A New Algivorous Heterolobosean Amoeba, Euplaesiobystra perlucida sp. nov. (Tetramitia, Discoba), Isolated from Pilot-Scale Cultures of Phaeodactylum tricornutum

ABSTRACT The diatom Phaeodactylum tricornutum is regarded as a prospective “cell factory” for the high-value products fucoxanthin and eicosapentaenoic acid (EPA). However, contamination with grazing protozoa is a significant barrier to its commercial cultivation. Here, we describe a new species of heterolobosean amoeba, Euplaesiobystra perlucida, which caused the loss of Phaeodactylum tricornutum in pilot-scale cultures. Morphological and molecular characteristics distinguish E. perlucida from the other species in the genus Euplaesiobystra. E. perlucida is 1.4 to 3.2 times larger than other Euplaesiobystra species in terms of average length/width and maximum length/width of the trophozoites. Unlike Euplaesiobystra salpumilio, E. perlucida has no cytostome; E. perlucida lacks a flagellate stage, whereas Euplaesiobystra hypersalinica and E. salpumilio both display a flagellate stage in their life cycle. The small-subunit rRNA gene sequence of E. perlucida shared only 88.02% homology with that of its closest relative, Euplaesiobystra dzianiensis, and had two distinctive regions. Its phylogenetic branch was clustered with one uncultured heterolobosean clone (bootstrap support/posterior probability = 100%/1.00). Results of feeding experiments demonstrated that E. perlucida could graze on various unicellular and filamentous eukaryotic microalgae (chlorophytes, chrysophytes, euglenids, and diatoms) and cyanobacteria. E. perlucida’s ingestion rate declined exponentially with increasing size of unicellular prey, and E. perlucida attained the highest growth rates on P. tricornutum. On the basis of its strong ability to graze on microalgae, capacity to form large populations in a short period of time, and capacity to form resistant resting cysts, this contaminant has the potential to cause severe problems in large-scale microalgal culture and merits further attention. IMPORTANCE Heteroloboseans have garnered considerable interest because of their extraordinary ecological, morphological, and physiological diversity. Many heteroloboseans have adapted to various extensive habitats, including halophilic, acidophilic, thermophilic, psychrophilic, and anaerobic habitats. Most heteroloboseans are bacterivores, with a few algivorous species reported. In this study, a new species of algivorous heterolobosean amoeba, Euplaesiobystra perlucida, is described as a significant grazer that causes losses in outdoor industrial Phaeodactylum cultures. This study provides phenotypic, feeding, and genetic information on a previously unknown heterolobosean, emphasizes the impact of contaminating amoebae in commercial microalgal cultures, and will contribute to the management strategies for predicting this kind of contaminant in large-scale microalgal cultivation.

T he diatom Phaeodactylum tricornutum Bohlin 1897 (1) is widely regarded as a promising "cell factory" for biomanufacturing multiple bioactive substances, such as fucoxanthin, eicosapentaenoic acid (EPA), and chrysolaminarin (2)(3)(4). These bioactive substances have antioxidant, antiobesity, antidiabetic, and anticancer properties and can reduce blood cholesterol and protect against cardiovascular and coronary diseases (5,6). However, commercial production of P. tricornutum has not yet been realized due to unstable cultures, low biomass yield, and low bioactive-substance content, which mostly arise as a consequence of protozoan contamination (7,8). Reports about the protozoan contamination of P. tricornutum pilot-scale cultures are much rarer. About 60 years ago, Ansell et al. were the first to find a predator flagellate, Monas sp. (Chrysophyceae, Stramenopiles), contaminating such P. tricornutum cultures, with the contamination affecting most of their 1,000-L open tanks (9). According to their findings, the appearance and development of Monas sp. were frequently linked to the rapid decline of the P. tricornutum population (9,10). In a recent study, seven protozoan strains were identified in the mass culture of P. tricornutum in open raceway ponds and tubular photobioreactors, with an undescribed heterolobosean (Heterolobosea, Discoba) amoeba being the most common and destructive predator, resulting in a considerable reduction in biomass and in fucoxanthin and EPA yields (11). However, the amoeba has not yet been precisely described and identified.
The taxon Heterolobosea was established in 1985 by Page and Blanton (12). Currently, this taxon is composed of ;150 species and 35 genera assigned to nine families and two main clades, Pharyngomonada and Tetramitia (13). Naegleria is the best known and most studied genus within the Heterolobosea (14). Naegleria fowleri causes primary amoebic meningoencephalitis (15). Naegleria gruberi has been studied mainly as a model organism for amoeba-to-flagellate transformation (16). Both species have been studied in detail for decades. The other heteroloboseans are considerably understudied and undescribed despite their enormous ecological and morphological diversity (14). Most heterolobosean species and genera (;116 species, ;32 genera) have been reported from soils, freshwater and marine habitats, or brackish sediments (13). Other species inhabit a wide range of habitats, including thermal springs (17), hypersaline brines (18), anoxic sediments (19), intestinal tracts of animals (20), and acidic rivers (21). However, reports of heteroloboseans from artificial systems such as microalgal culture systems are remarkably rare. Due to primer bias, the genetic material of these amoebae is not always amplified by the universal smallsubunit (SSU) primer, leaving many morphological species without corresponding molecular information (22). In addition, many heteroloboseans have not been successfully cultivated, so morphological and ecological details of these taxa are scarce.
The genus Euplaesiobystra is a member of the Tetramitia (Heterolobosea, Discoba) and is well known as a halophile genus (14). Studies of the genus started relatively late. In 2009, Park et al. reported the first comprehensively described species, Euplaesiobystra hypersalinica (the synonym of Plaesiobystra hypersalinica), via morphological and phylogenetic features and established the new genus Euplaesiobystra (18). In addition, their observations suggested that E. hypersalinica was an "extremely halophilic" eukaryote, with a salinity tolerance of 100 to 300% and an ability to grow well at 150 to 200% (18). Subsequently, Euplaesiobystra dzianiensis and Euplaesiobystra salpumilio were described, each successively expanding the genus Euplaesiobystra (23,24). Those studies, however, focused mostly on the salinity tolerance of Euplaesiobystra amoebae and less on their feeding characteristics. E. dzianiensis was the first found to be able to graze on Arthrospira filaments, which represented the most abundant cyanobacteria in Lake Dziani Dzaha (Mayotte island, France) (23). More recently, an uncultured Euplaesiobystra isolate, CBN AP20, was reported as a contaminant in Spirulina production cultures (25). Based on its high density and ability to graze Spirulina, isolate CBN AP20 appeared to be the contaminant most harmful to the Spirulina in that study (25). Even so, little is known about the feeding characteristics of Euplaesiobystra as a predator in microalgal culture systems. In addition, recent Illumina sequencing targeting the V9 region of SSU rDNA revealed six unclassified Euplaesiobystra sequences in the Eui-Seong solar saltern, Republic of Korea (26). This finding shows that the taxonomic inventory of Euplaesiobystra is far from complete and that the genus is more diverse than previously thought.
In this study, we report a new Euplaesiobystra amoeba from pilot-scale cultures of P. tricornutum that had a significant impact on the productivity of the culture. Our study focused on the detailed taxonomic description of the new amoeba, including its morphology, life cycle, feeding and digestive processes, and prey preference, together with phylogenetic analysis. Special attention was paid to how the amoeba became a significant contaminant in the Phaeodactylum culture, along with consideration of possible management measures to increase the resistance of microalgae to predation or to impair feeding.

RESULTS
Impact of the heterolobosean amoeba on mass culture of P. tricornutum in open raceway pond outdoor. In outdoor 13,000-L open raceway ponds, after 3 to 4 days of cultivation since inoculation, the color of healthy (control) P. tricornutum cultures changed from light brown to dark brown (Fig. 1A), whereas the color of contaminated P. tricornutum cultures changed from light brown to light yellow (Fig. 1B). Microscopic inspection revealed the presence of an algivorous amoeba in contaminated P. tricornutum cultures on the second day after inoculation, and this amoeba grazed extensively on the P. tricornutum cells (Fig. 2, "C" panels). On the fourth day of cultivation, the amoeba cells transformed into cysts, forming flocs with the algal cells, and healthy single P. tricornutum cells were rarely observed (Fig. 2, "C" panels). In uncontaminated culture, P. tricornutum cells grew well and the amoeba was rarely observed (Fig. 2, "H" panels).
The unknown amoeba had a devastating impact on the growth of P. tricornutum (Fig. 3). When the culture was uncontaminated by the amoeba or the amoeba's concentration was too low to be detected, the dry weight of P. tricornutum increased from 0.098 g L 21 to 0.18 g L 21 within 4 days. In contrast, once the concentration of the Identification of Euplaesiobystra perlucida sp. nov.
Microbiology Spectrum amoeba reached approximately 1.0 Â 10 4 cells mL 21 in P. tricornutum culture, the dry weight of P. tricornutum began to decrease significantly, eventually falling below that of the initial inoculation, and the final cell concentration of the amoeba reached values as high as 1.08 Â 10 5 cells mL 21 . Generally, the predatory amoeba contaminating the Phaeodactylum culture occurred from late spring to late autumn (highest temperature, 38°C; lowest temperature, 12°C). According to our observations, the amoeba could occur in any photobioreactors that were exposed to the air or could not be sterilized completely, including pilot-scale cultures located indoors or outdoors (Table 1). In addition to the amoeba, some other common protozoans also occurred in the pilot-scale culture systems of P. tricornutum, including Euplotes encysticus, Chilodontopsis sp., and a species of Scuticociliatia (Table 1).
Morphology and ultrastructure of the amoeba. (i) Light microscopy. The trophozoites displayed a cylindrical shape and possessed a marked hyaloplasmic region (also referred to as an anterior hyaline lobopodium), which moved relatively rapidly and was usually solitary but could also occur as two or multiple competing pseudopodia formed by eruptive movement (Fig. 4A to D; also, see Video S1 in the supplemental material).  These pseudopodia usually accounted for 10 to 30% of the cell length. Distinct unbranched uroidal filaments were sometimes observed in the trophozoites ( Fig. 4D and E). Occasionally, a caudal bulb with fine filaments was present when trophozoites moved (Fig. 4F). Cells were observed to quickly change the direction of movement at a 90°angle. The border between hyaloplasm and granuloplasm in trophozoites was sometimes obvious (Fig. 4A) and sometimes disappeared (Fig. 4F). The trophozoites' average length and width were 43.5 mm (range, 11.1 to 75.8 mm) and 25.2 mm (range, 8.9 to 44.6 mm), respectively (n = 50). The size of the trophozoites varied considerably, depending on the food ingested. As a general rule, the amoeba cell grew as it ingested more food (Fig. 4G to K). The average length/width ratio was 3.8 (range, 1.3 to 5.2) in active cells and 2.3 in slow-moving cells. Trophozoite cells had one to two nuclei ( Fig. 4L and M) and one or more contractile vacuoles (Fig. 4N)   The encystment of trophozoites in culture was very synchronous, occurring in less than 24 h once the food algae had been depleted. Cysts could form aggregations of numerous units (Fig. 5A). Young cysts were observed to have a round shape, with a thin wall and spherical, bead-like structures (Fig. 5B). The average diameter of the cysts was 16.6 mm (range, 10.4 to 19.8 mm; n = 50). Mature cysts had a thick double-layered wall that became increasingly irregular over time ( Fig. 5C and D). Meanwhile, the cytoplasm shrank and took on an irregular shape, leaving a small gap between it and the cell wall ( Fig. 5D to F). Most cysts had one central nucleus and at least one plugged pore in the cell wall (Fig. 5G).
(ii) Ultrastructure. In sections of the trophozoites, the electron-dense cytoplasm, the blunt round pseudopodia, a distinct nucleus, and an electron-dense nucleolus were observed (Fig. 6A). The nucleus was located close to the cell membrane (Fig. 6A). In well-fed trophozoites, Phaeodactylum cells were frequently seen in separate food vacuoles (Fig. 6B). Cysts were spherical. The number of layers in the cyst wall depended on the stage of cyst development, with mature cysts having two-layered walls (Fig. 6C). Cysts exhibited a central nucleus, a cyst pore, clusters of mitochondria, and clusters of electron-dense granules (Fig. 6C). A nuclear envelope and strands of condensed chromatin were evident in the nucleus (Fig. 6D). The mitochondria were surrounded by rough endoplasmic reticulum and possessed discoidal cristae (Fig. 6E). Electron-dense granules were spherical and had distinct boundaries (Fig. 6F).
The process of algal digestion in amoebae was studied using transmission electron microscopy. In general, uningested Phaeodactylum cells were found to have complete morphology and structure, with chloroplasts visible (Fig. 6G). In the early stage of digestion, after about 6 h, Phaeodactylum cells were deformed but relatively intact (Fig. 6H). After about 18 h, most of the cytoplasm of Phaeodactylum cells was hydrolyzed, but undigested cell walls remained (Fig. 6I). In the late stage of digestion, after about 24 h, the cytoplasm of Phaeodactylum cells had been fully hydrolyzed, but the cell walls still remained relatively intact ( Fig. 6J and K). Undigested algal cell walls expelled by the amoeba were detected throughout the sample sections ( Fig. 6J and K). Molecular phylogeny of the amoeba. The SSU rRNA gene sequence amplified from the unknown amoeba was 1,831 bp long. The two sequences most similar to this amoeba, as revealed by a BLAST search of the GenBank database, were a partial SSU sequence from an uncultured heterolobosean isolate, CBN AP20 (accession no. MF490458; 1,121 bp, 97.91% identity), and Euplaesiobystra dzianiensis (accession no. MN969059, deposited as Heterolobosea sp. VDe-2020a isolate DD2; 88.02% identity). A predicted helix 17_1 in the secondary structure of the SSU rRNA was present in the amoeba (Fig. S2). Our sequence, together with 60 previously published Tetramitia sequences and two sequences of Pharyngomonada as an outgroup (total of 63 sequences; .1,000 bp; 1,689 analyzed sites), was used to conduct comparative analyses.
Two distinctive regions were revealed in the sequence of the new heterolobosean amoeba. The first was nucleotides 816 to 824 in helix E23_11, and the second was nucleotides 859 to 865 in helix E23_14 (Fig. 7). The topologies generated by maximumlikelihood (ML) and Bayesian inference (BI) analyses were nearly identical, with the final result presented here being the ML topology (Fig. 7). According to phylogenetic analyses based on the SSU rRNA gene, our amoeba, a brackish-water species, belonged to the genus Euplaesiobystra and was closest to one environmental brackish-water sequence (MF490458; 1,121 bp; bootstrap support ML, 100%; posterior probability, 1.00). Both the latter sequence and our amoeba were part of the fully supported clade (unclassified clade in Tetramitia) that included E. dzianiensis (MN969059), Euplaesiobystra hypersalinica (FJ222604), Plaesiobystra hypersalinica (AF011459), and Euplaesiobystra sp. (KT210042). Heteramoeba was the sister group nearest to Euplaesiobystra.
Ability of the amoeba to graze on other microalgae and cyanobacteria. The 32 microalga/cyanobacterium strains tested as the prey of the amoeba were of various sizes, shapes, and motilities ( Table 2). Microscopic observation revealed that the amoeba could graze on the majority of the microalga strains and all the cyanobacteria (Table 2), showing a broad feeding spectrum of prey and algal cultures that it may potentially harm. Based on the ability of the amoeba to feed on the prey, the 32 microalga/cyanobacterium strains were separated into three groups.
One group of strains, described as "most suitable/rapid growth (11)," served as food organisms and supported rapid growth of the amoeba population. In cocultures with these strains, the amoeba depleted microalgal/cyanobacterial cells within 3 days. This group contained 20 microalgal/cyanobacterial strains, including seven chlorophytes (Chlorella, Nannochloropsis, and Dunaliella), two chrysophytes (Isochrysis), four diatoms (Phaeodactylum, Chaetoceros, Navicula, and Nitzschia), and seven cyanobacteria (Microcystis, Synechocystis, Desertifilum, and Synechococcus). Eight nonmotile, unicellular microalgal strains (Table 2) were chosen from this group in order to investigate the connection between the amoeba's feeding and growth rates and the size of the prey cell. The eight strains varied in size ( Fig. 8A to H). Microscopic results showed that there were numerous microalgal/cyanobacterial cells inside the amoeba cells during the experiments (Fig. 8a to h). The amoeba's ingestion rate on these eight microalgal/cyanobacterial species varied greatly, ranging from 9.02 to 1.23 prey predator 21 h 21 , with the highest ingestion rate being on Synechocystis, followed by Microcystis flos-aquae and Nannochloropsis oceanica (Fig. 9a). The ingestion rate declined exponentially with an increase in prey size (Fig. 9c) and was lowest when the prey was Phaeodactylum with a size of 27.23 mm. The growth rates of the amoeba on these eight microalgal/cyanobacterial species ranged from 0.053 to 0.020 h 21 , with the highest growth rate being on Phaeodactylum, followed by M. flos-aquae and Synechococcus (Fig. 9b). There were significant (0.01 , P # 0.05) or extremely significant (P # 0.01) differences between the growth rate achieved with Phaeodactylum as the prey and that achieved when the other microalgal/cyanobacterial species were the prey, except for M. flos-aquae (Fig. 9b). The growth rate of the amoeba increased exponentially with an increase in prey size (Fig. 9d).
The second group of strains, described as "suitable/limited growth (1)," also served as food organisms but were less well able to support the growth of the amoeba population. In this group, the amoeba reduced the prey population to nearly zero by the end of the 7day experiment. This group contained seven different strains, including one chlorophyte (Chlamydomonas), two chrysophytes (Poterioochromonas), three euglenids (Euglena), and one cyanobacterium (Arthrospira).
The third group of algae, regarded as "not suitable/no growth (2)," were not used as food organisms by the amoeba, and the amoeba showed no signs of interaction with this group. This group included only five chlorophytes (Scenedesmus, Haematococcus, and Ulothrix) ( Table 2). With these algae as food, the amoeba began to form cysts within hours during the experiment.   a Results were used to categorize amoebae as "most suitable/rapid growth (11)," "suitable/limited growth (1)," and "not suitable/no growth (2).

DISCUSSION
Rationale for the introduction of a new species in the phylum Heterolobosea. In this study, a new species in the phylum Heterolobosea is described. It has obvious unbranched uroidal filaments and a caudal bulb in the amoeba stage (Fig. 4), at least one plugged pore on the cyst (Fig. 5G and 6C), and a putative secondary structure of the helix 17_1 in the SSU rRNA gene (Fig. S2) and feeds by engulfment. All of the aforementioned characteristics are typical of the phylum Heterolobosea (13,14). Moreover, this amoeba was grouped with known members of the genus Euplaesiobystra in the phylogenetic tree (Fig. 7). We therefore concluded that this species belongs to the Heterolobosea, and we assigned it to the genus Euplaesiobystra with the name Euplaesiobystra perlucida.
According to previous studies, the genus Euplaesiobystra, which was established in 2009 (23), contains three well-described species, namely, E. dzianiensis (18), E. hypersalinica (23), and E. salpumilio (24). E. perlucida shows morphological and autecological characteristics that are distinct from those of these three species. Regarding average length/ width and maximum length/width in the trophozoite stage, E. perlucida appears to be 1.4 to 3.2 times larger than the three existing species of Euplaesiobystra (23,24 (Table S1) (18,24). Regarding feeding characteristics, E. perlucida displayed a broad spectrum of cyanobacterial and eukaryotic microalgal prey, whereas the other three species have been documented to feed only on prokaryotes/bacteria (Table S1) (18,23,24).
According to our molecular phylogenetic results, the new heterolobosean in our study did not group with any known species (Fig. 7) but clustered with one uncultured Identification of Euplaesiobystra perlucida sp. nov.
Microbiology Spectrum heterolobosean detected in a brackish environment (25), with which it shared two different sequence regions (Fig. 7). Given its morphological and feeding differences as well as a genetic divergence from the other recorded species in the genus Euplaesiobystra, we describe E. perlucida as a new species. This discovery of a new species, together with the report of the six unclassified Euplaesiobystra sequences from the Republic of Korea (26), supports the view that Euplaesiobystra is more diverse than previously realized and that the taxonomic inventory of Euplaesiobystra is still far from complete. Continuous investigation of Euplaesiobystra is therefore needed to better understand its phenotypic, ecological, and genetic diversity. Impact of E. perlucida on pilot-scale cultures of P. tricornutum in outdoor open raceway ponds. There are a few systematic studies of algivorous amoebae that contaminate commercial microalgal/cyanobacterial cultivation systems, and several important culprits have been identified. Investigations at Arizona State University revealed that the vampyrellid Vernalophrys algivore caused a decline in the quality and productivity of Scenedesmus dimorphus cultures cultivated in raceway ponds and photobioreactors, with contamination by the vampyrellid occurring at any time of the year (27). Zhang et al. reported another vampyrellid, Kinopus chlorellivorus, that resulted in the etiolation, flocculation, and collapse of Chlorella sorokiniana cultures grown in raceway ponds (28). The impact of amoebae is not, however, restricted to freshwater microalgal taxa. In Dunaliella salina ponds, the amoeba Naegleria sp. and a ciliate (Euplotes persalinus) were found to rapidly decimate the algal culture when the salinity dropped below 20% (wt/vol) NaCl (29). Outdoor mass cultures of Arthrospira sp. in the north of Chile were contaminated with an amoeba (Amoeba sp.) and a rotifer (Brachionus sp.), which generated cellular breakdown and the eventual death of the culture. According to the amoeba contaminations that were described, if an invasion was not controlled quickly, cultures would perish within days (30). Spirulina platensis cultures grown at an industrial scale of 500 to 2,000 m 2 in Brazil occasionally became suddenly contaminated with unknown amoebae (31). If the invasion was not dealt with early enough, the amoebae could destroy the culture within 3 days (31,32). All these studies suggest that algivorous amoebae play a critical role in the demise of microalgal mass cultures. In the context of ensuring productivity, grazing is a widespread problem and there is a growing volume of literature on the topic (33).
In this study, we report a heterolobosean amoeba, E. perlucida, causing significant damage to P. tricornutum culture in pilot-scale production. This predatory amoeba contaminating the Phaeodactylum culture was prevalent in diverse culture systems (Table 1). Our prior research showed that when P. tricornutum culture in a 1,300-L open raceway pond was significantly contaminated by this amoeba, the biomass productivity, cellular fucoxanthin concentration, and EPA concentration dropped by 84.62%, 96.80%, and 88.39%, respectively (11).
There are several possible reasons why E. perlucida causes catastrophic losses in P. tricornutum culture systems. The first is the strong grazing capability and high cell concentration of E. perlucida. E. perlucida was capable of moving and engulfing prey food using its pseudopodia, and one amoeba can engulf as many as 60 P. tricornutum cells within tens of minutes (11). The amoeba's cell size is quite malleable and can increase by 20 to 30 times throughout the feeding process (11). According to the results of this study and earlier studies, seven protozoan species have been observed to occur in P. tricornutum cultures, but only E. perlucida and the ciliate Euplotes encysticus have been found to graze on P. tricornutum cells, and microscopic examination reveals that the grazing ability of E. perlucida is stronger than that of E. encysticus. Furthermore, E. perlucida reproduces rapidly under conditions of adequate food. It can increase in abundance by a factor of 2 to 6 in just 24 h (11), and in the present study, the concentration of amoebae in pilot-scale cultures reached values as high as 1.08 Â 10 5 cells mL 21 (Fig. 3). In conclusion, E. perlucida can quickly consume a large number of algal cells after contaminating culture systems and has the capacity to achieve a high concentration.
The second reason is the capacity of E. perlucida to form cysts in adverse environments. Based on previous research, Euplaesiobystra can encyst, and these cysts remain viable for at least 60 days at 30% salinity (18). This indicates that the cysts may survive in P. tricornutum culture (20 to 30%) for a long time. Microalgal cultivation is a step-bystep, scale-up process in which the culture in the last step serves as the seed culture for the next cultivation (34,35). Therefore, once the seed culture is contaminated with cysts, the subsequent cultures are more susceptible to contamination and prone to collapse more quickly. Moreover, it is difficult to completely remove the cysts when cleaning photobioreactors, and this can easily induce a new round of infection. Contamination from the air may also occur; Cho suggested that the cysts of halophilic protozoa may be transported by aerosols (36). For cultures in open raceway ponds, there is a higher risk of becoming contaminated than in more closed systems, as many E. perlucida cells can enter the culture system at any time from the air as well as from the seed culture. We propose that this is the main reason why the contamination caused by the heterolobosean amoeba (the same species as E. perlucida) in our open pond systems was particularly pronounced.
Feeding characteristics of Euplaesiobystra amoebae. Previous studies on the autecology of the genus Euplaesiobystra have been devoted to salinity tolerance (18,23,24), with only a few, scant investigations on its feeding habits. Currently known described species in the genus Euplaesiobystra have the following feeding characteristics: E. hypersalinica is a bacterivore (18); E. salpumilio feeds on prokaryotes that have been grown on LB (Luria-Bertani) culture medium (15); E. dzianiensis grazes on the filamentous cyanobacterium Arthrospira fusiformis, and it has been suggested that this amoeba's distribution pattern follows that of Arthrospira in a thalassohaline lake (23). In this study, our amoeba, E. perlucida, fed on 27 strains of microalgae/cyanobacteria through phagocytosis, exhibiting a broad feeding range and rapid growth on both eukaryotic and prokaryotic taxa ( Table 2). Our results complement the known feeding characteristics of Euplaesiobystra and suggest that Euplaesiobystra may display more trophic complexity than previously thought.
Our amoebae exhibited different ingestion and growth rates on various-sized unicellular microalgae/cyanobacteria (Fig. 9). The ingestion rate of E. perlucida declined exponentially (Fig. 9c) and its growth rate increased exponentially (Fig. 9d) with increasing prey size. Many factors, including size, shape, swimming behavior, biochemical composition, surface structure variation, and physiology of the prey, have been reported to influence the feeding characteristics of protozoa on prey (37). Although the effect of shape, biochemical composition, and other factors on the ingestion rate of E. perlucida on different microalgae/cyanobacteria in our study could not be excluded completely, prey size has always been thought of as the first-order determinant in other studies (38). Dillon and Parry reported that the amoebae Echinamoeba sp. and Acanthamoeba sp. could ingest all synechococci presented to them but showed higher ingestion rates with the smaller Synechococcus strains than the larger strains (39). An inverse relationship between picocyanobacterial cell size and ingestion rate has also been recorded for heterotrophic nanoflagellates and ciliates (40). In addition, the grazing rate of Poterioochromonas malhamensis was found to decline exponentially as algal size increased, reaching an extremely low value when the prey size reached .8 mm (41). Reports of prey size versus predator growth rate, however, are relatively rare for amoebae, although they are available for other protozoan taxa. Phagotrophic dinoflagellates, for example, can grow at predator/prey size ratios between 0.15:l and 5.2:l but show optimal growth on prey approximately as large as themselves (42). Generally, the total nutritional value of microalgae/cyanobacteria increases with cell size. If the prey has high nutritional value, the predator's growth rate will increase and its ingestion rate will decrease (43). The mechanism of this phenomenon may be explained by the compensatory feeding response, where high ingestion on smaller prey compensates for lower nutritional quality to satisfy elemental content demand (44).
Studies on feeding can provide insights into possible management approaches that may be employed to reduce the impact of predators on microalgal cultures. For specific predators with a single feeding strategy, feeding selectivity is more of a function of prey properties, so changing prey properties can reduce the predator's feeding (45). Strom et al. (46) found that the deletion of the SwmA gene, which is involved with the synthesis of the S-layer in the cell wall of Synechococcus sp. strain WH8102, could enable this organism to resist predation by flagellates and ciliates. In addition, the presence of the giant SwmB protein in the cell surface of Synechococcus sp. strain WH8102 may confer resistance against Oxyrrhis marina predation (47). The majority of these resistance verification experiments were carried out in the laboratory, so whether these algal strains remain resistant to predation in outdoor environments needs to be further explored. Even so, this approach, whereby the properties of the alga are altered to increase its resistance to predators, is still considered the best solution for avoiding contamination of microalgal cultures, given that resistant strains are inexpensive, safe, and pollution free (33). With the commercialization of the biomass of P. tricornutum and its active substances, the prevention and control of contaminating organisms in large-scale cultures of this alga are likely to receive increasing attention. Our research provides empirical evidence and a source reference to facilitate the detection of contaminants and the sustainable cultivation of P. tricornutum.
Conclusions and significance. A new heterolobosean amoeba, causing devastating losses in commercial Phaeodactylum cultures by predation, is described. Based on microscopic and ultrastructural features, molecular phylogenetic analysis, and the presence of two distinctive regions in the SSU rRNA gene, we establish a new species, E. perlucida. Considering its high grazing ability, wide food spectrum, capacity to form large populations in a short period of time, and capacity to form resistant cysts, E. perlucida has the potential to cause severe problems in pilot-scale microalgal/cyanobacterial cultures. To avoid losses in the yield or quality of the algal crop, further research should concentrate on developing efficient control and monitoring methods specific to E. perlucida.

MATERIALS AND METHODS
Algal strain and mass cultivation conditions. The marine diatom P. tricornutum (strain UTEX640) was obtained from the Culture Collection of Algae at The University of Texas in Austin, TX, USA. P. tricornutum was maintained in F/2 medium (48) with a reduced salinity of 20%. The media for pilot-scale photobioreactors were prepared using nonsterile tap water. Cultivation of P. tricornutum was carried out at the R&D facilities of the State Development & Investment Corporation Microalgae Biotechnology Center, Hebei, China (N 39°57921.970, E 116°51935.950). Light intensity and temperature changed with local weather conditions during the experimental period.
Assessing the harmfulness of the amoeba on P. tricornutum mass culture in outdoor raceway ponds. Investigations into the biological contamination of mass cultures of Phaeodactylum were conducted from 2018 to 2020. During the cultivation period, some protozoan predators contaminated the Phaeodactylum culture system (Table 1). This contamination always included an unknown amoeba predator, which occurred in different culture systems, including 520-L indoor semiclosed tubular photobioreactors, 150-L outdoor open raceway ponds, 1,300-L outdoor open raceway ponds, and 13,000-L outdoor open raceway ponds (Table 1). More than 70% of crashes of P. tricornutum cultures were caused by contamination with this amoeba.
To accurately evaluate the impact of the unknown amoeba on the biomass productivity of P. tricornutum, we studied two 13,000-L outdoor raceway ponds: one was contaminated by amoebae and was considered the experimental pond; the other was not contaminated by amoebae, or the concentration of amoebae was too low to be detected, and this was considered the control. We monitored the dry weight of P. tricornutum and the abundance of predatory amoebae every day during cultivation. In both ponds, the initial dry weight of P. tricornutum was adjusted to approximately 0.1 6 0.01 g L 21 (approximately 3.5 Â 10 6 cells mL 21 ). The experiment lasted for 4 days. Light intensity and temperature changed with local weather conditions during the experimental period (Fig. S1). The cell dry weight (DW) of microalgal cultures was measured as described by Zhu and Lee (49). Samples (20 mL) were collected each day and filtered onto preweighed 0.45-mm GF/C superfine fiber membranes (1.2-mm pore size), with vacuum pressure differentials maintained at 35 to 55 mm Hg. The membrane was washed with 20 mL 0.5 M NH 4 HCO 3 to remove nonbiological adhering materials such as mineral precipitates and was then dried at 105°C until a constant weight was obtained (about 24 h), cooled to room temperature in a vacuum desiccator, and weighed. For enumeration of amoebae, samples (1 mL) were stained with Lugol's iodine (1% final concentration) (50). The cells (trophozoites and cysts of amoebae) were counted using 100-mL phytoplankton counting chambers (CC-F, China) under a phase-contrast microscope (BX53; Olympus, Japan) at Â400 magnification. Each sample was counted three times, and the mean value was used as the measure of abundance.
Isolation and maintenance of the predatory amoeba. Samples were collected from contaminated 13,000-L open raceway ponds to isolate the amoeba. Based on microscopic observations, single trophozoite amoebae were transferred into 12-well microtiter plates (catalog no. 40124; Beaverbio, China) containing P. tricornutum in F/2 medium to establish cocultures of amoebae and prey. These cocultures were maintained at 25°C under dim artificial light (a photon fluence rate of 5 to 20 mmol m 22 s 21 ) in a 12-h-12-h light-dark cycle. For long-term maintenance, 100 mL supernatant of a growing culture was transferred to algal material suspended in F/2 medium every 2 to 3 weeks. The amoebae and algal cultures are available from the Freshwater Algae Culture Collection at the Institute of Hydrobiology, Wuhan, China (http://algae.ihb.ac.cn/English/).
Light-microscopic observation. The morphological characters, life cycle, and feeding behavior of the amoeba, as well as the contamination process (described below), were observed with a Zeiss Axio Observer 3 inverted microscope (Zeiss, Oberkochen, Germany). In order to determine whether the amoeba has a flagellate stage, we set up experiments using various salinity gradients (3%, 15%, 30%, 50%, 75%, 100%, 150%, 200%, 250%, and 300%) and temperature gradients (5°C, 15°C, 25°C, 35°C, 45°C, and 55°C) to observe the cell morphology of the amoeba, as suggested by the previous studies (23,24,51,52). Different salinities were achieved by increasing the concentration of sea salt based on the formulation of the F/2 medium. The medium used in the temperature gradient experiment was F/2 medium (salinity, 20%). Well-growing cultures of P. tricornutum and amoebae were suspended in medium and distributed into 24-well microtiter plates (working fluid volume of 1 mL per well). All treatments were performed in triplicate. Experiments were carried out in a dark environment. The cellular morphology of amoebae was examined every 12 h throughout the 30-day experiment. To avoid cyst formation, food (P. tricornutum) was added every 3 days, and the culture system was refreshed every 7 days. High-resolution imaging and filming of real-time videos were performed with a Zeiss Axio Imager A2. Both microscopes were equipped with high-resolution differential interference contrast (DIC) optics and digital cameras (Zeiss Axiocam 506 color camera and SCMOS-pco Edge 4.2 LT). Adobe Photoshop CS5 (Adobe Systems, Munich, Germany) was used to adjust the color balance and contrast of light micrographs. Videos were analyzed and processed with Adobe Premiere Pro CC.
Transmission electron microscopy. The ultrastructure and digestion process of the amoeba were examined by transmission electron microscopy. Starved trophozoites, after-feeding trophozoites, and cysts of the amoeba were used to observe ultrastructure. The amoeba's digestion process was monitored using the following method: well-grown Phaeodactylum cells were added to a large amount of a culture solution containing starved amoebae, mixed, and sampled, and samples were taken every 3 h thereafter. All samples were harvested by centrifugation (3,000 Â g for 5 min; Eppendorf MiniSpin). Cells were then fixed with 2.5% glutaraldehyde overnight at 4°C. Following washing in phosphate buffer (0.1 M, pH 7.4) three times, cell samples were postfixed with 1% OsO 4 for 1 to 2 h at room temperature. Subsequently, cells were quickly washed with phosphate buffer three times and dehydrated with a graded series of ethanol-water mixtures (50% ethanol, 70% ethanol, 90% ethanol, 100% ethanol, 100% ethanol), followed by a graded series of ethanol-acetone mixtures (25% acetone, 50% acetone, 75% acetone, 100% acetone, 100% acetone). Cells were then infiltrated with Spurr's 812 epoxy resin, embedded, and cured in Spurr's epoxy resin for 48 h at 60°C. The resulting samples (representing the hapantotype material for the new species) were sectioned (73 nm) using a Leica Ultracut-R microtome and stained with 2% uranyl acetate and Sato's lead citrate (53). Sections were examined with a Hitachi HT-7700 (Japan) transmission electron microscope.
DNA extraction, amplification, and phylogenetic analysis. Starved individual trophozoites and cysts were isolated in a 0.2-mL PCR tube with 10 mL of Milli-Q water (one amoeba cell per tube, 60 tubes in total) and flash-frozen in liquid nitrogen. They were then subjected to single-cell PCR amplification. The SSU rRNA gene sequences were amplified using a combination of eukaryote primers (EukA, 59-AACCTGGTTGATCCTGCCAGT-39; EukB, 59-TGATCCTTCTGCAGGTTCACCTAC-39) (54). The samples were in a volume of 20 mL, containing 10 mL 2Â GoTaq green master mix (Promega Corporation, USA), 0.5 mL of each primer (10 mM), 2 mL template DNA, 1 mL dimethyl sulfoxide, and 6 mL double-distilled water (ddH 2 O). The cycling conditions were as follows: an initial denaturing step at 94°C for 5 min, followed by 35 cycles of 30 s at 94°C, 1 min of annealing at 55°C, and extension at 72°C for 2 min, with a final extension step at 72°C for 10 min. Amplicons were approximately 2,000 bp for the SSU rRNA gene. They were checked on an agarose gel and purified with a gel extraction kit (no. 28704; Qiagen, Germany), ligated into a pGEM-T vector system (Promega, USA), and then transformed into competent cells. Ten positive clones were sent to be sequenced by a sequencing company (Tianyi Huiyuan, Wuhan, China). The inserted fragments were sequenced using M13 forward and M13 reverse universal primers.
For the phylogenetic analysis, the SSU rRNA gene sequences from 62 representative heterolobosean species were retrieved from the National Center for Biotechnology Information (NCBI) database (https:// www.ncbi.nlm.nih.gov/). Multiple alignments of sequences were conducted using ClustalX 1.81 (55) and then manually arranged with Seaview (56). The alignment analyses were performed with DNAMAN (version 5.1) to determine any unique nucleotide regions for the amoeba, and the SSU rRNA secondary structure of Tetrahymena canadensis (accession number MW694332 [http://bioinformatics.psb.ugent.be/ webtools/rRNA/secmodel/Tcan_SSU.html]) was used as a reference. After nonalignable sites were manually excluded, a data set with 1,689 sites was analyzed for the SSU rRNA gene sequences. Data sets were subjected to maximum-likelihood (ML) and Bayesian inference (BI) analyses under the GTR1C1I model (general time-reversible model with correction for invariant sites) (Modeltest 3.7) (57), using PhyML 3.0 (http://atgc.lirmm.fr/phyml/) and MrBayes 3.0b (58), respectively. For ML analyses, 100 independent tree searches and 1,000 bootstrap repetitions were performed. For the Bayesian analysis, the chain length was 5,000,000 generations, with trees sampled every 1,000 generations, and 25% (1,250,000 generations) were discarded as burn-in. The resulting phylogenetic tree, based on the best ML topology, contained support values from bootstrapping (ML) and posterior probabilities (BI) at the branches.
Ability of the amoeba to feed on other microalgae and cyanobacteria. The capacity of the amoeba to feed on diverse prey was qualitatively assessed using 28 strains of eukaryotic microalgae and eight cyanobacteria (Table 2). Eukaryotic microalgal strains included 13 chlorophytes, four chrysophytes, three euglenids, and four diatoms. All species were obtained from the Freshwater Algae Culture Collection (FACHB) and the Center for Microalgal Biotechnology and Biofuels at the Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan, China. Well-growing cultures of microalgae and cyanobacteria were suspended in sterile F/2 medium and distributed into 12-well microtiter plates (approximately 3 mL per well). Starved amoebae were added, with the initial cell concentration of prey and starved amoebae being set at approximately 10 6 cells mL 21 and 3 Â 10 3 cells mL 21 , respectively, in each case. A positive control was established to assess the impact of the amoeba on P. tricornutum (UTEX640). For each algal strain, tests were conducted in triplicate, and the triplicates were independent biological replicates. Experiments were carried out at 25°C in a dark environment. The resulting cultures were monitored every day for a week to determine the following, as per More et al. (59): (i) whether amoebae were feeding, (ii) whether amoeboid populations were growing (assessed qualitatively), and (iii) whether the growth of amoebae was persistent (i.e., amoebae did not die before the food source was exhausted). Feeding was identified based on ingested prey cells or photosynthetic pigment material in trophozoites. The results were summarized in three categories: most suitable/rapid growth (11), suitable/limited growth (1), and not suitable/no growth (2).
In order to explore quantitatively the ability of the amoeba to graze on prey of different sizes, eight microalga/cyanobacterium strains (indicated in Table 2) that are nonmotile and unicellular and could be ingested well by amoebae were further selected as prey to assess the ingestion rate and growth rate of the amoeba. One of the strains was P. tricornutum UTEX640, which acted as a positive control. The remaining seven microalga/cyanobacterium strains included two chlorophytes (Chlorella sorokiniana CGMCC11801 and Nannochloropsis oceanica IMET1), one diatom (Chaetoceros sp. strain HR-CH301), and four cyanobacteria (Microcystis aeruginosa FACHB-905, M. flos-aquae FACHB-1028, Synechocystis sp. strain FACHB-898, and Synechococcus sp. strain FACHB-805). Cell diameters of the prey were measured using Zeiss microscope software, and the average diameter was calculated from 50 cells. For each strain, a suspension of the alga in sterile F/2 medium was added to 6-well microtiter plates (catalog no. 40106; Beaverbio, China; working fluid volume of 5 mL per well), with three control groups (containing only prey cells) and three experimental groups (containing both prey cells and amoebae) per strain. The initial cell concentration of algae in the control groups and experimental groups was always 2.0 (6 0.5) Â 10 6 cells mL 21 , and the initial cell concentration of the amoeba in the experimental groups was always 1.5 (6 0.2) Â 10 4 cells mL 21 . To eliminate the effect of photoautotrophic growth of algae on the analysis of ingestion rates, all cultures were cultivated in the dark. After cultivation/cocultivation for 36 h, cell concentrations of prey and the amoeba in all groups were counted as described in "Assessing the harmfulness of the amoeba on P. tricornutum mass culture in outdoor raceway ponds." The ingestion rate (I) of the amoeba on the algae and the growth rate (G) of the amoeba were calculated as described by Heinbokel (60): I = (g Â P)/D and G = [ln(D t1 ) 2 ln(D t0 )]/t, where g (specific grazing rate) is m n 2 m w . m is [ln(P t1 )ln(P t0 )]/t, in which m n and m w are the difference between net prey growth rate without and with grazers; P t0 and P t1 are the initial and final algal concentrations (in cells per milliliter), respectively; P (mean prey concentration) is (P t1 2 P t0 )/[ln(P t1 )ln(P t0 )]; and D (mean predator concentration) is (D t1 2 D t0 )/[ln(D t1 )ln(D t0 )], in which D t0 and D t1 are the initial and final E. perlucida concentrations (in cells per milliliter), respectively. t represents the time interval during which amoebae were in exponential growth phase. Here, the ingestion rate was used to evaluate the reduction in prey cells by grazing of per amoeba per hour, and growth rate represents exponential increase at a constant rate throughout time interval (60). Variation analysis of ingestion rate and growth rate were carried out by one-way analysis of variance (ANOVA) and Tukey's test. The feeding rate and growth rate of amoebae were fitted with the average diameter of prey cells, respectively, and the fitting equation was obtained using SPSS 26.0.
Nomenclatural acts. This published work and the nomenclatural acts have been registered in ZooBank (http://zoobank.org/), the official online registration system for the ICZN. The ZooBank Life Science identifiers (LSIDs) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix http://zoobank.org/. The LSID for this publication is as follows: urn:lsid:zoobank.org:pub: 8FE9E9FE-43F2-470C-B666-4E47BDDB070A.
Data availability. The sequence determined here was deposited in GenBank and is available under the accession number OM654558.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 0.3 MB. We are indebted to research assistants Yuan Xiao and Zhenfei Xing at the Center of Forecasting and Analysis, Institute of Hydrobiology, Chinese Academy of Sciences, for technical assistance on TEM. We also thank the National Aquatic Biological Resource Center, Chinese Academy of Sciences, for providing microalgal strains.
Hanwen Zhang designed and performed experiments, analyzed the data, and drafted and revised the manuscript. Qing He performed experiments and revised the manuscript. Xiaoying Jiang performed experiments. Hongxia Wang provided and cultured 32 algal strains. Yulu Wang performed experiments. Mingyang Ma assisted with data analysis. Qiang Hu assisted with writing. Yingchun Gong conceived the idea, revised the manuscript, gave final approval for submitting the article, and obtained the funding.