Non-sensory Assessment of Fish Quality
(Torry Research Station-Torry Advisor note No92)
What do the tests measure?
The eating quality of fish becomes less pleasant as a result of various spoilage and
deteriorative changes. Unfrozen fish become unacceptable to the consumer
principally because of the activities of bacteria; some of the chemical tests used to
estimate quality actually measure the products of bacterial growth. In the early stages
of spoilage, when the numbers of bacteria are still low, certain enzymes, or biological
catalysts, that have essential functions in the living fish remain active for some time
after death, and their effects can be measured and used as indicators of freshness.
Oily fish like herring and mackerel can become rancid; this is due to the reaction of
the oil with oxygen in the air to create unpleasant odours and flavours. Chemical tests
can measure the extent of such oxidation. Oxidation occurs quite slowly in iced fish,
and the products of bacterial growth can render fish inedible before oxidation
contributes much to the off odours and flavours.
In frozen fish, bacterial action is reduced to negligible levels but oxidation of the oil,
especially in oily fish, will continue during storage and will lead to a loss of eating
quality. Some of the proteins in fish undergo changes, not fully understood, during
long periods of frozen storage which lead to undesirable toughening of the flesh.
Do non-sensory tests have advantages?
Since sensory methods of assessment apply the same senses as the consumer uses
when deciding whether a piece of fish is pleasant to eat, they are likely to predict the
consumer’s reaction better than non-sensory methods; the latter do, however, have
certain advantages. Since sensory assessment may require the use of a taste panel,
non-sensory methods can be cheaper and often quicker. Taste panels need to be
trained and kept in training, which can be time-consuming. Non-sensory assessments
should give the same result no matter where they are carried out, whereas sensory
evaluations may depend on subjective responses of the panellists to the fish being
examined. Non-sensory methods can appear more objective and reliable than
sensory methods, although this need not be the case. When specifications are being
prepared it is easier to insert numerical limits based on non-sensory tests than on
sensory tests, especially in international trade, and this is widely done. Courts of law
may find it easier to accept the results of chemical or physical tests, being based on
impartial instrumental readings, than the results of sensory tests.
Non-sensory tests have disadvantages, too. They measure usually only one aspect of
spoilage or may even assess some change in the fish not directly related to spoilage:
sensory methods can take many aspects of quality into account in arriving at a single
value. Again, although non-sensory test results should be independent of the method
of measurement used, this is not always the case and can lead to disputes. Chemical
methods and some physical methods need laboratory facilities and trained staff and
are necessarily destructive, i.e. the fish once examined cannot then be sold.
What methods are there?
Most of the remainder of this Note describes the principles of a number of
non-sensory quality tests that are in common use. Many other tests have been
suggested over the years but few have found practical application: some that have
are briefly listed later.
It would unduly lengthen this Note to give full practical details for carrying out the tests
described; in some cases a choice of methods is available and to describe only one
method might give an unbalanced view. Further details of the methods employed at
Torry Research Station are available in publications in the scientific and technical
The tests are described separately for chilled, unfrozen fish and for frozen fish, but
these are not strict divisions. It is possible, for example, to apply some tests for chilled
fish to frozen products; the results will then indicate broadly the state of spoilage of
the fish before freezing, while additional tests may be applied to estimate the degree
of deterioration during frozen storage.
Methods for chilled fish
A substance adenosine triphosphate, ATP, is important in the utilisation of energy in
most living things. When fish die the ATP is broken down over a period of days by
enzymes present in the flesh, through a succession of different substances. The final
stage of this process is the formation of a compound called hypoxanthine, which
gradually increases in amount as time goes on and can be used as a measure of the
duration of icing. The rate of accumulation of hypoxanthine is not the same in all
species and this must be remembered when interpreting the results. The amount of
hypoxanthine present is measured either by an enzymic method that converts
hypoxanthine into uric acid, or by separating the hypoxanthine from any remaining
ATP and the intermediate compounds by a technique called high pressure liquid
chromatography (HPLC). In both cases the last stage is to measure how much of a
particular wavelength of uv light is absorbed by the solution of uric acid or
hypoxanthine itself; the instrument used is a spectrophotometer.
Like hypoxanthine, the K value measures the extent of the breakdown of ATP: it is the
percentage of the initial ATP present at death that has been converted by enzyme
action into hypoxanthine and its immediate precursor, called inosine, in the chain of
decomposition of ATP. The HPLC procedure used to measure hypoxanthine can
allow the K value, also, to be calculated.
Most marine fish contain a substance called trimethylamine oxide (TMAO). Certain
bacteria that occur naturally on the skin and in the guts of fish and in sea water can
break down TMAO to trimethylamine. The amount of TMA produced is a measure of
the activity of spoilage bacteria in the flesh and so is an indicator of the degree of
spoilage. TMA can be measured by a chemical method that produces a coloured
solution; the amount of the coloured product is measured using a spectrophotometer.
Alternatively, TMA can be separated from similar compounds, and its amount
measured, by gas chromatography (GC).
Bacteria can generate small amounts of ammonia in spoiling fish, mainly from free
amino acids; the amount of ammonia can give an indication, though not a particularly
accurate one, of the extent of spoilage. Much larger amounts of ammonia are
produced during spoilage of the elasmobranch fishes, skate and dogfish for example,
because they have large amounts of urea in their flesh. Shellfish, also, may develop
more ammonia than most marine fish and at an earlier stage. There are several
chemical and enzymic methods for measuring ammonia.
Total volatile bases (TVB)
Ammonia and trimethylamine are examples of bases; another base, dimethylamine
(DMA), can also be formed during spoilage of fish, together with traces of others.
These bases, other than ammonia, are known chemically also as amines. The
combined total amount of ammonia, dimethylamine and trimethylamine is called the
total volatile base content of the fish and is a commonly used estimate of spoilage.
The increase in the amount of TVB parallels the increase in TMA but the analysis is
easier to carry out than that for TMA.
Alternative terms used are total volatile base nitrogen (TVBN) and total volatile
nitrogen (TVN), since the results of the analysis are always given in terms of the
nitrogen content of the bases. Corresponding French and German names, sometimes
met with, are azote basique volatil total (ABVT) and flüchtiger Basenstickstoff. TVB
can be measured easily and quickly using relatively simple apparatus and, for this
reason, a TVB value is often used as a rejection limit in regulations and commercial
A range of methods are used to measure TVB. In all of them the fish, or an extract of
the fish, is made alkaline, the bases are distilled off and collected, then measured by
titration. Some of the substances used to make the fish, or the extract, alkaline can
convert other substances present in the fish to ammonia during the distillation, so that
the apparent amount of TVB increases as distillation proceeds. Several other factors
can affect the result so that the measured TVB depends quite significantly on the
details of the method used. It is possible, by two different methods, to get results on
the same sample that differ by a factor of 2. For this reason TVB, though widely used,
is not a particularly good index of spoilage; when a limiting value of TVB is included in
a specification or standard for any particular species, it is important that the method of
measurement to be used is described in detail.
Certain families of fish, notably the mackerel family, contain histidine, an amino acid,
in larger amounts than other families. During spoilage of these fish, especially if the
temperature rises to above 10°C, histidine may be converted to histamine. Histamine
is a substance that is produced by the body as part of the allergic response to foreign
substances, as in hay-fever. When spoiled mackerel is eaten, any histamine present
is usually inactivated in the stomach and rendered harmless (except in rare cases
where certain medicines are being taken). There is evidence, however, that some
other, unidentified substance is produced in the spoiling fish along with histamine and
this substance causes marked gastrointestinal disturbance. Measurement of the
amount of histamine in fish is used as a guide to the potential of the sample for
causing this form of food poisoning, the so-called scombroid poisoning.
To measure histamine a protein-free extract is first prepared; the histamine is
separated from interfering substances by extraction first into an organic solvent
followed by back extraction into an aqueous solution. The histamine is treated with a
substance that gives a fluorescent product and the amount of this product is measured using an instrument called a fluorimeter. Histamine can also be measured
by HPLC, along with certain other amines including putrescine and cadaverine; the
term “biogenic amines” is often used to describe these substances.
The electrical properties of fish skin and muscle change systematically after death
and can be used as the basis of an instrument; a few models are commercially
available, including the Torrymeter. The change in electrical properties is not caused
directly by bacterial action or other spoilage mechanism, but the instrumental
readings on iced fish can be correlated with the stage of spoilage, as measured by
sensory methods or by one of the non-sensory methods already described. The
instruments can be used only on whole fish or fillets with skin. Frozen fish, when
thawed, give no response to the meter and this can be used as a basis for checking
whether fish have been previously frozen.
What do the numbers mean?
It is not possible to lay down rules for deciding what values of any of the freshness
indices should be regarded as indicating any particular stage of spoilage or
acceptability. There are differences between species, the kinds of bacteria causing
spoilage may vary, the methods of analysis, as noted for TVB, can affect the values
and the mode of handling may influence the results. Ideally, the relationship between
the freshness as measured by sensory assessment and the various freshness indices
described should be derived for the species of interest, using well defined methods,
and for the particular handling procedure concerned. This is not always done: it is
common for a particular level of, say, TMA or TVB to be taken as indicative of an
unacceptable degree of spoilage in a range of species and without reference to the
handling procedure or the measurement technique.
Purely as a guide to the relative magnitude of certain indices in iced cod, values of
hypoxanthine, TMA and TVB, measured by methods used at Tony Research Station,
and readings on the Torrymeter (one of the instruments mentioned that measure the
electrical properties) are compared in the Table below with the time in ice and a
Days in ice means days in boxes since caught, well iced Sensory
score means the raw odour freshness score on the Torry wet fish scale
Hypoxanthine is measured in mg/100 g flesh
TMA is measured in mg N/100 g flesh
TVB is measured in mg N/100 g flesh (by a method that gives relatively high values)
A wide range of compounds arising from bacterial spoilage have been proposed as
freshness indices but only those in regular use are described above. Other indices
that may be met with include:
free fatty acids
putrescine and cadaverine
volatile fatty acids
volatile reducing substances.