1 INTRODUCTION

Phaeocystis is a genus of marine phytoplankton of world-wide distribution, belonging to the class of Prymnesiophyte (Rousseau et al., 1994, 2007). Some species can form extensive blooms and are recognized as both harmful algae and the keystone species that could shape the structure and function of marine ecosystems (Lancelot et al., 1994). For example, P. globosa, P. antarctica, and P. pouchetii can form colonies and accumulate a large amount of biomass in a short period, forming harmful algal blooms (HABs) (DiTullio et al., 2000; Shen and Qi, 2021). Blooms of Phaeocystis have exerted adverse impacts on marine ecological balance, global climate changes, carbon and sulfur cycling, and human activities such as fisheries, aquaculture, and tourism (Schoemann et al., 2005 and references therein).

Phaeocystis globosa blooms induced dramatic ecological disaster and caused economic losses in China. In October 1997, a large-scale P. globosa bloom occurred for the first time in the southeast coast of China, which caused direct economic losses of over 7.5 million US dollars (Qi et al., 2001). Since then, P. globosa blooms have occurred successively along the Chinese coastal area, including Fujian, Guangdong, Guangxi, Hainan, Hebei, Shandong provinces and Tianjin City (Chen et al., 1999; Dou et al., 2020). By the end of 2018, at least 97 times of P. globosa blooms have been reported, with a total affected area exceeding 20 000 km² (Wang et al., 2021a). Unlike the strains in high-latitude sea areas, P. globosa strains isolated from Chinese coastal waters can form giant colonies (maximum diameter up to 3 cm) (Chen et al., 1999). The produced hemolytic toxins by P. globosa blooms are toxic to co-occurring organisms (Yang et al., 2009; Liu et al., 2010), even lethal to fish and shrimp (Long et al., 2016). Moreover, the mucus released during the formation of colony increases the viscosity of seawater. Consequently, the colonies resulted in the blockage of the condensate water inlet pipes in nuclear power plants (He et al., 2019; Kang et al., 2020). At the decline stage of the bloom, senescent colonies generated a plethora of nuisance foam (Lancelot et al., 1987) and caused oxygen depletion, hemolytic toxin (Peng et al., 2005), dimethyl sulfide (DMS) (Shen et al., 2011) and halogenated hydrocarbon accumulation (Yan et al., 2019). Due to these unique features of P. globosa strain in China, scientists launched a series of research relevant to physiology, such as the mechanism of colonial formation (Zhang et al., 2020), the factors affecting morphological transition (Wang et al., 2010b) and the influence of marine environment on growth (Cai et al., 2011).

2 POLYMORPHIC LIFE HISTORY OF PHAEOCYSTIS GLOBOSA

Phaeocystis globosa has a polymorphic life cycle (Fig. 1) (Rousseau et al., 1994), alternating between free-living cells and large gelatinous colonies (Fig. 2). The diameter of solitary cells ranges from 2 to 8 μm (Peperzak et al., 2000). The solitary cells grow flagella first to cling to the solid substrate surface, then the flagella disappear, and continuously divide and produce mucus (Peperzak et al., 2000). The internal cells of colony proliferate through mitosis (Rousseau et al., 2007). Colonial cells without flagellates and scales are distributed homogeneously over the periphery of the colony. (Jahnke and Baumann, 1987). Colonies as "balls of jelly" have size diameters ranging from several hundred microns to several millimeters. With the continuous development of colony, the mature colony is broken by the agitated water stream and releases its internal cells to the extracellular matrix. Immobile cells released from colony undergo two consecutive cell differentiations to form a new colony (Rousseau et al., 2007).

Apart from gelatinous colony, aggregation is another formation generated by solitary cells (Fig. 2c) with a maximum diameter of 450 μm when co-cultured with Oxyrrhis marina (Wang and Wang, 2012). The aggregation was also found in P. antarctica in both laboratory environment and field (Gaebler-Schwarz et al., 2010). Similar to colony, aggregation is also formed by hundreds of diploid cells without flagella, but the cells are packed tightly without hollow structure, and the internal structure cannot be observed. Aggregation might be a new formation in P. globosa life history to overwinter and re-proliferate when conditions are suitable (Zheng, 2014). Albeit several descriptive life histories (Rousseau et al., 1994; Peperzak and Gäbler-Schwarz, 2012), there is still no coherent description of all cell types and their transitions in the life cycle of Phaeocystis (Veldhuis, 1987; Davidson and Marchant, 1992; Rousseau et al., 1994).

3 FACTORS INDUCE THE MORPHOLOGICAL TRANSITION

Colony is beneficial for P. globosa to occupy a favorable ecological niche. As the predominant form during bloom, colony possess strong competitive conditions. Compared with solitary cells, the flexible membrane of colony possessing elastic properties can protect the cells from rupture and other viral infections (Hamm et al., 1999), weaken the UVB damage (Davidson, 1996), enhance the light tolerance (Skreslet, 1988). The hollow structure of colony can provide buoyancy to extent (Davidson, 1996) and increase P. globosa adaptability in surface water. The organic matter contained in the colony support the cells grow continuously under adverse conditions (Liu et al., 2015). In addition, the formation of colony is in favor of avoiding grazing and reducing the mortality from zooplankton (Weisse et al., 1994).

Deciphering the factors and mechanisms of the colony formation is crucial to explain the outbreak of P. globosa blooms, which may further provide promising control strategies. The factors induced the morphological transition triggered tremendous interests in academia. So far, the factors such as nutrients, light, temperature, sessile matrix, and water flow (Hai et al., 2010) have been investigated extensively. Recently, grazing cue, an additional biological factor induced by predator, which might be considered as one type of infochemical in the intra-species communication intrigued great attention in marine science (Pohnert, 2010).

3.1 Temperature

Phaeocystis globosa blooms were recorded in almost all seasons. In general, 80.6% of the blooms happened in the South China Sea and 72.5% happened in the period from November to March of the following year. According to the integral analysis, temperature was ruled out as a trigger for the spatial and temporal distribution of P. globosa blooms (Wang et al., 2021a). Chinese strains of P. globosa show more adaptable to both high (the optimal range is 20–30 ℃) and low (9–12 ℃) temperature environment (Xu et al., 2017; Li et al., 2022a). 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. Compared with the strains of high latitude in Europe, the P. globosa bloom with an area of approximately 1 600 km² occurred in Tianjin coastal sea areas of Bohai Bay in October 2006 was attributed to the relatively high sea water temperature (Dou et al., 2020). The abruptly water temperature rise was the main factor for the P. globosa bloom occurrence in the Pearl River Estuary at the end of November 2009 (Wang et al., 2010a). Correspondingly, one of the key factors for the P. globosa bloom decay in the Beibu Gulf in February 2014 was the decrease of water temperature (Li et al., 2015). In the laboratory, temperature also exerts an influence on colonial formation (Wang et al., 2010b).

The conclusion relevant to the relationship between temperature and growth was controversial. Li et al. (2022b) found that high temperature (28 ℃) inhibited the growth of solitary cells and colonial size was larger than that at low temperature (20 ℃). However, Zhang et al. (2020) elaborated that solitary cells failed to form colonies at high temperature (32 ℃). Compared to 20 ℃, higher temperature inhibited the synthesis of glycosaminoglycan, which is required for the formation of colony, preventing the formation of cross-linked extracellular matrix (Zhang et al., 2020). Both experimental P. globosa strains were from the Beibu Gulf with different capture times, the differences between algal strains cannot be ruled out accurately (Hu et al., 2019; Xu et al., 2020). The temperature might be an important influence on colonial formation, but the effect of temperature on P. globosa blooms may not be consistent with other Phaeocystis blooms (Verity and Medlin, 2003).

3.2 Irradiance

Colonial formation was correlated with irradiance in situ and in the laboratory (Schapira et al., 2006; Cai et al., 2011). Adequate irradiance promotes the formation and enlargement of colonies (Peperzak, 1993; Peperzak et al., 1998). Under laboratory conditions, P. globosa reached its maximum growth rate when the irradiance was 60 μmol/(m²·s), and there was no obvious photoinhibition with the increase of light intensity under 500 μmol/(m²·s) (Xu et al., 2017). Under a high light condition of 500 μmol/(m²·s), colony density can reach 404.67± 24.58 colonies/mL; under the low light condition of 50 μmol/(m²·s), almost no colony was found (Wang et al., 2013). Integral analysis of the investigation shows conclusively that the daily irradiance threshold for colonial formation is 100 Wh/(m²·d) specifically to Dutch originated P. globosa train, and manifest a positive correlation with light density (Peperzak, 1993).

3.3 Nutrient

Anthropogenic or cultural eutrophication has been assumed as an important factor for colony formation. Li et al. (2015) reported that the occurrence of algal bloom was closely related to the decrease of the N/P ratio in sea water. Laboratorial observation showed that P. globosa was highly competitive for inorganic orthophosphates and dissolved organophosphates (Qin et al., 2018), but its growth was easily restricted by phosphorus (Wang et al., 2013). Nitrate is an important source of nitrogen during the growth of P. globosa, which plays an important role in maintaining cell growth and life cycle transformation (Wang et al., 2013). P. globosa colonial cells have a "preference" for NO₃^--N, which might be the key to the formation of colonies and the maintenance of blooms (Lv et al., 2019). Nitrate starvation induces flagellum abscission of P. globosa and colony formation, but motile capacity is recovered when nitrate levels increase (Chen et al., 2015). In the wild environment, the solitary cells will go through the oligotrophic stage after the diatom bloom eruption, and the low N environment will stimulate the formation of stationary solitary cells. When new nutrients are introduced, N restriction is disrupted and immobile solitary cells grow and develop rapidly, forming a large number of colonies and triggering the algal blooms (Peperzak, 1993).

3.4 Infochemicals-related with grazing pressure

Predation is the primary reason for the mortality of marine phytoplankton, and more than 50% of phytoplankton biomass is consumed by zooplankton (Tillmann and Hansen, 2009). For better survival and development, phytoplankton have evolved a number of defense strategies to resist ingestion. These strategies can be divided into constitutive defense (Agrawal, 1998) and induced defense (van Donk et al., 2011). Constitutive defense plays a defensive role regardless of whether a predator is present or not (Agrawal, 1998). Induced defense, however, is not expressed in the absence of grazing pressure and is activated only by touching or perceiving the scent of a graze information (van Donk et al., 2011). The constitutive defense consumes more energy than the induced feeding defense, so the induced feeding defense is used extensively by phytoplankton (Agrawal, 1998).

Predator-induced defenses are pervasively present in phytoplankton (van Donk et al., 2011), for example, in Alexandrium tamarense (Xu and Kiørboe, 2018), Skeletonema marinoi (Amato et al., 2018), P. globosa (Long et al., 2007) and many other harmful algal bloom species. Phytoplankton utilize different defense ways, such as thorn formation, thicker cell wall generation, toxin production, bioluminescence and chains or globular colony formation, increasing the deterrent effect of predators and reducing the mortality rate of prey feeding (Tollrian and Dodson, 1999; Xu and Kiørboe, 2018). Phytoplankton can adjust their defense strategies according to the size of the predator. Research on the morphological transition of P. globosa from solitary cells to colony facilitates the study of induced defense in phytoplankton.

Compared to those defended life forms, few free solitary cells are found to form algal blooms under natural conditions. The development of colonies effectively protected the interior cells from grazing due to their large size and tough skin, preventing the feeding or infecting from zooplankton, viruses, and bacteria, which facilitate the continued growth of the cells (Hamm, 2000). Grazing does not only reduce the competition from surrounding phytoplankton, but also releases chemical signals inducing the enlargement of colony (Huo et al., 2020). In the wild environment, P. globosa blooms usually outbreak after diatom blooms in the North Sea and English Channel (Delegrange et al., 2018). Further studies found that changes in zooplankton populations may also contribute to the formation of algal blooms (Weiße, 1983; Turner et al., 2002). Compared with natural environment, the colonial size of P. globosa was reduced in laboratory culture. This may be attributed to the lack of interaction with predators or competing algae such as diatoms. In the presence of copepod predators that prefer to feed on lager algae, P. globosa tends to occur as flagellated cells; while P. globosa are more prone to form colony when coexisting with ciliates that feed on solitary cells (Long et al., 2007). In brief, P. globosa choose the morphological formation to avoid predation. However, other studies have shown that copepods, ciliates and heterotrophic dinoflagellates all promote the formation of colony (Tang, 2003; Wang et al., 2015). The predator species that can promote colony formation are debatable. Taken together, the proneness to form large colony by P. globosa can be interpreted as improving the ability to avoid a wider feeding range, thus achieving greater ecological advantages (Wei and Wang, 2013).

In marine ecosystem, planktonic communication involves a remarkably efficient chemical language. As prey, phytoplankton respond to kairomones released by grazers and alarm pheromones from conspecifics. Infochemicals released by zooplankton during feeding play an important role in regulating the growth and morphological transformations of P. globosa (Wei and Wang, 2013). Long et al. (2007) reported that water-borne cues filtered from the seawater where macrocopepods live increased the proportion of solitary cells, while the cues from the ciliate, a small protozoan, increased the proportion of colonial cells. The experiment not only presented that kairomones affected the formation mechanism of colony, but also demonstrated that the response of P. globosa was consumer-specific.

Phaeocystis globosa is not an ideal food source for herbivores (Turner et al., 2002) since lack of unsaturated fatty acids, DHA, EPA, and other nutrients that are needed for plankton growth (Wei et al., 2020; Li et al., 2022c). However, in the absence of other food supplies, some zooplankton can efficiently feed on solitary cells or colony (Turner et al., 2002). According to the particle size spectrum theory, there is an optimal particle size ratio between predator and prey (Hansen et al., 1994). Nevertheless, many zooplankton have a wide feeding range. Under laboratory conditions, the maximum feeding rate of zooplankton on P. globosa occurs in the range of the predator-to-prey size ratio of 4–16 (Nejstgaard et al., 2007). We summarized the effects of grazing influence on the formation mechanism of P. globosa in recent decades. As shown in Table 1, the different conclusions drawn from the identical predator conditioned experiments may be due to the divergence of algae strains, experimental methods, and physical conditions of predators.

Infochemicals that can elicit P. globosa defensive traits may come from the metabolites of conspecific injured cells or/and heterospecific herbivores. The pheromones released from phytoplankton after mechanical damage are used to fulfill the conspecific communication. In addition, kairomones, specific odorants associated with herbivores, and compounds closely related to feeding activities (van Donk et al., 2011), are used by prey in inter-species communication. So far, a lipid soluble substance has been provn to affect the formation mechanism of colony. Unfortunately, the chemical structure of the lipid soluble substance has not been identified (Long et al., 2007). A lot of factors, such as minute amount of sample, low proportion of active compounds, difficulty in enrichment and separation from seawater, uncertainty of synergistic mechanism, hindered the characterization of plankton semiochemicals. Identification based on bioassay-guided fractionation has been extensively used to characterize allelochemicals in marine organisms. The development of nuclear magnetic resonance (NMR), gas chromatography-mass spectrometry (GC-MS), and ultra-high liquid chromatography-mass spectrometry (UPLC-MS) as well as the promotion of multi-omics methodology, are in favor of the separation and analysis of allelochemicals. Using a strategy of activity tracing, eight copepodamides 1‒8 (Figs. 3–4) have been identified during predation by copepod (Calanus finmarchicus) at concentrations of 10^-12 to 10^-9 mol/L, which caused the amount of paralytic shellfish toxin released by A. minutum to increase more than 20 times (Selander et al., 2015). Compound 9, another defense-inducing amide (Fig. 4), was effective in S. marinoi. The unique amide mixtures released from different species of copepods could induce the defense of algae (Grebner et al., 2019).

The predator chemically induced effects on marine plankton P. globosa exemplified the predator-induced defense. Understanding the mechanism of automatic balance of ecosystem mediated by chemical communication and filling the knowledge gap on chemical communication in marine ecological systems manifest scientific significance. In addition, the induced defense between phytoplankton and zooplankton interpreted the diversity of morphological mechanism at the population level, which helps to understand the important role of minute endogenous infochemicals in unicellular-dominated food webs such as marine plankton, and shed light on algal bloom control.

4 PERSPECTIVE

Phaeocystis has become one of the most intensively studied marine phytoplankton species and the occurrence of Phaeocystis blooms drew attention for a century (Bullen, 1908). In the last two decades, P. globosa caused harmful algal blooms almost every year and has become the primary causative species in Chinese coastal waters. Chinese scientists investigated the outbreak scenario and summarized the influence factors of P. globosa blooming intensively (Xu et al., 2017; Wang et al., 2021a, b).

Factors for the triggering of bloom, such as temperature, salinity, freshwater run-off, and eutrophication, has been proposed, but none of them has been conclusive. The formation of colony reduces feeding pressure from herbivorous zooplankton and plays a unique role in the life history of P. globosa, but the formation mechanism, internal physiological structure, and the difference of morphological transformations among diverse P. globosa strains need to be further explored. This implies that colony formation may be a multipurpose strategy. The investigation of colony formation mechanism mainly focused on the physiological and ecological levels in past two decades and has been restricted by research methods and technical strategies. Many multidisciplinary problems such as the physiological mechanism, external regulatory factors, and the influence of biological processes of P. globosa colony formation should be addressed in future.

Chemical communication in the sea provides significant insights into the ecology and evolution of marine populations, the organization of marine communities, and the function of marine ecosystems. It strongly influences the behavior of organisms, including foraging strategies, feeding choices, selection of mates and habitats, competitive interactions, and transfers of energy and nutrients within and among ecosystems. Hence, the chemical cues mediate population structure, community organization, and ecosystem function. Chemical cues are omnipresence from bacteria to phytoplankton to benthic invertebrates and fish in the sea. However, the chemical nature of the chemical cue is inadequately recognized. If the chemical essence is understood, our fundamental understanding of the ecological processes that structure marine systems, interaction of biotics in selecting for present day traits, and how these processes combine to determine the structure of marine communities will advance more rapidly.

Infochemicals exert a crucial impact on phytoplankton growth, reproduction, and development, which provides a pathway for information exchange between intra- or interspecies. Chemical cues released by grazers can be recognized by P. globosa and linked to potential predation risks. The solitary cell reduces the risk of being preyed by herbivorous zooplankton by forming colony. Reduced feeding pressure on P. globosa means increased feeding pressure on other algae, which further lifts the competitiveness of P. globosa. Although the changes are insignificant in individuals, large biomass can amplify this effect and could alter energy flow, nutrient cycling, and patterns of carbon sequestration. Thus, chemical cues affect not only individual behavior and population-level processes, but also community organization and ecosystem function. The indirect effects of chemical signals on behavior have as much or even more impact on community structure and function as the direct effects of consumers.

The formation mechanism of colony is species-specific and related to ecological niche of grazer and special life activities. Comprehensive understanding on the chemical interaction between grazers and algae could provide a strategy to predict and control the P. globosa bloom. The pheromones and all potential signals involved in the communication of a plankton community constitute a complex research system. Pohnert (2010) argued that a comprehensive inventory of plankton signals and large scale bioassays addressing the role of selected metabolites in whole plankton communities might lead to a further understanding of infochemicals beyond laboratory systems. Nevertheless, an untargeted approach is not helpful in advancing research and would end up easily. Therefore, we propose the comparative chemical analysis of blooms in situ or in mesocosms and even plankton samples under non-bloom conditions, which will be beneficial for monitoring chemical fluctuation and screening possible chemical factors. Moreover, determination of the source, chemical structure, function pattern, and spatial and temporal distribution of infochemicals in the sea and the revealing of physiological and biochemical pathways led to significant advancement in further explanation about the induced defense evolution mechanism, the ecological interactions between and across populations, and the mechanisms and magnitude of predator effects in the trophic structure of marine food webs.

5 DATA AVAILABILITY STATEMENT

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.