Jian GAO , , Yuelei DONG ,3 , , Xiaoyu ZHOU , Lei CUI ,3, Songhui Lü ,2,3 ,

1 Research Center of Harmful Algae and Marine Biology, College of Life Science and Technology, Jinan University, Guangzhou 510632, China

2 Southern Marine Science and Engineering Guangdong Laboratory, Zhuhai 519000, China

3 Key Laboratory of Eutrophication and Red Tide Prevention of Guangdong Higher Education Institutes, Jinan University,Guangzhou 510632, China

Abstract A large-scale algal bloom, caused by Aureococcus anophageff erens, has plagued the coastal embayment of Qinhuangdao, China since 2009. The bay scallop agriculture industry in this area has been adversely aff ected. Researchers claimed that the poor nutritional value of brown tide cells might be responsible for the detrimental eff ects on bivalve mollusks. To verify whether brown tide cells are nutritionally inadequate food sources, the biochemical composition (total extractable lipids, amino acids,fatty acids, and monomeric carbohydrates) of the Chinese strain A. anophageff erens was determined during the late logarithmic growth phase when culturing in diff erent nitrogen sources (nitrate, urea and nitrate-urea mixture). Cells cultured in nitrate contained 39.12% protein, 21.99% total extractable lipid,10.25% total carbohydrates, and a relatively high amount of polyunsaturated fatty acid (PUFA) (51.98%,percentage of total fatty acids), including eicosapentaenoic acid (EPA) (4.81%) and docosahexaenoic acid(DHA) (14.56%). The gross biochemical composition and PUFA content in A. anophageff erens in nitrate cultivation are comparable with values found in the literature of frequently used species in bivalve feeding.Nine monomeric carbohydrates were signif icantly reduced when cultivated in urea and nitrate-urea mixture( P<0.05). The DHA, EPA, and PUFA contents signif icantly decreased when cultivated in urea ( P<0.05).Although the nutritional value of A. anophageff erens dropped when cultured in urea, it is still comparable with certain favorably used algal species in bivalve feeding (i.e., Skeletonema costatum), indicating that A.anophageff erens is not a nutritionally inadequate food source.

Keyword: brown tide; Aureococcus anophageff erens; fatty acid; carbohydrate; nutritional value

1 INTRODUCTION

Summer algal blooms, caused by the rapid growth and proliferation of eukaryotic picoplanktonAureococcusanophageff erens, have become a recurring phenomenon in many coastal embayments,e.g., the east coast of the United States, since 1985 and Saldanha Bay, South Africa, since 1997 (Sieburth et al.,1988; Bricelj and Lonsdale, 1997; Probyn et al., 2001,2010). In 2009, a large-scale bloom caused by a similar picoplankton (～2 μm), which was later conf irmed to beA.anophageff erens, occurred along the coast of Qinhuangdao, China (Zhang et al., 2012). Since then,such blooms have recurred in this area, making China the third country to suff er from brown tides.

The maximum cell density of brown tide along the Qinhuangdao coast can reach over 106cells/mL,covering more than 3 000 km2and normally lasting for 1–3 months (Zhang et al., 2012; Zhen et al., 2016).Both the feeding activities and growth of shellf ish ,especially the bay scallopArgopectenirradians, the main maricultural species cultivated in this region,were severely aff ected (Zhang et al., 2012). From 2009 to 2016, the annual economic loss caused by brown tide exceeded USD 14.5 million (Ou et al.,2018b).

Brown tide has recurred in coastal ecosystems of the United States for over three decades and has also signif icantly damaged the agriculture industry along the East Coast. Extensive studies focusing on the adverse impacts and ecological consequences ofA.anophageff erenshave been conducted. It has been well conf irmed thatA.anophageff erenscan severely impact the feeding activity, growth, survival, and reproduction of bivalve mollusks, such asMercenariamercenaria(Bricelj et al., 2001; Harke et al., 2011),Mytilusedulis(Tracey, 1988; Bricelj et al., 2004), andArgopectenirradians(Gallager et al., 1989; Griffi th et al., 2019). Beyond that, detrimental eff ects on microzooplankton were also observed (Lonsdale et al., 1996; Caron et al., 2004; Smith et al., 2008; He et al., 2018). However, the exact reason and mechanism remain ambiguous thus far.

Hypothetical mechanisms have been proposed. One claimed that cellular toxicity contributes to noxious eff ects. According to Sieburth et al. (1988), organic material forms a diff use exocellular layer associated with the cell membrane ofA.anophageff erens. After being hydrolyzed by amylase, the reaction product of the outer polysaccharide layer could reduce the beating frequency of the excised lateral cilia ofM.edulis(Gainey and Shumway, 1991). The eff ect was similar to that of dopamine and could also could be blocked by pretreatment with the dopamine antagonist ergometrine. Thus, the reduced grazing rates may be caused by the impact of a dopamine-mimetic compound associated with the cell membrane ofA.anophageff erens(Gainey and Shumway, 1991). In addition, the ciliary movement of bivalves may also be physically aff ected by sticky outer polysaccharides(Sieburth et al., 1988; Liu and Buskey, 2000).

Nevertheless, to date, a precise chemical substance related to the noxious eff ect of brown tide has never been identif ied. In addition, Caron et al. (2004)found that when serving as a sole food source,A.anophageff erenscould support the good growth of three protist species (Oxyrrhismarina,Uronemaspp. andEuplotesspp.). Slipper limpets (Crepidulafornicata) can eff ectively clearA.anophageff erensat biomass-specif ic rates (Harke et al., 2011). In addition,according to Padilla et al. (2006),A.anophageff erenscan support a faster development of certain bivalve mollusk species (i.e.,M.mercenarialarvae) when mixed withIsochrysisgalbana. Moreover, when moderate-density brown tide cells mixed withI.galbanaserved as the diet, there was no distinct eff ect compared with the eff ect of a single species-diet ofI.galbana, and in certain cases, enhanced growth was generated (Padilla et al., 2006). These f indings are all serve as counterevidence to the cellular toxicity ofA.anophageff erens.

Another hypothesis, which was proposed after Lonsdale et al. (1996) provedA.anophageff erensto be a nutritionally inadequate food source for some copepods, declared that the nutritional inadequacy ofA.anophageff erensmight account for its detrimental eff ects. Smith et al. (2008) found that although the ingestion rates of nauplii of a calanoid copepod,Acartiatonsa, were not restrained byA.anophageff erens(CCMP 1708), the development of nauplii was severely depressed byA.anophageff erenscells. A similar case was discovered when the Chinese strainA.anophageff erensserved as a monoalgal diet for nauplii of the copepodPseudodiaptomuspoplesia.Nauplii exhibited a high ingestion rate for brown tide cells, however, they did not metamorphose and died in the late naupliar phase, in a similar way to those under starvation (He et al., 2018). Evidence for the suspected nutritional inadequacy ofA.anophageff erenswas also found when it served as a food source for bivalves. Conspicuous brown lipid droplets in the hard clamM.mercenarialarvae were observed when fedI.galbana, while no visible lipid stores were found in larvae whenA.anophageff erensserved as the sole food source (Padilla et al., 2006).

However,A.anophageff erenscultured in f/2 medium was affi rmed to contain eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), which are highlighted as two vital nutritional constituents for the development of zooplankton and bivalves(Bricelj et al., 1989).

Although the total lipid content and fatty acid composition ofA.anophageff erenscultured in f/2 culture medium were detected, the biochemical composition of brown tide cells has not been fully elucidated. In addition, no previous study has been conducted regarding the protein, amino acid, and carbohydrate composition ofA.anophageff erens.These constituents are also important ingredients related to the nutritional quality of food sources for bivalve mollusks (Hemaiswarya et al., 2011).Additionally, the biochemical composition of microalgae is signif icantly aff ected by the source of nitrogen (Levasseur et al., 1993; Lourenco et al.,2002). The brown tides ofA.anophageff erensoften bloom when the ambient dissolved organic nitrogen(DON) level is elevated (Mulholland et al., 2009;Probyn et al., 2010; Gobler and Sunda, 2012), which was also observed along the coast of Qinhuangdao during bloom periods (Zhang et al., 2021). Thus, more information is needed on the biochemical composition ofA.anophageff erenscultivated in diff erent nitrogen sources.

Nitrate and urea are dominant nitrogen sources among the various forms of nitrogen that are available for phytoplankton communities in nature(Durmaz, 2007), including coastal ecosystems in whichA.anophageff erensblooms. In this study, an elaborate analysis of the eff ect of diff erent nitrogen resources (urea, nitrate, and a mixture of both) on the biochemical composition of the Chinese strainA.anophageff erenswas conducted. The results were also compared with species that are frequently used to feed bivalves (values from the literature), to determine whetherA.anophageff erensis nutritionally inadequate for f ilter predators.

2 MATERIAL AND METHOD

2.1 Algal strain and culture

The specie of microalgae used for the experiment was the Chinese strain ofA.anophageff erens(No.AA-1). The strain was originally isolated from the bloom water column of the Qinhuangdao coast in 2012 (119°37.911?E, 39°54.111?N) and preserved at the Research Center of Harmful Algae and Marine Biology, Jinan University, Guangzhou, China. It was grown in sterilized artif icial sea water (ASW;Cavanaugh, 1956) enriched with f/2 culture medium(Guillard and Ryther, 1962) and subcultured routinely.The culture was grown at 20±0.5 °C in an illumination intensity of 100 μmol photons/(m2·s) along with a 12-h?12-h light?dark cycle.

2.2 Nitrogen source

Three conical f lasks, containing 1 000-mL medium and each with nutrient added at a quarter of the f/2 culture medium concentration, were inoculated with approximately 0.6×106cells/mL ofA.anophageff erens. When the cultures grew to the mid-exponential phase, algal cells were harvested by centrifugation (4 500×g, 10 min), re-suspended in 3 000-mL medium containing diff erent nitrogen sources. Nitrogen conditions were designed as follows: Treatment A: 220-μmol/L NaNO3; Treatment B: 110-μmol/L urea; and Treatment C: 110-μmol/L NaNO3and 55-μmol/L urea. Each treatment was conducted in triplicate.

2.3 Growth phase

Cell number counting was conducted every other day for cell growth rate calculation. After cultivation,algae biomass was harvested by centrifugation, freezedried, and grounded before biochemical analysis.Treatment A was harvested on Day 7, Treatments B and C were harvested on Day 5 for total lipid, fatty acid, amino acid, and carbohydrate analysis.

2.4 Ash, moisture, and total lipid content

The total ash and moisture contents of Treatment A were detected according to Van Wychen and Laurens(2016).

The total lipid contents of all the treatments were measured by a modif ied method based on Bligh and Dyer (1959).

2.5 Amino acid, fatty acid, and carbohydrate analysis

The amino acid content of all treatments was determined by an external, commercial testing organization according to the standard GB 5009.124-2016 acid hydrolysis method.

Fatty acid analysis was adapted from Van Wychen et al. (2016), see Supplementary material for details.

Carbohydrate analysis was conducted using the method modif ied from Van Wychen and Laurens(2020), see Supplementary material for details.

2.6 Statistical analysis

All results are reported as the mean±standard deviation (SD) of three replicate groups. A comparison between means was analyzed by one-way ANOVA,followed by least-signif icant diff erence (LSD)test, and signif icance was accepted whenP<0.05.Statistical analysis was conducted by using the SPSS program, version 25 (SPSS Inc., Chicago, IL, USA).

3 RESULT

3.1 Growth rate of A. anophageff erens

Fig.1 Exponential phase growth rates of A. anophageff erens cultivated in nitrate, urea, and nitrate-urea mixture

Table 1 Gross biochemical composition of Aureococcus anophageff erens cultivated in diff erent nitrogen sources (% by dry weight, n=3)

The growth rates in the exponential phase(calculated from Days 1–5) are shown in Fig.1.Growth rates for cultures grown in urea were highest(0.32/d), followed by the nitrate-urea mixture(0.30/d). There were no signif icant diff erences among growth rates (P>0.05).

3.2 Lipid content and gross biochemical composition

The total extractable lipid content is shown in Fig.2. ForA.anophageff erenscells grown on urea, the total lipid content was the highest (24.31%), followed by the cells grown on nitrate (21.99%). There were no signif icant diff erences between these two groups(P>0.05). The total lipid content of the cells grown on the nitrate-urea mixture was the lowest (19.00%),which was signif icantly lower than that of other two groups (P<0.05).

Fig.2 Total extractable lipid contents (% of dry weight) of A. anophageff erens cultivated in nitrate, urea, and nitrate-urea mixture

The gross biochemical composition, expressed as a percentage of dry weight, showed the total extractable lipid, protein, and total carbohydrate contents ofA.anophageff erensunder diff erent nitrogen cultivations (Table 1). According to Laurens (2016),the protein content was determined as the summation of amino acids, and the total carbohydrates (including starch) in algal biomass were determined as the summation of monosaccharides. The protein content of the culture on nitrate (39.12%) was not signif icantly diff erent from that of the culture on urea (40.35%,P>0.05) and was signif icantly lower than that of the culture on the nitrate-urea mixture (40.88%,P<0.05).There was no signif icant diff erence between the two organic nitrogen addition groups (P>0.05). The total carbohydrate contents ofA.anophageff erenswas signif icantly higher (10.25%) when cultured in nitrate(P<0.05), and there was no signif icant diff erence between cultures on the urea (7.87%) and the nitrateurea mixture (7.80%).

3.3 Fatty acid prof iles

The fatty acid compositions ofA.anophageff erenscultivated in diff erent nitrogen conditions are shown in Table 2.

Gas chromatography-mass spectrometry analysis revealed that, as a percent of total fatty acids,A.anophageff erenscultured in diff erent nitrogen sources contained 25.96%–28.55% saturated fatty acids (SFAs), 18.05%–19.89% monounsaturated fatty acids (MUFAs), and 51.98%–55.29% PUFAs.The major fatty acids identif ied inA.anophageff erenswere C14?0 (myristic acid, 18.81%–22.43%),C16?1n-7 (palmitoleic acid, 11.24%–15.13%),C18?4n-3 (stearidonic acid, 22.98%–27.77%),and C22?6n-3 (DHA, 7.69%–14.56%). C14?0 and C16?1n-7 comprised 22.43% and 15.13% of the total fatty acids of the culture on nitrate and were signif icantly higher than those of the culture in urea(18.81% and 11.24% respectively). The culture in the nitrate and the nitrate-urea mixture contained a higher percentage of EPA and DHA than the culture in urea(P<0.05). However, regarding the percentages of all C18 mono- and poly-unsaturated fatty acids, culture in the urea had signif icantly higher concentrations than both cultures in the nitrate and nitrate-urea mixture, except for C18?4n-3 in which no signif icant diff erence between the culture in the urea and nitrateurea mixture was shown. The culture in urea contained the highest percentage of MUFAs. The culture in the nitrate-urea mixture contained the highest percentage of PUFAs. There was no signif icant diff erence among the three cultures in the percentage of SFAs. For all three treatments, no eicosatetraenoic acid (C20?4n-6,ARA) or docosapentaenoic acid (C22?5n-6, DPA)was detected.

Table 2 Fatty acids composition of Aureococcus anophageff erens cultured in diff erent nitrogen sources shown as content per dry weight (mg/g) and percentage of total fatty acids (%)

3.4 Amino acid prof iles

The amino acid content (% by dry weight) ofA.anophageff erenscultivated in diff erent nitrogen conditions is shown in Table 3.

The percentage composition of nonessential amino acids varied from 23.47% to 24.60%. The percentage composition of essential amino acids varied from 15.65% to 16.28%. The percentage composition of total amino acids varied from 39.12% to 40.88%.The principal fraction of amino acids in the treatment consisted of leucine (3.50%–3.61%), aspartic acid(4.20%–4.39%), and glutamic acid (5.08%–5.43%).

The nonessential amino acid content of the culture in the nitrate was signif icantly lower than that of the cultures in the urea and the nitrate-urea mixture(P<0.05). Both the essential amino acid and total amino acid contents of the culture in nitrate were not signif icantly diff erent from those of the contents of the culture in the urea (P>0.05) and were signif icantly lower than those of the contents of the culture in the nitrate-urea mixture (P<0.05).

Table 3 Amino acids composition of Aureococcus anophageff erens cultured in diff erent nitrogen sources (% by dry weight, n=3)

3.5 Carbohydrate prof iles

The monomeric carbohydrate compositions ofA.anophageff erenscultivated in diff erent nitrogen conditions are shown in Table 4.

Eleven monomeric carbohydrates were detected inA.anophageff erens, and the major monomeric carbohydrates identif ied inA.anophageff erenswere galacturonic acid (1.60%–2.90%), glucose (1.30%–1.70%), and galactose (0.96%–1.59%). Except for glucose and galactose, the contents of nine other monomeric carbohydrates were signif icantly higher in the cultures in nitrate (P<0.05). The glucose content was highest in the culture in urea, and the galactose content was highest in the culture in the nitrate-urea mixture.

4 DISCUSSION

4.1 Growth of A. anophageff erens cultured in diff erent nitrogen sources

In this study,A.anophageff erensgrew wellwhen urea served as the sole nitrogen supplement.Cultures with urea and the nitrate-urea mixture were associated with higher growth rates and entered the stationary phase two days earlier than the culture with nitrate. A similar phenomenon occurred duringA.anophageff erensblooms in the coast of Qinhuangdao, China: inorganic nitrogen was negatively correlated with the abundance ofA.anophageff erens, and half of the total dissolved nutrient pools were contributed by DON and DOP(Yao et al., 2019; Zhang et al., 2021). The results also matched previous nutrient enrichment studies in Quantuck Bay and Narragansett Bay, USA, in which,both mesocosms and f ield measurements showed that dissolved inorganic phosphorus (DIN) concentration was inversely correlated withA.anophageff erenscell density, and that the initiation ofA.anophageff erensblooms was usually associated with increasing DON levels (Keller and Rice, 1989; Kana et al., 2004;Gobler and Sunda, 2012).

Table 4 Monomeric carbohydrates composition of Aureococcus anophageff erens cultured in diff erent nitrogen sources (% by dry weight, n=3)

4.2 Assessment of gross biochemical composition

According to Guedes and Malcata (2012),microalgae grown to the late exponential growth phase usually consist of 5%–15% carbohydrate,10%–20% lipid, and 30%–40% protein. TheA.anophageff erenscultured in diff erent nitrogen sources contained 7.80%–10.25% carbohydrate, a relatively high amount of protein (39.12%–40.88%,around the upper limit of the average value), and a high amount of total extractable lipid (up to 24.31%in urea cultivation). The protein and carbohydrate contents were basically within the range, but the total extractable lipid content was higher than that of the average range value. Bricelj et al. (1989) also found that when cultured in f/2 medium, the lipid content ofA.anophageff erenswas quite high (26.4%±6.4%of dry weight). The result of the present study is consistent with Bricelj’s work.

Table 5 Literature values for gross biochemical composition(% of dry weight) of seven phytoplankton species:Skeletonema costatum, Chaetoceros calcitrans,Isochrysis galbana, Pavlova lutheri, I. galbana(T-ISO), Thalassiosira pseudonana, and Tetraselmis suecica

When served as food for bivalve mollusks, the nutritional value of the algal diet depends largely on the gross biochemical components (Marshall et al., 2010). Saucedo et al. (2013) discovered that elevated algal diet protein levels were associated with enhanced growth and respiration of juvenile lion-paw scallop,Nodipectensubnodosus. Uriarte and Far??as(1999) discovered that both the growth and survival of the Chilean scallop,Argopectenpurpuratus, are positively correlated with dietary diet protein levels.Tang et al. (2006) proved that the feeding behavior of hard clamMeretrixmeretrixis positively correlated with the total lipid content of algal diet. Wikfors et al.(1992) found that the gross biochemical composition of algae aff ects the growth of hard clamsM.mercenaria,one of the most adversely aff ected species byA.anophageff erens. To support rapid growth, both dietary protein and lipids must be present in suffi cient quantities. For bivalve mollusks, neutral lipids are the principal energy-providing constituent in algal diets, followed by protein (Fernández-Reiriz et al.,2011; Matias et al., 2011). From this perspective, the relatively high amount of lipids and proteins should makeA.anophageff erensan appropriate bite.

Skeletonemacostatum,Chaetoceroscalcitrans,Pavlovalutheri,lsochrysisgalbana,Thalassiosirapseudonana, andTetraselmissuecicaare some of the commonly used microalgae species in bivalve rearing (Kaparapu, 2018). When compared with these species, the gross biochemical composition ofA.anophageff erensis comparable with all (Table 5).Among them,Isochrysissp. (T-ISO),P.lutheri, andC.calcitrans, are the most commonly used species when feeding the larva, early juvenile, and broodstock stages of bivalve mollusks (Brown, 2002). Moreover,S.costatum,T.pseudonana, andI.galbanasupport excellent growth of a great number of bivalve mollusks, includingM.mercenaria(Wikfors et al.,1992; Greenf ield et al., 2004; Bricelj and MacQuarrie,2007).

4.3 Composition of fatty acids in A. anophageff erens

Fatty acids, especially PUFAs, are vital constituents in membrane f luidity and function maintenance and serve as precursors to some bioactive molecules involved in metabolism and reproduction (Da Costa et al., 2015; Cheng et al., 2020). Most animals, including bivalve mollusks, cannot synthesize either n-3 or n-6 family PUFAs de nova (Zhukova, 2019). The ability to produce long-chain PUFAs from shorter chain precursors is also limited (Langdon and Waldock,1981; Delaunay et al., 1993). The primary source of long-chain PUFAs for bivalves in the f ield is through the ingestion of algae (Cheng et al., 2020). Therefore,for most bivalves, the content of long-chain PUFAs is emphasized when evaluating the nutritional value of algae.

Bricelj et al. (1989) detected the fatty acid composition ofA.anophageff erensisolated from the East Coast of the United States. They reported that the amount and composition of fatty acids were remarkably constant between the exponential growth phase and the early stationary phase when cultured in nitrate. The levels of saturated fatty acids(SFAs), monounsaturated fatty acids (MUFAs), and polyunsaturated fatty acids (PUFAs) in cells in the early stationary phase were 23.2%, 17.5%, and 59.2%respectively. EPA and DHA accounted for 5.1% and 14.0% of the total fatty acids, respectively. In thepresent study, the SFA level was higher (28.55%),and the PUFA level was lower (51.98%). The MUFA,EPA, and DHA levels of the present study were all comparable to the values of Bricelj’s results.

Table 6 Literature values for essential fatty acids composition (% of total fatty acids) of six phytoplankton species:Skeletonema costatum, Chaetoceros calcitrans, Pavlova lutheri, Thalassiosira pseudonana, I. galbana (T-ISO), and Chaetoceros muelleri

When compared with f ive favorably used microalgae species in bivalve feeding,A.anophageff erensis similar to, and, in certain cases higher than, these algal species regarding their content of stearidonic acid (C18?4n-3), α-linolenic acid (C18?3n-3), DHA,and total PUFA (Table 6). Although the EPA content ofA.anophageff erensis lower than that of most of these species, the content of stearidonic acid (SDA,C18?4n-3), which is an important precursor for EPA(Jónasdóttir, 2019), surpassed those of all the species.

Hard clams (M.mercenaria), marine mussels(M.edulis) and bay scallops (A.irradians) are three bivalve mollusk species that are severely aff ected by brown tide. Since bivalve nutritional requirements are diff erent among species, a detailed comparison is indispensable. When served as a solo food supplement, bothS.costatumandI.galbana(T-ISO)consistently supported rapid growth ofM.mercenaria(Helm and Laing, 1987; Wikfors et al., 1992; Bricelj and MacQuarrie, 2007). The C18-PUFA (except C18?2n-6) and DHA contents ofA.anophageff erensin the nitrate and nitrate-urea mixture were higher than those ofS.costatum, whereas the EPA level was lower. When compared withI.galbana(T-ISO),the C18?4n-3 and EPA levels were higher, while the C18?3n-3 and DHA contents were lower. Portilla et al.(2015) found that an exogenous supplement of EPA and DHA is indispensable for hard calms acclimation to declining temperatures during the overwinter period. The present study proved that these two PUFAs are contained inA.anophageff erenscultured in diff erent nitrogen sources.

ForA.irradians, Milke et al. (2006) reported that,to support fast growth of postlarval and juvenile bay scallops, in addition to EPA and DHA, n-6 PUFAs such as ARA and DPA are also essential. APavlovasp. (CCMP 459) andChaetocerosmuellericombination diet containing high amounts of ARA and DPA, outperformed all other diets. Similarly,excellent growth of marine musselM.edulislarvae was supported byC.muelleri, an algae that contains a high amount of ARA (Leonardos and Lucas, 2000).Although n-6 PUFAs were not detected in the present study or in Bricelj’s report (Bricelj et al., 1989), the presence of both ARA (2.35%) and DPA (4.55%) inA.anophageff erenswas detected by Bigelow et al.(2013), upon culture in L1 medium. Moreover, the fast growth ofM.edulislarvae was also supported byS.costatum, which contained no ARA or DPA(Leonardos and Lucas, 2000).

The fatty acid composition of algae is signif icantly aff ected by the culture medium. Bigelow et al. (2013)reported thatA.anophageff erenscontained less C14?0(10.99%), C16?1n-7 (4.55%), and MUFAs (7.38%),and more C18?2n-6 (3.92%), C18?4n-3 (28.57%),and PUFAs (68.76%) when cultured in L1 medium.The present study found that both EPA and DHA levels were signif icantly lower when cultured in urea(P<0.05). Moreover, when expressed as per unit of algal dry weight (mg/g), cultures in urea contained signif icantly lower levels of EPA, DHA, SFA, PUFA,and TFAC than cultures in nitrate and the nitrateurea mixture (P<0.05). The nutritional value ofA.anophageff erensdropped when cultured in urea;nevertheless, both EPA and DHA contents were still comparable withS.costatum, which support good growth ofM.mercenariaandM.edulislarvae. The lack of n-6 PUFAs may makeA.anophageff erensa nutritionally inadequate food source forA.irradians.Paradoxically, if this is true, instead of expecting feeding cessation and starvation, a low growth rate and delayed development of the bivalve should be expected.

4.4 Amino acid composition of A. anophageff erens

In total, 11 amino acids are essential for bivalve mollusks: threonine, valine, methionine, isoleucine,leucine, phenylalanine, lysine, histidine, arginine,proline and tryptophan (Brown, 1991). Except for tryptophan, which was destroyed by hydrolysis during analysis, the other 10 essential amino acids were detected inA.anophageff erens. Moreover,when cultivated in urea, the levels of all 10 amino acids were either not signif icantly or signif icantly higher than those of cells cultured in nitrate. Ou et al. (2018a) found that, compared with nitratecultivated cells,A.anophageff erenscultured in urea contained signif icantly higher amounts of protein in the stationary phase. Dong et al. (2014) reported that many transcripts encoding enzymes involved in amino acid synthesis increased whenA.anophageff erenswas cultured in urea. The results of the present study are in agreement with Ou et al. (2018a) and Dong et al. (2014)’s observations.

When compared with favorably used species in mariculture, the amino acid composition ofA.anophageff erensis similar to that of these species.Unlike the amount of protein, the quality of the protein is unlikely to be an element that contributes to diff erences in nutritional value among microalgae,since it is generally accepted that the amino acid composition is rather similar among species (Guedes and Malcata, 2012). Therefore, when applied to bivalve feeding, the amino acid composition of algae is unlikely to be responsible for diff erential growth performance.

4.5 Carbohydrate composition of A. anophageff erens

Glucose is the principal sugar in most species used in mariculture (Brown, 1991). However, the present study showed that galacturonic acid is the primary monomeric carbohydrate in theA.anophageff erenshydrolysate. Although available research does not suggest a close connection between the monomeric carbohydrate composition and the nutritional value of microalgae (Wikfors et al., 1992), the proportion of glucose in the readily hydrolysable carbohydrate may be related to the nutritional quality of microalgae(Knauer and Southgate, 1999). The glucose content is 21%–87% in species that are frequently used in aquaculture, when expressed as a percent of the total carbohydrates (Brown, 1991). Compared with them,A.anophageff erenscontained relatively lower amounts of glucose (12.65%–21.74%).

The best growth of juvenile oysters and larval scallops was associated with high levels of carbohydrates provided in algal diets, when suffi cient protein and essential fatty acids were also supplied(Guedes and Malcata, 2012). Compared with cultures in nitrate, the carbohydrate concentration ofA.anophageff erenssignif icantly dropped when cultured in urea and nitrate-urea mixture, which may reduce the nutritional value. Nevertheless, it is still comparable with some of the favorably used species in bivalve feeding (Table 5).

4.6 The speculated reason for bivalve starvation caused by A. anophageff erens

To serve as an appropriate diet for bivalves, apart from having high nutritional value, microalgae should be easily digested, nontoxic and of adequate size(Cheng et al., 2020). Recruitment failure, reduced f ilter feeding rate, and cessation of feeding were observed whenA.anophageff erensreached a certain cell density (Tracey, 1988; Bricelj et al., 2001; Gobler and Sunda, 2012). The retention effi ciency of bivalves usually decreases with decreasing particle size(Cranford, 2019). The small size ofA.anophageff erensmay adversely aff ect bivalve growth. However, it seems not be an adequate explanation that accounts for starvation. In addition, the high absorbance effi ciency (up to 90%) (Bricelj and Kuenstner, 1989)ofA.anophageff erensbyM.edulisandA.irradiansapparently rules out indigestibility. Moreover, the relatively low nutritional value ofA.anophageff erenscultured in urea is still comparable with algae such asS.costatumthat supports fast bivalve growth. A plausible cause of detrimental eff ects is the toxicity of cells, which should be seriously evaluated.

5 CONCLUSION

This study found that the gross biochemical compositions ofA.anophageff erensare comparable with the values found in the literature for species that are frequently used to feed bivalves. When cultured in nitrate and a nitrate-urea mixture, fatty acid levels,especially PUFAs, including EPA and DHA, are comparable to and in certain cases higher than, the values found in the literature. Although the DHA,EPA, and PUFA contents were signif icantly decreased when cultivated in urea, they are still comparable to those ofS.costatum, which supports good growth ofM.mercenariaandM.edulislarvae. In summary, we found thatA.anophageff erensis not a nutritionally inadequate food source for bivalve mollusks.Cytotoxicity is the more likely cause of detrimental eff ects.

6 DATA AVAILABILITY STATEMENT

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