Tuna Metabolism and Physiology in Captive Yellowfin Tuna (Thunnus albacares Abstract At the Tuna Research and Conservation Center, in Monterey California, captive yellowfin tuna (Thunnus albacares ) demonstrated an increase in lipid accumulation and mortalities when fed a high calorie and high fat diet. To determine what influence a high fat diet has on tuna physiology, one tank (T2) holding 28 fish was designated for diet manipulation while oxygen consumption and stomach temperatures were used to measure changes in metabolism. No significant differences in metabolic rate were measured between diet treatments. The low fat diet (LFD; «1% fat content in kcal) was composed of squid and smelt while the high fat diet (HFD; 8% fat content in kcal) only included sardines. Samples were sent away for bomb calorimetry analysis and based on these values, an equivalent total caloric level (28 kcal/kg/day) was maintained between diets. Individual respirometry trials were performed in a Closed-system respirometer. Three trials were run for each diet treatment and swimming velocity was measured for each run to calculate standard metabolic rate (LFD: 135.70 +/- 42.79 m, 02/kg/hr; HFD: 147.54 +/- 8.89 mg 02/kg/hr; p-0.80). Whole tank metabolic rate was measured using T2 as an Open-system respirometer with help from the dissolved oxygen monitoring system (Program Logic Control) operated by Monterey Bay Aquarium. Stomach temperature was measured by minilog temperature probes and no significant difference between diets was measured (pre-feed: p=0.91; post-feed: p-0.80). Introduction Yellowfin tuna (Thunnus albacares) are apex predators in an environment sparsely populated with prey items (Olson and Boggs, 1986). They have evolved the physiology and anatomy necessary for high rates of digestion, high rates of growth in somatic and gonadal tissue, high rates of oxygen-debt recovery and efficient swimming (Brill, 1996; Korsemeyer, 1996). These adaptations, including a large gill surface area, large blood volume, large amount of red muscle, and an elevated body temperature (Brill, 1987; Block, 1994), require large amounts of energy to maintain. Diet is the main source of metabolic substrates, such as fat, protein, and carbohydrates, for fuel and biosynthetic precursors (Jobling, pg. 37-38). Fat is a valuable metabolite for yellowfin tuna who have high energy expenditures but unreliable prospects for energy consumption, a bioenergetic situation often described as speculative (Olson and Boggs, 1986). It is rich in energy, a nutritional requirement and a compact, anhydrous storage molecule (Jobling, pgs. 10-20, Schmidt-Neilsen, pgs. 178-181, Stryer, pg. 469-472). It is intuitive that the amount of lipids ingested should be proportionate to the amount available for absorption, utilization, and storage. Studies on the stomach contents of commercially netted yellowfin tuna, have shown that fish constitute a large volume of ingested prey items (Olson and Boggs, 1986; Maldeniva, 1996). This may overlook seasonal fluctuations in prey species available given that tuna are general and opportunistic, rather than selective, feeders (Olson and Boggs, 1986). The percentage of fish found in tuna stomachs is large enough (280%) to suggest that yellowfin tuna are feeding on a diet with adequate lipid content, as most fish species are higher in fat content than other prey, such as cephalopods (Chuck Farwell, personal communication) A study in sled dogs suggests that a high fat diet will increase the plasma concentrations of free fatty acids, especially after exercise. Äfter a high level of aerobic activity, the levels of free fatty acids in blood have been shown to increase in actively training sled dogs (Reynolds et al., 1994) and in another fish, the Arctic Char (Barton et al., 1995). Both studies suggest that an increase in free fatty acid concentration in circulating blood promotes uptake and oxidation especially in animals with high energy demands. At the Tuna Research and Conservation Center (TRCC) in Monterey, California, it has been a challenge to design a diet that fits the nutritional needs of confined yellowfin tunas. Initially (September, 1994, to March, 1995), the tunas were fed a diet of squid, sardines, and anchovies (total calories = 80 kcallkg/day) twice daily. This diet, high in both fat and calories, resulted in an increase in lipid accumulation (in muscle tissue) and related mortalities. By decreasing fat content and total calories fed, a decrease in intra-muscular lipid accumulation and mortality rate was demonstrated (Perla, unpublished - 1995). Currently, the diet consists of 82% squid and 18% smelt by wet mass and is low in both total calories and dietary fat content (total calories = 28 kcallkglday, fat content « 1%). The low fat diet fed to captive yellowfin tuna at the TRCC could result in a decrease in the availability of lipid metabolites to use and store. This change in substrate concentration suggests a change in the metabolic cost of digestion and storage as different substrates follow different biochemical pathways (Schmidt-Neilsen, pgs. 145-150, 157-159). By increasing the lipid content but maintaining a low total caloric value, the effect of a high fat diet on tuna metabolism can be determined. Oxygen consumption and internal heat increment are two indicators of metabolism (Schmidt-Neilsen, pgs. 178 - 179) and can be measured using respirometry (Cech, pgs. 335 - 336) and minilog temperature probes to gauge any effect of diet treatment. Materials and Methods At the TRCC, yellowfin tuna, caught off of Southern California, have been successfully maintained in captivity. One tank, T2, (29,000 gallons) holds twenty-eight experimental fish for Hopkins Marine Station. The water temperature is kept at 21-22°C and the tunas are fed roughly 3 - 4% of their body weight (Chuck Farwell, personal communication), three times a week. Body mass was estimated from fork length measurements: Weight = 0.0000407752 fork length?3.02 Eight smaller fish (mean mass = 3.55 kg, mean fork length = 56.51 cm), designated as respirometry fish, were marked with yellow tags for identification. Diet manipulation To determine the effect of a high fat diet on tuna metabolic rate, two diet treatments were designed. The composition of the low fat diet (LFD) was a continuation of the regular TRCC diet - 82% squid and 18% smelt by wet mass. The high fat diet (HFD) was composed of 100% sardines. Twenty-five specimens each of squid, sardines, and smelt were weighed to calculate average mass. To correct for a large variation from mean mass, squid and sardines were separated into different size classes. Five specimens of each group were sent to the Michelson laboratories for bomb calorimetry analyses which measured total caloric content (kcall100 g) and percent fat, protein, and carbohydrate. Based on these values and biomass of tank, treatments were modified to keep total calories constant across diets (28 kcallkg/day) while differences in fat content (LFD « 1%; HFD = 8%) could be manipulated. At 21 - 22°C, the calorie level was low enough to ensure a minimal growth rate over experiment duration (Chuck Farwell, personal communication). Consideration was taken in food preparation to account for size differences of fish. Food was cut into pieces suitable for the smallest fish in T2 to ensure that all fish were eating. Closed-System Respirometry A Closed System respirometer (Gooding et al, 1981, Cech, pgs. 343-345) was used to measure routine metabolic rate of tuna under the different diet treatments. An oxygen probe (Yellow Springs Instrument 5739 probe, attached to a YSl 52 Dissolved Oxygen Meter) measured dissolved oxygen within the oval tank (1,237 L). To minimize errors due to gas exchange at the waterlair interface, all air bubbles were removed before sealing the acrylic lid to the closed-cell foam gasket. Oxygen data was logged in spreadsheet form on a computer. Approximately 36 hours after a feeding, a yellow tagged fish was netted or captured in a sling from T2, placed in the respirometer and allowed to recover from any handling stress for about three hours. Outside stimuli were controlled by shielding the tank with a black curtain and keeping the noise level to a minimum. Before the high fat trials, a blue gel was placed over the light that illuminated the tank to further decrease stress levels. Six trials were run (n = 3 per diet treatment). Metabolic rate was measured by stopping flow through the tank and allowing the dissolved oxygen level to fall over time (60 minutes). Oxygen level was maintained above 80% saturation at all times. Between runs, flow to the tank was restored for 30 minutes and the dissolved oxygen levels were allowed to return to at least 90% saturation to avoid stressing the fish (Cech, pg. 345). Total trial duration averaged twelve hours for each fish. The first two runs of each trial were disregarded due to the elevated rate of oxygen consumption attributed to stress of capture and placement in the respirometer. A regression was calculated from dissolved oxygen level as a function of time for each trial using the last six runs - the first two runs ignored due to interfering stress response. The slope of the regression was used to calculate rate of oxygen consumption: Vo2 (mg Olkg per hour) - slope * time' volume mass mass = body mass of fish time = trial duration - 60 minutes volume = tank volume - 1,237 liters lo correct for any increase in oxygen consumption due to increased activity, runs were videotaped and swimming velocity was measured as time in seconds it took a fish to swim a marked distance on the tank (average of ten passes). Open-System Respiromety 12 was used as an Open or Flow Through respirometer to measure oxygen consumption of the entire population of fish in the tank (Cech, pgs. 347 - 348). A monitoring system (Program Logic Control), installed in each tank at the TRCC by the Monterey Bay Aquarium, provided data on the levels of dissolved oxygen in the out-flowing water of T2. The values were subtracted from in¬ flowing oxygen levels to calculate oxygen consumption. In-flowing oxygen levels were approximated from the relationship drawn between temperature and dissolved oxygen. Appendix 1. In-flowing water, from the Monterey Bay Aquarium water tower, was assumed to be 100% saturated (Chuck Farwell, personal communication) and oxygen levels were calculated accordingly, dependent on water temperature. Whole tank respiration was calculated using the following equation (Gooding et al., 1981): Vo2 (mg 021 kg per hour) = (102 inl -102 outl)* Flow N* mass“ (1-exp(-Elow ' time) volume Flow = rate of exchange volume = tank (109,765 liters) Time = 1 hour N = number of fish mass = mean body mass of fish (4.9 kg) Background metabolic rate was subtracted and a comparison of post-feeding increase in oxygen consumption made between diets. Feeding event was plotted as hour Ö and the areas under each curve were used to approximate total oxygen consumed. Stomach temperature measurements Minilog temperature probes were used to record stomach temperature. The probes were inserted into pieces of food, either sutured into the mantle cavity of a squid or into a piece of sardine, to insure ingestion upon introduction to the tank. The tuna would periodically evacuate their stomachs and the probes could then be retrieved from the bottom of the tank and the temperature measurements downloaded. Temperature probes recorded measurements at 3 minute intervals and remained in the fish for a mean duration of 2 days. The maximum time between ingestion and evacuation was 6 days. The stomach temperature measurements were converted to thermal excess by subtracting the temperature measurements recorded from a minilog probe suspended in the tank to record ambient water temperature. Fluctuations in water temperature, due to the heating system, did not affect the internal temperature of the fish immediately (Heidi Dewar, personal communication). A lag period of 24 minutes, between changes in water and stomach temperatures, approximated a best fit. Calibrations All oxygen probes were calibrated against Hach Winkler titrations which are accurate to 0.2 mgll. Miniloggers and all other temperature probes used were calibrated in a water bath against a thermometer previously corrected to the Bureau of Standards thermometer. Results Diet Analysis Based on data from the bomb calorimetry analysis, the total caloric content and percent composition of different food types was shoun to vary greatly towards a higher percentage of fat in sardines. Figure 1. The LFD was composed mainly of squid and smelt while the HFD was composed only of sardines. Though each diet was of equivalent caloric value (28 kcallkg/day), the LFD was 1% fat (s1 kcal of fat/kg/day) as compared to the HFD which was 8% fat (2.56 kcallkg/day). Figure 2. Closed-system Respirometyy Rate of uptake (mg O2/ kghr') was averaged from the last six runs of each trial. Table 1. Differences in activity between fish were corrected for by graphing oxygen consumption as a function of velocity (body lengths/second). Velocity was measured as both centimeters per second and body lengths per second. Oxygen consumption increased with increasing swimming speed. Figure 3. A regression was calculated for each trial (due to erratic behavior and swimming speeds, one trial from each treatment was discarded from the comparison (n - 2)) and extrapolated to O swimming speed (cm/s) to obtain standard metabolic rate (LFD : 135.7 +- 30.26 mg Oø/hr.kg; HFD : 147.54 +- 6.29 mg O2/hr.kg) A comparison across diets shows no significant differences due to treatments (p = 0.80). Table 2. The trial averages from each diet were combined and expressed as oxygen consumption against swimming velocity which shows no difference between the two treatment levels. Figure 4. The results from each individual trial was also compared to previously published results by Dewar and Graham, 1994, taken in a water tunnel where swimming speed could be controlled. The values for these six trials were lower than others at equivalent swimming speeds. Size may be influencing results (mean fork length = 54.2 cm: Dewar and Graham mean fork length (largest size class) = 51 cm). Figure 5. Open-system Respirometyy The LFD returns to baseline metabolic rate at approximately 34 hours after feeding while the high fat feeding does not return to original level before 36 hours after feeding. Figure 6. The ratio of the area under each curve for HFD and LFD whole tank respiration is 1.003. The high fat feeding shows a trend towards a longer recovery period or longer duration of elevated oxygen consumption. Stomach Temperature Measurements Though pre-feeding and post-feeding measurements within each diet were statistically different, LFD and HFD means of the thermal excess measurements were not significantly different. (LFD pre-feed: 0.26 +/- 004 ?C thermal excess; LFD post-feed 0.46 +-0.03 °C; HFD pre-feed = 0.26+/- 0.025 •C. HFD O.44 +0.02 °C) (p value 0.08) Figure 7. Discussion By using oxygen consumption and stomach temperature as indicators of metabolism, it was determined that no significant difference could be measured between diet treatments. When an analysis of variance (ANÖVA) was calculated, the variation within groups was found to be the majority of the total variation of individual respirometry trials and thermal excess measurements. Because of such a large variance within groups, the power of these statistical analyses is very small (Sokal and Rohlf, pgs. 260-265). A larger sample size will increase the sensitivity of statistical tests in determining whether the insignificant differences are actually due to no effect from treatments or if the difference is too small for detection without a high power test. There was a trend in the whole tank respiration that indicated a longer period of elevated oxygen consumption after a HFD feeding that could indicate a higher oxygen consumption during digestion for a HFD. This trend in post-feeding oxygen consumption (n = 1) can not be accepted as representative without further trials. Because of the ambiguities due to variables such as small sample size, gas exchange at the air and water interface, and outside stimuli, the results of a high fat diet change on the metabolism of a captive yellowfin tuna are inconclusive Digestion of different metabolites are only few of many variables that will influence metabolism, or the total use of energy in a system. Öther variables that will affect the rate of energy utilization in a captive yellowfin tuna include activity levels, oxygen availability, body size, hormones, and temperature in an abbreviated list (Jobling, pgs. 121-145, Schmidt-Neilsen, pgs. 177-218). Even a post-feeding increase in oxygen consumption and thermal excess is due to multiple factors such as the energy dependent processes of metabolite degradation, storage of excess calories, and to a small extent, mechanical work of digestive organs (Carey et al., 1984). Äfter an increase in dietary fat intake, there are three possible ways in which we may expect metabolism to change; it may increase, decrease or remain constant. The metabolic rate of the captive tuna may increase due to the higher energy cost of fat oxidation and the availability of high energy metabolites circulating in the blood stream, promoting the use of lipids rather than glycogen as fuel. Since the degradation of fatty acids is dependent on low ATP levels, high activity levels may increase the energy spent in oxidation which in turn may increase metabolic rate (Stryer, pg. 628 - 629). Metabolism may remain constant due the higher vield of energy from fat; it requires the oxidation of less fat than other metabolites to gain an equivalent amount of energy (Schmidt-Neilsen, pgs. 178-180), Finally, there is a possibility that metabolism will decrease due to the lower energy cost of converting ingested fat into free fatty acids (Jobling, pgs. 66-68) in comparison to protein or carbohydrate being first degraded into intermediates (Acetyl Coenzyme A) and then converted to free fatty acids (Lehninger, Nelson and Cox, pgs. 642-685). Since the metabolic rate remained fairly constant across treatment levels, then the lack of measurable difference could suggest that an equal amount of total energy is being metabolized on both diets. This would increase the amount of oxygen consumed on the high fat diet which may be offset by the fact that more energy is being expended on the low fat diet to store excess calories as fat. As a metabolite, fat contains twice the caloric content as protein or carbohydrate (fat = 9.4 kcallg, protein = 4.3 kcallg, carbohydrate = 4.2 kcallg). But, it also requires less energy to convert dietary fats than proteins or carbohydrates into lipid stores (Schmidt-Neilsen, pg. 179). This is one reasonable explaination for the insignificant results measured. It is possible that the lack of any measurable differences in oxygen consumption or stomach temperature may be due to the short length of the diet treatment. The duration of the high fat feedings (5 weeks) may have been too brief to elicit a physiological change in tuna metabolism. If there was a change, the difference may have been too small for the sensitivity of the experiment and statistical tests to quantify. Data from free fatty acid assays of blood serum (Fletcher, 1997), sampled from T2 at the end of each treatment, does indicate a physiological effect of the increase in dietary fat content. As compared to samples taken from the low fat feedings, levels of free fatty acids increased in samples taken after a 36 hour fast after the 5 weeks of high fat feedings. This would suggest that more free fatty acids are available both for storage and utilization and may influence the metabolism of the captive tuna at levels of food stuff oxidation and energy storage in lipid accumulation. Again, we must consider the multiple factors that cause an increase in free fatty acids concentrations in blood serum, including increased activity and starvation. (Stryer, pgs. 634 - 640) Further Research To elucidate the effect of a high fat diet on tuna physiology, one suggestion for further research is a longer treatment duration and a larger sample size. There are other questions that need to be pursued to determine the effect of high dietary fat content on the metabolism of a captive yellowfin tuna. In a captive tuna, what amount of fat ingested is actually absorbed? What amount is immediately utilized and what amount is stored? And how does this compare to the fat utilization of wild tuna? And finally, is the threshold at which excess calories are stored as lipid or glycogen similar in captive tuna as compared to wild tuna? The study of fat utilization in yellowfin tuna is pertinent to their feeding strategy. Especially in the wild, where seasonal variations in prey availability and nutritional content compound the bioenergetic gamble that yellowfin tuna play, there is potential for diet to influence metabolism and other physiological processes. Literature Cited Adams, Marshall and James Breck. "Bioenergetics" pgs. 389 - 415 in Methods in Fish Biology. Carl Schreck and Peter Moyle, editors. 1990. American Fisheries Society, Bethesda, Maryland. Barton, Kimby, Martin Gerrits, and James Ballantyne. 1995. "Effects of Exercise on Plasma Nonesterified Fatty Acids in Arctic Char (Salvihnus alpinus)". Journal of Experimental Zoology, v. 271: 183 - 189. Block, Barbara and John Finnerty. 1994. "Endothermy in Fishes: a Phylogenetic Analysis of Constraints, Predispositions, and Selection Pressures". Environmental Biology of Fishes. v. 40: 283-302. Carey, Frank, John Kanwisher and Don Stevens. 1984. "Bluefin Tuna Warm Their Viscera During Digestion". Journal of Experimental Biology. v. 109: 1-20. Cech, Joseph. "Respirometry" Pgs. 335 - 348 in Methods in Fish Biology. Car Schreck and Peter Moyle, editors. 1990. American Fisheries Society. Bethesday, Maryland. Dewar, Heidi and Jeffrey Graham. 1994. "Studies of Tropical Tuna Swimming Performance in a Large Water Tunnel". Journal of Experimental Biology. v. 192: 13-31. Jobling, Malcolm. Fish Bioenergetics. 1994. Chapman and Hall, London. Korsemeyer, K., and H. Dewar, N. Lai and J. Graham. 1996. "The Aerobic Capacity of Tunas: Adaptation for Multiple Metabolic Demands". Comparative Biochemistry and Physiology. v. 113 (1): 17 -24. Lehninger, Albert, David Nelson, and Michael Cox. Principles of Biochemistry 2“ edition 1993. Worth Publishers, New York. Olson, Robert and Christofer Boggs. "Apex Predation by Yellowfin Tuna (Thunnus albacares): Independent Estimates from Gastric Evacuation and Stomach Contents, Bioenergetcs, and Cesium Concentrations. Canadian Journal of Fisheries Aquatic Science. V. 43:1760 - 1775. Perla, Bianca. 1995. "Food requirements and Lipid Accumulation in Captive Yellowfin Tuna: Thunnus albacares .“ Hopkins Marine Station unpublished. Reynolds, Arleigh, Laurent Fuhrer, Harris Dunlap, Mark Finke and Francis Kallfelz. 1984. "Lipid Metabolite Responses to Diet and Training in Sled Dogs". Journal of Nutrition, supplement: 2754S - 27588. Schmidt-Neilsen, Knut. Animal Physiology: Adaptation and Environment. 3“ edition. 1975. Cambridge University Press, Cambridge. Sokal, Robert and James Rohlf. Biometry. 3“ edition. 1995. W.H.Freeman and Company, New York. Stryer, Lubert. Biochemistry. 3“ edition. 1988. W.H.Freeman and Company, New York. Acknowledgements This project was dependent on the generous help of many workers from Hopkins Marine Station and the Monterey Bay Aquarium, to whom I owe much gratitude. I would especially like to thank my advisors, Dr. Barbara Block and Chuck Farwell, for their quidance, Ellen Freund for her advice and the use of her equipment, Heidi Dewar for the use of her minilog probes, and Doug Fudge for his patience and technical support. Appendix 1 : Relationship of Temperature to Dissolved O, Levels Tank 2 7.5 7.4 7.3 oo 0 7.2 7.1 o 00 00 7.0 6.9 6.8 6.7 O,]=-0.117(C) + 9.62 6.6 P=0.03 6.5 - 23.0 22.0 22.5 21.5 21.0 Ambient Water Temperature ?0 ON 80 - 3888 + Vu 3 C a 2 . + + oo O 8. . - o a1 * * 5 IE * * 28 53 . 55 Figure Legends Figure 1: Caloric composition (kcal/100g) of different size classes of squid, smelt and sardines. Results from bomb calorimetry analysis. Figure 2: Total calories and caloric composition of Low Fat and High Fat Diet treatments in kcal/kg/day. Figure 3: Results from the Closed-system respirometer for Low Fat and High Fat Diets are plotted as oxygen consumption to activity (or swimming velocity). Two fish, short yellow and black base, were repeat trial subjects across the change in diet. (Each individual fish is identified by tag description.) Figure 4: The average oxygen consumption and swimming velocity for each diet treatment from the Closed-system respirometer. Figure 6: Results from Closed-system respirometer are compared with previously published results (Dewar and Graham, 1984) collected in a water tunnel where swimming velocity could be controlled. The average fork length of Diet-manipulated fish was 54.2 cm. Figure 6: By using T2 as an Open-system respirometer for over 36 hours and subtracting out baseline (non-absorbtive, inactive respiration, oxygen consumption demonstrates the sharp increase after feeding (absorbtive state) that gradually returns to baseline values for both Low Fat (LFD) and High Fat Diet (HFD). Duration may be too short to observe full return to baseline in HFD. Figure 7: Stomach Temperature measurements for the Low Fat and High Fat Diets demonstrate a significant difference between pre¬ and post-feeding (LFD p=0.003; HFD p=0.05) but no significant differences between diet treatments (pre-feed p=0.91; post-feed p-0.80). 30 28 26 24 22 20 18 16 14 12 10 - 8 - 6 - Figure 1: Bomb Calorimetry Results Fat Protein Carbohydrate E 30 29 28 - 27 26 § 25 4 - Figure 2 : Distribution of Calories Across Diets Low Fat Diet High Fat Diet 28.00 4.94 3.96 2.57 0.59 0.24 0.21 Total Protein Carbohydrate Fat Kilocalories 350 300 250 200 - 150 100 50 0.6 Figure 3 : Individual Respirometry Trials - black base long grey - black base 1 short yellow spot Q short yellow Low Fat Diet • High Fat Diet 0.8 1.0 Speed (body lengths second") 1.2 350 300 - 250 200 150 100 Figure 4 : Metabolic Rate of Captive Yellowfin Tuna Between a High and Low-fat Diet O Low Fat Diet • High Fat Diet 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Swimming velocity (body lengths second") Figure 5 : 500 400 300 42cm 21 Sen 32cm 14* 5 0 228 880 800 0 . 100 125 150 50 25 Swimming velocity (cms*) Dewar and Graham, 1994 200 150 100 50 Figure 6 : Whole Tank Respiration High Fat Diet LoW Fat Diet -50 5 10 15 20 25 30 35 40 Time (hours) Feeding Event Figure 7 : Thermal Excess in Tuna: A Comparison Across High and Low-Fat Diets 0.7 pre - feeding (p) = 0.9109 not significant post - feeding (p) = 0.8007 not significant 0.6 0.5 2 0.4 L 0.3 0.2 0.1 0.0 Post-Feed Pre-Feed Pre-Feed Post-Feed High Fat Diet Low Fat Diet