Halibut PEI

FINAL REPORT
AIMAP-2008-G03

Table of contents

• EXECUTIVE SUMMARY
• INTRODUCTION AND GOALS OF THE PROGRAM
• FIBERGLASS TANK TRIALS
• Logistics
• Protocols
• RESULTS
• Water Quality, All Groups
• Survival and Health
• Growth and Feeding
• DISCUSSION
• Water Quality
• Survival and Health
• Growth and Feeding
• CONCRETE TANK TRIAL
• METHODS
• RESULTS
• Lobster Tank Modification
• Water Quality
• Fish Arrival and Initial Assessment
• Survival and Health
• Growth and Feeding
• DISCUSSION
• CONCLUSIONS

EXECUTIVE SUMMARY

Halibut PEI Inc set out to determine if a halibut feedlot industry was viable using low salinity well water available in the province. The natural advantages of well water including stable geothermal temperatures (10°C) and under ground filtration through sandstone to avoid diseases, and the significant reduction of initial capital expenditures by modifying existing lobster holding tanks, were anticipated to give economic advantages over other means of halibut culture. Beginning in September of 2008, with funding from both federal and provincial governments, equipment and expertise from both private and academic institutions and the cooperation of the plant owner, halibut were raised in PEI for the first time.

Through a series of evaluations and growth measurements of three different ages of young halibut, the project demonstrated excellent survivability and satisfactory growth of the fish and a variety of adaptive techniques to facilitate their successful culture in lobster tanks while promoting a high degree of environmental protection. Throughout the trial period, no treatment of any kind was necessary.

Fish from the system were sampled by local chefs and received rave reviews.

The saltwater wells on PEI can biologically support an efficient industry to grow halibut juveniles to market weight. The business model for success still requires extensive development. Fingerling costs must be reduced, growth discrepancy between males and females must be addressed, new markets must be found for different sizes of fish, and feeds have to be evaluated for efficiency in this type of water. Lobster facilities around the Island need to be made aware of the basic parameters for saltwater wells, tank modification and growth estimates in order to assess their potential for growing halibut as second product when not holding lobster. Halibut PEI Inc believes that capture of effluents needs to be refined to ensure the best environmental protection possible.

INTRODUCTION AND GOALS OF THE PROGRAM

Halibut PEI was formed in 2004 with the objective of developing a land based halibut aquaculture industry on PEI. Based on the scientific literature and first hand knowledge of the salt water wells on PEI, the principals sought a suitable location to initiate the project and managed to obtain use of the MorningStar Fisheries plant in Victoria PEI. After obtaining funding from the AIMAP initiative of Fisheries and Oceans, Canada and from the PEI Department of Fisheries, Aquaculture and Rural Development as well as in kind support from the same provincial department, Scotian Halibut Limited, Atlantic Veterinary College (AVC) and Ocean Choice International, the project began in September of 2008.

The objective of the project was to determine the capability of PEI salt water wells to support an Atlantic halibut feed lot industry using existing equipment and facilities from lobster pounds. Several factors were inherent in this objective.

The geothermal regulation of water temperature from salt water wells was expected to be ideal for halibut culture. Water from deep wells was considered virtually abiotic containing no parasites, bacteria or known viruses but with a relatively low salinity (20 to 26 parts per thousand or ppt). The initial factor to be determined was the viability of the fish in this salt water well environment.

Halibut PEI intended to use existing facilities to minimize initial capital investment in the early stages of the province’s halibut grow out industry. Lobster pounds could provide the infrastructure but would not allow year round production making it necessary to determine the size of fish to be placed into the tanks in order to provide a marketable product in the time frame available. This strategy recognized the availability of a ready supply of juveniles and young halibut from Scotian Halibut Limited but required an understanding of halibut growth rates under these water and tank conditions. This was a second factor to be determined.

Inherent to the philosophy of the company is an unwillingness to compromise the PEI environment which is so attractive to tourists. The project therefore also included plans to reduce any ecological foot print as much as possible. This was achieved at the MorningStar facility with the installation of a waste extraction system that gathered and removed solids and allowed analysis of effluent. The facility also allowed Halibut PEI to pursue its original intent of developing a holding system strategy which did not rely on the use of antibiotics, parasiticides, or hormones because of its abiotic salt water well supply. 

Assessment of these factors was the focus of evaluation when the project started in September 2008. Two set ups were used. The first consisted of shallow fiberglass tank trials allowing close, detailed observation in order to determine the viability and growth of the fish in salt water well water.  The second consisted of pre-commercial rearing of halibut in a modified concrete lobster holding tank. This part assessed the feasibility of using existing infrastructure and would ascertain that growth rates obtained in the fiberglass tanks could be achieved at commercial densities.

FIBERGLASS TANK TRIALS

The survivability and growth of Atlantic halibut, in water provided by salt water wells characteristic of PEI, was determined by rearing three groups of halibut, each representing different size classes, in research scale fiberglass tanks. These allowed close observation of the fish in order to determine their behavior, feeding activity and any demonstration of disease or stress. They also permitted scheduled measurements of total fish biomass to give precise determinations of growth.

Three sizes of tanks were used, because of their availability: 3.65m X 1.83m (2.34m3), 3.65m X 1.52m (1.94m3), and 3.65m X 1.22m (1.56m3). These tanks were equipped with gas scrubbers previously used for lobster holding. These were modified with the addition of two water distribution plates supporting 2.5cm Bio Barrels (Aquatic Eco-Systems, Inc., FL) to optimize gas exchange. A line from a regenerative blower was connected to each scrubber to increase the air flow through the incoming water. In case of water flow shut down, oxygen was available at each tank delivered via fine pore diffusers for oxygen (Aquatic Eco-Systems, Inc., FL). The tanks all had rigid hardware cloth (“Vexar”) wall extensions to allow for 40 cm of extra height to prevent jumping. Eventually it became necessary to completely close in the top of the tanks with similar mesh to prevent jumping over the top of the walls.

Logistics

Group 1 (150g)
On October 2, 2008, 743 fish, at approximately 150g mean weight, were transferred from Scotian Halibut Limited to Halibut PEI’s facility (MorningStar lobster plant) in Victoria, PEI. The fish were trucked to the facility by Scotian Halibut Limited in oxygenated sea water held in modified 1 m3 plastic fish tote boxes. Upon receipt, the fish were netted out of the tanks into self draining tubs which were emptied into eight fiberglass holding tanks.

Group 2 (650g) in Fiberglass Tanks
On December 20, 2008, 1459 halibut, with a mean weight of 642g, were successfully transferred from Scotian Halibut Limited to Halibut PEI’s facility in Victoria, PE using fiberglass smolt transfer tanks on a trailer. A subset of 82 of these fish was allocated to two fiberglass tanks (2.34m3 each) previously used for Group 1 fish, after disinfection. These fish provided a comparison in growth with Group 1 in adjacent fiberglass tanks and with the majority of Group 2 fish housed in the concrete tank 1B.

At the end of the growth trial, the Group 2 fish held in the fiberglass tanks were moved into the concrete tank (1B). Prior to the transfer, the fish from the fiberglass tanks were individually tagged with Visual Implant Elastomer (VIE) tags (Northwest Marine Technology, WA). At this time, the fish were anesthetized with MS-222 (180mg/L) (Syndel Laboratories Ltd., BC), individually weighed and the presence or absence of cherry belly was determined. The fish were tagged to differentiate those with and without this condition.

Group 3 (13g)
On February 12, 2009, 178 juvenile halibut with a mean weight of 13.5g were transferred in buffered, oxygenated water in sealed plastic bags in styrofoam coolers from Scotian Halibut Limited to Halibut PEI’s facility in Victoria, PE. They were all released into one fiberglass tank after a one hour temperature acclimation was performed by floating the sealed bags in the recipient tank. 

Protocols

Group 1 (150g)
Water quality was checked and fish were fed to satiation a minimum of twice daily. Fish were offered Corey AquaSea diet for marine fish in a 5mm pellet: 50% protein, 14% fat, 18.9MJ/kg digestible energy. Feed offered each day was recorded along with oxygen, temperature, salinity and differential gas pressure. Oxygen levels were maintained at optimum (greater than 8.5mg/L or 88% saturation) to enhance appetite and reduce any stress on the fish. Any debris (feces, uneaten food) in the tanks was swept into the drain daily or as required. Mortalities were removed immediately. Mortalities were sent to AVC for detailed disease testing or frozen to provide retrospective reference.

Group 1 fish status was very closely monitored. The system was set up for early disease detection as individual fish could be seen and examined visually with a pencil light. The bright aquamarine color of the tanks was remediated with low ambient lighting: approximately 1 lux or less as measured by a light meter. The lower light level did not seem to inhibit their predation for pellets and all fish appeared quiet and unperturbed in the tanks.

The fish were not weighed when the tanks were initially stocked to reduce stress at the time of transfer. This was particularly important because the transport truck was delayed for 6 hours on the New Brunswick side of the Confederation Bridge due to a freak storm causing high winds in excess of 100km/hr. An initial assessment of fish biomass in each tank was conducted October 16, 2008.  Subsequent tank biomass determinations were conducted at 2 week intervals from this point to December 11, 2008, and monthly thereafter. Biomasses were determined by netting fish out of the fiberglass tanks into self draining tubs that were then placed in water filled receptacles on a tared scale. 

On December 16, 2008, the fish were graded by size into six tanks to free two tanks for use for another cohort of fish (Group 2). February 3, 2009, the Group 1 fish were again redistributed but into five tanks to make room for the smallest size cohort of fish in the study (Group 3).

Group 2 (650g) in Fiberglass Tanks
The fish put into the fiberglass tanks were weighed at two week intervals January 6 to March 3, 2009. Adjustments to the biomass determination procedure were required to accommodate the larger fish. Water quality was checked and fish were fed to satiation a minimum of twice daily.  Fish were offered Corey AquaSea diet for marine fish in a 10mm pellet: 47% protein, 19% fat, 19.1MJ/kg digestible energy. Feed offered each day was recorded along with oxygen, temperature, salinity and differential gas pressure. Any debris (feces, uneaten food) in the tanks was swept into the drain daily or as required. Vinyl mats (1.9m2) were added to one of the tanks February 24, 2009 to determine their effect, if any on incidence of “cherry belly”. 

Group 3 (13g)
These fish were fed every half hour throughout the working day. Fish were fed in excess in an attempt to reduce cannibalism which is common in very young halibut.  Fish were offered Gemma Wean Diamond diet for marine fish in a 1.8mm pellet: 55% protein, 15% fat. Feed offered each day was recorded along with oxygen, temperature, salinity and differential gas pressure. The tank bottoms were swept with a soft broom daily to move uneaten feed and feces to the drain. A biomass determination was done twice in 6 weeks on these fish. 

RESULTS

Water Quality, All Groups

All tanks were subjected to the same water quality conditions since they were supplied with the same water. Oxygen levels were maintained above 85% saturation by adjusting water flow. Total gas pressure hovered around 100% and pH was not a concern in the tanks because of the high water turnover and low stocking densities. It was measured on one occasion and all tanks were within the range of 7.70 to 7.71.  Water clarity remained excellent. 

Survival and Health

Group 1 (150g)
Once transferred to the small fiberglass tanks in MorningStar, the fish settled quickly and could be observed individually. The 6 hour transportation delay of the 743 fish from Scotian Halibut Limited in NS to Victoria, PEI, on October 2, 2008 contributed to 16 fish dying on Friday, October 3 as a result of stress and injury from the transport. These showed massive effusion of gill mucus, pale color of the gills and gaping mouth and opercula to indicate hypoxia (asphyxiation) and subsequent respiratory distress leading to death. Other changes noted included some transportation related damage to fins and eyes. 

One fish found several days later, describes evidence of bacteria in the vessels and elsewhere. This fish was in the tank undetected for at least a day and the bacteria described were interpreted as being a post mortem contaminant typical of autolysis. The fish in the tank were monitored intensively and no others were lost or became ill. All other reports indicate no evidence of infectious or any other disease. The cause of death was not determined for the submitted fish but was likely transfer related.  

No other in tank mortalities occurred for the entire study, more than a six month period. Two fish surviving transport difficulties had ulcerative traumatic damage to one eye as a result of the shipment. One fish was removed, culled and sent for necropsy while the other had the eye surgically removed and went on to heal uneventfully. Two additional fish were late in the study (March 13, 2009) due to their apparent weak state (pale and small) to use for disease monitoring and a further fish was culled April 3, 2009 due to an abrasive growth on its jaw. No indications of infectious or environmental diseases were found in any examined fish. 

In week two after their arrival, fish in several tanks began flashing and demonstrating intermittent irritability. The tanks were flushed with increased water flows and no further manipulation. The problem never returned so no diagnosis was made. Likely the incident involved protozoan infection which is common after transport; and the reduced salinity and higher water flow controlled any parasites on the skin. Throughout the entire demonstration project no treatments of any kind were used.

Despite the securing of tanks with jump nets and covers, 15 fish losses occurred over the six month period due to fish jumping out of tanks. 

Mild lesions were noted on the abdominal surface of the fish at weighing and grading, a condition known in the aquaculture and fishing industries as “cherry belly”. This condition seemed to lessen as tank densities were increased after grading into another tank and remained very mild in this group of fish. Halibut PEI embarked on a preliminary investigation of this dermatitis.

Group 2 (650g) in Fiberglass Tanks
No in tank mortalities occurred for this group. One fish died as a result of jumping out of the tank. The only other fish losses were a result of culling for disease testing.  Five fish were culled specifically to examine a multifocal dermatitis labeled as “cherry belly”. Handlers noticed these small, red, elevated lesions about one month after receipt. The lesions caused no noticeable irritation to the fish and feeding appeared normal. A new substrate was put in Tank 1 in the form of a vinyl anti-fatigue mat on February 24, 2009. When the fish were examined six weeks later (April 9, 2009) incidence of cherry belly in tank 1 fish had reduced dramatically compared with tank 2 which had only a fiberglass substrate. At this time, the tank with the vinyl mats had a 5% incidence of the condition whereas the tank without the mats had an 86% incidence. Fish with the lesions were lethally sampled March 27, 2009 and April 3, 2009 for detailed histopathologic study of the lesions. Though commonly referred to by others, this condition has never been examined in detail. Initial histopathology indicated severe edema around the scales, papilloform proliferation of the skin and proliferation of blood vessels adjacent to the lesions and under the skin accounting for the red color. The lesions are being examined by electron microscopy and a report will follow.

Group 3 (13g)
One very small fish (less than 1g) was found dead 2 days after the shipment arrived. Three obvious “runts” were noted upon receipt of shipment. These remained much smaller than the rest of the population. One of the small fish was noted to have a freshly injured eye on April 11, 2009. The eye subsequently healed but the fish was still often observed swimming at the surface while other fish rested on the tank bottom the vast majority of the time. 

Growth and Feeding

Group 1 (150g)
Feeding began within a day of transport and remained strong throughout, though the decrease in water temperature later on did inhibit feed consumption. Handling for biomass determination also inhibited feed consumption the following day.  Feeding rates ranged from 0.43% to 0.79% body weight per day. Feed conversion ratios averaged 1.2 over the sampling period. Specific growth rates were affected by temperature. There was no significant tank effect on fish performance (growth and food conversion).

The Group 1 fish achieved a mean weight of 389g by the end of the study. At this time, a broad size distribution was noted in the fish with the smallest fish estimated at being less than 100g and the largest fish greater than 500g. 
 
Group 2 (650g) in Fiberglass Tanks
Feeding began within a day of transport. A high feeding response was never noted in this group of fish. Handling for biomass determination inhibited feed consumption the following day.  Feeding rates ranged from 0.35% to 0.56% body weight per day with feeding rate dropping over time.  Initial feed conversion ratios averaged 1.5 until mid February. After this point, they increased considerably in conjunction with a reduction in growth of the fish. There were not enough replicates to determine tank effect on fish performance.

In the fiberglass tanks, the Group 2 fish achieved a mean weight of 914g by the end of the study. Initial growth was excellent. It tailored off in the latter part of the study (post mid February). A broad size distribution was noted in the fish at the end. We quantified this by weighing individuals on April 9, 2009. At this time size ranged from 523g to 1584g. 

A line from the air blower was also hooked into the scrubber to allow air to be blown through its center trays to increase the gas/liquid ratio and maximize its degassing ability. 

Air Infusion
Prior to fish entry, trials were conducted to determine the usefulness of the intact water aeration system. The system tested is typical of lobster holding tanks and consists of a regenerative air blower connected to a piping manifold sitting on the bottom of the tank.  Holes are drilled in the pipes to allow air flow. 

Oxygen Use
From December 22 to December 31, small amounts of oxygen were infused into the tank water using fine pore diffusers for oxygen. There was no oxygen flow meter on the system and inflow was not regulated such that the oxygen tank was drained rapidly (9 days). Feeding was reduced slightly until water flow was increased to accommodate for the oxygen supply loss.  Oxygen was not used in the tank after this date except when the pump was shut down.

Foam Fractionator
A foam fractionator (RK25-PE Model from RK2 Systems), rated for 100-150 lpm water flow, was installed in early February and turned on February 10, 2009 to extract mid stream water from the tank for clarification of the fine suspended solids. A water fall pump was inserted in the base of a 20cm diameter pipe sleeve with holes drilled midway. The pump and sleeve were placed 2/3 way back in tank, midway between the catwalk and the wall.

This pump provided the foam fractionator with 115 lpm of water.  e decided to try it at many different settings and discovered that balancing air flow and internal water flow gave the best results. This balance often had to be reset after re-starts. It was suggested that the tank water was not dirty enough for optimal foam fractionator performance. The water always appeared clear when viewed from the walkway except at the morning feeding. More evaluation will be required to demonstrate the advantage of maintaining foam fractionation for fish at this density.

Waste Characteristics
Water which passed through the drum filter had very little suspended solids. Because the post filter water was very clear with minimal fines present and it took so long to collect sufficient material to look at, our effluent study concentrated on the material that was filtered by the drum filter and sent to a separate collection tank. Solids from the drum filter effluent separated quickly into two parts- heavy material that settled quickly into a tote box and light material which decanted off the tote box then out of the tank. This material was white to brownish and stayed suspended in the water for a long time. We made several unsuccessful attempts to quantify the solids collected by the drum filter over a 24 hour period. These attempts involved flowing the waste through a series of screens finishing with a nitex mesh. In the end, we collected solids accumulated in one hour intervals. This was repeated for 5 sampling periods. Of interest was the short term clogging of the nitex mesh with no visible debris on the mesh. When cleaned off the mesh there was a tiny amount of gel like material. Analysis of this material showed a large proportion of mucus, extensive growth of long chain algae and water mold as well as normal sewage worms. These were not the characteristics expected with normal waste removal systems in aquaculture facilities using other types of fish. Changes in collection strategy will require substantial modification to the originally proposed effluent clarification process.

Water Quality

Tank 1B was supplied with water from the highest producing well on the premises - Well #3 was very similar to that of the other well tested - Well #1, suggesting a similar source from two different aquifers.

The water temperature of Well #3 increased then decreased slightly over the winter months remaining between 7.9°C and 9°C. The salinity slowly fluctuated up and down, keeping between 23 and 26 ppt. 

Fish Arrival and Initial Assessment

The fish arrived on December 20, 2008. They were delivered in square smolt delivery tanks. The fish were unloaded by dip netting out of the transfer tanks into porous totes and carried by hand to Tank 1B. 

Initial attempts were made to weigh and count fish prior to entry to the tank but the time taken was excessive so that finally only weighing was done. The fish count done by Scotian Halibut Limited would be used. The biomass determined by Halibut PEI was 937 kg. This is very close to the biomass of 948 kg reported by Scotian Halibut Limited. At the time of unloading, 82 fish were put into 2 of the fiberglass (1A and 2A) tanks. Initial stocking density of concrete Tank 1B was estimated at 113% coverage or 18 kg/m2. The fish appeared very comfortable in this environment.

Survival and Health

One fish died as a result of jumping out of the tank. Only 2 fish were found dead in the tank.  One was submitted for necropsy and disease testing with no evidence of infectious disease found. Both “in tank” mortalities were likely latent mortalities from the fish transfer. 

Growth and Feeding

Fish were initially on reduced rations and feeding was increased slowly over the next few weeks to assess system performance. Fish fed ravenously within days. They were generally fed twice daily; an occasional third feeding was attempted if feeding was slower than normal or if activity around the tank disrupted a feeding. In late January, a red light was used as operative conditioning to cue the fish that the feeding session would begin. This allowed them to differentiate from the many visitors using the overhead walkway.

Feeding was slightly reduced for a week in March in order to collect feces. For this period, a 90% ration was offered in an attempt to ensure that all feed was eaten so that collected matter would only be feces and no uneaten feed. 

When we returned to feeding to satiation again in April, feed offerings were lower than other months, despite an expected increased biomass due to growth. This was likely the result of staff being better trained to assess the end point of feeding and reduced appetite of fish due to a decrease in temperature.

Fish grew very well in the concrete tank. A preliminary size assessment suggested a mean weight of 1007g by the end of the study. This represents better growth than the fish in the fiberglass tanks when temperature is considered. An overall FCR of 1.5 was estimated.

DISCUSSION

The optimum culture set up would be inexpensive, promote self cleaning of the tank and require minimal maintenance while sustaining optimal water quality for commercial densities of halibut.   The concrete tank used in this study, with the implemented modifications, met all of the above criteria.

Only minor modifications of the tank were required in order to sustain all parameters of water quality examined.  It is notable that more work on tank design is required in order to increase recirculation of water and reduce required make up water flow. The benefits would include:  application of a similar system in other facilities where copious amounts of water are not available, use of the recirculation system as a back up measure to hold fish in the event of emergency water shut down or use of the recirculation system to increase holding capacity for fish when awaiting market.

Fish survival was excellent. Only three fish died over a four month period, with one death being caused by the fish jumping out of the tank before it was fully secured with jump netting.  The other two mortalities were most likely the result of the transport to the facility. 

The fish grew very well in this setting, better than the analogous group in the fiberglass tanks. The feed conversion ratio was not as good as anticipated but staff experience will improve this drastically. Improvements in growth and feed conversion are possible and should be investigated to maximize profitability of similar ventures. These include advances through stock selection and feed optimization.

The fish grew well enough for us to harvest a small number in order to gauge the response.  The product received rave reviews by all recipient chefs and other individuals.  Some restaurants even requested to add the product to their menu. This is promising but additional work in this area requires doing a formal review of the quality of the product and promoting its place in the local market.

Effectiveness of the drum filter for removing wastes produced in the system was evidenced by the clarity of the post filter water and our ability to recirculate it back to the culture tank with no loss of water quality. The potential environmental impact of this culture method still needs to be assessed however with more precise assays. In order to best protect the environment in future operations, Halibut PEI will filter all water leaving the facility through a 50 micron drum filter, separate the solids in a normal septic tank from which the sewage can be removed periodically, and design a self cleaning mesh filter to remove most of the remaining debris. Halibut PEI also intends to evaluate bioscrubbing. Shell fish feed normally on seston (organic debris in sea water) and Halibut PEI intends to pass the effluent water through shellfish to determine if they will consume the fine suspended solids.

CONCLUSIONS

Based on the results of this project, salt water wells on PEI can reasonably support an Atlantic halibut grow out industry in lobster pounds. Fish survival was excellent with growth rates being on par or better than published studies. Only minimal modifications of a concrete lobster tank were required in order to sustain commercial densities of halibut.

Sufficient information was collected to estimate required stocking size of lobster pounds to achieve harvestable fish during the time of year they are not in use. These estimations were based on the application of a thermal growth coefficient.

Additional investigations are necessary to answer questions raised during the course of this study.  They include:

• Optimizing growth of halibut in low salinity water by comparing growth of all female versus mixed sex halibut populations and using fish pedigrees to predict growth characteristics
• Evaluating feeds and alternative feed sources
• Determining the ecological footprint of commercial production including collection and analysis of fecal solids, and monitoring effluent environmental effects with a shellfish bioassay.
• Assessing recirculation technology to alleviate well water limits including predicting disaster time in the event of well failure, determining carrying capacity at 80% recirculation of water, and determining how long over- density conditions can be maintained to adjust for market fluctuations.
• Assessing the taste of halibut produced to determine optimum taste from feeding strategies implemented, and promoting PEI farmed halibut to the food service industry.