Project Title: Final Report (10-AM-02) to the. NC Fishery Resource Grant Program. Co-Principal Investigators:

Transcription

1 Project Title: Pilot Commercial Scale Testing of Promising Diets for Intensive Cultivation of Black Sea Bass and Southern flounder in North Carolina Using an Alternative Protein Source Final Report (10-AM-02) to the NC Fishery Resource Grant Program Co-Principal Investigators: Dr. Wade O. Watanabe University of North Carolina Wilmington, Center for Marine Science (UNCW-CMS) 601 S. College Rd., Wilmington, NC Tel: , Fax , Dr. Md. Shah Alam UNCW-CMS, Tel: , Fax: , Commercial Collaborators: Ted M. Davis (Aqua Plantations, LLC), Shawn Longfellow (Blue Ocean Farms, LLC), Keith Hairr (Carolina Flounder, LLC), E: August 14,

2 Abstract Under commercial scale conditions and practical growout durations, the growth performance of hatchery-reared black sea bass and southern flounder fed UNCW-formulated diets in which fish meal protein was replaced by soybean meal protein at the levels of 30% (southern flounder) and 50% (black sea bass) was compared with the performance of fish fed a high fish meal-based (0% soybean meal) control diet. Diets were prepared by a NC feed manufacturer (Carolina Feeds, Ayden) and were tested at three startup commercial marine fish farms in NC: Aqua Plantations (Wilmington), Carolina Feeds (Wallace), and Blue Ocean Farms (Jacksonville) to transfer technology to commercial practitioners. For both species, fish were grown from advanced juvenile stages (~ g) to marketable stages (~ g) on the high soybean meal-based diets, with no apparent impairment of growth performance compared to the high fish meal-based diet. Biochemical analyses of fish whole bodies and fillets revealed that differences in proximate and fatty acid profiles between fish fed high fish meal-based and high soybean meal-based diets were minimal and were not expected to affect product quality. Fish were marketed live, in-the-round on ice, or filleted to various wholesale and retail outlets in NC, or metropolitan markets on the eastern US seaboard, with excellent consumer acceptance. Compared to the high fish meal-based control diet, the substitution of 30 and 50% soybean meal protein for fish meal protein and producing feeds in North Carolina, lowered feed prices by 15.6 and 26.0%, respectively. Results are important toward lowering the cost of finfish production and reducing the incorporation of fish meal in aquaculture feeds, while providing the opportunity for additional growers to enter into marine finfish production. Results will promote the development of the nascent finfish mariculture industry in NC by providing fundamental, practical information on a key input component to commercial production - i.e., nutritious, cost-effective, and environmentally-friendly practical 2

3 diets for production of black sea bass and southern raised in intensive recirculating aquaculture systems. Introduction Background and Rationale Protein is the most expensive nutrient in fish diets, representing from 30 to 60% dry wt. Fish meal (FM) is typically used as the main protein source in formulated diets due to its high nutritive value and palatability. Finding alternatives to FM is needed to reduce costs and reliance on this expensive and finite protein source. Soybean meal (SBM) is considered to be one of the most suitable alternative ingredients for replacing FM in fish feed (Webster et al. 1995, Kikuchi 1999, Lunger et al. 2006). Our laboratory studies have shown that in black sea bass (BSB) and in southern flounder (SF), FM protein may be successfully replaced by SBM protein (without amino acid supplements) at levels of 70% and 40%, respectively, in a 45% crude protein (CP) diet (Alam et al. 2007, 2008b, 2009c; Sullivan et al. 2008). These laboratory findings revealed a potential for substantially reducing feed costs for commercial aquaculture of BSB and SF using sustainable alternatives to fish meal. Studies are needed to test the most promising diets at the pilot and commercial scales. The black sea bass Centropristis striata (family Serranidae) is a high-value marine finfish that inhabits continental shelf waters from Maine to northern FL (Wenner et a. 1986). Black sea bass are heavily exploited south of Cape Hatteras and fishery regulations are becoming more restrictive (NCDMF 2009b), with commercial landings unable to satisfy increasing consumer demand. There is increasing interest in aquaculture production of black sea bass due to its high market price and the increasing level of details on its culture. Based on research supported by NOAA, NC Sea Grant and USDA, technology for producing embryos (Watanabe et al., 2003; Berlinsky et al. 2005), rearing 3

4 larvae through juvenile stages (Atwood et al. 2001, 2003, 2004; Berlinsky et al. 2000, 2001; Berlinksy et al. 2004, Copeland and Watanabe 2006) and raising fingerlings to full marketable sizes in intensive RAS have been developed (Copeland et al. 2002, 2003, 2005; Alam et al. 2008a, 2009b, Watanabe 2011). Cultured black sea bass are being test-marketed throughout NC and in metro markets of the northeast (Philadelphia, NYC) and is considered to be of exceptional quality (Wilde et al. 2008, Dumas and Wilde 2009). Black sea bass is now in the transition phase between research and commercial production in NC. A pilot demonstration project (Aqua Plantations LLC) is in progress at the UNCW Aquaculture Facility on Wrightsville Beach, and a startup farm (Coastal Aquafarms) has been established in Hubert, NC. Commercial production has also started in Virginia (Mid Atlantic Aquatic Technology LLC). The southern flounder Paralichthys lethostigma is a high-valued recreationally and commercially over-harvested marine flatfish inhabiting estuarine and shelf waters of the south Atlantic and Gulf coasts of the U.S. (Gilbert 1986, NCDMF 2009a). Declining natural populations and wide temperature and salinity tolerances make southern flounder a versatile candidate for intensive cultivation (Berlinsky et al. 1996; Daniels and Borski 1998; Jenkins and Smith 1999; Smith et al. 1999; Daniels and Watanabe 2003). Under NC Sea-Grant- and USDA-supported research, technology for commercial scale production of southern flounder embryos (Watanabe et al. 2001; Watanabe et al. 2006) and for rearing larvae through juvenile stages (Daniels and Watanabe 2003; Henne and Watanabe 2003; Moustakas et al. 2004; Mangino and Watanabe 2006; Alam et al. 2009a) is available. Juveniles have been grown to full marketable sizes in intensive recirculating aquaculture systems (RAS) in both fresh water and seawater using commercial diets (Daniels et al. 2006). Startup commercial operations are underway in NC, including Carolina Flounder LLC in Wallace, Little River Trails Aquaculture in Bunnlevel, and Coastal Aquafarms at the Sturgeon City Environmental Education Center in Jacksonville. 4

5 Objectives The goal is to accelerate the development of the nascent marine fish culture industry in NC by developing a key input component to commercial production - i.e., a nutritious, cost-effective, and environmentally-friendly practical diet for commercial production of BSB and SF raised in intensive RAS. Specific objectives are as follows: (1) Test the most promising diets from previous laboratory studies in pilot-commercial scale RAS, (2) Compare the growout performance to full marketable sizes of hatchery-reared SF and BSB fed diets in which FM is replaced by SBM at the levels of 30% (SF) and 50% (BSB), (3) Transfer technology to commercial aquaculturists by using their facilities, (4) Source diets in NC by involving a local feed manufacturer, and (5) Determine the effects of alternative feed-based diets on product quality. Methods Experimental Systems and Animals The collaborating North Carolina fish farms included one production facility for black sea bass (Aqua Plantations, Wilmington, NC) and two for southern flounder (Carolina Flounder, Wallace and Blue Ocean Farms, Jacksonville). At the Blue Ocean Farms and Aqua Plantations, four 4.61-m diam. tanks (vol. = 16.7-m 3 ) were used for the experiments, while at the Carolina Flounder facility, four 4.61-m diam tanks (vol. = 24-m 3 ) were used. At these facilities, the production tanks were supported by a recirculating aquaculture system (RAS), including particle traps, micro-particulate filter, low space bioreactor, UV sterilizer, oxygen saturator cone, and heat pump. At Aqua Plantations, juvenile black sea bass Centropristis striata were produced at UNCW Aquaculture Facility from photothermally conditioned broodstock and induced to spawn using established procedures (Watanabe 2011). At Carolina Flounder, juvenile southern 5

6 flounder Paralichthys lethostigma were raised from eggs supplied from the UNCW Aquaculture Facility (Wrightsville Beach, NC) where adult broodstock were photothermally conditioned and induced to spawning using established procedures (Watanabe et al. 2006). Southern flounder juveniles at Blue Ocean Farms were sourced from the Carolina Flounder hatchery facility. Experimental Design and Feed Formulation Based on our previous laboratory studies at UNCW, a control, high fish meal-based diet and two cost-effective, optimized formulations of sustainable diets containing 45% crude protein (CP) and 11% crude lipid (CL) were formulated to incorporate reduced levels of fish meal by substituting combinations of alternative protein sources (Table 1). A control diet (Diet 1) contained a high percentage of fish meal (60%) as the primary protein source. Diet 2 formulated for southern flounder replaced 30% fish meal protein by alternative terrestrial animal (poultry meal) and vegetable (soybean meal and corn meal) feedstuffs. Diet 3 formulated for black sea bass replaced 50% fish meal protein by these alternative feed stuffs. All other ingredients were according to nutrient requirement information available for black sea bass and southern flounder (Alam et al. 2011, 2012) (Table 1). At each facility, two replicate tanks were assigned to each of two dietary treatments, consisting of a high fish meal-based control diet (diet 1) and a high soybean meal-based diet (diet 2, 30% fish meal protein replacement for southern flounder and diet 3, 50% fish meal protein replacement for black sea bass). Slow-sinking extruded diets (8 mm pellet size) were prepared by a collaborating North Carolina feed manufacturer (Carolina Feeds, Ayden) and then supplied to the 3 collaborating fish farms, where their growth performance was compared on the UNCW-formulated test diets under commercial scale conditions. The first order of approximately 7,368 lbs of experimental feed (Diets 1, 2 and 3) was delivered to the UNCW Aquaculture Facility on August 16, 2010 and 6

7 stored at 21 C. Feed was distributed to the Blue Ocean Farms and Carolina Flounder facilities as required. The second order of 11,600 lbs of feed was delivered to UNCW on April 04, 2011 and then distributed to the Blue Ocean Farms and Carolina Flounder facilities. Feeding Trials and Protocols To begin the feeding trial at each farm, initial fish weights and lengths were measured in a sample of 100 fish from each tank. Three to five fish from each farm were sacrificed for biochemical analyses (proximate composition and whole body fatty acid profiles) and stored in a freezer at 20 C. A feeding protocol was developed and provided to each farmer at the start of the feeding trial to ensure that administration of feed, rearing conditions, water quality, monitoring of growth performance and mortalities, and collection of data were standardized among facilities. Fish were fed two times a day as much as they could consume (without wastage) in about min. The daily ration was adjusted according to this observation and recorded. Feed consumption was calculated as a percentage of tank biomass consumed per day for the duration of the rearing period. Feed conversion ratios (FCR) were calculated as the total biomass gain during the rearing period (not adjusted for biomass lost through mortality) divided by the total feed administered during that period. A standardized data sheet was distributed to each farm to monitor and record ammonia, nitrite, and nitrate. In addition, magnesium, calcium, and potassium levels were recorded at both southern flounder growout facilities, which utilized very low to moderate salinity brackishwater (2-15 g/l) for their culture operations. Electrolytes were not measured at the black sea bass growout facility, which used full strength seawater (32 g/l) from natural sources for culture. Sampling was conducted monthly during the 10-month feeding trial at Aqua Plantations and at 7

8 Blue Ocean Farms. At Carolina Flounder, sampling was conducted monthly, but for duration of only 4 months as the experiment was terminated early due to high mortality. Black Sea Bass Growout (Aqua Plantations) The feeding trial at the Aqua Plantations facility was started on September 15, Black sea bass (mean initial weight = 145 g) were stocked into four16.7-m 3 tanks at a density of approximately 1,200 fish per tank. Approximately 150 fish were sampled monthly to monitor the growth. The amount of feed administered to each tank and the observed mortalities were recorded daily. The experiment was terminated on July 19, 2011 after 10 months of feeding, and a sample of three fish from each tank were collected and stored at -20 o C for biochemical analysis as described below. On July 27, 2011, all of the fish in each of the four treatment tanks were counted to determine final survival. Southern Flounder Growout (Blue Ocean Farms and Carolina Flounder) Due to availability of fingerlings, the feeding trials at Blue Ocean Farms (Jacksonville, NC) were conducted as two separate trials, one (Trial 1) starting on January 28, 2011 and the second (Trial 2) starting on April 24, Each trial used two production tanks, one fed diet 1 (high fish meal-based control) and the other fed diet 2 (30% replacement of fish meal protein with soybean meal protein). To begin Trial 1, southern flounder juveniles (mean initial weight = 255 g) were stocked into each of two 16.7-m 3 tanks at a density of 650 fish per tank. To begin Trial 2, juveniles (mean initial weight = 201 g) were stocked into each of two additional 16.7-m 3 tanks at a density of 455 fish per tank. For both feeding trials, fish were sampled for growth monthly. Both of the feeding Trials 1 and 2 were terminated on December 2, 2011 after 10 and 7 months of feeding, respectively. Six fish from each tank were collected for biochemical 8

9 Table 1. Composition of diets (g/100 g). Control (Diet 1) S. flounder (Diet 2) Black sea bass (Diet 3) Percent FM protein 0% 30% 50% replacement: Ingredients Menhaden fish meal Solvent extracted soybean meal Poultry meal Wheat mids Ground corn Fish oil Binders Mineral premix DiCal phosphate Vitamin premix Stay-C Total Proximate composition Control S. flounder Black sea bass (% dry basis) Protein Lipid

10 analysis. On December 7, 2011, all fish in each of the four tanks were counted to determine final survival. The feeding trial at the Carolina Flounder facility was started on September 15, 2010 when each of four 24-m 3 tanks were stocked with juvenile southern flounder (mean initial weight = 119 g). However, heavy mortalities affecting approximately 60% of the fish were observed within 30 days of stocking, presumably related to mineral deficiencies in the groundwater, a problem that had previously affected production of flounder in this low salinity (8-10 g/l) system at this facility. Given these mortalities, it was decided to restart the experiment using a different group of fish from the same cohort. On October 6, 2010 the salinity of the recirculating system was increased to 12 g/l and fish health improved. On December 1, four tanks 24-m 3 were each stocked with 950 southern flounder using additional healthy fish from the same cohort. Fish remained healthy through January 19, 2011 when the experiment was re-started. Biochemical Analyses Proximate composition of all experimental diets and terminal samples of body tissues of fish for all three experiments were analyzed at UNCW Center for Marine Science. Crude protein was determined by the Kjeldahl method with a Labconco Kjeltec System (Rapid Digestor, Distilling Unit-Rapid Still II and Titration Unit, Labconco Corporation, Kansas city, MO, USA) using boric acid to trap ammonia. Crude lipid (Soxhlet by ether extraction), ash (BARNSTAD Thermolyne Muffle Furnace, IA, USA) and moisture (Fisher Scientific Isotemp Oven, Pittsburgh, PA, USA) contents in the diets were analyzed by standard methods (AOAC, 2000). 10

11 Moisture contents in whole bodies were determined by drying the fish in a freeze dryer (Labconco Freeze Dryer, Kansas City, MO, USA). Fatty acid composition of the diets and fish fillets was determined by first extracting total lipids into chloroform: methanol (Folch et al. 1957). One ml of a g ml -1 solution of C19:0 fatty acid was added to each sample as an internal standard. Lipid fatty acids were converted to their methyl esters (FAMEs) for GC analysis by refluxing the concentrated lipid sample in 1.0 ml of 0.5 M NaOH/MeOH for 30 min., followed by addition of 1.5 ml of boron trifluoride-methanol (BF 3 ) and refluxing for an additional 30 min. The FAMEs were extracted into hexane, concentrated and dissolved in 1 ml of chloroform. GC analysis was performed on a HP-6890 Gas Chromatograph using a 25 meter x 0.25 m HP-5 capillary column with FID detection (Department of Chemistry and Biochemistry, UNCW). Helium was used as the carrier gas. The column temperature profile was: 195 C, hold for 8 min, ramp to 270 at 15 C min -1 and hold at 270 C for 2 min. FAME peaks were integrated using the HP Chemstation software package and individual FAMEs were identified by comparison of retention times to standards: GLC-84 (Nu Chek Prep U.S.A) as well as individual standards of steariodonic, eicosapentaenoic and arachidonic acid methyl esters (Sigma-Aldrich, MO, U.S.A). FAMEs from all samples were quantified using their peak areas compared to the peak area of the C19:0 internal standards. 11

12 Results Growth Performance and Feed Utilization Black Sea Bass (Aqua Plantations) Growth of black sea bass at the Aqua Plantations farm over a 328-d study period on the high fish meal-based (control) diet 1 and the high soybean meal-based diet 3 (50% FM protein replaced by soybean protein) is compared in Fig. 1. At stocking, mean fish weight was 145 g in both treatments. Growth of fish under these two diet treatments was very similar for the duration of the experiment. However, after d 84 post-stocking, mean weights of fish fed the high soybean meal-based diet 3 were consistently higher than in fish fed the high fish meal-based control diet 1. At terminal sampling on d 328 post-stocking, no differences in growth were evident between fish fed the high fish meal-based control diet 1 (mean weight + SD = g, weight gain = 296 g) and fish fed the high soybean meal-based diet 3 (mean weight = g, weight gain = 315 g). Fish fed a high fish meal-based (control) diet 1 had an average survival of 82% (range = 70-93%) compared to 95% (range = 93-96%) for fish fed a high soybean-based diet 3. The mortalities observed during this study were primarily related to mechanical problems, including cavitation of the high-pressure centrifugal pumps causing nitrogen super saturation, or to excessive dissolved oxygen levels that were related to operation of the high-pressure oxygen cones used in the recirculating system. Feed consumption (% body wt./d) of fish fed the high soybean meal-based diet 3 (mean = 0.64%/d, range = %/d) was similar to fish fed the high fish meal-based diet 1 (mean = 0.53%/d, range = %/d). FCR was 1.58 (range = ) for fish fed the soybean mealbased diet 3 and 1.45 (range = ) for fish fed the high fish meal-based control diet 1. 12

13 Fig. 1. Growth of black sea bass fed a high fish meal-based diet 1, or a high soybean meal-based diet 3 in 16-m 3 recirculating seawater tanks (N = 2 per treatment) for 10 months at the Aqua Plantations farm. Each tank was stocked with 1,167 hatchery-reared fish. Plotted points represent means of 150 fish. The ranges of water quality parameters during the growout period were as follows: temperature , ph , salinity g/l, dissolved oxygen mg/l, ammonia mg/l and nitrite mg/l. Southern Flounder Growout (Blue Ocean Farms and Carolina Flounder) Blue Ocean Farms At Blue Ocean Farms, two tanks (one per treatment) were stocked in January 2011 (Trial 1) and fed their respective experimental diets (Diets 1 and 2) for 10 months (Fig. 2), while two additional tanks were stocked in April 2011 (Trial 2) and were fed these diets for seven months (Fig. 3). 13

14 Trial 1 (10-month growout) For trial 1, initial fish weight was 255 g for both treatments (Fig. 2). Mean fish weights for both treatments remained similar, with no departures between these diet treatments throughout the study. On d 307 post-stocking, mean fish weight was 437 g for fish fed the high fish meal-based (control) diet 1 and 433 g for fish fed the high soybean meal-based diet 2 (Fig. 2). In trial 1, survival of fish was 76% and 61% for the fish fed the high fish meal-based and high soybean meal-based diets 1 and 3, respectively. In trial 1, feed intake (% body wt./d) was 0.33 and 0.28% for fish fed the high fish meal-based and high soybean meal-based diets, respectively. FCR of the high soybean meal-based diet 2 (1.40) was slightly lower than the control diet 1 (1.82). Fig. 2. Growth of southern flounder fed a high fish meal-based (control) diet 1 or a high soybean meal-based diet 2 in 16-m 3 recirculating brackishwater (8-10 g/l) tanks over 10 months at Blue Ocean Farms. Each tank was stocked with 650 hatchery-reared fish. Plotted points represent means of 150 fish. 14

15 Trial 2 (7-month growout) For Trial 2, initial fish weight was g for both treatments and remained similar throughout the study (Fig. 3). On 226 day post-stocking, mean fish weight was 423 g for fish fed the high fish meal-based (control) diet 1and 393 g for fish fed the high soybean mealbased diet 2 (Fig. 3). In Trial 2, final survival was 64% and 66% for fish fed the high fish mealbased diet 1 and the high soybean meal-based diet 2, respectively. In trial 2, feed intake was 0.32 and 0.28% for fish fed the high fish meal-based and high soybean meal-based diets, respectively. FCR of the high soybean meal-based diet 2 (1.40) was slightly lower than the control diet (1.45). Fig. 3. Growth of southern flounder fed a high fish meal-based (control) diet 1 or a high soybean meal-based diet 2 in 16-m 3 recirculating brackishwater (8-10 g/l) tanks (N = 2 per treatment) over 7 months at Blue Ocean Farms. Each tank was stocked with 455 hatchery-reared fish. Plotted points represent means of 150 fish. 15

16 The ranges of water quality parameters during feeding trials 1 and 2 at Blue Ocean Farms were as follows: temperature , ph , salinity 6-20 g/l, dissolved oxygen mg/l, ammonia mg/l, and nitrite mg/l. The analyzed electrolytes level (ppm) at Blue Ocean Farms during the feeding trials were recorded as follows: sodium 2,674-3,089, phosphorous , potassium , calcium , magnesium , iron , manganese , sulfur , zinc , boron , chloride 11, Total alkalinity and hardness (ppm) during the feeding trial were and 1,610-1,441, respectively. Carolina Flounder Growth trials on southern flounder at the Carolina Flounder farm over a 4-month study period on the high fish meal-based (control) diet 1 and the high soybean meal-based diet 2 were compromised by poor fish health, presumably related to deficient ionic composition of the groundwater sourced at the site. Starting from mean initial weights of 158 ± 8.7 and 147 ± 8.03 g in the high fish meal-based and high soybean meal-based diet treatments, respectively, final mean weights of 221 ± 1.41 g and 202 ± 14.8g, respectively, were attained by 120 d poststocking. Chronic mortalities caused survival to decline to approximately 30% in all four tanks. Hence, fish from the duplicate tanks for each diet treatment were combined into one tank to enable the treatment comparisons to continue with a single replicate tank per treatment. Mortalities persisted during this final phase of the study, and survival (measured from restocking) declined to around 50% by d 60 post-stocking when the study was terminated. A final sampling on d 60 post-stocking showed fish fed the high fish meal-based diet 1 with a mean weight of 230 g, while fish fed the high soybean-based diet 2 had a mean weight of 211 g. 16

17 The ranges of water quality parameters during the feeding trials at Carolina Flounder were as follows: temperature , ph , salinity g/l, dissolved oxygen mg/l, ammonia mg/l, and nitrate mg/l. The analyzed electrolytes level (ppm) at Carolina Flounder during the feeding trial were recorded as follows: sodium 2,410-2,905, phosphorous , potassium , calcium , magnesium , iron , manganese , sulfur , zinc , boron , chloride 1,515-2,556. Total alkalinity and hardness (ppm) during the feeding trial were and 954-1,800, respectively. Diet Effects on Whole Body Proximate Composition Black Sea Bass Growout (Aqua Plantations) Whole body and muscle samples from black sea bass fed a high fish meal-based diet 1and a high soybean meal-based diet 3 at Aqua Plantations were analyzed for initial (Table 2) and final (Table 3) proximate composition. Initial (pre-treatment) moisture content of whole body tissue from fish in both treatments ranged from % (Table 2). Initial protein, total lipid and ash ranged from %, % and %, respectively (Table 2). After 10 months of feeding black sea bass on the high fish meal-based (control) diet 1 and the high soybean meal-based diet 3, final proximate composition was similar between diet treatments: moisture ( %), protein ( %), lipid ( %) and ash ( %) (Table 3). Final moisture ( %) was lower than initial ( %), while final lipid ( %) was higher than initial ( %) (Table 3). 17

18 Table 2. Initial (pre-treatment) whole body proximate compositions contents of black sea bass (% wet weight basis) fed a high fish meal-based (control) diet 1 and a high soybean meal-based diet 3. Data are means + SD (N = 3). Diets Moisture Protein Lipid Ash D1 (FM) 61.4 ± ± ± ± 0.17 D3 (SBM) 63.2 ± ± ± ± 0.12 Table 3. Final whole body proximate compositions of black sea bass (% wet weight basis) after 10 months of feeding on the high fish meal-based diet 1, or a high soybean meal-based diet 3. Data are means + SD (N = 3). Diets Moisture Protein Lipid Ash D1(FM) 58.4 ± ± ± ± 0.56 D3 (SBM) 57.4 ± ± ± ± 0.43 Southern Flounder Growout (Blue Ocean Farms) Initial and final proximate composition of whole body samples from southern flounder reared at Blue Ocean Farms is presented in Table 4. In both Trials 1 and 2, initial (pre-treatment) moisture content of whole body tissue from fish in both treatments ranged from % (Table 4). Initial protein, total lipid and ash ranged from %, % and %, respectively (Table 2). After 10 months of feeding on the high fish meal-based (control) diet 1 and the high soybean meal-based diet 2, final proximate composition of southern flounder was similar between diet treatments: moisture ( %), protein ( %), lipid ( %) and ash ( %). Final proximate composition was similar to initial values for all parameters. 18

19 Table 4. Initial and final whole body proximate composition of southern flounder (% wet weight basis) at Blue Ocean Farms in Trials 1 and 2. Data are means ± SD (N = 3), except for initial moisture (N = 1). Blue Ocean Farms (Southern Flounder) Initial (10-month study, Trial 1) Diets Moisture Protein Lipid Ash FM ± ± ± 0.66 SBM ± ± ± 0.40 Initial (7-month study, Trial 2) Diets FM ± ± ± 0.20 SBM ± ± ± 0.11 Final (10-month study, Trial 1) Diets FM 68.1 ± ± ± ±0.14 SBM 67.6 ± ± ± ± 0.47 Final (7-month study, Trial 2) Diets FM 67.6 ± ± ± ± 0.58 SBM 66.3 ± ± ± ±

20 Total Lipid in Fish Fillets Total lipid contents (% dry weight) in black sea bass (Aqua Plantations) and southern flounder (Blue Ocean Farms) fillets after the feeding trials are presented in Table 5. In black sea bass, total lipid in fillets was similar in fish fed the high-fishmeal-based control diet 1 (27.2%) and in fish fed the high soybean meal-based diet 3 (26.4%). Total lipid contents in southern flounder fillets (14.2%) from fish fed the high fish meal-based control diet 1 were slightly higher than in fish fed the high soybean meal-based diet 2 (12.4%). The total lipid contents in black sea bass fillets ( %) were approximately twofold higher than in the flounder fillets ( %). Table 5. Total lipid (% dry wt.) of fish fillets after the feeding trials comparing a high FM-based diet (control diet 1, 0% FM protein replacement with SBM protein) and a high soybean mealbased diet 2 (30% FM protein replacement with SBM protein) in southern flounder and diet 3 (50% FM protein replacement with SMB protein) in black sea bass. Values are means (N = 3). Black Sea Bass (Aqua Plantations) Diets Total Lipid Diet 1 (control) 27.2 ± 3.97 Diet 3 (SBM) 26.4 ± 2.66 Southern Flounder (Blue Ocean Farms, Trial 1, 10-month study) Diet 1 (Control) 14.2 ± 1.37 Diet 2 (SBM) 12.4 ± 0.33 Fatty Acid Composition of Diets and Fish Fillets Fatty acid compositions (mg/g dry wt.) of diets 1 (control, 0% FM protein replacement with SBM protein), 2 (30% FM protein replacement with SBM protein) and 3 (50% FM protein 20

21 replacement with SMB protein) are presented in Table 5. The highly unsaturated omega-3 fatty acid EPA (20:5n3) ranged from in diet 3 and in diet 1, but was lower at 6.45 in diet 2. DHA (22:6n-3) levels were higher in the control diet 1(11.13) than in diet 2 (5.88) or diet 3 (7.87). Total saturated fatty acids (SFA) were higher in diet 1 (30.52) and diet 3 (34.51) than in diet 2 (28.62). Total monounsaturated fatty acids (MUFA) were also higher in the control diet 1 (30.99) and diet 3 (34.62) than in diet 2 (25.74). Total n-3 PUFAs was higher in diet 1 (28.87) and in diet 3 (24.25) than in diet 2 (15.01). Total n-6 PUFAs ranged from among diets. The ratio of n-3 to n-6 PUFAs decreased from 2.76 in the control diet 1 to 1.62 in the high soybean meal-based diet 3. Fatty acid analysis of fish fillets from southern flounder (Blue Ocean Farms, Trial 1, 10 month study) and black sea bass (Aqua Plantations) after the feeding trials are also presented in Table 5. For southern flounder, post-treatment fatty acid profiles of fillets were similar between fish fed the high fish meal-based control diet 1 and the high soybean meal-based diet 2. For fillets of southern flounder fed the control and high soybean meal-based diet, respectively, concentrations (mg/g dry wt.) of total SFA (17.56 and 15.98), total MUFA (17.93 and 15.32), EPA (4.19 and 3.87), DHA (8.59 and 7.22, total n-3 PUFAs (16.19 and 14.3), and total n-6 PUFAs (7.27 and 5.40), and ratios of n-3/n-6 PUFAS ( ) and DHA/EPA ( ) were similar between diet treatments. For black sea bass, some notable differences in fatty acid profiles of fillets were evident between fish fed the high fish meal-based control diet 1 and the high soybean meal-based diet 3 at the end of the feeding trials (Table 5). For post-treatment n-3 PUFAs in fillets of fish fed the control and high soybean meal-based diets, respectively, concentrations of EPA (22.17 vs ), DHA (24.09 vs. 18.4), total n-3 PUFAs (57.5 vs. 48.2) and ratios of n-3/n-6 PUFA (

22 vs. 1.35) were generally higher in fish fed the control diet 1 than in high soybean meal-based diet 3. However, post-treatment concentrations of total n-6 PUFAs (35.51 vs ) and total MUFAs (98.23 vs ) were higher in fillets of fish fed the high soybean meal-based diet 3 than in fish fed the control diet 1. Total SFA ( ) and ratios of DHA/EPA ( ) were similar between these diet treatments. When post-treatment fatty acid profiles in fillets were compared between southern flounder and black sea bass, striking differences were observed (Table 5). In general, total SFA levels in black sea bass fillets fed both the high fish meal-based control diet 1 and the high soybean meal-based diet 3 (93.76 and 91.39) were markedly higher than those measured in southern flounder (17.56 and 15.98) for these treatments. Total MUFAs were also higher in black sea bass (92.24 and 98.23) than in southern flounder fillets for both treatments. The n-3 PUFA levels in black sea bass fillets fed both the high fish meal-based control diet 1 and the high soybean meal-based diet 3 (57.5 and 48.02) were much higher than those measured in southern flounder fed diet 2 post-treatment (16.19 and 14.3). Likewise, post-treatment n-6 PUFA levels in black sea bass fillets fed both the high fish meal-based control diet 1 and the high soybean mealbased diet 3 after the feeding trials (27.96 and 35.51) were markedly higher than those measured in southern flounder (7.27 and 5.40). Ratios of n-3/n-6 PUFA and DHA/EPA were lower in black sea bass (1.35 and 0.93) than in southern flounder (2.65 and 1.87). 22

23 Table 5. Fatty acid analysis (mg/g dry sample) of southern flounder and black sea bass muscle (fillet) after the feeding trials. Values are means (N = 2). Diets Southern Flounder Black Sea Bass Fatty Acid Control Flounder BSB Fillet Fillet Fillet Fillet Diet 1 Diet 2 Diet 3 0% 30% 0% 50% 14: :1n : :4n :2n :1n : :5n-3 (EPA) :1n :6n-3 (DHA) :5n Σ SFA Σ MUFA Σ n-3 PUFA Σ n-6 PUFA n-3/n-6 PUFA DHA/EPA

24 Discussion Black Sea Bass Growout In this study, a 10-month feeding trial conducted under pilot commercial scale conditions demonstrated that a UNCW-formulated diet (developed through controlled laboratory studies) (Alam et al. 2012) replacing 50% fishmeal (FM) protein by soybean meal (SBM) protein (diet 3) did not decrease growth performance, feed utilization and survival of black sea bass compared to a high fish meal-based control diet (diet 1). These results are consistent with our laboratory findings (Alam et al. 2012) that black sea bass juveniles are able to efficiently utilize diets replacing up to 58% FM protein with soybean meal protein for growth (Alam et al. 2011). The earlier laboratory studies were conducted under controlled-environment conditions and raised small juveniles (5-10 g initial weight) in triplicate 75-L aquaria to maximum sizes of g over relatively short time periods (40-70 d). While the same level of replication and control of experimental conditions could not be practically achieved at the commercial facilities participating in the present study, the results validate these laboratory findings by demonstrating that black sea bass can be raised on a high soybean meal-based diet (based on laboratory diet formulations) under commercial scale rearing conditions over an extended time period from advanced juvenile (~ 145 g) to full marketable sizes (> 450 g) without a diminution of performance compared to fish fed a high fish meal-based control diet. In the present study, the successful replacement of 50% FM protein with SBM protein in practical diets for black sea bass was similar to or higher than the maximum replacement levels determined in the laboratory for other marine finfish species such as Japanese flounder Paralichthys olivaceus (45%, Kikuchi, 1999); gilthead sea bream Sparus aurata (45%, Martinez- Llorens et al. 2008); yellowtail Seriola quinqueradiata (20%, Shimeno et al. 1993), cobia 24

25 Rachycentron canadum (50%, Zhou et al. 2005) and Atlantic cod Gadus morhua (50%, Walker et al. 2010), but lower than that found for freshwater fish, such as carp Cyprinus carpio (100%, Viola et al. 1982) and Nile tilapia Oreochromis niloticus (100%, Deyab et al. 2002). The ability of black sea bass to grow efficiently on a diet replacing 50% FM protein with SBM protein indicates a similar ability to digest and assimilate SBM protein as cobia and Atlantic cod (50% replacement) and a superior ability compared to other marine finfish such as yellowtail (Shimeno et al. 1993). This may indicate that black sea bass possess specific digestive enzymes, but further investigation is needed. Protein digestibility of the test feeds was not measured in the feeding trials; however, in several species it has been suggested that the high growth performance and high feed utilization of fish fed high substitution levels of soybean meal for fish meal could be due to higher digestibility of nitrogen and energy in soybean meal than in fish meal (Storebakken et al. 2000; Hendricks, 2002; Martinez-Llorens et al. 2008). Although relative e feed consumption (FC) and feed conversion ratios (FCRs) of black sea bass fed the high fish meal-based control diet 1 and high soybean meal-based diet 3 in this study could not be compared statistically due to inadequate replication, consumption rates (FC = 0.53%/d and 0.64%/d, respectively) and feed conversion ratios (FCRs =1.45 and 1.58, respectively) were similar between these diet treatments, suggesting that appetency and growth efficiency of black sea bass were not affected by replacing 50% FM protein by soybean meal protein (diet 3) during growout from large juvenile (~150 g) to marketable stages (~ 450 g). FC and FCR in the present study (0.53%/d and 0.64%/d, 1.45 and 1.58) were comparable to those (FC = 0.34%/d, FCR = ) reported for wild-caught black sea bass raised from subadult ( g) to large marketable sizes (873-1,051 g) over 221 days on commercially prepared diets with different crude protein ( %) and crude lipid ( %) levels (Copeland et 25

26 al. 2002). These authors reported that FC was not constant throughout the 221 day study, but declined from %/d during the first 56 d to %/d from days (Copeland et al. 2002). In a subsequent study by these authors (Copeland et al. 2003), much higher feed consumption rates of %/d (at similar FCRs of ) were reported for wild-caught black sea bass raised from subadult ( g) to large marketable sizes ( g) over 201 days on a commercially prepared diets with 50% crude protein and 12% crude lipid. While the reasons for these differences in FC are not clear, they could also be related to ontogenetic variation in FC with life stage, as well as to differences in gross energy levels in the diets, or other environmental (e.g. temperature) or rearing conditions (e.g. stocking density) used in these studies (Copeland et al. 2002, 2003). In the present study, FCRs of fish fed the high fish meal-based control diet 1 (1.45) and the high soybean meal-based diet 3 (1.58) were somewhat lower than the FCRs ( ) previously reported for laboratory scale studies of early juvenile black sea bass raised on 44% crude protein and 13% crude lipid diets in which FM protein was replaced with SBM protein at levels ranging from 0 to 60% (Alam et al. 2012). These differences in FCRs among studies could also be related to differences in temperature or life stage which are known to influence FCR in black sea bass. Copeland et al. (2002) reported that FCRs in black sea bass ranged from 1.0 to 2.7 for wildcaught subadults raised in a recirculating system, depending on temperature (Copeland et al. 2002). FCRs of fish fed the high fish meal-based control diet 1 (1.45) and the high soybean meal-based diet 3 (1.58) in this study were also higher than those previously achieved with black sea bass reared from small juvenile (276 g) to full marketable sizes (682 g) in a recirculating system (FCR = ) using a commercially prepared diet (Skretting, Vancouver, BC, Canada) containing 55% protein and 18% lipid (Watanabe 2011). However, the diets used in the current study 26

27 contained much lower levels of both crude protein (44%) and crude lipid (13%), and these results are not directly comparable. Furthermore, since FCRs in the present study were calculated without adjustment for fish mortality (which ranged from 18% for the high fish meal-based control diet 1 to 5% for fish fed a high soybean-based diet 3), and an adjustment for biomass losses would further reduce these FCR values, calculated as the weight fed divided by weight gain (including biomass lost through mortality). Southern Flounder Growout The 7- and 10-month commercial feeding Trials 1 and 2 with southern flounder conducted under pilot-scale commercial scale conditions at Blue Ocean Farms showed that a UNCW-formulated diet (Alam et al. 2011) replacing 30% fishmeal (FM) protein by soybean meal (SBM) protein (diet 2) did not reduce growth performance, feed utilization and survival of southern flounder compared to a high fish meal-based control diet 1). These results are also consistent with our previous laboratory findings that southern flounder juveniles are able to efficiently utilize diets replacing up to % FM protein with soybean meal protein for growth (Alam et al. 2011). In these controlled laboratory studies, small juveniles ( g initial weight) were raised in triplicate 75-L laboratory aquaria to maximum sizes of g over short time periods (42-60 d) (Alam et al. 2011). The pilot scale results reported here validate these laboratory findings by demonstrating that southern flounder can be raised on a high soybean meal-based diet under commercial scale rearing conditions over an extended time period from small subadult (~ g) to near marketable sizes (> 433 g) without compromising performance compared to fish fed a high fish meal-based control diet. 27

28 Species-specific responses to replacement of dietary FM protein by SBM protein are well known (Refstie et al. 2000; Davis et al. 2005), and varying levels of SBM protein have been incorporated successfully into the diets of marine and freshwater finfish species in replacement of FM protein. The replacement level of FM protein with SB protein (without supplemental amino acids) for juvenile southern flounder of 30% as tested in this study is lower than reported for cobia (40%) (Zhou et al. 2005) and in gilthead sea bream (45%) (Martinez-Llorens et al. 2008). In Japanese flounder, 32% extruded soybean meal could be added to the diet of Japanese flounder without compromising growth when the diet contained 6% squid powder and 4% krill meal, but no supplemental amino acids (Saitoh et al. 2003). Among carnivorous marine finfish species, effective SBM protein substitution levels vary significantly. In cobia, for example, up to 40% of FMP can be replaced by defatted SBP in the diet without compromising fish growth and feed efficiency (Zhou et al. 2005), similar to results of the present study. However, in yellowtail, reduced growth and feed efficiency were observed when FM protein replacement by SB protein was greater than 20% (Shimeno et al. 1993). In contrast, juvenile red drum and Atlantic cod tolerated a replacement of FM protein with SBM protein up to 50% (Reigh and Ellis 1992, Walker et al. 2010). In Trials 1 and 2 at the Blue Ocean Farms, feed consumption (FC) and feed conversion ratios (FCRs) of southern flounder fed the high fish meal-based control diet 1 and high soybean meal-based diet 2 (FC = and %/d, FCR = and , respectively) were similar between these diet treatments, suggesting that appetency and growth efficiency of southern flounder were not affected by replacing 30% FM protein by soybean meal protein during growout from small subadult (~ g) to near marketable stages (~ g). FCRs in the present study ( for Trial 1 and for Trial 2) were higher than 28

29 those reported (FCR = ) for southern flounder raised from small fingerling (1 g) to marketable sizes (~ 600 g) over a 16-month period on prepared primarily fish meal-based diets with 50-55% crude protein and 6-14% crude lipid (Daniels et al. 2010). FCRs of fish fed the high fish meal-based control diet 1 ( ) and the high soybean meal-based diet 2 (1.40) are within the ranges (FCR = ) previously reported for laboratory scale studies of early juvenile southern flounder raised on 45% crude protein and 12% crude lipid diets in which FM protein was replaced with SBM protein at levels ranging from 0 to 70% (Alam et al. 2011). These differences in FCRs among studies could be related to differences in environmental or rearing conditions or life stage. Survival in this study in both Trials 1 and 2 was similar between diet treatments, ranging from 64-76% for the high fish meal-based control diet 1 from 61-66% for fish fed a high soybean-based diet 2. Hence, mortalities did not appear to be related to the diet treatments, but rather to mechanical (pump cavitation and gas supersaturation), density dependent (interfish aggression) and operational (fish jumping from tanks) problems that were encountered during the growout period. FCRs in the present study were calculated without adjustment for fish mortality, and an adjustment for biomass losses would further reduce these FCR values. At Carolina Flounder, extended growout of fish on both the high fish meal-based control diet 1 and the high soybean meal-based diet 2 at the Carolina Flounder facility was hindered by chronic mortalities. A potential cause of these mortalities at Carolina Flounder is hypothesized to be deficiencies in the mineral composition of the rearing water at this farm site, where salinities ranges from g/l during the study. This is lower than the salinity range recorded at Blue Ocean Farms (6-20 g/l), where overall survival was higher. The relationships between water salinity, mineral composition, and fish performance, as well as the potential for 29

30 supplementing important minerals through the diet as well as through the rearing medium is an active area of current investigation. Proximate Composition In both black sea bass and southern flounder, whole body moisture, protein, lipid and ash contents post-treatment were not affected by diet, similar to what was reported for yellow croaker Pseudosciaena crocea (Richardson) (Ai et al. 2006), rainbow trout (Bureau et. al. 2000) and Indian major carp rohu Lebeo rohita L. (Khan et al. 2003) fed high soybean meal-based diets. Whereas in southern flounder, final proximate composition was similar to initial values for all parameters, in black sea bass, final moisture ( %) was lower than initial ( %), while final lipid ( %) was higher than initial ( %). The increase in lipid levels for black sea bass fed both diets indicates that fat intake may have been excessive, causing increasing fat deposition with growth. A lack of foraging activity under the intensive culture conditions used may have reduced metabolism and energy requirements. The reduction in moisture in both diets was related to an inverse relationship between whole body lipid and moisture reported in a number of fish species (Catacutan and Pagador 2004). Total Lipid in Fish Fillets Whereas in black sea bass, no post-treatment differences in lipid content (% dry wt.) of fish fillets were evident, in southern flounder, fillets of fish fed the high fish meal-based control diet 1 had a higher lipid content (14.2%) than those fed the high soybean meal-based diet 2 (12.4%). This may indicate a greater use of lipid for energy in fish fed diets high in SBM (Krogdahl et al. 2003) possibly related to the higher level of fish oil included in the high soybean 30

31 meal-based diets. Mangrove red snapper also had a lower whole body lipid level when 50% FMP was replaced by SBP in the diets (Catacutan and Pagador 2004). Total lipid contents in black sea bass fillets ( %) were approximately twofold higher than in the flounder fillets ( %). These differences may reflect the higher initial whole body lipid levels in black sea bass ( %) than in southern flounder ( %) as well as greater fat deposition in black sea bass compared to southern flounder during the feeding trials. Fatty Acid Composition of Diets and Fish Fillets Diets In general, SFA, MUFA, EPA, DHA and n-3 PUFA were higher in high fishmeal-based control diet 1 and in diet 3 (replacing 50% FM protein with soybean meal protein) than in diet 2 (replacing 30% FM protein with soybean meal protein. In this study, the ratio of DHA to EPA in the diets ( ) was within the range of required by gilthead sea bream (Sargent et al. 2002). Although the DHA or EPA requirements for juvenile black sea bass are not known, it appears that sufficient DHA and EPA was provided in all diets. Fish Fillets In general, the fatty acid composition of the fish fillets reflected the composition of their diets (Table 6). For black sea bass, EPA (22.17 vs ), DHA (24.09 vs. 18.4), total n-3 PUFAs (57.5 vs. 48.2) and ratios of n-3/n-6 PUFA (2.06 vs. 1.35) were higher in fish fed the control diet 1 than in the soybean meal-based diet 3. These differences reflect the higher levels of fish meal (rich in n-3 PUFAs) in the control diet 1. On the other hand, post-treatment concentrations of total n-6 PUFAs (35.51 vs ) and total MUFAs (98.23 vs ) were 31