ELSEVIER Aquaculture 123 (1994) 309-322 Aquaculture The potential use of lipid microspheres as nutritional supplements for adult Ostrea edulis Horacio Herasa,l, J. Kean-Howie", RG. Ackman?" "Canadian Institute ofFisheries Technology, Technical University ofNova Scotia, P.O. Box 1000, Halifax, N.S., B3J 2X4, Canada bMolluscan Section, Department ofFisheries and Oceans Canada, Halifax, NiS; Canada Accepted 5 January 1994 Abstract A method of preparing lipid microspheres is described. These have shown potential as a diet alternative suitable for supplementing algae as sources of essential fatty acids and other lipids in the culture of marine suspension-feeders such as oysters. The preparation technique is fast, not labour intensive and uses inexpensive raw materials. The method is by sonication ofa mixture offish oil, soy lecithin, vegetable oil and vitamin E in the ratio 50: 20: 29: 1 (w/w/w/w). Fish oil was also successfully replaced with a concentrate of n­ 3 fatty acid ethyl esters, thereby increasing the amount of the essential fatty acids eicosa­ pentaenoic (EPA) and docosahexaenoic (DRA). The microspheres have a log-normal particle size distribution of approximately 1 to 20 /lm, which is in the range accepted by adult bivalves. With minor modifications, the size of the particles could be adjusted to match the food of other developmental stages of bivalves. Actual analysis for lipid class composition revealed that more than 75% (w/w) of the mass was energy-rich fish oil triac­ ylglyceridesor ethyl esters of essential fatty acids with a polyunsaturated fatty acid content of up to 64% (w/w). Microspheres were stable in recirculating seawater with temperatures of up to 21 0 C. The oxidation stability of stock emulsions was assessed over a storage pe­ riod of 8 days and oxidation products measured as anisidine values did not increase. Bac­ terial growth was also not a problem over 8 days of storage. The water quality of a recir­ culating system was not degraded by the microspheres. Particle concentrations were in the range of 1.8-1.9 X 108 ml- 1 for stock emulsions, which could be diluted to any desired dispersion in the culture media. To test the ingestion and digestion processes under labo­ ratory conditions, fluorescent beads were encapsulated into the microspheres and fed to adult oysters together with algal cultures. Oysters were able to ingest and digest the micro­ spheres in a concentration of 50% of the total particles supplied as an algal-rnicrosphere mixed diet. *Corresponding author. Tel. (+ 1-902) 420-7734; Fax (+ 1-902) 420-0219. 'Present address: Universidad Nacional de La Plata, INIBIOLP, 60 y 120, (1900) La Plata, Argentina. 0044-8486/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0044-8486 (94 )00005-9 personal Rectángulo personal Rectángulo personal Rectángulo 310 1. Introduction H. Heras et al. / Aquaculture 123 (1994) 309-322 Bivalve hatcheries routinely use combinations of unicellular algal diets for both broodstock and larvae (Brown et al., 1989). The production of live feeds is time­ consuming and labor intensive, and there is great variability in the quantity and quality ofthe nutrients supplied. Moreover, in terms ofbivalve hatchery produc­ tion, the cost of algal food is a major expense. Commercially available feeds for adult fish or crustaceans costs as little as one one-hundredth of bivalve feeds. All of these facts have led to increased research seeking alternative diets. One ap­ proach has been to alter the algal nutritional value by changing growth medium, photoperiod and other variables. This has proved valuable but is still costly and time-consuming. Another alternative is the development of artificial diets and in particular much work has been done on the development of microencapsulated diets suitable for filter-feeding bivalves (Epifanio, 1979; Langdon et al., 1985). These techniques have been increasingly used to examine nutrient requirements and to supplement algal foods. Artificial feeds can be produced with minimal demands for time, space, labour and money. They can also have a long shelflife and can theoretically be tailored to meet the nutrient requirement of any culti­ vated filter-feeding species. In bivalve hatcheries, it is essential to provide good broodstock nutrition in order to ensure the quality of subsequent eggs, larvae and spat. This is especially important for larviparous species, including Ostrea edulis, since the larvae are brooded by the parent. Both Helm et al. (1973) and Napolitano et al. (1988a,b) reported the importance of the neutral lipid content of newly liberated O. edulis larvae and linked this to the subsequent success oflarvae with a high neutral lipid content when liberated from broodstock. These larvae performed better than the ones containing less neutral lipid. Trider and Castell (1980) demonstrated the importance of highly unsaturated fatty acids, such as EPA (eicosapentaenoic) and OHA (docosahexaenoic) in lar­ val bivalves. Recent work by Napolitano et al. (1988a) has shown that two spe­ cific fatty acids, EPA and OHA, comprise a particularly high proportion of the fatty acids of O. edulislarvae. In order to increase the amount of essential lipids, especially 20: 5n-3 and 22: 6n-3, nutritional supplements can be used in oyster diets. In work on broodstock conditioning of the Pacific oyster Crassostrea gigas, Robinson (1991, 1992) used lipid microspheres to increase the dietary levels of lipid. Her results were promising and stimulated our research which addresses the use of modified lipid microspheres to increase dietary lipid, in particular the content ofpolyunsaturated fatty acids (PUFA), for filter-feeding molluscs. Lipid as an encapsulation medium for other materials maximizes the energy input since the high energy density of the outer capsule contributes to the efficiency of the process. We describe the methodology developed, and the results ofusing fluores­ cent materials incorporated into the micro spheres in preliminary feeding experi- personal Rectángulo H. Heras et al. / Aquaculture 123 (1994) 309-322 311 ments with adult oysters. Process parameters were varied in order to optimize particle size and the incorporation of lipid into microspheres. 2. Materials and methods 2.1. Preparation oflipid microspheres The technique developed was based on the thesis work of Robinson (1991), with modifications. The method involves emulsification of an oil mixture (con­ taining the desired essential fatty acids) with seawater to make the microspheres. An oil mix (Table 1) was sonicated for 10 min at full output with a 12 mm flat tip probe (Biosonik III, Bronwill Science, Rochester, NY) to thoroughly blend the components. The operating temperature was kept at 40°C, above the solidi­ fication point ofthe major lipids employed. Three g of this mixture, warmed to approximately 40°C, was sonicated with 20 ml ofa 2% solution of polyvinyl alcohol (PVA) (Sigma Chemical Company, St. Louis, MO) in filter-sterilized seawater preheated to 40-50 °C. Sonication was continued for 6 min at 75% full output with the same 12 mm flat tip probe placed 4 mm into the lipid dispersion. The lipid emulsion was immediately cooled by pouring it into ice and seawater with a ratio of lipid emulsion: seawater of 1:2 (v: v). This stock emulsion was stored at 4°C. A nitrogen atmosphere was pro­ vided at every step of the preparation and storage. Process parameters such as mixing method and time, percentage of PVA, molecular weight of PVA, and phosphatidylcholine sources were varied in order to determine their effect on particle size and percentage incorporation oflipid into microspheres. Incorpora­ tion of marker particles was evaluated. Methods ofmixing In order to optimize selection of compact and inexpensive equipment readily available for a hatchery situation, three pieces ofdispersion equipment were tested for the mixing of the oil mix with the PYA solution. A tissue homogenizer Table 1 Composition of the oil mix used to prepare the microspheres Ingredient Menhaden oil (I ) or n-3 concentrate (2) Vitamin E Vegetable oil Lecithin %w/w 50 29 20 Source ( 1) Zapata Haynie Corp, Reedville, VA (2) Supplied as ethyl esters, Laboratory of Aquaculture, Reference Center, Rozier, Belgium Covi-Ox T70, a natural mixture of tocopherols 70% w/w, Henkel Corp., La Grange, IL Crisco Shortening, The Procter and Gamble Co., Cincinnati,OH Soya lecithin granules, Trophic, Toronto, Ont. personal Rectángulo 312 H. Heras et al. / Aquaculture 123 (1994) 309-322 (Polytron Kinematica, Brinkman Inst., Rexdale, Ont.), a household mixer (Braun), and a sonicator (Biosonik III) were compared. PVA PYA was added in the process as an emulsifier and it also helps to produce microspheres with a narrower particle size distribution. PYA concentrations of 2, 1, and 0.5% PYA in seawater were tested. Also tested were two molecular weights ofPVA: 110000 (VWR Scientific, Dartmouth, NS) and 70 000-100 000 (Sigma Chemical Co., St., Louis, MO); the latter was chosen because it produced a more uniform size of particles. Lecithin sources Lecithin was added to help in the production of oil-in-water emulsions. Phos­ phatidylcholine capsules (Quest, Vancouver, BiC.}, liquid lecithin (Fearn Nat­ ural Products, Milwaukee, WI), soya lecithin granules (Trophic, Toronto, Ont.), soy lecithin, refined (US Biochemicals, Cleveland, OH), and dipalmitoylphos­ phatidylcholine (Serdary Research Labs, London, Ont.) were tested. Markers For the uptake and digestion experiment, 0.618 /.lm fluorescent marker beads made of polystyrene (Duke Science Corp., Palo Alto, CA) were included. A sus­ pension of0.2 ml containing 1.8X 109 beads ml- I was added to 3 g ofoil mix and sonicated as described above for 5 min. The rest of the procedure followed the general technique described above for lipid microspheres. 2.2. Chemical analysis Microsphere size distributions were determined with a Coulter Counter (Mul­ tisizer II, Coulter Electronic Ltd, Luton, UK) using 70 /.lm and 30/.lm aperture tubes. Data from triplicate analyses of 500 III of sample were stored individually in a matrix using a personal computer interfaced to the counter. The matrix was then transferred onto a spreadsheet and graphic programs for final analysis. Total lipids were extracted by the method of Bligh and Dyer (1959) and gra­ vimetrically determined. Lipid classes were analyzed by thin-layer chromato­ graphy with flame ionization detection (TLC-FID) by the procedure of Parrish and Ackman (1985). An latroscan TH-lO Mark III analyzer (distributor, Sci­ entific Products Equipment, Ltd, Concord, Ont.) was used for quantitation. The air flow was 2 1 min- 1 and the detector hydrogen flow rate 160 ml min- I. The scanning speed was 0.42 em S-I. Recorder peak area measurements were deter­ mined using an SP-4200 integrator. The silica gel Chromarods employed were type S-III. The lipid sample was dissolved in chloroform, and 1/.ll of this solution was spotted on the rods using 1 /.ll microcap pipettes (Drummond Scientific Co., Broomall, PA). Phosphatidylcholine, tristearin, monoglyceride, free fatty acids and free cholesterol were used for the calibration of the Chromarods for quanti­ tative analyses. Before developing the rods, samples were focused for 1 min in personal Rectángulo H. Heras et al. / Aquaculture 123 (1994) 309-322 313 acetone and dried in a humidity chamber for 5 min. The developing solvent mix­ ture employed was hexane: diethylether:formic acid (97: 3: 1 v[v Iv) for 45 min. After development, the rods were dried for 4 min at 110°C and then fully scanned. For the analysis of fatty acids, methyl esters were prepared with BF3-MeOH according to the method of Morrison and Smith (1964). Analysis by gas chro­ matography (GC) was carried out in a Perkin Elmer 8420 with FID on an Ome­ gawax 320 column (30 m, 0.32 mm ID, Supelco Inc., Oakville, Ont.). The GC conditions and temperatures were: FID 270°C; injector 250°C. The column tem­ perature was programmed as follows: initial temperature 175 °C held for 8 min; increased at a rate of 3°C min- 1; final temperature 220 °C, held for 10 min. The carrier gas was helium at a pressure of 95 kPa, and the split ratio was 1: 32. The weight percentage of individual fatty acids was obtained from peak areas accord­ ing to Ackman and Eaton (1978). An internal standard of 23: 0 methyl ester was used to quantify EPA and DHA (Joseph and Ackman, 1992). Possible oxidation during storage ofthe microsphere stock was monitored over a period of 8 days by measuring the anisidine value in lipid recovered from sam­ ples stored at 4°C (AOCS, 1990). 2.3. Oyster feeding experiments In feeding experiment 1, three European oysters were left unfed for 24 hand then placed into each of three 3 1 beakers containing 2 1 of seawater. To each container both algae (Chaetoceros muelleri; 80 X 103 ml- 1 ) and the microspheres containing fluorescent beads (120 X 103 ml :") were added. Feeding chambers were closely monitored and fecal material collected for microscopic examination. Fecal material was examined under a Leitz Aristoplan microscope equipped with an HB050 watt mercury lamp. Photographs were taken with a Wild Leitz camera using Ektachrome 400 HC film. Six more adult oysters were similarly fed and monitored in experiments 2 and 3. In all experiments, seawater filtered at 0.22 J..lm was used and the temperature was kept at 10°C. During each experiment, aeration was provided to the beakers. 3. Results anddiscussion The Biosonik sonicator gave the best results in the form of better microsphere size distribution and better emulsions, and formation of a primary emulsion in the homogenizer followed by different times of sonication allowed manipulation ofthe particle size distribution (Fig. 1). In all cases, microsphere sizes were found to follow a log-normal distribution, but with shorter sonication times a shift to­ wards larger droplets was observed, with an increase in the mean particle diame­ ter and a decrease in the total number ofparticles in the emulsion. When only the Polytron tissue homogenizer was employed (Fig. 1) the concentration of micro­ spheres increased significantly from 0.75 X 108 particles ml- 1 to 1.12 X 108 with personal Rectángulo 314 H. Heras et al. I Aquaculture 123 (1994) 309-322 T.H + 0 sec SONICATION T.H. + 45 sec SONICATION ~ 350 '§ 300 0.75x108 ± 3.2 x105 part./ml .e 250 ~ 200 (/) ~ 150 o 'e 100 co ~ 50 Z 0 5 10 15 20 25 30 Jim ~ 350 r.-----r--,--,---,----,---,---, (/) :§ 300 0.84x108 ± 8.1x105 part./ml .e 250 ~ 200 (/) Ql 150 U 't 100 co Q. 50 o Z 0 5 10 15 20 25 30 Jim T.H. + 90 sec SONICATION T.H. + 270 sec SONICATION 1.04x108 ± 7.7x10 5 part./ml 1.12x10 8 ± 11x10 5 part./ml 5 10 15 20 25 30 Jim ~ 350 r-.----r--,--,---,----,---,---, (/) 'E 300 :::J.e 250 ~ 200 (/) ~ 150 o 't 100 co Q. 50 o Z 0 5 10 15 20 25 30 Jim Vi 350 r-.----r--,--,---,----,---,---, +-' 'c 300 :::J .e 250 ~ 200 (/) ~ 150 o 't 100 co ~ 50 Z 0 Fig. 1. Effect of sonication time on the particle size distribution of lipid microspheres. Histogram values are the mean of triplicate analysis with a Coulter Counter. T.H., Tissue homogenizer. sonication time of4.5 min, and to 1.9X 108 after the standard 6 min ofsonication (Table 2). Only lecithins from Trophic and Serdary gave acceptable emulsions. Soya lec­ ithin granules from Trophic were chosen because of the lower price. Microscopic examination of the microsphere products also showed that the particles prepared with this material were the most uniform in size. The 2% level was finally chosen because it yielded a more uniform size of particles (results not shown). 3.1. Physical and chemical characteristics ofmicrospheres Microsphere size distribution was in the range of 1-20 }lm with more than 50% in the range of 1-3}lm (Fig. 2) which is suitable for bivalves. The log-normal distribution found is typical of emulsion droplet distribution (Groves and Fresh­ water, 1968). Values were determined with a Coulter Counter and checked by light microscopy, although light microscopy analysis seemed unsuitable for these emulsions because particles smaller than 1.5 }lm were difficult to measure and were missed, biasing the values toward a higher mean size. The stock solution personal Rectángulo ll. Heras et al. / Aquaculture 123 (1994) 309-322 Table 2 Physical and chemical characteristics of the algae and the two types of microspheres 315 Concentration (particles ml : I ) Dry weight (mg ml- I ) Lipid (mg 100ml-l) EPA (mg 100 mg- 1 lipid) DHA (mg 100 mg-' lipid) Bacteria increase (8 days) Chaetoceros muelleri 5.31 X 106 ±4.5X 104 14.7±2.5 2.9±0.3 l3.3±3.35 1.2±0.45 NS Microspheres Menhaden oil 1.86 X 108 ±2.9x 106 44.0±2.0 3780± 199 6.7±0.32 3.6±0.12 NS n-3 cone. 1.91XI08 ± 1.5 X 106 45.6±0.9 3954±86 7.6±0.58 23.2±2.46 NS JOO 200 EPA=20:5n-3, DHA=22:6n-3. Values are the means of two determinations ± s.d. NS = not significant. C/l 400 .~ C ::l e ~....:e Ctl Chaetoceros o o ~ ~ 1000 :E 800 Ctl 5 10 15 20 25 JO j.Jm I ----,-------,----'l Microspheres l 200 400 C/l 600 ~ u "E Ctl Q.-o -liE 5 10 15 20 25 JO j.Jm Fig. 2. Particle size distribution of Chaetoceros muelleri and the lipid microspheres used in the feeding experiments. Values are the mean of triplicate analysis with a Coulter Counter. had a milky-white color governed by the particle size, although the presence of pigments in the lipid mixture could give the dispersion a slight yellowish color. Visual inspection showed that microspheres tended to coalesce and agglomer- personal Rectángulo II. llna. , " Q/. I Aq"QCullu,~ / 2J(1994) 309- 111 Fie.. 3. (a) lipid miefl»phern prepw,n:d with l1umn ccnl bcadHlf d Iamete r 0.618 J.IfI1 . (b) Fluores­ (enl M"d, of diameter 0.6 18~m ronlai nni in1idc 0Yllet fttal mllenal. personal Rectángulo H. Heras et al. / Aquaculture 123 (1994) 309-322 317 80 60 OJ o o 40.----OJ 20 o PL CHO TG FFA Fig. 4. Major lipid classes of the menhaden microspheres determined over 8 days of storage at 4 0 C in amber bottles. Values are the mean of three determinations ± 1 s.d.• 0 Day; tfJ 4 day; 0 8 day; PL, polar lipids; CHO, cholesterol; TG, triacylgiycerol; FFA, free fatty acids. Table 3 Biochemically important fatty acids oflipid microspheres prepared with menhaden oil or n-3 concen­ trate as essential fatty acid sources Fatty acid group 14:0 16:0 18:0 I 16: 1 I 18: I 18: 2n-6 18: 3n-3 18:4n-3 20:4n-6 20: 5n-6 22: 5n-3 22:6n-3 I Saturates I Monounsaturated I Polyunsaturated Microsphere type Menhaden oil 6.34±0.07 16.94±0.26 5.99 ± 0.11 8.28±0.12 23.08±0.45 1O.74±0.22 1.44±0.04 1.80±0.04 0.40±0.01 7.42±0.13 1.09±0.04 4.04±0.14 30.82±0.61 32.69±0.65 36.49±2.38 n-3 concentrate 0.22±0.02 7.79±0.25 4.53±0.20 0.79±0.03 17.23±0.63 1O.98±0.28 1.00±0.01 1.07±0.03 0.54±0.01 10.08±0.13 2.39±0.07 31.85±0.30 13.42±0.49 22.33± 1.70 64.25 ± 1.12 Values are expressed as % w/w± s.d. of total fatty acids (n=3). ate on freezing and thawing, decreasing the quality of the product and the con­ centration of particles. However, the results were not as adverse as expected and this procedure could be used if necessary (results not shown). On the other hand, temperature stability compatible with bivalve holding was found to be good and temperatures of up to 21 0 C did not affect the integrity of the microspheres (as­ sessed by visual inspection). This makes the micro spheres, even those with high personal Rectángulo 318 H. Heras et al. / Aquaculture 123 (1994) 309-322 PUFA levels, a product suitable for conditions normally encountered in bivalve hatcheries. The water quality in a recirculating system also seems not to be af­ fected by the addition of the microspheres. They dispersed themselves freely, did not form a surface layer, and did not adhere to glass, metal or plastic surfaces. The technique for incorporation of fluorescent marker beads into the lipid microspheres was successful. Clearly, within each microsphere there were two or more beads, and very few beads were left outside (Fig. 3a). This combination provides a very useful tool to quantify either food intake or digestion processes (Fig. 3b) in bivalves. All-lipid microspheres differ from seawater in specific gravity and can separate during storage, but as no coalescence or agglomeration was observed the emul­ sion was easily redispersed by gentle shaking. The number of particles in the stock emulsion varied between 1.86x 108 and 1.91X 108 ml- I (Table 2), regardless of the n-3 fatty acid source employed. The efficiency of incorporation of lipid into the emulsion was ~ 99% based on the starting amounts and the amount of lipid recovered from the microsphere emulsion. The lipid class composition (Fig. 4) shows that triacylglycerols are the major component of the microspheres prepared from menhaden oil and accounted for more than 75% (w/w) of the total. Fish oil triacylglycerols are a useful and in­ expensive energy source for developing bivalve larvae. The shortening provided a CI8 fatty acid (18: 2n-6) not usually a major component of fish oils, although common in algae, and also some isomeric 18: 1monounsaturated fatty acids (Ta­ ble 3), all contributing to the high energy value of the microspheres. They can be particularly enhanced for nutrition purposes by elevated amounts of EPA and OHA offish oil concentrates, essential fatty acids in oyster larvae (Langdon and Waldock, 1981; Delaunay et al., 1989). This last fact led us to change the source of EPA and OHA and prepare the microspheres with a fatty-acid concentrate of ethyl esters containing (w/w %) 18% and 58% of EPA and DHA, respectively. Compared to other approaches to increase the EPA and DHA in the diets, such as changing the irradiance flux density or the phytoplankton cultures (Thompson et al., 1990), or lowering the temperature (Ackman et al., 1968), the present technique is less costly and time consuming. The amounts of lipids provided are several orders of magnitude higher than any level achieved by manipulation of the live algal diets. 3.2. Oxidative stability Oxidation of the micro spheres was minimized by the addition of a mixture of natural tocopherols as antioxidants in the lipid mixture, and the use of an inert atmosphere throughout the process (see Methods). However, there was concern that the high level ofPUFA, which are susceptible to oxidation, could eventually decrease the nutritional quality ofthe microspheres. To assess this possibility and that of lipid hydrolysis, samples stored at 4 0 C were followed for oxidation and composition over a period of 8 days, which is a practical and likely period of time for storage at a hatchery before use. The results (Table 4 and Fig. 4) show that personal Rectángulo H. Heras et al. I Aquaculture 123 (1994) 309-322 319 Table 4 Anisidine values of the two types of lipid microspheres stored for 8 days at 4 0 C in amber glass bottles Microsphere Menhaden oil n-3 concentrate Day 0 23.7± 1.37 34.2± 1.06 Day 4 22.6± 1.18 33.4±2.4 Day 8 23.6± 1.03 37.5±3.66 Values are the mean of three determinations (menhaden) or two determinations (concentrate) ± s.d. Values expressed as 100 times the absorbance at 350 nm in a I em cell of a solution containing I g of oil in 100 ml of solvent. neither the anisidine value nor the lipid class composition varied significantly over that period oftime. Iflipids were hydrolyzing, a decrease oftriacylglycerides and/or phospholipids and an increase in diacylglycerides would occur, but these changes were not observed. Moreover, bacterial growth, when followed for 8 days, showed no significant change in the stock emulsions of micro spheres held under conditions similar to those above, so addition of an antibiotic was not required. Gamma and delta tocopherols have a longer lifetime than alpha (Lambedet and Loliger, 1984), a characteristic found helpful for the long-term antioxidant ef­ fects needed for the storage of the lipid microspheres. In addition to their antiox­ idant effect, tocopherols serve also as a vitamin supplement for the oysters, al­ though in the wild these animals contain only alpha tocopherol (Ackman and Cormier, 1967). 3.3. Feeding experiments In all experiments with animals, microscopic observations revealed that the mixed diet of algae and microspheres was both ingested and digested. Weights and concentrations of the stocks of algae and microspheres employed in the feed­ ing experiments are shown in Table 2. Pseudofeces production was minimal and did not include fluorescent beads. Intact microspheres with fluorescent beads can be seen in Fig. 3a. Fig. 3b de­ picts fecal material from two of the oysters. Here it is apparent that the fluores­ cent beads were fully liberated, indicating that the microspheres had been di­ gested, at least partially. Remnants ofalgal material are the dominant material in this photograph. Macroscopic examination of feces of oysters fed algae revealed that normal fecal material was dark, brittle and compact in form, with long strands inside. Oysters fed the lipid micro spheres produced feces that were much lighter in color (golden brown), softer, and less compact. These differences can be ob­ served at a quick glance and are useful in determining whether the oyster has consumed the presented food. Clearly, the lipid microspheres were both ingested and digested. The unique advantage that these lipid microspheres have over other micropar­ ticles used in aquaculture is that lipases are readily available in the extracellular digestive system of the bivalves (Langdon et aI., 1985), facilitating the digestion of the lipid wall. Thus, not only the encapsulated material, but the outer coating personal Rectángulo 320 H. Heras et al. / Aquaculture 123 (1994) 309-322 of the microsphere is digested and presumably incorporated, increasing the amount of energy-providing material per microsphere. In support ofthis hypoth­ esis, Parker and Selivonchick (1986) reported uptake and distribution of radio­ labeled lipids entrapped in liposomes fed to oysters. Other advantages of these lipid microspheres over other and synthetic non-biological microcapsules include their near-neutral buoyancy, proper size range for efficient filtration by bivalve gills, and a composition of solely non-toxic and digestible materials. They have the potential to be used as carriers for many lipid-soluble compounds such as vitamins, pigments, and phagostimulants as well as hormones. This could make them useful tools in metabolic studies (e.g. vitellogenesis). The gametogenic cycle of marine bivalves is linked to synthesis of lipid during vitellogenesis at the expense of stored glycogen (Gabbott, 1975; Thompson and Macdonald, 1990). Furthermore, Gallager and Mann (1981, 1986) reported that lipids are an important energy source in the development of ova to normal larvae. Increasing the amount of essential lipids available to broodstock will increase both ova and larval quality. Alternatively, the particles can be fed to independent veliger larvae as well. Filter-feeding molluscs are known to reject particles greater than 100 J.lm (Langdon et al., 1985) during sorting. Very little is known of phagostimulants in bivalves and no trials were done to include such materials in the microspheres. In general, bivalves have a wide range of particle size preferences, ranging from 2 to 100 tux: We optimized the microsphere size to between 2 and 20 J.lm which is where maximum clearance is generally observed (Bayne, 1983). The method can be altered to provide size distributions which are optimal for different species or stages. Long-term effectsof the microspheres on condition of oysters are currently being investigated. The microspheres are also currently being tested as carriers of EPA and OHA in an in vitro system with cultured mammalian blood cells by Dr. O. Mills of Waterloo University. Acknowledgements The study was partially supported by a grant from the Natural Sciences and Engineering Research Council of Canada. The assistance of S.F.T. Venture Ltd. and the financing from the National Research Council of Canada are gratefully acknowledged. We thank Mr. G. Maillet from the Aquatron Laboratory, Dalhou­ sie University, for help in the use of the Coulter Counter. Dr. J.O. Castell from the Department of Fisheries and Oceans Canada cooperated in providing the n­ 3 fatty acid concentrate developed by the Artemia Reference Center, Belgium. We also would like to thank the Fish Health Unit (DFO) for monitoring bacterial growth. Dr. A. Robinson advised on the microsphere preparation. Mr. O. Jack­ son from the BiologyDepartment, Dalhousie University, provided assistance with the fluorescence microscopy photography. personal Rectángulo References H. Heras et al. / Aquaculture 123 (1994) 309-322 321 Ackman, R.G. and Cormier, M.G., 1967. (¥-Tocopherol in some Atlantic fish and shellfish with par­ ticular reference to live-holding without food. J. Fish Res. Board Can., 24: 357-373. Ackman, R.G. and Eaton, C.A., 1978. Some contemporary applications of open tubular gas liquid chromatography in analyses of methyl ester oflonger chain fatty acids. Fette Seifen Anstrichmittel, 80: 21-37. Ackman, R.G., Tocher, D.R. and McLachlan, J.M., 1968. Marine phytoplankter fatty acids. J. Fish Res. Board Can., 25: 1603-1620. AOCS, 1990. Method Cd 18-90. In: Official and Recommended Practices of the American Oil Chem­ ists' Society, AOCS, Champaign, IL. Bayne, B.L., 1983. Physiological ecology of marine molluscan larvae. In: K.M. Wilbur (Editor), The Mollusca. Vol 3. Development. Academic Press, New York, pp. 299-343. Bligh, E.G. and Dyer, W.J., 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol., 37: 911-917. Brown, M.R., Jeffrey, S.W. and Garland, C.D., 1989. Nutritional aspects of microalgae used in mari­ culture: a literature review. CSIRO Report No. 250,44 pp. Delaunay, F., Marty, Y. and Moal, J., 1989. Changes in the composition of the lipid classes during larval development of Pecten maximus. Oceanis, 15: 519-530. Epifanio, C.E., 1979. Comparison of yeast and algae diets for bivalve molluscs. Aquaculture, 16: 187­ 192. Gabbott, P.A., 1975. Storage cycles in marine bivalve molluscs: a hypothesis concerning the relation­ ship between glycogenmetabolism and gametogenesis. In: H. Barnes (Editor), Proc. 9th Eur. Mar. BioI.Symp., Aberdeen University Press, Aberdeen, pp. 191-211. Gallager, S.M. and Mann, R., 1981. Use of' lipid-specific staining techniques for assaying condition in cultured bivalve larvae in large scale cultures. J. Shellfish Res., I: 69-73. Gallager, S.M. and Mann, R., 1986. Growth and survival oflarvae of Mercenaria mercenaria (L) and Crassostrea virginica (Gmelin) relative to broodstock conditioning and lipid content of eggs. Aquaculture, 56: 105-121. Groves, M.J. and Freshwater, D.C., 1968. Particle-size analysis of emulsion systems. J. Pharm. Sci., 57: 1273-1291. Helm, M.M., Holland, D.L. and Stephenson, R.R., 1973. The effect of supplementary algal feeding of hatchery breeding stock of Ostrea edulis (L.) on larval vigor. J. Mar. BioI. Assoc. U.K., 53: 673­ 684. Joseph, J.D. and Ackman, R.G., 1992. Capillary column gas chromatographic method for analysis of encapsulated fish oils and fish oil ethyl esters - Collaborative study. J. AOAC Int., 75: 488-506. Lambedet, P. and Loliger, J., 1984. The fate of antioxidant radicals during lipid autoxidation. I. The tocopheroxyl radicals. Chern. Phys. Lipids, 35: 185-198. Langdon, CiJ. and Waldock, M.J., 1981. The effect of algal and artificial diets on the growth and fatty acid composition of Crassostrea gigas spat. J. Mar. BioI. Assoc. U.K., 61: 431-448. Langdon, C.S., Levine, D.M. and Jones, D.A., 1985. Microparticulate feeds for marine suspension­ feeders. J. Microencapsulation, 2: I-II. Morrison, W.R. and Smith, L.M., 1964. Preparation of fatty acid methyl esters and dimethylacetals from lipid with boron fluoride-methanol. J. Lipid Res., 5: 600-608. Napolitano, G.E., Ratnayake, W.M.N. and Ackman, R.G., 1988a. Fatty acid components of larval Ostrea edulis (L.): the importance of triacylglycerols as a fatty acid reserve. Compo Biochem. Physiol., 90B: 875-883. Napolitano, G.E., Ratnayake, W.M.N. and Ackman, R.G., 1988b. Importance oftriacylglycerides as a fatty acid reserve in larvae of the European oyster, Ostrea edulis. J. Shellfish Res., 7: 171. Parker, R.S. and Selivonchick, D.P., 1986. Uptake and metabolism oflipid vesicles from seawater by juvenile Pacific oysters (Crassostrea gigas). Aquaculture, 53: 215-228. Parrish, c.c. and Ackman, R.G., 1985. Calibration of the Iatroscan-Chrornarod system for marine lipid class analyses. Lipids, 20: 521-530. personal Rectángulo 322 H. Heraset at. I Aquaculture123 (1994) 309-322 Robinson, A., 1991. The effects of dietary algal and lipid supplements on the gonadal and larval de­ velopments of Crassostrea gigaskumatoto (Thunberg). Ph.D. Thesis, Oregon State Univ., 103 pp. Robinson, A., 1992. Dietary supplements for the reproductive conditioning of Crassostrea gigasku­ mamoto (Thunberg). II. Effects on glycogen, lipid and fatty acid content of broodstock oysters and eggs. J. Shellfish Res., 11: 443-447. Trider, D.J. and Castell, J.D., 1980. Effect of dietary lipids on the growth, tissue composition and metabolism ofthe oyster Crassostrea virginica. J. Nutr., 110: 1303-1309. Thompson, R.J. and Macdonald, B.A., 1990. The role of environmental conditions in the seasonal synthesis and utilisation of biochemical energy reserves in the giant scallop, Placopecten magellan­ icus. Can. J. Zoo!., 68: 750-756. Thompson, R.J., Harrison, P.J. and Whyte, J.N.C., 1990. Influence of irradiance on the fatty acid composition of phytoplankton. J. Phycol., 26: 278-288. personal Rectángulo