by N.

Tomlinson
(with appendices by E. Bilinski, A. G. Comer,
S. E. Geiger, R. E. E. Jonas, H. L. A. Tarr, and
N. Tomlinson)
Bulletin No. 150
......
J
eSA3arch Board of Canada. Ottawa 1 9 6 6 1 ' ~
('.
~ .
. "
GREENING IN TUNA AND
RELATED SPECIES
Bulletins of the Fisheries Research Board of Canada are designed to assess and interpret
current knowledge in scientific fields pertinent to Canadian fisheries. Recent numbers in this
series are listed on the inside back cover of this Bulletin.
Editor:
J. C. STEVENSON
Associate Editor: G. I. PRITCHARD
Assistant Editor : R. H. WIGMORE
Production: R. L. MacIntyre
Fisheries Research Board of Canada
Sir Charles Tupper Building
Ottawa 8, Ontario, Canada
The Board also publishes the Journal of the Fisheries Research Board of Canada in annual
volumes of twelve monthly issues , and an Annual Report. Fisheries Research Board of Canada
publications are for sale by the Queen's Printer, Ottawa. Remittances must be in advance, payable
in Canadian funds to the order of the Receiver General of Canada. Publications may be consulted
at Board establishments located at Ottawa, Ont.; Nanaimo and Vancouver, B.C. ; Sault Ste. Marie
and London, Ont. ; Ste. Anne de Bellevue and Grande-Riviere, Que.; St. Andrews, N.B.; Halifax
and Dartmouth, N.S.; Ellerslie, P.E.I.; and St. John's, Nfld.
BULLETIN No. 150
Greening in tuna
and related species
by N. Tomlinson
Fisheries Research Board of Canada
Vancouver Laboratory, Vancouver 8, B.C.
(with appendices by E. Bilinski,
A. G. Comer, S. E. Geiger, R. E. E. Jonas,
H. L. A. Tarr, and N. Tomlinson)
FISHERIES RESEARCH BOARD OF CANADA
Ottawa, 1966
c _
Crown Copyrights reserved
Available by mail from the Queen's Printer, Ottawa,
and at the following Canadian Government bookshops :
OTTAWA
Daly Building, Corner Mackenzie and Rideau
TORONTO
Mackenzie Building, 36 Adelaide St. East
MONTREAL
LEterna-Vie Building, 1182 St. Catherine St. West
WINNIPEG
Mall Center Bldg. , 499 Portage Avenue
VANCOUVER
657 Granville Street
or through your bookseller
A deposit copy of this publication is also available
for reference in public libraries across Canada
Price $1.25 Catalogue No. Fs 94-150
Price subject to change without notice
ROGER DUHAMEL, F.R.S. C.
Queen' s Printer and Controller of Stationery
Ottawa, Canada
1966
iv
Contents
ABSTRACT, vi
INTRODUCTION,
DESCRIPTION AND IDENTIFICATION, 1
OCCURRENCE, 1
COMPOUNDS RESPONSIBLE FOR COLOR OF COOKED TUNA, 2
V ARIABILITY IN THE OCCURRENCE OF GREENING, 5
DETECTING SUSCEPTIBILITY TO GREENING IN RAW FLESH, 9
Metmyoglobin, 9
Concentration of an alcohol-soluble substance producing a yellow color with
ninhydrin, 10
Color of an aqueous extract of raw kidney, 10
Trimethylamine oxide (TMAO) concentration, 10
Redox indicator decolorizing power, 10
Color of precooked small flesh samples, 11
MEANS OF PREVENTING GREENING, 11
ACKNOWLEDGMENTS, 11
REFERENCES, 12
ApPENDIX I. The haem pigments of normal and of green albacore tuna and the
influence of certain additives on color. By N. Tomlinson and
s. E. Geiger, 15
ApPENDIX II. Greening in albacore tuna in relation to the trimethylamine oxide
(TMAO) and volatile carbonyls content. By E. Bilinski and
R. E. E. Jonas, 18
ApPENDIX III. Relationship between the reducing sugars of albacore tuna flesh
and greening. By H. L. A. Tarr and A. G. Comer, 21
v
ABSTRACT
The literature regarding the development of undesirable tan to brownish-green
colors ("greening") in tuna during precooking has been reviewed.
The undesirable color is caused by the oxidation of haemochromes derived
from myoglobin and haemoglobin normally present in tuna flesh and occurs irregu-
larly. The available evidence indicates that a relatively large number of factors can
influence the occurrence of greening, and it seems that apparent contradictions
between the findings of different investigators regarding the reason for the occurrence
of greening have arisen as a consequence of the complexity of the system involved.
A number of workers have proposed various tests for predicting whether or not
tuna will yield green meat when cooked. These tests are described briefly and their
limitations indicated.
The information available concerning the use of certain additives (reducing
agents and nitrogenous haemochrome-forming compounds) during canning of tuna
as a means of eliminating greening has been considered, and local experience with
the use of some of these compounds (ascorbic acid, sodium nitrite, and nicotinamide)
is described.
VI
INTRODUCTION
The occurrence in canned tuna of an off-color known as "greening" is a matter
of concern and of financial loss to the fish processing industry. Local canners of
albacore tuna imported from Japan usually encounter a small percentage of so-
called "green" fish, but recently one processor met with an unusually high incidence
of fish showing the fault and requested assistance in finding some means of reducing
losses occasioned by it.
A fairly extensive literature has accumulated regarding green tuna, and it
appeared a compilation of the information available on the subject would be of
value. This has been prepared and is presented below, together with appendices
summarizing results of certain investigations made in this laboratory. The report
does not constitute a complete review of the literature, but is an attempt to bring
together the information that seems pertinent to the problem.
DESCRIPTION AND IDENTIFICATION
In preparing tuna for canning, the eviscerated fish are precooked at 215-225 F
for a period of 2-6 hr, the length of time depending on the size of the fish. After
being precooked, the flesh, if it is to yield a high-quality canned product, should
be white or more or less pink in color. However, the flesh of some fish takes on
undesirable tan to brownish-green colors during the precook. This phenomenon
has been termed "greening" and is a serious matter to the industry since, depending
on the severity of the condition, green fish are rejected (Miyauchi, 1950; Brown
et aI., 1958; Naughton et aI., 1958; Dollar et aI., 1961; Amana, 1965). Hirao et ai.
(1958b) have published an excellent color plate illustrating the difference in ap-
pearance between green and good precooked yellowfin tuna flesh. Doctor Hirao's
original illustration was not available for reproduction, but he has kindly supplied
another which is reproduced here as Fig. 1.
Brown et ai. (1958) suggested that other off-colors may occur in tuna, but
state that the green condition can be identified by treating precooked off-color
tuna meat with sodium hydro sulfite. If the flesh then becomes pink, at least part
of the off-color is the result of greening.
OCCURRENCE
Greening has been encountered in albacore (Germo alalunga) (Miyauchi,
1950; Brown et aI., 1958; Sasano et aI., 1961; Aldrin, 1962), big-eye (Thunnus
obesus) (Amano, 1965), blue-fin (Thunnus orientalis) (Amana, 1965), yellowfin
(Neothunnus macropterus) (Naughton et aI., 1956, 1958; Brown et aI., 1958; Dollar
et aI., 1961; Sasano et aI., 1961), and skipjack (Katsuwonus pelamis) (Brown et aI.,
1958).
A good deal of variability in the incidence of occurrence of green fish has
been experienced in Japan (Miyauchi, 1950), in the United States (Dollar et aI.,
1961), in Africa (Aldrin, 1962), and in British Columbia (A. Woodland, personal
communication). Miyauchi (1950) stated that green albacore are winter-caught fish,
between 30 and 40% of albacore tuna landed in Japan during winter months
yielding a green product. The percentage of such fish in the catch was higher during
the period October to December and March through April than in January and
February. Aldrin (1962) reported a landing of 100 tons of albacore at Abidjan,
Cote d'Ivoire, Africa, of which 60% became green when cooked. These fish were
caught in African waters apparently in Mayor early June.
COMPOUNDS RESPONSIBLE FOR COLOR
OF COOKED TUNA
It has been established, largely through the efforts of Brown, Tappel, and
Olcott of the Department of Food Technology, University of California, and of
Naughton, Frodyma, and Zeitlin of the Department of Chemistry, University of
Hawaii, that the desirable pink color of good-quality cooked tuna flesh is caused
by the presence of haemochrome(s) derived from the heamoproteins myoglobin
and haemoglobin normally present in the flesh. Myoglobin and haemoglobin are
similar compounds and are responsible in mammals for the red color of muscle and
blood, respectively. In the ordinary (or light) muscle of tuna, myoglobin usually
accounts for much more than 50% of the total haemoprotein (Matsuura and
Hashimoto, 1959; Brown, 1962) , the remainder being haemoglobin with a very
small amount of cytochrome (Matsuura and Hashimoto, 1954) . The first two
groups of investigators mentioned above have also shown that the undesirable tan
or brownish-green colors that occur in so-called green flesh are associated with
the appearance of haemichrome(s) formed by the oxidation of the haemochrome(s)
responsible for the normal pink color. Similar color difficulties have been known in
cured meat products for some time and were investigated earlier (see review by
Watts, 1954) . The chemistry of the haem compounds, and of other metallo-por-
phyrins and related porphyrins and bile pigments, is of great interest and has been
the subject of literally thousands of scientific publications (see reviews by Lemberg
and Legge, 1949; Falk, 1964).
Matsuura and Hashimoto (1954) investigated the pigments of raw tuna flesh
and found them to be principally haemoglobin and myoglobin, the latter apparently
accounting for much the greater portion of the total haem compounds of the
red muscle (both superficial lateral and deep-seated) than of the ordinary muscle,
in which the haem was nearly all present as haemoglobin. Naughton et al. (1958)
also reported that the principal pigment of the ordinary muscle was apparently
haemoglobin. Of the two muscle types, the ordinary muscle is of greater interest
to the North American processor. Two criteria were used by Matsuura and Hashi-
moto (1954) and by Naughton et al. (1958) in distinguishing between myoglobin
and haemoglobin. The first was the insolubility of haemoglobin in 3 M phosphate
buffer atpH 6.6 in contrast to the solubility of myoglobin under the same conditions.
2
s. Hirao
FIG. 1. Color of precooked yellowfin tuna. Left - green, Right - normal.
The second was the difference in the spectral characteristics of the two compounds,
the absorption peaks of myoglobin in the visible portion of the spectrum being
shifted a few millicrons higher in wavelength than those of haemoglobin. However,
although these criteria apply in the differentiation of mammalian myoglobin and
haemoglobin, it is now known that they do not do so in the differentiation of the
tuna compounds. Matsuura and Hashimoto (1955) found that the spectral pro-
perties of tuna myoglobin were the same as those of tuna and mammalian haemo-
globin, and Rossi-Fanelli and Antonini (1955) and Matsuura and Hashimoto
(1956) reported that the solubility properties of tuna myoglobin and haemoglobin
were different from those of mammals. Matsuura and Hashimoto (1959) devised
a new method of measuring tuna myoglobin and, using it, found that nearly the
whole of the haemoproteins of ordinary muscle consisted of myoglobin. Brown
(1961) subsequently developed a chromatographic method for the separation of tuna
myoglobin and haemoglobin and in 1962 he demonstrated that the myoglobin
content of the ordinary muscle of tuna usually greatly exceeded the haemoglobin
content. Tuna myoglobin differs from mammalian myoglobin in that the protein
portion (globin) contains the amino acid cysteine (Konosu et aI., 1958; Brown
et aI., 1962), but the two compounds have certain important functional properties
in common (Rossi-Fanelli et aI., 1960; Matsuura and Hashimoto, 1961).
Naughton et ai. (1958) have pointed out that it is probably only the haem
portion of the molecule, which is common to both compounds, that is concerned
in the color changes observed, and consequently, in the present state of our knowl-
edge, it does not appear to be of importance whether haemoglobin or myoglobin
is the precursor of the haemichromes. However, this view may eventually require
revision. In view of the evidence presented above, and for the sake of convenience
and clarity, the haemoproteins of ordinary tuna muscle are referred to below as myo-
globin, although they may have been called haemoglobin in the papers mentioned.
The haemochromes and haemichromes are complexes of haem with nitrogenous
bases or denatured proteins (Falk, 1964). In the haemochromes the iron of haem
is in the reduced or ferrous form, whereas in the haemichromes it is in the oxidized
or ferric form. In the precooking of tuna, conditions are suitable for the produc-
tion of haemochrome(s) and haemichrome(s) by complex formation between
haem and either denatured proteins or nitrogenous bases (e.g. nicotinamide)
(Brown and Tappel, 1957). Since the haemoproteins, haemochromes, and haemi-
chromes exhibit characteristic light absorption spectra (Lemberg and Legge,
1949 ; Falk, 1964), and since the principal known pigments of raw tuna flesh are
haemoproteins (Matsuura and Hashimoto, 1954), Brown et al. (1957, 1958) and
Naughton et al. (1956,1957,1958) used the spectral properties of these compounds
in studying the colors of raw and cooked tuna flesh. As the pigment of cooked
flesh, unlike that of raw flesh, was found to be insoluble, the investigators made
use of light reflectance spectra prepared from raw or cooked meat.
The spectra of several samples of raw tuna flesh (Naughton et aI., 1958) were
those expected for mixtures of metmyoglobin and oxymyoglobin, showing absorp-
tion maxima at 500, 540, 575, and 630 m,u in addition to a very strong maximum
3
between 400 and 440 mj.t characteristic of all haem compounds (the Soret band).
Judging from the relative intensities of the different maxima (those at 500 and
630 mj.t are characteristic of metmyoglobin, those at 540 and 575 mj.t of oxymyo-
globin) the relative proportions of metmyoglobin and myoglobin varied considerably
from sample to sample. Metmyoglobin is derived from oxymyoglobin by deoxygen-
ation of the latter compound and oxidation of the iron of haem from the ferrous
to the ferric state. It was found that oxymyoglobin was converted to metmyoglobin
during frozen storage of the muscle (temperature not given) (Naughton et aI.,
1958). However, Ono and Tawara (1961) found the reflectance spectrum of some
samples of very freshly thawed tuna to be that of myoglobin, with an absorption
maximum at 550 mj.t, and that this spectrum changed to that of oxymyoglobin
during short exposure to air, and more recently Bito (1964) has shown that the
rate of metmyoglobin formation in frozen tuna meat is temperature-dependent,
being somewhat slower at -18 C than at -5 to -14 C and very much slower at
-78 C. The rate of formation was faster in the deeper portion of the flesh than near
the surface.
During cooking, the reflectance spectra of raw flesh altered to those charac-
teristic of haemochrome(s) or haemichrome(s) or mixtures of the two (Brown
and Tappel, 1957; Brown et aI., 1958; Naughton et aI., 1958). Tuna loins of good
color exhibited a spectrum with distinct absorption maxima near 528 mj.t and
558 mj.t characteristic of haemochrome(s), while green loins had a spectrum with
a broad maximum between 530 and 540 mj.t characteristic of haemichrome(s).
Both spectra showed the Soret band absorption typical of haemochrome(s) and
haemichrome(s). It should be noted that Naughton et al. (1958) detected some
haemichrome formation in raw flesh as a result of protein denaturation that presum-
ably occurred during frozen storage. Oxidation of a good cooked tuna loin by
exposure to air resulted in the appearance of the green color and of the haemichrome
reflectance spectrum, while the reduction of such oxidized meat, or other green
tuna meat, by means of a suitable reducing agent (e.g. sodium hydro sulfite spread
thinly on the surface) resulted in the appearance of the pink color of good meat
and of the reflectance spectrum of haemochrome(s). All of this evidence taken
together clearly indicates that oxidation of haemochrome(s) is responsible for the
greening of tuna flesh, but the degree to which the oxidation proceeds is not quite
so well established and at this point the situation becomes complex. Both reversible
and irreversible oxidations of the porphyrin ring structure of the haem portion of
the molecule, in addition to oxidation of the iron, are known, and bring about the
formation of green compounds. Haemichromes containing an unoxidized porphyrin
ring are brown in color. The chemical structure of some of the green compounds is
known, but that of others, particularly of those formed by reversible oxidation,
is obscure (Lemberg and Legge, 1949; Watts, 1954). Brown et al. (1958) reported
finding in the reflectance spectra of some samples of cooked green tuna an absorp-
tion maximum at about 650 mj.t in addition to the others mentioned above, while
Naughton et ai. (1958) similarly reported observing relatively minor absorption
maxima between 610 and 630 mj.t. According to Naughton et al. (1958) the intensity
of the maxima they observed increased with increasing greenness of the cooked
4
-
flesh. These maxima, and that at about 650 mAL observed by Brown et al. (1958),
are located at or very close to wavelengths at which the green oxidation products
mentioned above are known to exhibit absorption maxima (Lemberg and Legge,
1949) . The very strong absorption in the Soret region of the spectrum of green
precooked flesh observed by both groups that of
the haem pigments present can have been oXidized to the pomt of the openmg of
the porphyrin ring. The orange color sometimes met with in precooked tuna (Naugh-
ton et aI., 1957, 1958) might perhaps be the result of such an oxidation, but has not
been extensively investigated.
V ARIABILITY IN THE OCCURRENCE OF GREENING
While the mechanism involved in greening appears to be fairly well under-
stood, the irregularity with which it occurs is not, although rather numerous
investigations have been made.
It has been suggested that the high incidence of greening in winter-caught
al bacore is related to the method of catching - winter fish being taken by longline,
summer fish by pole and line - but Miyauchi (1950) reported that a comparison
of canned fish caught by the two methods showed no distinct difference between
them in the color of their flesh. Naughton et al. (1958) found no detectable differ-
ence in the metmyoglobin content of fish killed in different states of exhaustion.
Aldrin (1962) expressed the opinion that the apparent effect of the method of catch-
ing may in fact be related to a selection in the size of the fish caught, the larger the
fish the greater the susceptibility to greening (but see Amano (1965) below).
Miyauchi (1950) stated that some Japanese believed the occurrence of the
condition to be related to decomposition of the flesh, whereas others believed it
had no relation to the degree of freshness. In this connection, Aldrin (1962) reported
a distinct tendency for off-odours and flavours to accompany greening, but Brown
et al. (1958), although they detected off-odours in some green fish, did not find
any consistent relationship, and Amano (1965) reported that greening can occur in
very fresh fish, even having been observed in an experimental pack of albacore
canned on board immediately after the fish were caught. This latter finding is also
contrary to Aldrin's (1962) observation that greening only occurs in previously
frozen fish.
According to Amano (1965), Japanese investigators have found that the
following variables are not significantly related to greening: the locality of the
fishing ground; the time between catching and freezing of the fish; whether the fish
are alive or dead when boated; the degree of damage to the fish during struggling;
the weight (contrast with Aldrin (1962) above) or sex of the fish; the condition
of the viscera. Amano stated that the fatness of the fish, its liver weight, and loss
in weight during precooking are significantly related to the appearance of off-
color. Naughton et al. (1958) also observed a tendency for greening to be associated
with a greater fat content of the flesh, and Miyauchi (1950) reported a correlation
between fatness, liver weight, and color, but Dollar et al. (1961) and Aldrin (1962)
5
found no reliable correlation between total fat and greening. Hirao et al. (1958a)
found a tendency toward a higher percentage of moisture in green than in good
precooked flesh.
Several other apparently conflicting findings have been recorded. Miyauchi
(1950) described results of Japanese investigators that showed a clear relation
between flesh pH and greening, low pH being associated with good color, but
Dollar et al. (1961) found that, while green meat tended to have a higher pH than
meat of good color, the difference was not significant (cf. Appendix I). Brown
and Tappel (1957) added alkali to tuna flesh before cooking it in order to alter its
pH, but found this did not affect haemochrome formation as judged by the pinkness
of the flesh. Similarly, although Hirao et al. (1958b), Morizono (1960), Ono and
Tawara (1961), and Aldrin (1962) found a close relation between greening in
yellowfin and albacore tuna and the color of an aqueous suspension prepared from
the kidney of the fish, Dollar et al. (1961) and Sasano et al. (1961) did not.
There has been somewhat greater agreement between investigators regarding
one or two aspects of the problem. Naughton et al. (1956), Yamashita et al. (1957),
Yamamota (1960), and Nagaoka and Suzuki (1962, 1964) found a tendency for a
higher concentration of metmyoglobin than of myoglobin in the raw flesh to be
associated with greening. Dollar et al. (1961) made a similar observation, but
reported that green flesh tended to contain less total haem pigment. Naughton
et al. (1958) found tuna metmyoglobin was more readily extracted from raw flesh
than was oxymyoglobin, and suggested that leaching during thawing and precook-
ing might result in a loss of pigment from flesh with a high metmyoglobin content.
Dollar et al. (1961) observed that flesh from fish yielding green meat contained
more lipid peroxide and less water-soluble material, volatile base, and sulfhydryl
than did the flesh offish yielding meat of good color. Naughton et al. (1957,1958),
like Dollar et al. (1961), found a tendency for greening (and other off-colors) to be
associated with elevated lipid peroxide content.
Sasano et aI., (1961), Sasano and Tawara (1962), and Ono and Tawara (1961)
found an apparent association between greening and the occurrence in the raw
flesh of an alcohol-soluble substance which gave a yellow color on treatment with
ninhydrin. This compound was tentatively identified as a peptide. Mixed with
raw meat of good quality, it produced greening when the meat was subsequently
cooked. In some instances appreciable amounts of the substance were found in
good quality meat. Nagaoka and Suzuki (1962, 1964) reported finding a positive
correlation between the trimethylamine content of cooked tuna, the trimethylamine
oxide content of raw tuna, and greening. A few exceptions were observed (also
see Appendix II). Koizumi and Hashimoto (1965a, b) have shown that the alcohol-
soluble substance producing a green color with ninhydrin (Sasano et aI., 1961), and
which can be associated with greening, is trimethylamine oxide (TMAO). Koizumi
and Hashimoto also found that at least part of the color changes brought about in
tuna meat precooked with added TMAO is not reversible by treatment with a reduc-
ing agent. However, when a suitable reducing agent was added in addition to the
TMAO prior to cooking the development of off-color was prevented.
6
A few interesting findings have been recorded which, as far as we are aware,
have not yet been re-examined independently by other investigators. Ono and
Tawara (1961) found a relationship between greening and the ability of flesh
to decolorize a redox indicator in hot water, the weaker the reducing power under
these conditions the greater the tendency of the flesh to become green when cooked.
This ability did not correlate with either the reflectance spectra or the redox po-
tential of the raw flesh. In meats having color grades intermediate between good
and green there was a considerable range in their ability to decolorize the indi-
cator. Brown et aI. (1958) had earlier examined the possibility that there might
be a relationship between the redox potential of raw flesh and its tendency to
become green, but had found none. Finally, Hirao et aI. (1958a) examined the con-
centration of certain B vitamins and minerals in tuna flesh. A tendency toward a
lower concentration of thiamine and a higher concentration of vitamin BI2 was
found in green meat, but exceptions occurred.
The results of some in vitro investigations of the properties of pure tuna
myoglobin appear to be pertinent to the problem of variability in the incidence of
greening. Thus the rate of autoxidation of tuna oxymyoglobin to metmyoglobin,
which may have some relation to greening, has been shown to be influenced by
the partial pressure of oxygen (Matsuura et aI., 1962), temperature (Sano and
Hashimoto, 1958; Matsuura et aI., 1962 ; Brown and Dolev, 1963a, b) , buffer
concentration (Matsuura et aI. , 1962; Brown and Dolev, 1963a), and pH (Sano
and Hashimoto, 1958 ; Matsuura et aI., 1962). The findings of these investigators
may be summarized as follows:
1. The rate increases as the partial pressure of oxygen is decreased to
about 3 mm Hg, near which point it reaches a maximum.
2. The rate decreases with decreasing temperature and reaches a minimum
at about - 5 C. The rate then increases sharply as the solution solidifies
at slightly lower temperatures. This increase is related to the solidifying
of the solution rather than to the temperature change, and the rate
decreases once more as the temperature is lowered below the freezing
point. Repeated freezing and thawing does not greatly affect the rate.
3. There is some lack of agreement on the effect of buffer concentration.
Matsuura et ai. (1962) found an increase in rate with increasing buffer
concentration, while Brown and Dolev (1963a, b) found very little
effect at all.
4. The rate increases with decreasing pH (from pH 6.8 to 5.65) at tempera-
tures between 20 and 40 C.
Such influences could be effective in raw tuna flesh, and there is some evidence
that at least 1 and 2 above are. This evidence has been presented by Naughton
et aI. (1958), who obtained spectral data indicating rapid formation of metmyoglobin
in tuna flesh during frozen storage and also in tuna flesh SUbjected to lowered
partial pressure of oxygen, and by Tanaka (1961) and Bito (1964) who found a
7
reduction in the rate of formation of metmyoglobin in tuna flesh as the tempera-
ture of frozen storage was reduced. There might be some relation between 3 above
and the effect of freezing. The available data regarding pH of the flesh, described
earlier, does not permit the drawing of any definite conclusion regarding the effect
of 4 above, but does appear to be at variance with expectations based on 4 above.
The information in the preceding paragraph refers to the oxidation of haem
compounds in the raw flesh. In cooked flesh the haem compounds are present as
haemochromes and it is known from earlier work with mammalian flesh (see
review by Watts, 1954) that conditions favouring their oxidation are different
from those favouring oxidation of haem compounds in raw flesh, although the
oxidation products of haem are the same in both raw and cooked flesh. Thus
when the globin portion of myoglobin is denatured in cooking, the haem portion
loses its ability to combine with oxygen to form oxymyoglobin and consequently
becomes very susceptible to oxidation to the ferric form (haemichrome), the rate
increasing with increasing oxygen pressure. It is interesting that in cured meats
(Watts, 1954) certain influences, such as increased acidity and freezing, which
accelerate metmyoglobin formation in fresh meat, have either no effect on pig-
ments or actually improve color fixation. Akoyunoglou et al. (1963) have recently
shown that the rate of formation of ferrihaemochromes and ferrohaemochromes
from haeme and haematin is strongly influenced by pH.
Koizumi (1961) conducted an interesting in vitro experiment in which he
demonstrated that pyridenehaemichrome could be destroyed by peroxidized
skipjack head oil. Younathan and Watts (1959) investigated the relation of haem
pigments to lipid oxidation and found evidence that the ferric form is the active
catalyst in the rancidification of oil. Contact of haem pigments with unsaturated
fat in the presence of oxygen results in rancidity and discoloration. These findings
support the view that the apparent relation between greening and lipid peroxide
content mentioned above is probably significant. Again, Watts (1954) has pointed
out that the conditions for such a reaction are not necessarily the same for fresh
and cured meat. She cites the weaker effect of free hydrogen peroxide in fresh
meat, in comparison with that in cured meat, as an example. Tarladgis (1961)
has presented a detailed discussion of the mechanism of haem catalyzed lipid
oxidation in animal tissues.
While there is still a good deal of uncertainty regarding the number and identity
of the factors that can influence the occurrence of greening, some progress has
been made in clarifying a situation that appears to be very complex. Of the possible
factors that have been investigated, those for whose actual participation there is
the most supporting experimental evidence are the ones that would tend to influence
the oxidation of the haemoproteins and haemochromes. However, as Brown
et al. observed in 1958, it is not yet established whether the absence of the proper
reducing conditions in flesh yielding a green cooked product is a characteristic
property of the particular fish , of its handling after capture, or a combination
of both.
8
In conclusion, the number of factors that can apparently influence the occur-
rence of greening is such that it seems to be reasonable to suppose that anyone
of them might be dominant in different fish, or in different groups of fish, at dif-
ferent times. Consequently it appears probable that contradictions between the
findings of different investigators, and even within those of individual investigators,
occur because of the complexity of the system involved.
DETECTING SUSCEPTIBILITY TO GREENING IN RAW FLESH
As pointed out earlier, greening takes place during the precooking of tuna,
and, although it has been said fish yielding green and good-colored meat on cook-
ing can be distinguished with 85% certainty by the appearance of their raw flesh
alone (e.g. Matzuzaka and Takahashi quoted by Naughton et al. , 1956), it appears
from an overall examination of the literature that this is not generally true. The
result has been that various tests have been proposed for use in predicting the
color of the meat to be expected when a given fish is cooked. Although several
of these tests apparently provide reasonably accurate predictions, none seems to be
completely reliable, and all appear to be too time-consuming to apply in a large-
scale canning operation. However, it has been suggested that certain of them
could be useful in assessing the quality of a large group of fish by applying them
in the examination of a representative sample (e.g. Nagaoka and Suzuki, 1964).
The tests that have been proposed are summarized below. It will be obvious
that they are based on certain correlations between greening and particular proper-
ties of raw flesh that have been discussed above.
M ETMYOGLOBIN
In this test, the percentage of the total haem pigments accounted for by met-
myoglobin is determined, and in Japan a simplified instrument has been developed
for use in making the measurements required (Yamashita et al., 1957 ; Nagaoka and
Suzuki , 1964), but a description of the instrument has not been available to the
writer. In any event, Nagaoka and Suzuki (1964) state that experiments with the
instrument on a factory scale were not very successful, and that efforts should
be made to improve it. The instrument presumably makes use of reflectance meas-
urements. Sano et al. (1958, 1959) have described an extraction procedure for
use in making the required determinations by absorption measurements, but
this procedure would be too time-consuming for routine use. However, as pointed
out earlier and in Appendix I, the correlation between percentage metmyoglobin
in the raw flesh and greening is by no means absolute, so that this test is at best
of li mited usefulness. Nagaoka and Suzuki (1964) have proposed that it be used
in a preliminary selection of samples to be subsequently examined by the trime-
thylamine oxide test below. Even applied in this way, all fish capable of yielding
green meat will not be detected and some good fish may be rejected (Nagaoka
and Suzuki, 1964).
9
CONCENTRATION OF AN ALCOHOL-SOLUBLE SUBSTANCE PRODUCING
A YELLOW COLOR WITH NINHYDRIN
Sasano et aI. (1961) have proposed a procedure in which an alcoholic extract
of raw tuna flesh is prepared, chromatographed on paper, and the chromatogram
then sprayed with ninhydrin. The intensity of a particular yellow spot on the
sprayed chromatogram may be utilized in predicting the color of the meat to be
expected when the fish from which the extract was prepared is cooked. While
the reliability of this test seems to be quite good, it is not absolute, and would be
too time-consuming for use on a large scale.
COLOR OF AN AQUEOUS EXTRACT OF RAW KIDNEY
Hirao et aI. (1958b) first proposed this test. A 1-g sample of kidney is homo-
genized in 50 ml of water and the color of the resulting suspension is compared
with a set of standards in order to predict the color to be expected when the fish
from which the kidney sample was taken is cooked. This is a relatively simple
test, but while some workers have found it to be useful and reliable (Hirao et aI.,
1958b; Morizono, 1960; Ono and Tawara, 1961; Aldrin, 1962) others have not
(Dollar et aI., 1961; Sasano et aI., 1961). The latter investigators indicated the test
was not reliable for albacore but was fairly reliable with large, but not with small
yellowfin. Morizono (1960) found that the amount of blood in the kidney could
influence the results, and that consequently damage to the tuna by sharks would
have to be taken into account. Morizono noted that the color of the kidney of tuna
yielding green meat was that of "wet-biscuit" and proposed that visual inspection
of the kidney color would provide a practical means of examining a large number
of fish.
TRIMETHYLAMINE OXIDE (TMAO) CONCENTRATION
Nagaoka and Suzuki (1962, 1964) have proposed that the TMAO content
of the raw flesh be measured and used in predicting the color to be expected when
the flesh is cooked. The method used in making the measurement is a slight modi-
fication of that of Dyer et al. (1952) . Again, while the method has given evidence
of being quite reliable, exceptions do occur, and it would be difficult to apply it to
a large number of samples, as Nagaoka and Suzuki have pointed out. They suggest
that the number to be examined be limited by making a preliminary selection
using the metmyoglobin method above, only those fish with a high metmyoglobin
percentage being selected for further testing.
REDOX INDICATOR DECOLORIZING POWER
Ono and Tawara (1961) proposed the use of Lauth's violet (thionin) as an
indicator of the tendency of raw flesh to subsequent greening. In conducting the
test, a 5-g sample of flesh was suspended in a test tube in 20 ml of 0.2 M acetate
buffer solution of pH 5.65 containing Lauh's violet at a concentration of 2 mg/ lOO
mI. The test tube was loosely plugged with a glass stopper and heated, with occa-
10
sional stirring, for 30- 45 min after the flesh became coagulated and the degree
of decolorization was then estimated. Meat unable to decolorize the indicator was
classified as destined to become green or at least to be of low grade, while meat
that strongly decolorized the indicator was usually of good grade. Exceptions
occur, and there was a good deal of intermingling of values in flesh of intermediate
grades.
COLOR OF PRECOOKED SMALL FLESH SAMPLES
Sasano et al., (1961) proposed that precooking of small samples presented
the most practicable means of predicting the tendency of a fish to become green.
It is possible that variability within the flesh of individual fish might present a
problem in this test, although Sasano et al. indicate it is quite reliable. They pointed
out that too short a time of cooking could result in failure of the appearance of
greening.
MEANS OF PREVENTING GREENING
Brown et al. (1957, 1958) and Yamashita (1961) appear to be the only investi-
gators who have previously considered means of treating tuna during canning in
order to prevent greening. Brown et al. made use of a model system with haemin
as the haematin compound in order to test a number of nitrogenous compounds,
in conj unction with various reducing agents, for their ability to form haemochromes.
It was found that nicotinamide was superior to others in terms of haemochrome-
formi ng ability. Nitrite also brought about the formation of pink pigment, but
the absorption spectrum of the solution was not typical of a haemochrome. Sodium
hydro sulfite was the most effective reducing agent examined, but sodium ascorbate
was also satisfactory. On the basis of these findings, and the results of some explo-
ratory experiments with small tuna samples, Brown et al. (1957) concluded that
the addition of 0.05% nicotinamide or 0.05% sodium ascorbate or, better, of a
combination of the two would almost certainly improve color in commercial
packs. They suggested that the testing of even smaller additions, of the order of
0.01%, under commercial conditions would be warranted. Yamashita (1961)
reported that the addition of sodium hydrosulfite together with sodium nitrite
or nicotinamide during canning improved the appearance of "blue" tuna, and
Tomli nson and Geiger (see Appendix I) have found that the addition of ascorbic
acid with or without sodium nitrite or nicotinamide during canning greatly im-
proved the color of green albacore.
ACKNOWLEDGMENTS
It is a pleasure to acknowledge the generous help of Dr Tomoo Nakano,
Holy Ghost Junior College, Akita City, Japan, National Research Council of
Canada Post Doctoral Fellow, 1964- 65, who translated a number of the papers
referred to in the review. I wish to thank Dr S. Hirao of the Fisheries Agency,
Ministry of Agriculture and Forestry, Tokyo, for permission to reproduce the
color photograph contained in this Bulletin.
11
REFERENCES
AKOYUNOGLOU, J . A., H. S. OLCOTT, AND W. D. BROWN. 1963. Ferrihemochrome and fer-
rohemochrome formation with amino acids, amino acid esters, pyridine derivatives, and
related compounds. Biochemistry, 2: 1033- 1041.
ALDRIN, J. F. 1962. Considerations pratiques sur Ie verdissement du thon tropical. Revue de
la conserve, Sept.- Oct., p. 147- 151.
AMANO, K. 1965. Technological problems in the Japanese fish-freezing industry, p. 91- 93.
In The technology of fish utilization. R. Kreuzer [ed. ] Fishing News !Books) Ltd., London.
BITo, M. 1964. Studies on the retention of meat color of frozen tuna. 1. Absorption spectra
of the aqueous extract of frozen tuna meat undergoing discoloration. Bull. Japan. Soc. Sci.
Fish., 30: 847- 857.
BROWN, W. D. 1961. Chromatography of myoglobin on diethylaminoethyl cellulose columns.
J. BioI. Chem., 236 : 2238- 2240.
1962. The concentration of myoglobin and hemoglobin in tuna flesh. J. Food Sci. ,
27: 26- 28.
BROWN, W. D., AND A. DOLEV. 1963a. Autoxidation of beef and tuna oxymyoglobins. Ibid.,
28 : 207- 210.
1963b. Effect of freezing on autoxidation of oxymyoglobin solutions. Ibid. , 28: 211- 213.
BROWN, W. D., M. MARTINEZ, M. JOHNSTONE, AND H. S. OLCOTT. 1962. Comparative bio-
chemistry of myoglobins. J. BioI. Chem., 237 : 81- 84.
BROWN, W. D. , AND A. L. TAP PEL. 1957. Identification of the pink pigment of canned tuna.
Food Res., 22 : 214- 221.
BROWN, W. D. , A. L. TAPPEL, AND H. S. OLCOTT. 1958. The pigments of off-color cooked
tuna meat. Ibid., 23: 262- 268.
DOLLAR, A. M., A. M. GOLDNER, W. D. BROWN, AND H. S. OLCOTT. 1961. Observations on
"green" tuna. Food Technol. , 15(5) : 253- 255.
DYER, W. J. , F. E. DYER, AND J. M. SNOW. 1952. Amines in fish muscle. V. Trimethylamine
oxide estimation. J. Fish Res. Bd. Canada, 8: 309- 313.
FALK, J . E. 1964. Porphyrins and metallophorphyrins. Elsevier Publishing Co., Amsterdam,
London, New York. 266 p.
HIRAO, S., S. MURAYAMA, M. YANASE, J. YAMADA, R. KIKUCHI, K. TABEI , K. KISAKA, AND
Y. ENOMOTO. 1958a. Study on the green meat of tuna. I. Quantitative differences of
vitamin B group and minerals between the green and the normal meat of precooked tuna.
Bull. Japan. Soc. Sci. Fish., 24: 671- 675.
1958b. Study on the green meat of tuna. III. A simple method to distinguish greened
body from normal one before cooking. Ibid., 24: 679- 681.
KOIZUMI, C. 1961. Studies on the shirata in katsuobushi. II. Spectrophotometric studies into
the discoloration of katsuobushi under storage. Ibid., 27: 166- 170.
KOlZUMI , C., AND Y. HASHIMOTO. 1965a. Studies on "green" tuna-1. The significance of
trimethylamine oxide. Ibid., 31: 157- 160.
1965b. Studies on "green" tuna-2. Discoloration of cooked tuna meat due to trimethyla-
mine oxide. Ibid. , 31: 439- 447.
KONOSU, S., K. HASHIMOTO, AND F. MATSUURA. 1958. Chemical studies on the red muscle
(chiai) of fish. XI. Amino acid composition of tuna myoglobin. Ibid., 24: 563- 566.
LEMBERG, R., AND J. W. LEGGE. 1949. Hematin compounds and bile pigments. Interscience
Publishers, Inc. , New York and London. 749 p.
MATSUURA, F., AND K. HASHIMOTO. 1954. Chemical studies on the red muscles (chiai) of
fishes. II. Determination of the content of hemoglobin, myoglobin and cytochrome C in
the muscles of fishes . Bull. Japan. Soc. Sci. Fish., 20: 308- 312.
12
1955. Chemical studies on the red muscle (chiai) of fishes. IV. Preparation of crystalline
myoglobin from the red muscle of fishes. Ibid., 20: 946- 950.
1956. Chemical studies on the red muscle (chiai) of fishes. VI. Solubilities of red muscle
myoglobins and blood hemoglobins. Ibid., 22: 408- 412.
1959. Chemical studies on the red muscle (chiai) of fishes. X. A new method for determin-
ation of myoglobin. Ibid., 24: 809- 815.
1961. Chemical studies on the red muscle (chiai) of fishes. XIII. Oxygen dissociation
curve of tuna myoglobin. Ibid., 27: 162- 165.
MATSUURA, F. , K. HASHIMOTO, S. KIKAWADA, AND Y. YAMAGUCHI. 1962. Studies on the
autoxidation velocity of fish myoglobin. Ibid., 28: 210- 216.
MIYAUCHI, D. T. 1950. Some processing and technological methods in the Japanese fisheries.
Comm. Fish. Rev., 12(10) : 1- 20.
MORIZONO, T. 1960. Detecting green meat of tuna on board. Tuna Fishing, No. 65, p. 32- 36.
NAGAOKA, C., AND N. SUZUKI. 1962. A method for predicting green meat of tuna. Refrigera-
tion, Tokyo, 37: 338-346.
1964. Detection of green-meat tuna before cooking. Food Technol., 18: 183- 187.
NAUGHTON, J. J., M. M. FRODYMA, AND H. ZEITLIN. 1956. Nature of green or off-color condi-
tion in precooked yellowfin tuna. Special Sci. Rep., Fisheries No. 197. U.S. Dept. of the
Interior, Fish and Wildlife Service, Washington, D.C. 7 p.
1957. Further studies on green or off-color condition in precooked yellowfin tuna. Ibid.,
No. 247, 13 p.
NAUGHTON, J. J., H. ZEITLIN, AND M. M. FRODYMA. 1958. Spectral reflectance studies of the
heme pigments in tuna fish flesh. Some characteristics of the pigments and discoloration of
tuna meat. J. Agr. Food Chem., 6: 933- 938.
ONO, H., AND T. TAWARA. 1961. Studies on the green meat of albacore and yellowfin tuna.
II. Redox indicator decolorizing power of meat as a new measure for predicting "green
meat." Bull. Japan. Soc. Sci. Fish., 27: 1066- 1074.
ROSSI-FANELLI, A., AND E. ANTONINI. 1955. Purification and crystallization of the myoglobin
of some salt water fish. Arch. Biochem. Biophys., 58: 489- 500.
ROSSI-FANELLI, A., E. ANTONINI, AND R. GIUFFRE. 1960. Oxygen equilibrium of myoglobin
from Thul1l1uS thYl1l1us. Nature, 186: 896- 897.
SANO, Y., AND K. HASHIMOTO. 1958. Studies on the discolouration in fish meat during freezing
storage. I. A spectrophotometric method for the simultaneous determination of ferrous
and ferric forms of myoglobin in their mixed solution. Bull. Japan. Soc. Sci. Fish., 24:
519-523.
SANO, Y., K. HASHIMOTO, AND F. MATSUURA. 1959. Studies on the discolouration in fish
meat during freezing storage. II. A spectrophotometric method for the simultaneous deter-
mination of ferrous and ferric forms of myoglobin in tuna meat. Ibid., 25: 285- 289.
SASANO, Y. , H. ONO, T. TAWARA, AND K. HIGASHI. 1961. Studies on the green meat of albacore
and yel\owfin tuna. Ibid., 27 : 586- 592.
SASANO, Y., AND T. TAWARA. 1962. Studies on the green meat of albacore and ye\lowfin tuna.
III. Relationship between a substance producing yellow color with ninhydrin and the green
meat of tuna. Ibid., 28: 722- 725.
TANAKA, K. 1961. Fundamental studies on the prevention of discolouration of frozen yellow-fin
tuna meat. Refrigeration, Tokyo, 36: 962- 978.
TARLADGIS, B. G. 1961. An hypothesis for the mechanism of the heme catalyzed oxidation
in animal tissues. J. Amer. Oil Chern. Soc. , 38: 479- 483.
WATTS, B. M. 1954. Oxidative rancidity and discoloration in meat. Adv. Food Res., 5: 1- 52.
YAMAMOTO, T. 1960. Problems of colour condition of frozen tuna meat in cooking. Refrigera-
tion, Tokyo, 35: 137- 144.
YAMASHITA, R. 1961. Blue meat of tuna. Ibid., 36: 1018- 1019.
YAMASHITA, R., N. TAKAHASHI, AND Y. MATSUZAKI. 1957. Studies on green meat of tunas.
Ibid., 32: 358.
YOUNATHAN, M. T., AND B. M. WATTS. 1959. Relationship of meat pigments to lipid oxidation.
Food Res. , 24: 728- 734.
13
APPENDIX I
THE HAEM PIGMENTS OF NORMAL AND OF GREEN
ALBACORE TUNA AND THE INFLUENCE OF CERTAIN ADDITIVES ON COLOR
By N. Tomlinson and S. E. Geiger
Several samples of frozen tuna loins which, when precooked, varied in color
from good to poor were obtained from a local processor. The appearance of pre-
cooked portions of poor color was that usually associated with greening. That
this was indeed the cause of the off-color was shown by a change in the color from
greenish-tan to pink when the precooked meat was treated with sodium hydro-
sulfite (Brown et aI. , 1958) .
The pigments of the raw flesh were examined using the extraction and measure-
ment procedure of Sano et aI. (1959). The absorption spectra of the extracts indi-
cated that the color of the solution was due to haem pigments, but that there was
a large variation in the myoglobin content of muscle samples taken from the same
loin, extreme ranges of the order of 20- 100 mg/ IOO ml being observed. The percen-
tage of myoglobin present in the extracts as metmyoglobin was also variable. For
example, in two samples rejected for poor color it was 31 % and 82%, in a good
sample, 83%, and in a sample whose color was considered borderline, 22%. This
seems to throw further doubt on the possibility of distinguishing good from poor
individual samples on the basis of the percentage of the haem pigments present as
metmyoglobin. The pH values of raw muscle samples from these fish were 5.98 and
6.18 for the rejected fish; 6.64 for the good one; and 6.24 for the borderline. This
order is the reverse of that reported from Japan by Miyauchi and is in accordance
with that to be expected if oxidation of myoglobin is of importance in greening,
but is apparently not in agreement with the actual measurements of percentage
metmyoglobin given above. The content of myoglobin in these extracts, which
were prepared from samples taken from the same part of the lateral-dorsal muscle,
was 65, 76, 23, and 23 mg/ IOO ml, respectively, and suggests that possibly the total
amount of myoglobin, in conjunction with the pH of the muscle, might have been
of more importance in this instance during the precook than the percentage metmyo-
globin in the raw flesh. Far too few samples have been examined to draw any
definite conclusions in this regard.
In view of the findings and recommendation of Brown and Tappe! (1957), the
effect of certain additives on the color of canned tuna was examined. Some repre-
sentative results are tabulated in Appendix Table I.I. All the additives provided some
improvement in appearance. The combinations of ascorbic acid with sodium nitrite
or nicotinamide yielded a more intense pink color than did ascorbic acid alone, while
ascorbic acid yielded a product with better appearance than did either nitrite or
15
nicotinamide alone. There is evidence that nitrite plus ascorbic provided a product
whose color was more stable on exposure to air than did the other additions. Sodium
nitrite added alone in this experiment was not particularly effective, but in some
experiments with very green fish it has given remarkably good results. There is no
obvious explanation for this variation, but it might be related to the pH of the
flesh. The appearance of good flesh was not affected by the treatments investigated.
The use of nitrite and ascorbic acid in combination is a procedure used in
commercial curing of meat products and has consequently been extensively inves-
tigated. The development of a satisfactory color requires the reduction of the haem
pigment to the ferrous form and the formation of nitric oxide by reduction of nitrite.
Ascorbic acid is commonly used as the reducing agent, but there is evidence that
mammalian muscle contains enzymes capable of effecting both reductions (Walters
and Taylor, 1963, 1964). Tarladgis (1962) considers that at least 80% of the pigment
of heat-cured meat is present as the soluble nitric oxide myochrome, a pigment
free from protein and in which both co-ordination positions of the iron of haem
are occupied by nitric oxide. The remainder of the pigment appears to be metmyo-
ApPENDIX TABLE 1.1. The influence of certain additives on the color of canned albacore tuna."
Color of meat Color of meat after
Treatment immediately after being stored 24 hr
no. Addition Concentration
b
opening the can at 32 F in the open can
ppm
l. None Mostly tan, but Dark tan externally,
pink inside some pink inside
2. Sodium 3 Some improvement Like (1)
nitrite over (1) but stilI
some tan coloration
3. Ascorbic 100 Light cream to Possibly a little
acid light pink lighter tan than (1)
externally, some
pink inside
4. Nicotinamide 500 Light , quite pink Like (1)
with some tan evident
5. Sodium nitrite
{
3 Light cream to pink Tan with some pink
plus ascorbic Pink much more externally, pink
acid 100 intense than in (3) inside
6. Nicotinamide
{
500 Like (5) Like (3)
plus ascorbic
acid 100
BThe fish used in this experiment was on the borderline of rejection for color. After precook,
the flesh was tan in color with some pink areas internally. Precook : 2 hr at 216 F. Cooling: 3
days at 32 F in air. All treatments done in duplicate. Canning: vacuum-packed, then processed
for It hr at 240 F.
bExpressed as parts per million calculated on the weight of the total contents of the can
(155 g meat with 30 ml water and 15 ml oil added). Additives dissolved in water and added to
the cans just before closing.
16
chromogen. Certain factors that hasten oxidation of oxymyoglobin to metmyoglobin
in raw flesh (e.g. freezing, salt, acid) accelerate the formation of nitric oxide myo-
globin from metmyoglobin (Watts, 1954).
The effect of light is of importance in color changes in both raw and cured
meat. Exposure of raw meat to fluorescent light increases the rate of oxidation of
myoglobin to metmyoglobin (Lane and Bratz1er, 1962), while exposure of cured
meat to light brings about the dissociation of nitric oxide from the ferrous-porphyrin
co-ordination complex and leaves the ferrous ion very susceptible to oxidation
(Tarladgis, 1962).
In Canada, consideration is being given to altering food and drug regulations
to permit the addition of ascorbic acid, but not of nitrite, in the canning of tuna.
REFERENCES FOR APPENDIX I
BROWN, W. D., AND A. L. TAPPEL. 1957. Identification of the pink pigment of canned tuna.
Food Res., 22: 214- 221.
BROWN, W. D., A. L. TAPPEL, AND H. S. OLCOTT. 1958. The pigments of off-color cooked
tuna meat. Ibid., 23: 262- 268.
LANE, J. P., AND L. J. BRATZLER. 1962. Spectrophotometric estimation of metmyoglobin in frozen
meat extracts. J. Food Sci., 27: 343- 346.
SANa, Y., K. HASHIMOTO, AND F. MATSUURA. 1959. Studies on the discoloration in fish
meat during freezing storage. II. A spectrophotometric method for the simultaneous deter-
mination of ferrous and ferric forms of myoglobin in tuna meat. Bull. Japan. Soc. Sci.
Fish., 25: 285- 289.
TARLADGIS, B. G. 1962. Interpretation of the spectra of meat pigments. II. Cured meats.
The mechanism of colour fading. J. Sci. Food Agr., 13: 485- 491.
WALTERS, C. L., AND A. McM. TAYLOR. 1963. Biochemical properties of pork muscle 10
relation to curing. Food Technol. , 17: 354- 359.
1964. Nitrite metabolism by muscle in vitro. Biochem. Biophys. Acta, 86: 448-458.
WATTS, B. M. 1954. Oxidative rancidity and discoloration in meat. Adv. Food Res., 5: 1- 52.
17
APPENDIX II
GREENING IN ALBACORE TUNA IN R ELATION TO THE
TRIMETHYLAMINE OXIDE (TMAO) AND VOLATILE CARBONYLS CONTENT
By E. Bilinski and R. E. E. Jonas
In view of the reported relation between the TMAO content of raw muscle,
the trimethylamine (TMA) content of precooked muscle, and greening (Nagaoka
and Suzuki, 1964) , several samples of raw tuna, some known to yield meat of good
color, others green, were examined for their content of TMA and TMAO. It is
known that haemoglobin can catalyze the reduction of TMAO (Vaisey, 1956).
EXPERIMENTAL PROCEDURE
One 25-g sample of dorsal muscle, free from dark muscle and skin, was taken
from each fish. Trimethylamine oxide (TMAO) was reduced to trimethylamine
(TMA) by the procedure of Dyer et al. (1952). Determinations were made in triplicate
by the colorimetric method of Dyer (1959). The results are expressed as mg N / 100 g
muscle (wet basis).
RESULTS
Representative data are presented in Appendix Table 11.1. Clearly a high
concentration of TMA and TMAO in flesh is not necessary for the development of
serious greening. On the other hand, the results for the fish yielding precooked
meat of borderline color suggest that an increased concentration may be a factor
capable of influencing development of greening. The flesh of this particular fish,
unlike the two rejected for poor color, had a relatively low concentration of myo-
globin, as did the good sample, but in the latter the TMA- TMAO concentration
was relatively low. However, differences in the pH of the flesh could also be a
factor in this instance (see Appendix I for myoglobin and pH data on the same fish) .
The possibility that volatile carbonyls might be exerting an influence on off-
color development (Mendelsohn and Steinberg, 1962) was also investigated.
EXPERIMENTAL PROCEDURE
Two 35-g samples of the dorsal muscle, free from dark muscle and skin, were
taken from each fish. Protein was removed from the aqueous extract of the muscle,
obtained after homogenization and centrifugation, by following the Ba(OH)z-ZnS0
4
method (Somogyi, 1945). Volatile carbonyl compounds were collected by steam
distillation (Ota, 1958) and quantitatively measured by the color formed with
2,4-dinitrophenylhydrazine, using a modification of the method of Mayes and
Robson (1957). Determinations were made in triplicate and the results are expressed
as equivalents of acetaldehyde per gram of muscle (wet basis).
18
ApPENDIX TABLE Il.L Trimethylamine and trimethylamine oxide in albacore tuna flesh in
relation to greening.
Fish
no.
2
3
4
Color of
precooked
flesh
Good
Very green,
rejected
Very green,
rejected
Green,
borderline
TMA+TMAO
2. 90
2. 93
3.05
Mean: 2.96
2.40
2.50
2. 41
Mean: 2.44
1. 87
1. 91
1.93
Mean: 1.90
8. 78
8. 50
Mean: 8. 64
TMA
mg N / 100 g muscle
0.16
0. 18
0.17
0. 17
0.13
0.15
0 . 13
0.14
0. 24
0. 24
0.20
0.23
1. 30
1. 34
1. 32
1. 32
ApPENDIX TABLE II.II. Volatile carbonyls in albacore tuna flesh In
relation to greening.
Fish
no.
2
3
4
Color of
precooked
flesh
Good
Very green,
rejected
Very green,
rejected
Green,
borderline
Sample
no.
1
2
Average:
1
2
Average:
1
2
Average:
1
2
Average:
19
Acetaldeyde equivalents
(}J.g/ g muscle)
Mean values
15.8
15.4
15.6
16. 7
21. 3
19.0
12. 9
15. 4
14 . 1
11.8
12 . 3
12.0
TMAO
2.79
2.30
1.67
7. 32
RESULTS
Representative data are presented in Appendix Table 11.11. There was no
apparent relation between greening and the concentration of volatile carbonyls in
the raw flesh.
REFERENCES FOR APPENDIX II
DYER, W. J. 1959. Report on trimethylamine in fish. J .A.O.A.c., 42: 292- 294.
DYER, W. J., F. E. DYER, AND J. M. SNOW. 1952. Amines in fish muscle. V. Trimethylamine
oxide estimation. J. Fish. Res. Bd. Canada, 8: 309- 313.
MAYES, P. A., AND W. ROBSON. 1957. The determination of ketone bodies. Biochem. J.,
67: 11- 15.
MENDELSOHN, J. M., AND M. A. STEINBERG. 1962. Development of volatile carbonyls in
haddock (Melanogrammus aeglefinus) flesh during storage at 2C. Food Techno!., 16(5) :
113- 115.
NAGAOKA, C., AND N. SUZUKI. 1964. Detection of green-meat tuna before cooking. Ibid ..
18(5) : 183- 187.
OTA, F. 1958. Carbonyl compounds in fish as related to the deterioration. 1. Detection of
volatile compounds formed in fish flesh. Bull . Japan. Soc. Sci. Fish., 24; 334- 337.
SOMOGYI, M. 1945. Determination of blood sugar. J. Bio!. Chern., 160: 69- 73.
VA ISEY, E. B. 1956. The non-enzymic reduction of trimethylamine oxide to trimethylamine,
dimethylamine, and formaldehyde. Canadian J. Biochem. Physiol., 34: 1085- 1090.
20
APPENDIX III
RELATIONSHIP BETWEEN THE REDUCING
SUGARS OF ALBACORE TUNA FLESH AND GREENING
By H. L. A. Tarr and A. G. Comer
The possibility that the reducing sugars ribose and glucose might be of signifi-
cance in the production of off-colors in tuna (albacore and bluefin) and in skipjack
as a consequence of nonenzymic browning (Maillard reaction) has been considered
previously (Tarr, 1954). It was found that the concentration of ribose in tuna
flesh was too low to be of any importance in nonenzymic browning, but in skipjack
the concentration was sufficient (190-200 f.l.g/ g muscle) to be of significance. Glucose
concentration in each of the three species was considerably higher than that of
ribose, but was still such that it was unlikely to occasion serious Maillard reactions.
Tomlinson et al. (1962) also examined albacore and found much the same con-
centrations of the two sugars. Dollar et al. (1961) measured the orcinol reducing
substances (pentose sugars) in both green and good samples of yellowfin tuna, but
found no relation with the color of the precooked flesh. Since the reducing sugars
of green albacore had not previously been examined, it appeared to be worthwhile
to do so. The method used was similar to that previously described (Tarr and
Leroux, 1962). The results are recorded in Appendix Table lIU. None of the
samples contained sufficient ribose or glucose to occasion appreciable nonenzymic
browning.
ApPENDIX TABLE IILI. Ribose and glucose in relation to
greening in albacore tuna.
Color of
Fish precooked Ribose Glucose
no. flesh (J.Lg/ g muscle) (J.Lg/ g muscle)
1 Good 28 23
2 Very green, 44 77
rejected
3 Very green, 48 314
rejected
4 Green , 47 74
borderline
REFERENCES FOR APPENDIX III
DOLLAR, A. M., A. M. GOLDNER, W. D. BROWN, AND H. S. OLCOTT. 1961. Observations on
"green" tuna. Food Techno!., 15(5) : 253- 255.
TARR, H. L. A. 1954. The Maillard reaction in flesh foods. Food Techno!., 8(1): 15-19.
TARR, H. L. A., AND M. LEROUX. 1962. Acid-soluble phosphorus compounds and free sugars
in fish muscle and their origin. Canadian J. Biochem. Physio!. , 40: 571- 589.
TOMLINSON, N. , S. E. GEIGER, AND E. ROBERTS. 1962. Frozen albacore tuna. The influence
of storage conditions prior to freezing. Fish. Res. Bd. Canada, Prog. Rep. Pac. Coast Stns.,
No. 114, p. 19- 21.
21