and the burn is discovered relatively quickly, within 12 hours of catch­ ing. However, if it is exported, Burnt Tuna: Conditions butchering usually occurs after the Leading to Rapid Deterioration fish is sold to the last and most dis­ in the Quality of Raw Tuna tant retailer, from 36 to 60 hours after catching. Transportation costs have thus been incurred before dis­ J. L. CRAMER, R. M. NAKAMURA, covery of the condition. A. E. DIZON, AND W. N. IKEHARA Upon discovery of burn, the fisherman is required to refund a portion of the auction price to the wholesalers which is then used to defray a fraction of the costs in­ Introduction recreational fishermen who troll off volved in the marketing and ship­ Hawaii's Kona coast. The problem ping. Rebates on burnt fish were In Hawaii, raw tuna flesh which is now a major economic concern of reported to range from 5 to 75 per­ seems paler and softer than normal the night handline fishery of the cent of the original selling price in is characterized by buyers as being same island. This highly effective, 1977. Annual losses to the night "burnt." High-quality tuna should cost efficient industry, found only handline fishermen were estimated be translucent, red, and firm. Burnt in Hawaii and in the Philippines to be about 16 percent of the total tuna, because of its poor texture, (Yuen, 1979), would be an excellent value of the catch (Cramer et al.)'. color, and slightly sour taste, while candidate for fishery development Clearly the reduction of burn edible, is undesirable for raw con­ in export-poor Pacific island na­ through improved fishing and sumption (as "sashimi"). It there­ tions if the burn problem could be handling methods is an important fore commands only a fraction of controlled. Whatever the causes of goal and some method must be the selling price of high quality burn, we suspect that the problem is developed to identify the condition tuna, depending on the extent and exacerbated by the limited chilling of the fish early in the marketing se­ the severity of burn. However, the facilities found aboard most night quence before filleting. burnt condition is not absolute in handline and recreational fishing These requirements formed the extent or degree; the affected area boats. motivation for our work. We can vary from 5 to 100 percent of In Hawaii, traditional marketing reasoned that our best strategy to in­ the total marketable meat of the fish practices delay the discovery of vestigate burn was to correlate and range in severity from marginal burn. A typical fish changes hands quality of the fish with those to severely burnt. The variability in at least twice in the first 48 hours measurable variables that would be occurrence compounds an already after death. The fisherman consigns likely to be responsible for, or at difficult diagnostic problem. his catch to a wholesaler who then least indicate burn. Those correlates The condition periodically occurs either sells it locally or ships it to a fell into three classes: 1) Fishing in the large (45-136 kg) tropical more distant market. If the fish is variables such as time of the year, tunas-the yellowfin tuna, Thunnus sold locally, the fish is butchered temperature of the water, fighting albacares, and bigeye tuna, T. time, and care of the catch; obesus-and was first brought to 2) biological characteristics such as the attention of the National Marine sex, weight, species, and body Fisheries Service (NMFS) in 1974 by temperature at death and after; and 3) biochemical, histological, and J. L. Cramer was with the Honolulu pH samples. We hoped that burnt Laboratory, Southwest Fisheries Center, Na­ fish would exhibit some differences tional Marine Fisheries Service, NOAA, 2570 in key characters relating to exercise Dole Street, Honolulu, Hawaii; present ad­ ABSTRACT-Burnt tuna is raw tuna dress: Northwest and Alaska Fisheries or flesh quality. Thus, differences which is paler and softer than normal. Center, National Marine Fisheries Service, between burnt and normal tuna This study indicates that the burnt tuna NOAA, P.O. Box 21, Mukilteo, WA 98275. condition results from muscle cell R. M. Nakamura is with the Department of degeneration which begins prior to the Animal Sciences, University of Hawaii death of the fish and proceeds more (Manoa), 1800 East-West Road, Honolulu, HI 96822. A. E. Dizon is with the Honolulu rapidly after death than in normal tuna. Laboratory, Southwest Fisheries Center, Na­ 'Cramer, J. L., R. s. Shomura, and H. S. H. Female sex, longer fighting times, and tional Marine Fisheries Service, NOAA, 2570 Yuen. 1978. The problem of burnt tuna in the less efficient chilling are positively cor­ Dole Street, Honolulu, HI 96812. W. N. Hawaiian fishery. National Marine Fisheries related with the occurrence and severity Ikehara is with the Department of Zoology, Service, Southwest Fisheries Center Admin. of the burnt tuna condition. University of Hawaii, Honolulu, HI 96822. Rep. IlH, 17 p.

12 Marine Fisheries Review would indicate the causes and sug­ superficial (just under the skin) tis­ Statistical Analysis gest remedies. sues of fish. Fish held in captivity or trans­ Correlations of quantitative var­ Materials and Methods ported live to KRF were placed in iables were determined by simple Fish ice water for 4-8 hours after killing regression analysis. Correlations in­ to simulate normal commercial pro­ volving one or more nonparametric Bigeye and yellowfin tunas were cedures for treatment of fish prior variables were determined by using studied in this report. Three tunas to off-loading. Spearman's correlation rank test were obtained from commercial The handling of fish during the (Snedecor and Cochran, 1967). fishermen on the island of Hawaii. first 24 hours after off-loading from The significance of differences Seven fish were caught off Kona, the boats was evaluated and catego­ between mean values were deter­ Hawaii, and transported live in the rized. The fish were designated to mined using a t test for the differ­ baitwell to the Kewalo Research have received: 1) Excellent treat­ ences of means (Snedecor and Facility (KRF) in Honolulu and two ment when stored in ice water at Cochran, 1967). fish were held in captivity at KRF O°C, 2) fair treatment when stored Results for over 6 months before being used in refrigerators at 5°C, and 3) poor in these studies. treatment when stored at ambient The fighting time and other rele­ The date, area of catch, fighting temperature of 28°C. vant data on the 12 fish used in this time, weight, species, and sex of the After the fish were auctioned and study are recorded in Table 1. fish were recorded. The prechill in­ quartered, the flesh was graded and Significant (P<0.05) positive cor­ terval (the time from landing to scored on the basis of color and tex­ relations were found between chilling) was also recorded. ture as: Excellent, 1 point; good, 2 fighting time and concentrations of points; marginal, 3 points; poor, 4 blood lactate (Fig. 1) and blood Sampling and Testing points; and very poor, 5 points. glucose (Fig. 2) at death and be­ Temperature, blood, and muscle These subjective evaluations were tween fighting time and tissue lac­ samples were recorded when the fish corroborated by discussions with tate (Fig. 3) when fish were being were: 1) First landed on the boat, the auctioneer and correlated with off-loaded from the boats. A 2) off-loaded from the boats, and the price received per pound of fish. negative correlation was found be­ 3) at 24 hours after off-loading. tween fighting time and relative Blood was obtained by cardiac acidity of muscle at off-loading puncture immediately after boating Tests (Table 2). The correlations of these the fish using silicone-coated vacu­ Blood lactate and glucose concen­ parameters are also indicated in tainers (Kimble-Terumo, Elkton, trations were determined by en­ Table 2. MD.).2 Captive fish were bled just zymatic analysis utilizing the con­ There was a significant difference prior to killing, and 2 ml of whole version of NAD to NADH as (P<0.05) between sexes in quality of blood were immediately mixed with described by Burgmeyer (1974). A flesh of fish over 30 kg in weight 2 ml of 15 percent perchloric acid creatine phosphokinase (CPK) assay (Table 3). The flesh of female fish for glucose and lactate determina­ was also conducted following the was more often of low quality. An tions. The serum was separated method of Burgmeyer (1974). The interaction was found between from the remaining blood and fro­ above assays were performed at the zen for later testing. University of British Columbia, A coring tool was used to obtain Vancouver, with the direction and tissue samples from the deep muscle assistance of P. Hochachka and J. Table 1.-Data on fish studied. near the vertebral column. Tissues Balantyne. Fight· were frozen on dry ice or fixed in The pH of muscle tissue was Date ing caught Area time Wt. Dietrich's solution. The latter tis­ determined by titration and the (1978) caught (min.) (kg) Species Sex sues were processed to obtain amount of potassium carbonate re­ 21 Aug. Hila' 9 115.8 T. a/baeares M Hemotoxylin and Eosin stained sec­ quired to lower the pH of the muscle 21 Aug. Hila' 10 63.6 T. a/baeares F tions for microscopic examination. suspension to a pH of 5.6-6.0. A pH 27 Sept. Kewalo' 15 10.8 T. a/baeares F 27 Sept. Kona2 9 5.2 T.obesus M An electronic thermometer with a meter with a 4 mm diameter probe 27 Sept. Kana3 9 4.4 T.obesus M 27 Sept. Kana3 4 3.4 T. a/baeares F 12 cm probe was used to obtain was used to determine the pH of the 27 Sept. Kana) 4 3.5 T. a/baeares F temperatures of deep (11 cm) or midsection and tail muscle tissue of 27 Sept. Kana) 12 3.7 T. a/baeares M 27 Sept. Kana) 18 3.6 T. a/baea,es M yellowfin or bigeye tuna on the auc­ 27 Sept. Kana3 31 3.5 T. a/baeares F tion floor. The pH was also deter­ 17 Oct. Kewalo' 0 9.4 T. a/baea,es F 23 Oct. mined by pH meter for macerated Kana' 15 50.4 T.obesus F 'Reference to trade names or commercial muscle tissue suspended in distilled 'Sampled at sea. firms does not imply endorsement by the Na­ 'Captive fish. tional Marine Fisheries Service, NOAA. water. 'Transported live to KRF.

June 1981,43(6) 13 1.6 6.0

5.5 • 1.4

5.0 1.2 • 4.5 0> • "- • '" • '"0 4.0 E - E 1.0 ••• "- • .3 '" .c "0'" 3.5 • E 0 .3 • "0'" 0.8 3.0 a • 2 0 '" • '"0 U u .!2 2.5 ::> "0 0> 0.6 0 Q) .!2 2.0 ::> CD '" f='" 1.5 • •• 0.4 • 1.0 • 0.2 0.5

0 0 5 10 15 20 25 30 35 oOL---L5----l.10---15-----J20---21..5---3LO---.J35

Catching time (minutes) Catching time (minutes) Figure I.-Correlation between catching time (fighting Figure 2.-Correlation between catching time (fighting time) and concentration of blood lactate at death. Line time) and concentration of glucose at death. Line fitted fitted by least squares methods. by least squares methods.

Table 2.-Coneletlona between fighting time (FT), Table 3.-The effect of alze and aex on f1eah Table 4.-lnteraetlona between blood laetatele.ela at blood lactate le.el at death (BL), tlaaue glueoae le.al temperature and fleah quality. death and 24·hour treatmenta after off·loadlng. at death (TGa), tlaaue lactate le.el at off·loadlng (TLb), relatl.e acidity In tuna tlaauea at off·loadlng (TAb), Body Temperature ('C) Flesh quality' Quality at and quality of flah at quartering (0). wI. quartering' (kg) Sex Average Range Average Variance Blood lactate Quality (24 h after Independent Dependent micromole/ml at death' Treatment' death) .arlable variable Correlation <11 M 27.5 26.9·28.4 2.4 1.3 <11 F 27.7 26·29.5 2.8 0.7 0.91 1 Excellent 1 >30 M 29.6 29·31 1.5 1.0 1.42 1 Excellent 1 FT BL '0.64 >30 F 29.7 27·31 2.8 1.6 1.42 1 Poor 3 FT TGa '0.69 1.47 1 Fair 2 FT TLb '0.59 '1 = Excellent, 2 = good, 3 = marginal, 4 = poor, 1.82 2 Fair 2 FT TAb '·0.66 and 5 = very poor. 2.58 2 Fair 2 BL Q '0.82 2.67 1 Fair 2 3.20 2 Poor 3 'Significant at the 0.05 percent level. 2.83 1 Poor 4 'Significant at the 0.01 percent level. 4.25 3 Poor 4 4.49 2 Fair 3 5.64 2 Poor 4 treatment after off-loading and con­ In some cases, there was excellent 'Quality was determined histologically and was scored as follows: 1 = Excellent, 2 = good, 3 = centrations of blood lactate at death correlation between pH of muscle marginal, 4 = poor, and 5 = very poor. tissue and quality of tuna flesh 'Excellent = stared in ice water at 0 'C, fair = stored (Table 4). Low lactate and good in refrigerators at 5 'C, and poor = stared at ambient treatment resulted in excellent qual­ (Table 5) while in other instances temperature of 28 ·C. ity flesh; low lactate and poor treat­ lower pH could not be correlated ment or high lactate and fair treat­ with poor quality flesh (Table 6, ment resulted in marginal quality Fig. 4). These differences are prob­ No significant correlations were flesh; and high lactate and poor ably due to differing temperatures found between deep body tempera­ treatment produced poor quality and durations before auction tures and quality of fish or results of flesh. (Nakamura et al., 1977). biochemical tests. Nor could signifi-

14 Marine Fisheries Review 60 0 • •• • 0> '- V> C ., B "0 E A ::t

0> .!:: 2 0 "0 0 E >- 30 0 • ::> -0 C -; ~ 3 0 "0 • E 20 ~ ~ ::> ::2' 4 10

0'---"'---'-----'----'---'---"'---' 5'----'------__-'-_-' o 5 10 15 20 25 30 35 5.5 6.0 Catching time (minutes) Average pH

Figure 3.-Correlation between catching time (fighting Figure 4.-Correlation between pH and auction quality time) and muscle lactate at off-loading. Line fitted by (1 = excellent, 2 = good, 3 = marginal,4 = poor, and least squares methods. 5 = very poor) of: A) all fish sampled, B) fish sampled on 28 July, C) Honolulu fish sampled on 10 August, and D) Hilo fish sampled on 10 August.

Tabl. 5.-Quallty and pH ot mldaactlona and talla 01 yallowlln tuna at tha Honolulu Table 6.-Quallty and pH 01 mldaactlona end talla 01 Ilah auction, 28 July 1978. yellowtln tuna at the Honolulu and Hllo tlah auctlona, 10 Auguat 1978. Tuna quality Tuna quality Excellent Good Marginal Poor Very Poor Excellent Good Poor Mids Tall Mlds Tall Mlds Tall Mlds Tall Mids Tall Mlds Tail Mids Tail Mids Tall 6.0 6.1 6.0 6.2 5.6 5.7 5.6 5.6 5.4 5.5 6.0 6.0 5.9 5.6 5.7 5.8 5.5 5.6 Honolulu fish 6.0 5.7 5.9 5.9 5.8 5.9 5.7 5.8 5.8 5.9 5.8 5.9 5.7 5.9 5.6 5.6 5.5 5.6 6.0 5.5 6.0 5.6 5.6 5.7 5.6 5.7 5.7 5.7 5.9 5.8 5.8 6.0 5.9 5.9 5.6 5.7 5.9 6.0 5.6 5.7 5.7 5.8 Mean 5.93 5.90 5.90 5.90 5.68 5.78 5.53 5.63 Variance 0.01 0.03 0.01 0.06 0.01 0.02 0.03 0.03 Mean 5.68 5.74 5.60 5.67 5.95 5.65 Variance 0.2 0.01 0.01 0.00 0.08 0.02

Hilo fish

5.8 5.9 5.8 5.8 5.7 5.8 6.0 5.9 5.7 5.8 5.9 5.8 cant correlations be found between shortly before or immediately after 5.9 5.9 5.7 5.8 the pre-chill interval and quality of death; 2) edema was seen in muscle 5.9 5.9 fish or results of biochemical tests obtained at the time of boating; 6.0 6.0 5.8 5.8 of blood and tissues. 3) extensive edema and muscular 6.0 6.0 Histopathologic studies of muscle degeneration were seen in burnt fish 6.0 6.1 6.0 5.9 from a limited number of fish at the time of auctioning; and 6.1 6.1 revealed the following: 1) Inflam­ 4) tissue gram stain showed that 5.9 5.9 matory changes were not seen, sug­ bacteria was associated with the Mean 5.90 5.90 5.90 5.92 5.80 5.80 gesting that all changes occurred muscular degenerative changes. Variance 0.02 0.00 0.02 0.01 0.02 0.00

June 1981, 43(6) 15 Discussion 3) Chilling of fish after catching study (12) resulted in possible leads determines the temperature of the and solutions, but precludes strong As suspected by tuna fishermen, deeper portions of the fish muscles, and definitive statements on this the results of this study suggest that which are least affected by external condition. It is likely that improved the care and handling of fish after cooling procedures. handling and chilling procedures being caught is partly related to the 4) Poor chilling procedures allow from the time of capture may help subsequent quality of the flesh of higher temperatures in muscle, to reduce the incidence of burnt the tuna (Table 4). However, the allowing bacteria to grow more pro­ tuna. pathogenesis of burnt tuna may also fusely and resulting in increasing be related to physiological changes degrees of burn in the deeper areas Acknowledgments in the fish during the struggle while of the muscle mass. This research was funded by a being caught. Fighting time was The most reliable diagnostic National Marine Fisheries Service related to blood lactate and glucose method may be histopathologic ex­ contract. Technical assistance and levels at death (Fig. 1, 2). The amination of biopsy specimens ob­ laboratory space were provided by edema (abnormal accumulation of tained from deep muscle tissues, the University of Hawaii (Manoa), fluid) in muscle seen in histopatho­ flash frozen, sectioned with a University of British Columbia, and logical studies of biopsies obtained cryostat at the auction, stained and the NMFS Southwest Fisheries at death also suggest that the strug­ examined. All of this can be ac­ Center's Honolulu Laboratory. Ex­ gle of fish being caught may complished in 0.5-1 hour. With the pertise in fishing and fish handling, predispose the flesh to subsequent exception of tests made on blood without which this study could not degeneration. The lack of inflam­ samples collected at the time of have been completed, were provided matory changes at that time in­ catching, biochemical tests were not by fish wholesalers and fishermen in dicates that the condition is not useful in prediction and diagnosing Hilo, Kona, and Honolulu, Hawaii. developed prior to being hooked. the burnt tuna condition of un­ A hypothesis on the mechanism quartered tuna. Literature Cited for the development of burnt tuna is A stress-induced abnormality as follows. which occurs in hog muscle known Bergmeyer, H. U. 1974. Methods of en­ zymatic analysis. 2:437, 787. Acad. Press., 1) The struggle of fish while be­ as porcine stress syndrome (Win­ N.Y. ing caught may result in muscle stanley, 1979) seems similar in some Nakamura, K., Y. Fujii, and S. Ishikawa. edema, the severity depending on characteristics to burnt tuna. Both 1977. Turbid meat of tuna - I. Study on its possible causes. Bull. Tokai Reg. Fish. the intensity of the struggle. (Longer conditions result in pale and soft Res. Lab. 90. and more intensive struggling results muscles. It is possible that there may Snedecor, G. W., and W. G. Cochran. 1967. Statistical methods. Iowa State in higher blood levels of glucose and be related causes for the two condi­ Univ. Press, Ames, 593 p. lactate.) tions. Winstanley, M. 1979. A cure for stress. 2) The edema fluid of muscle pre­ It is obvious that further studies New Sci. 84:594-596. Yuen, H. S. H. 1979. A night handline sents a good medium for bacterial on burnt tuna are necessary. The fishery for tunas in Hawaii. Mar. Fish. growth. small number of fish sampled in this Rev. 41(8):7-14.

16 Marine Fisheries Review