The New England Aquarium is currently engaged in a captive breeding program for blue fin tuna (Thunnus thynnus). We have maintained wild caught tuna in captivity for the past two years. These fish are housed in a semi-closed 60,000 gallon recirculating sea water system and are fed a diet of previously frozen, vitamin supplemented mackerel, squid, herring and capelin. In order to effectively maintain a species which has not previously been held in captivity, it is necessary to develop techniques to safely handle the animals for routine medical procedures.
Blue fin tuna are exceptionally delicate animals that can become severely bruised even with the most gentle handling techniques. We are currently examining techniques utilizing tricaine methane sulfate baths for induction of anesthesia followed by forced ventilation with water pumps. With these methods we have been able to successfully force feed anorexic animals, collect blood samples, and perform minor surgical procedures and ultrasound examinations. We will compare wild versus captive blue fin tuna blood parameters.
Blue fin tuna (Thunnus thynnus) have only recently come under veterinary care. The focus of the past has been how to most effectively catch, kill, and serve these magnificent creatures. As testament to our fishing proficiency, the blue fin tuna population has plummeted over the last 20 years. The catch of blue fins has dropped 52% between 1970 and 1991 (ICCAT, 1992). General studies of the population at large are continuing while other studies such as larval stock assessment, remote sensing, and captive rearing programs have begun. Large (giant) tuna can bring as much as $30,000 at market with some small tuna caught earlier this year bringing $23 - $27/LB (Seafood Business, 1994). High prices and decreased supplies have increased economic incentives for farming blue fins. As domestication efforts intensify the need for veterinary care of the valuable animals increases.
The size, athletic nature and unique physiology of blue fins present unusual challenges to veterinary staff. Substantial logistical problems occur when handling even small individuals (40 - 80 lbs). Captive pathologies are often unique and extensive with severe implications for long-term captive holding [see paper: A Preliminary Study of Pathologies Associated with the Maintenance of Blue fin tuna (Thunnus thynnus Linnaeus) in captivity]. Blue fins possess unique adaptations for their lifestyle, counter-current heat exchangers to raise muscle temperature (Carey and Teal, 1969), a unique hemoglobin that resists a temperature related Bohr shift (Carey and Gibson, 1983), and ram-ventilation (Hughes, 1984) just to name a few. Some of these adaptations may help the tuna in their daily life in the sea but can make life very challenging for the animal health staff trying to diagnose a health problem. Anesthesia, usually an acceptable risk factor in most fish, becomes a high risk procedure due to the necessity of ram ventilation. Tuna swim constantly with the mouth slightly agape, allowing water to pass over the gills and exit the operculum. The respiratory system has evolved so that gilling alone may not provide sufficient ventilation to overcome metabolic demands (Hughes, 1984; Brill, 1987). Hence, a simple skin scrape has proved to be a near fatal procedure for them. The focus of this paper is to describe some of the problems we have encountered in handling these unique animals, our solutions, the results of some of our testing, and direction for future studies.
Materials and Methods
Collection and Animal Holding
The tuna in this study are part of the New England Aquarium's Blue fin Tuna Project. Fish were collected off Cape Cod, MA (Woods Hole, MA, 1992) and Wachapreague, Va (Virginia Institute of Marine Sciences, 1993). All were caught by rod and reel with angling time kept as short as possible (under 10 min.). Fish were brought on board, PIT (Passive Integrated Transponder) tagged, then put into an elliptical holding tank. At the Cape Cod site, the fish were immediately transported to a temporary holding facility: an open ocean pen. In Wachapreague, the fish were also transported from the collection site to an outdoor land based temporary holding facility. This system was a semi-closed, 12000 gal, 20' dia. tank. Water was supplied from an adjacent estuary. Temperature was maintained at 19 C±5 C by water changes and an inline chiller. Water chemistry (pH, dissolved oxygen (D.O.), Ammonia, and nitrite) was checked every 2 hours. Once the animals were stable and eating well (1-4 weeks) they were transported via a refrigerated box cargo truck to the long-term holding tank in the Animal Care Center at the New England Aquarium. The animals to be shipped were isolated from the others, herded into a black vinyl capture bag (81"x42"x48"), then maneuvered into a black vinyl stretcher and quickly transferred to the elliptical transport tank that had been installed into the truck. The fish were transported in pairs to minimize injury. During truck transport, the water was filtered through a rapid sand filter and supplemental oxygen added to keep O2 levels slightly above saturation (102-104%). Water chemistries (pH, D.O. and Ammonia) were checked hourly during the entire transport and changes made depending on results. Upon arrival at New England Aquarium, the animals were herded into the black vinyl stretcher, and quickly moved to the permanent holding tank in the Animal Care Center. Each animal was scanned for PIT tag number, checked quickly for transport abrasions and released into the holding tank. The holding tank is 28'x 40'x 8' holding 63,000 gallons of filtered seawater. Temperature was maintained by water changes and by the addition of steam. Particulate and biological filtration was achieved through use of two pressure sand filters. Ozone was used to oxidize organic waste load. e tuna received natural lighting all year long, but had additional lighting during the winter months to simulate a 12 hr light period.
Two methods of anesthesia were utilized: both used Tricaine methanesulfonate (MS222, Argent Chemical Co). In the first method the desired animal was isolated from the school with a 1/2" seine net and herded into a black vinyl capture bag. The bag was then closed, supplemental oxygen added, and the water immediately dosed with MS222 at 75 ppm. Once the animal reached stage III, plane 1 (Stoskopf,1993) of anesthesia it was brought closer to the surface and taken out for various short procedures such as weighing, measuring, and force feeding. Recovery was achieved by placing a 1" hose from a 1/3 hp submersible pump into the mouth directing water over the gills. A diver with SCUBA accompanied the animal at all times. Once the animal became reoriented itself it was released back into the main pool. The second method was used for longer procedures. The desired animal was isolated as before, but this time it was herded into a stretcher unanesthetized, and quickly transported to a 6' diameter pool containing 200 gal of water dosed with MS222 at 75ppm. Once the fish lost equilibrium it was moved to an exam table and a double 1/2" hose was placed into the mouth directing flow of water born anesthetic agent over both sets of gills. Depth of anesthesia was monitored by EKG and regulated by varying the amount of water without MS222. When the procedure was finished, the anesthetic was stopped and only anesthetic free water was pumped into the buccal cavity. As the stage of anesthesia was reduced to stage 2, plane 2, the animal was moved to the elliptical transport tank for recovery. The fish was held, pump effluent flowing across the gills to ensure proper ventilation. The animal was supported until it could swim upright on its own.
Ultrasound was used on one animal to determine the location of the lateral retia artery and for general exam of an anesthetized individual. An Aloka 1 500v with a 7.5 MHz probe was used for this study with a water soluble gel used as a conducting medium. Scans were recorded on a Panasonic cam-corder model #CCDV9.
An electrocardiograph (Datascope 870) was used on one animal during one procedure. The snaps at the patient end of the leads were replaced with standard alligator clips and soldered. One clip was attached to each pectoral fin and one to the gular area. ECG size was set at 2 and speed at 25mm/sec.
Blood samples were drawn just before exsanguination on wild caught fish. One animal was sampled when caught and placed into the live holding in Wachapreague. All captive samples were taken as part of a tagging experiment or when survival of the individual was in question and the option of euthanasia was selected. Samples were generally taken with Vacutainer blood drawing equipment. A 20 ga needle was used to gain access to the caudal vein or the lateral artery. Attached to the needle was a 20" extension set to minimize needle movement while changing tubes. Vacutainer tubes (red tops, sodium heparin, sodium citrate and potassium oxalate/sodium fluoride) were then pushed on to the adapter and the vacuum would draw in the blood. At times a syringe was used to apply a slight vacuum to the needle so "flash" could be seen easier. Blood gases were drawn into either sodium heparin tubes or into a heparin loaded syringe and run immediately. At one trial, a 20 ga Terumo in-dwelling catheter was placed in the lateral artery, an extension set attached, and blood was drawn into Vacutainer tubes. Blood was held to clot retraction on ice then spun to obtain the serum. Sodium heparin tubes were well mixed and put on ice, packed cell volume and cell counts being determined once on shore. Lactate samples were drawn into grey-top tubes (potassium oxalate/sodium fluoride), well mixed and then either spun immediately or held on ice. Blood gases were run immediately, compensating for temperature (45 C above ambient) on either a Nova Biomedical STATProfile 3, STATProfile 9, or a Stat-Pal Blood gas analyzer. Nova controls were within range on all but the Stat-pal for which the Nova controls have no values. The default in this case was the Nova Statprofile 3 values. On board ship the Stat Profile 9 and the Stat-Pal were used although high ambient temperatures above 97°F and high seas made use of the Nova STP 9 difficult. Lactates were either determined on the Nova STP 9 or were done using a Kodak Ektachem DT60 II analyzer. Serum chemistries were run by an outside lab (Tufts Veterinary Diagnostic Lab, Grafton, MA) on a Hitachi 747 blood analyzer. Some values were confirmed using the Kodak Ektachem Analyzers (DT60 II, DTSC, and DTE). Osmolalities were calculated using a standard equation (lurk and Bistner, 1981; Kaplan and Pesce, 1989).
Only those parameters that were unaffected by hemolysis were compared. Greg Skomal 3 and Brad Chase 4 of the Massachusetts Division of Marine Fisheries (MDMF) kindly lent us a data set of wild caught blue fin tuna blood for use in comparisons. The MDMF samples were sampled under similar conditions as the NEAq wild caught tunas with the exceptions that the animals were released and fight time varied. The NEAq data were first organized into two distinct groups: NEAq wild caught and NEAq captives. Wild caught were in "captivity" for 0 (zero) days. Captives on the other hand were held for longer than 1 (one) day. NEAq data were then compared to MDMF data. NEAq wilds were compared to MDMF data first using a Student's "T" test (95%, 2 tailed). If the NEAq wild population was similar to the MDMF wild population then the NEAq captive data was compared to the MDMF wild data set using the same criteria. Those chemistries that were not different in the NEAq wild data set yet significant in the NEAq captive data set were then subjected to a regression analysis to determine if there was any linear relationship between the chemistry and time in captivity. If a particular chemistry was different in the NEAq wild data set then the NEAq captive portion was not compared as it was assumed to be an invalid comparison. Sample means, standard deviations, variation, standard error of the mean and where appropriate, F statistics were calculated (Zar, 1974). All statistics were calculated in Lotus 123 ver. 4 for Windows.
Both methods of immersion anesthesia worked well. Induction was rapid and deep. Recovery, though long, seemed to be complete without side effects. The dose of 75ppm worked well with stage III, plane 1 anesthesia achieved within five minutes.
Obtaining sonograms proved to be quite informative. The goal was to visualize the lateral artery that supplies the retia (Figure 1). In this figure the artery is quite easily seen. The liver and other internal organs were also easily visualized. After visualizing the artery, an arterial puncture could be made to obtain a blood sample from the animal.
Summaries of blood results are shown in tables 1 and 2. Creatinine (P= 0.0492, 05, 2), uric acid (p= 0.00018,.05, 2), C02 (p=0.0075,.05, 2), triglycerides (p=0.0148,.05, 2), ALT (p=0.00061,.05, 2), and CK (P=0.0306, 0.05, 2) were significant captive chemistries by Students T test when compared to MDMF data.
Indirect bilirubin was also significant but was dropped do to the influence of hemolysis on the values. The regressions done on these chemistries (examples, Figures 2 and 3) have r 2 values that are low, only Uric acid into double digits and it was still not significant by F statistic (F=2.54, 0.05(2),l, 7,P=0.5> x >0.2).
Lateral retial artery located for arterial puncture. The gradations along the edge are in CM. the artery is the dark band at approximately 2. mm.
MS222 was used in this study because of its ease of use and its apparent margin of safety. This was of prime concern since these animals were difficult to collect and expensive to replace. There were no outward effects from the anesthesia itself but there were problems encountered during capture. The act of crowding the animals and separating them appears to be quite stressful. When the animal had herded into the capture bag they would often rub against the side of the bag and rush into the supporting poles. This happened frequently in the excitement phase of induction. The result would be seen several days later - a large surface abrasion on both laterals despite the use of a smooth vinyl bag. Although injuries of this type would sometimes occur during the initial capture, it seemed that the severity increased in the captive fish.
The anesthesia pump worked quite well supporting the animal for a medical procedure that lasted 17 min. During that time, the animal was monitored with ECG, allowing the heart rate to be visualized. Anesthesia was stopped and anesthesia free water was pumped over the gills when the heart rate decreased. Traces were weak when compared to mammals but were clearly visible on screen. The ECG appears to have been a valuable indicator of depth of anesthesia. It has since been successfully used on other fish since that time. Caution should be exercised when setting up the unit, making sure there is a good ground connection, and that a minimal amount of water splashes onto the "electrodes".
Ultrasound was used to visualize the lateral artery for blood sampling (Figure 1). Several fishermen reported that they would normally bleed them by lacerating them just caudal to the gill operculum. We were looking for another blood sampling site that might provide a more consistent sample and safer for the investigator, when sampling a larger fish, than sampling at the caudal vein. Necropsy evidence showed the presence of the artery and was localized on the individual by ultrasound. It was found approximately 2.5mm below the epidermis and approximately 2mm in dia. The punctures went smoothly with minimal hemolysis in the blood samples. A larger bore needle could be potentially used for obtaining samples in shorter time, and larger volumes. The lateral vessel was also accessed with a 20 ga catheter.
The use of the data set obtained from Skomal and Chase (pers. comm.) proved to be very helpful in assessing the data generated. It gave us the ability to test our data against a larger set of "wild caught" tunas. The result (Student's "T" test) was that NEAq wild caught data is similar when compared to MDMF data, with some exceptions, and that the NEAq captive data has some significant differences. Of the exceptions; lactate, BUN, total bilirubin, direct bilirubin, cholesterol, and albumin; the bilirubins would have been affected by hemolysis. Albumin values, although statistically different, appeared grossly similar between MDMF and NEAq wilds. Lactate values were much higher in the MDMF data set suggesting that the variations in fight time had a significant effect on this parameter. In the NEAq captive data set, several chemistries were significantly different from the 2 MDMF data set: creatinine, triglycerides, uric acid, C02, ALT, and CK. All had r values less than 0.10, except uric acid (0.2661), and none of the regressions were significant by F statistic. Therefore there was no relationship between the parameter in question and time in captivity. Hemolysis presented difficulties especially on wild caught tunas. Some tests (alkaline phosphatase, GGT, and bilirubin) are affected by even small amounts of hemolysis and can preclude their interpretation. We found that keeping the blood collection and centrifuge tubes at about 10 C helped reduce the incidence of hemolysis. Other factors that confounded the data interpretation were length of time in captivity and health status at the time of sampling. To decrease variability, future experiments could include a controlled study with set blood sampling intervals. Other methods of anesthesia will be examined, namely an injectable anesthetic and possibly electronarcosis, in an attempt to obtain blood samples that more closely reflect an unstressed animal. Baseline blood parameters for the species could then be determined to help in the diagnosis of disease and for general life history information. The use of ultrasound could be of great benefit in biopsy deep tissue or assessing reproductive status.
I'd like to thank Greg Skomal and Brad Chase for the use of their data set, as well as Captain Jack Stallings, Paul McFarland, Leah Faretra, Joshua Singer, and Michelle Lee. And special thanks to my wife Terry.
1. Brill, R.W. 1987. On the standard metabolic rates of tropical tunas, including the effect of body size and acute temperature change. Fishery Bulletin: Vol. 85 (1), pp. 25 - 35.
2. Carey, F.G., and J.M. Teal. 1969. Regulation of Body Temperature by the Blue fin Tuna. Comp. Biochem. Physiol. Vol. 28, pp. 205 - 213.
3. Carey, F.G., and Q.H. Gibson. 1983. Heat and oxygen exchange in the rete mirabile of the blue fin tuna Thunnus thynnus. Comp. Biochem. Physiol. Vol. 74a (2), pp. 333 -342.
4. Hughes, G.M.. 1984. General Anatomy of the Gills. in: Fish Pathology. (Hoar, W.S. and
5. D.J. Randall eds.). Vol. 10, pp. 20, 101.
6. Kaplan, L.A. and A.J. Pesce. 1989. Clinical Chemistry: Theory, analysis, and correlation. 2nd ed. C.V. Mosby Company, St. Louis, Missouri. p. 879.
7. Kirk, R.W. and S. 1. Bistner. 1981. Handbook of Veterinary Procedures and Emergency Treatments. 3rd ed. W. B. Saunders, Philadelphia, PA. p. 114.
8. National Marine Fisheries Service. 1992. National Report of the United States: 1992. ICCAT Working Document SCRS/92/124. pp. 19.
9. Seafood Business. 1994. Seafood Update. (Perkins, Caroline. ed.). Vol. 13 (2), pp. 24.
10. Stoskopf, M. J.. 1993. Clinical Examination and Procedures. in: Fish Medicine. W.B. Saunders, Philadelphia, PA. p. 81.
11. Zar, J.H.. 1974. Biostatistical Analysis. Prentice Hall, Englewood Cliffs, NJ. pp. 15-18, 92 - 93, 107 - 108, 198 - 213.
12. Massachusetts Division of Marine Fisheries, State Lobster Hatchery, Vineyard Haven, MA.
13. Massachusetts Division of Marine Fisheries, Cat Cove Marine Laboratory, Salem, MA.