of P. globosa populations occurring in the South China Sea are quite different from those in other geographic areas. For example, it was reported that water temperature reached 30℃ during a P. globosa bloom in coastal regions of the Guangdong Province in July 1999 (Xu et al., 2003), while suitable temperatures for other strains are below 16℃ (Jahnke and Baumann, 1987; Riegman and van Boekel, 1996). Furthermore, colony diameter in the Guangdong Province was up to 30 mm, much greater than previously reported (8–9 mm) (Jahnke and Baumann, 1987).

Harmful algal blooms (HABs) are often linked to nutrient enrichment and the consequent eutrophication of coastal waters (Nixon, 1995; Paerl, 1997; Anderson et al., 2002; Zhou et al., 2003; Heisler et al., 2008; Howarth, 2011). Although widespread eutrophication is increasing in global coastal waters, Phaeocystis blooms only occur in some euthrophic areas (Garrison et al., 1987; Lancelot et al., 1987; Al-Hasan et al., 1990; Chen et al., 1999), which also have large differences in other environmental factors. It is therefore necessary to study the ecological characteristics of this species, isolated from different geographic areas of the world, to explore their occurrence mechanisms and the potential developing trend of the widespread blooms. We hypothesize that P. globosa populations occurring in Chinese coastal waters may belong to a new ecotype, and possess a competitive advantage in their environmental adaptation over other phytoplankton species, and thus can dominate and form blooms. In this study, we examined the effects of temperature, salinity, irradiance and nutrients on the growth of a strain of P. globosa, isolated from the South China Sea, under nutrient-replete laboratory conditions.

2 MATERIAL AND METHOD 2.1 Culture and culturing conditions

The strain of P. globosa used in the study was isolated from Junk Bay, Hong Kong in 1999. Phylogenetic analysis of its partial sequence of 18S rDNA indicated that this strain is P. globosa (Chen et al., 2003).

Artificial seawater (salinity 30.5; Harrison et al., 1980) enriched with silicate-free f/2 culture medium (Guillard, 1975) was used as the culture medium. Before the experiments, the culture was maintained at 24℃ and an irradiance of 120 μmol/(m²∙s) in a 12-h:12-h (light:dark) photoperiod cycle.

2.2 Experiments for the effect of irradiance

The irradiance experiments were conducted in an artificial climate incubator (CC275TL2H, Hangzhou, China) following the method of Kim et al. (2004). The pre-culture lasted for 1 week at 24℃, and 120 μmol/(m²∙s) from cool-white fluorescent bulbs with a 12-h light: 12-h dark cycle. The culture was inoculated with ~5 000 cells/mL in triplicate, in 50-mL capped test tubes (ø25 mm×150 mm) containing 35 mL culture suspended in f/2 modified medium, and they were gently shaken twice daily. Eleven different irradiance levels (2, 4, 7, 15, 30, 60, 90, 110, 140, 170 and 230 μmol/(m²∙s)) were obtained by wrapping the tubes with UV-absorbing vinyl screens. A quantum light meter (Model LI-190SA; LI-COR Biosciences, Lincoln, NE, USA) was used to confirm the irradiance levels, by wrapping the receiver sensor with UV-absorbing vinyl screens. The experiment continued for one to two weeks. Every 1–2 days, an aliquot of 100 μL of culture was sampled from each test tube and cell numbers were determined using a Sedgwick-Rafter counting chamber under a compound light microscope. A fluorometer (Model TD-700; Turner Designs Co., CA, USA) was used to determine in vivo chlorophyll a. The specific growth rates (μ, /d) of cultures in the exponential growth phase were calculated according to the method of Guillard (Guillard, 1973), by the least-squares fit of a straight line to the growth data logarithmically transformed.

Equation 1, modified from Lederman and Tett (1981), was used to describe the relationship between growth rate and irradiance:

where μ is the specific growth rate (/day), μ_m the maximum specific growth rate (/day), I the irradiance (μmol/(m²∙s)), I₀ the compensation irradiance (μmol/(m²∙s)), and K_s the irradiance at μ_m/2 (halfsaturation light intensity).

2.3 Experiments for the effects of temperature and salinity

The growth experiments were conducted using a crossed factorial design with 42 different combinations of six temperatures (10, 15, 20, 24, 27, 31℃) and seven salinities (10, 15, 20, 25, 30, 35, 40) under a constant light intensity (about 120 μmol/(m²∙s)). The f/2 media with salinities < 30.5 were prepared by diluting the initial artificial seawater (salinity 30.5) with de-ionized water, and the f/2 media with salinities >30.5 were prepared by evaporating the initial artificial seawater to the desired salinities. To avoid salinity and temperature shocks, cultures were acclimated to the desired experimental conditions by stepwise transferring over 1–4 weeks according to the method outlined by Yamaguchi and Honjo (1989) and Kim et al. (2004). If the cells could not grow under an experimental regime, no further growth experiments were carried out at that particular combination of temperature and salinity, and the growth rate was regarded as zero.

For each experimental regime, acclimated stock cultures were inoculated into triplicate capped test tubes (ø25 mm×150 mm) containing 35 mL of modified f/2 medium (Guillard, 1975) without silicate. Growth rates during the exponential growth phase were calculated as described above.

2.4 Experiments for the combined effects of nutrient composition and irradiance

To clarify the combined effects of irradiance and nutrient composition on the growth of P. globosa, exponential-phase algal cells were inoculated into 50 mL capped test tubes (ø25 mm×150 mm) containing 35 mL culture suspended in modified f/2 medium with different DIN/DON ratios (1:4, 2:3, 3:2, 4:1). NO₃^- and urea were used as inorganic or organic nitrogen sources, respectively, and whole NO₃^- and whole urea treatments were used as controls. The total nitrogen was 300 μg/L. The experiment was conducted in an artificial climate incubator at 24℃ under both high (100 μmol/(m²∙s)) and low irradiance levels (5 μmol/(m²∙s)) with a 12-h light:12-h dark cycle, using the same method as outlined in Section 2.2. All treatments were conducted in triplicate and lasted 13 d. Prior to the experiment, P. globosa cultures were acclimated to the corresponding nitrogen composition and irradiance by at least three transfers. A fluorometer (Model TD-700; Turner Designs Co., CA, USA) was used to determine in vivo chlorophyll a everyday. The specific growth rates (μ, /d) of cultures was calculated according to the Eq.2.

where μ is the specific growth rate (/day), N_t the fluorescence value at time t, N₀ the initial fluorescence value, and T the incubation time.

2.5 Statistical analysis

The software SPSS 11.5 for windows (SPSS, Chicago, IL, USA) was used for the analysis of variance (ANOVA), to determine the effects of temperature and salinity on growth rate. Cubic polynomial equations were determined based on the ANOVA results (Yamaguchi and Honjo, 1989; Yamaguchi et al., 1991, 1997; Ellegaard et al., 1993; Kim et al., 2004; Nagasoe et al., 2006; Matsubara et al., 2007; Xu et al., 2010).

3 RESULT 3.1 Effect of irradiance on the growth of P. globosa

During the experiment, no colonies were found and P. globosa cells remained in the solitary flagellate morphology. A specific P. globosa growth rate of ~0.22/d was maintained at the lowest irradiance level of 2 μmol/(m²∙s), while the growth of this organism appeared to become saturated at 60 μmol/(m²∙s) with a growth rate 0.83/d (Fig. 1). Photo-inhibition was not observed at an irradiance of 230 μmol/(m²∙s), the maximum irradiance used in this study. Thus, the optimum irradiance for growth was ≥60 μmol/(m²∙s).

The following hyperbolic equation could be used to describe the exponential growth phase, according to the data shown in Fig. 1:

The compensation irradiance (I₀) was 0.5 μmol/(m²∙s). The maximum growth rate (μ_m) and half-saturating irradiance (K_s) were 0.87/d and 8.73 μmol/(m²∙s), respectively.

3.2 effect of temperature and salinity on the growth of P. globosa

Under experimental conditions, P. globosa existed as solitary flagellate cells. The dependence of the specific growth rate of P. globosa on temperature and salinity is plotted in Figs. 2–4. There was no growth potential under temperatures ≤15℃ at any salinity, while P. globosa was able to grow at temperatures between 20℃ and 31℃, depending on the salinity. A maximum growth rate of 0.80/d was obtained at the following combinations: 24℃ and 35, 27℃ and 40. In addition, specific growth rates >0.60/d occurred when the combinations of temperature and salinity were 20℃ to 27℃ and 25 to 40, respectively. The range of salinities conducive to growth tended to be narrower when temperatures were lower than 24℃ or higher than 27℃ (Fig. 4).

Results from two-way ANOVA indicated significant effects of temperature, salinity and the interaction of temperature and salinity on the growth rates of P. globosa (P < 0.001; Table 1). Of the total sum of squares, 38%, 37% and 24% were accounted for by the sum of squares for temperature, salinity and the temperature-salinity interaction, respectively. This indicated that, within the ranges of temperature and salinity applied in this study, the contributions of temperature and salinity to growth were approximately equal and there was a significant interaction between these two factors.

Based on the ANOVA results, a cubic equation could be fitted in a form as shown in Eq.4:

where μ is the specific growth rate, T the temperature, S the salinity, and b_nn the regression coefficients, fitted by means of the stepwise forward regression method (SPSS11.5). The multiple regression of the specific growth rate of P. globosa on temperature and salinity is shown as follows:

Indeed, in situ growth rates of alga were influenced by complicated components, including irradiance, nutrients, and grazing pressure. Given suitable light, nutrients and other conditions suitable for the growth of P. globosa, this regression equation provides a useful reference to estimate in situ growth rates of P. globosa based on available temperature and salinity data.

3.3 Combined effects of nutrient composition and irradiance on the growth of P. globosa

Our results showed that the growth rates of P. globosa were significantly higher at high light intensity than at low light intensity, suggesting that irradiance was one of the key factors influencing nitrogen utilization. More specifically, the specific growth rates (/d) of P. globosa increased with the increase in DIN percentage composition at high light intensity (Fig. 5a), and the maximum growth rate of 0.83/d was obtained in the 100% DIN treatment. At low light intensity (5 μmol/(m²∙s)), a maximum growth rate (~0.70/d) of P. globosa was observed in the whole-urea treatment. Statistical analysis showed that growth rate had a significant inverse relationship with increasing DIN percentage composition (Fig. 5b). Our results suggested that P. globosa preferred organic nitrogen when ambient light intensity was low.

4 DISCUSSION

Eutrophication and other anthropogenic alterations have been linked to global increases in HAB frequency and intensity in the past few decades (Paerl, 1997; Anderson et al., 2002; Zhou et al., 2003; Heisler et al., 2008; Fu et al., 2012). Phaeocystis globosa has been one of the most important HAB causative species, causing severe damage to fisheries, tourism and aquatic ecosystems in Chinese coastal waters, and arousing great public attention (Chen et al., 1999; Wang et al., 2007; Liu et al., 2010). Our laboratory experiments also revealed that P. globosa required mid-nitrogen levels (half saturation constant of NO₃^-, 1.71 μmol N/L) and high phosphorus levels (half saturation constant of PO34ˉ, 1.08 μmol P/L), and was adapted to eutrophic waters (unpublished data, authors). Phaeocystis globosa blooms usually occurs at low to medium temperatures in coastal waters (Riegman and van Boekel, 1996; Schoemann et al., 2005). However, in China, a bloom spread from the Bohai Sea to the South China Sea, bloomed in different seasons and lasted a long period, particularly in the South China Sea (Wang et al., 2007). The population in Chinese coastal waters must therefore have unique ecological characteristics that allows it to adapt to variable environments. From our laboratory experiments, we confirmed that the P. globosa strain from the South China Sea not only prefers higher temperature and higher salinity, but is also adapted to lower irradiance levels.

Irradiance has been suggested to influence cell morphology of P. globosa; flagellate cells are better competitors for light than colonial cells, owing to their superior uptake characteristics (Riegman and Boekel, 1996). During the irradiance experiments, P. globosa was present as solitary cells. Our study showed the compensation irradiance (I₀) of P. globosa was 0.5 μmol/(m²∙s), which is similar to Karenia mikimotoi (Yamaguchi and Honjo, 1989), Prorocentrum donghaiense (Xu et al., 2010), and Pseudo-nitzschia pungens (author, unpublished) (Table 2), which are representative HAB causative species in the coastal waters of China in the spring. The saturating irradiance (I_s) of P. globosa was 60 μmol/(m²∙s), while the saturating irradiances for most HAB species were found to be higher than 90 μmol/(m²∙s). This relatively lower I_s suggested that P. globosa may become dominant under turbid environments, where solar irradiance is attenuated rapidly (Table 2). However, photo-inhibition was not observed at the highest irradiance used in this study (230 μmol/(m²∙s)), suggesting a wide tolerance range of irradiance for P. globosa.

Field investigations indicated that Phaeocystis species could survive from the sea surface to 150-m-depth water layer, exhibiting its extraordinary flexibility in response to light environments (El-Sayed et al., 1983). Further studies have indicated that it could adapt to a wide irradiance range of 16–1 600 μmol/(m²∙s) (Palmisano and Sullivan, 1985). It has been reported that Phaeocystis can absorb blue-green light to adapt to low irradiance, and form amylase-containing colonies to protect the cells from the damage of UV-B under high light intensity (SooHoo et al., 1987). Therefore, the ability to transform between flagellate cells and colonies is a superior competing strategy for irradiance. The unique adaptation to extremely low-or high-light environments enables Phaeocystis to develop blooms under turbid environments in the spring, and strong irradiance in the summer.

Light is thought to be one of the most important factors influencing the photosynthesis of phytoplankton. Further, as the main driver of phytoplankton growth in turbid estuaries, its availability would also affect nutrient utilization for phytoplankton (Harris and Lott, 1973; Chang et al., 1992; Domingues et al., 2011). An in situ incubation experiment on three size components, micro-(20– 200 μm), nano-(2–20 μm) and pico-plankton ( < 2 μm), showed that nitrate (NO₃^-) uptake was most light-sensitive, followed by ammonium (NH₄⁺) and urea (Chang et al., 1992). Phaeocystis globosa also had higher affinity for urea than Scrippsiella trochoidea or Skeletonema costatum at 120 μmol/ (m²∙s) (Hu et al., 2010). In our culture experiments, when different DON (as urea)/DIN (as NO₃^-) ratios were provided, the corresponding specific growth rates was always higher at high light intensity (100 μmol/(m²∙s)) than at low light intensity (5 μmol/ (m²∙s)). Meanwhile, at low light intensity, the specific growth rates were significantly higher with higher DON ratios than lower DON ratios (Fig. 5). Thus, it can be suggested that irradiance is one of the key factors influencing nitrogen utilization, and P. globosa possess a flexible nutrient strategy adapting to variable light conditions.

Phaeocystis globosa was able to grow within the ranges of 20–31℃ and salinities of 20–40, and the maximum growth rate (0.80/d) was obtained at two combinations, 24℃, 35 and 27℃, 40. Specific growth rates >0.60/d (75% of the maximum growth rate) were observed at 20–27℃ and salinities of 25–40 (Fig. 4). It appears that this strain of P. globosa prefers higher water temperature and salinity, whereas lower temperatures (≤15℃) and salinities (≤15) are unfavorable. Field evidence revealed that in situ temperature and salinity during P. globosa blooms in the South China Sea were both significantly higher than usual (Chen et al., 1999; Xu et al., 2003), which was consistent with our culture experiment. The first P. globosa bloom recorded in Zhelin Bay, Guangdong Province in October 1997, was associated with abnormally elevating water temperature (18–21℃) and salinity (30–33), while the average water temperature during the bloom in the summer of 1999 in the same area was 25–26℃ (Xu et al., 2003). Therefore, the ranges of temperature and salinity observed during the P. globosa blooms fall into the ranges for optimal growth observed in our laboratory experiments (Fig. 4). Phaeocystis globosa blooms have been observed from temperate to tropical regions and could grow at temperatures of 4–22℃, with an optimum range of 4–15℃ (Grimm & Weisse, 1985; Jahnke, 1989; Jahnke & Baumann, 1987; Lancelot et al., 1991; Medlin et al., 1994; Riegman and Van Boekel, 1996; Schoemann et al., 2005). It seems that the population in the South China Sea exhibits different temperature requirements. The consistency between our experimental results and the field observations in which P. globosa favors higher temperature and higher salinity, suggests that the P. globosa population from the South China Sea is a separate ecotype.

The unique characteristics of P. globosa, adapting to high temperature and high salinity, implies a higher probability in developing blooms under hightemperature and low-precipitation conditions. As for the coastal waters of the Guangdong Province, the annual average temperature is around 21℃, with monthly average temperatures of above 20℃ from May to November, during which the temperature is suitable for the growth of P. globosa. Therefore, it was not surprising that P. globosa blooms in the South China Sea were frequent and of a long duration (Bulletin of Marine Environmental Quality of China, 2000–2008, State Oceanic Administration, China). Furthermore, the growth of this organism requires high levels of nutrients, especially phosphorus, and its maximum growth rate is much higher than those of most other HAB species. Thus, along with eutrophication and global warming, the potential risk of P. globosa bloom in coastal waters of China will persist for a long period of time.

Some models have been used to identify blooms and investigate the dominant factors controlling phytoplankton production (Carstense et al., 2007; Camacho et al., 2015). Owing to differences in geographic environments and algal bloom causative species, these models should be adjusted to the specific situations. In this study, we provide a useful regression equation (Eq.5) to estimate in situ growth rates of P. globosa, based on temperature and salinity data available. This equation can also be used in the prevention and control of P. globosa blooms. In addition, nutrients are also important factors affecting the formation of blooms. In further studies, a more comprehensive model, including nutrients, should enable more accurate predictions and control of P. globosa blooms.

5 CONCLUSION

Our study demonstrates the ability of P. globosa isolated from the South China Sea to acclimate at higher temperatures, higher salinities and to lower light environments. Therefore, the Chinese population should belong to a separate ecotype, and there is a high risk of P. globosa blooms developing in the South China Sea. It is important to quantify the relationship between the growth rate of HAB species and factors such as temperature and salinity, to enable understanding of the outbreak timing of harmful algal blooms. The regression equation (Eq.5) obtained in the study can be used to estimate growth rates and predict the occurrence of blooms, based on in situ water temperature and salinity data. The present study provides important information for understanding the formation mechanisms of P. globosa blooms in the coastal waters of China, and for developing techniques to predict the occurrence of blooms in the field.