Meat quality

[behnam63]

chapter seven
Meat quality: sensory and
instrumental evaluations
Brenda G. Lyon and Clyde E. Lyon
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Sensory quality attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Evaluating food with the five senses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Aroma and taste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Sight, touch, and hearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Other characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Sensory methods to evaluate poultry quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Laboratory/analytical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Affective methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Determining which type of test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Considerations in conducting sensory tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Sample presentations and preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Testing room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Specific sensory test formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Difference/discriminative tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Triangle test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Duo-trio test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Two-out-of-five test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Paired-comparison test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Ranking tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Category scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Descriptive analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Flavor Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Texture profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Other profiling methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Rating scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Instrumental methods of analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Selected texture instrumental methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Shear test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Warner-Bratzler shear device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Kramer Shear Press (KSP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Texture profile analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Sample considerations for shear or profile tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Relationships between instrumental procedures and sensory panels
for texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Intact muscle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Ground poultry meat texture studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Color . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Flavor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Factors that influence or contribute to meat quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119


Introduction
There are several dimensions to quality. Quality products are those that meet some need
or expectation of consumers and are safe and wholesome as well. Products that can be
produced and sold to meet a demand at a profit for producers are quality products.
Products that meet processing and handling guidelines set by agencies charged with protecting
the commercial food supply are quality products. Quality has several dimensions,
depending on whose viewpoint is neede regulatory personnel, producer, and ultimately
consumer.
Consumers are interested in appearance, aroma/odor, taste, texture, and sound, which
are all quality characteristics measured by use of the senses. Human testers measure these
characteristics (sensory attributes) by evaluating products and marking their responses on
paper or electronic scoresheets. Instruments can measure characteristics that are directly
related to the physical or chemical components of the product. These two types of measurements
are used together to draw conclusions and make assumptions about quality. This
chapter deals with quality factors perceived and measured by consumers (appearance,
aroma/odor, taste, texture, and sound) and how these factors relate to chemical or physical
component characteristics that can also be measured.


Sensory quality attributes
Sensory evaluation is analysis of product attributes perceived by the human senses of
smell, taste, touch, sight, and hearing. People (consumers or users of the product) are used
to assessing the sensory characteristics and providing a response. Instruments are used to
measure some physical or chemical characteristic that influences the sensory stimulus perceived
and responded to by the human. Instruments do not measure sensory characteristics.
However, instruments are sought that provide a corollary measurement that can
predict or relate to the anticipated sensory experience. Both human and instrumental methods
are critical when assessing sensory quality. Human assessment is more complicated.
People differ in their innate ability to sense stimuli. They differ in the experiences with
foods that allow a base for the neurological categorizing of a stimulus and the subsequent
varieties of responses that can be given. Instruments on the other hand can be calibrated
and programmed to respond consistently in a given way, but the meaning of the responses
has to be interpreted by humans and validated by the human sensory experience.

Evaluating food with the five senses
The five senses are taste, smell, sight, touch, and hearing. The responses to food are shown
in Figure 7.1.


Aroma and taste
The senses of smell and taste are interrelated and assess the quality attribute known as flavor.
Volatiles are small molecules released from the food (during heating, chewing, etc.)
that react with receptors in the oral or nasal cavities. Signals are sent to the brain where they
are processed. This processing results in responses that indicate whether the sensation was
sweet, sour, salt, or bitter (four basic tastes) and whether the sensation can be identified
more specifically (e.g., brothy, chickeny, fruity, etc.). Primary receptors for the four basic
tastes are on the tongue and other surfaces of the oral cavity. Receptors for volatiles are
located in the various sections of the nasal cavity. Sniffing is a technique used to collect a
concentration of the volatiles and force them to the receptors in the nasal cavity for processing
and identification.


Sight, touch, and hearing
The senses of sight, touch, and hearing are related to the structure and state of product
components. With the sense of sight, the sensory attributes of color and appearance are
evaluated. Receptors in the eyes are stimulated by light waves, causing signals to be sent
to the brain for processing. Therefore, appearance and color of foods involve the eyes as the
sense organ of the body and the components of the object (food) that reflect or transmit
light. Instrumentally, color is measured with instruments that determine the amount of
light reflected by the object at each wavelength. Color is very complicated. Humans measure
color as a composite, whereas instruments break the color into individual wavelengths.
Examples of texture characteristics perceived by sight are smoothness and bumpiness.
The physical characteristics of texture are the mechanical and geometrical characteristics
that are related to structure. These include strength, size, shape, and type of components
perceived as the product breaks down due to some force applied. The force could be the
teeth or it could come from an instrument. Other characteristics such as oily, greasy, wet,
and dry relate to mouthfeel and the sense of touch. The sense of hearing can also be used
to evaluate texture. For example, crunchiness may be an important quality in the batter and
breading of poultry products.



Other characteristics
Chemical and thermal mouthfeels such as cool, warm, hot, and cold are the other characteristics
perceived by the senses. These are called the trigeminal sensations and are related
to responses to stimuli on the cells of the linings of the mouth, tongue, and throat.


Sensory methods to evaluate poultry quality
There are two general types of sensory methods. Laboratory/analytical methods use a
small number of panelists to determine if a difference exists between samples and the
nature, direction, and intensity of the difference. Consumer affective methods involve a
larger number of panelists and include tests that measure how consumers feel or react to
the product to provide a measure of preference, acceptance, and like/dislike. There are different
panel criteria for laboratory and affective methods.


Laboratory/analytical methods
Methods that focus on detecting whether differences exist in products and how those differences
might be described are called laboratory/analytical methods. Small panels (6 to 12
assessors) of people who have been screened for their sensory acuity and ability to describe
products are used. Laboratory panels may be composed of staff or of outside persons paid
to attend sensory training and testing sessions. The key factor is that the panelists have
been screened and trained to evaluate products for specific characteristics, not for whether
they like or dislike the product. Therefore, the focus in these tests is the product attributes
using panelists as the measuring tools or instruments. Performance of the panel must be
measured to determine if their responses are reliable and consistent. Some panels have
been described as trained, semi-trained, or experienced. Trained panels have gone through
orientation and specifically designed sessions to screen for acuity. They have spent many
hours learning and applying descriptive language. Their results have been tested to determine
performance in finding sample differences. Some researchers may shorten the
process of training or only provide instructions to the panelists and call them semi-trained.
That is not appropriate. Experienced panelists are those who have been trained, have participated
on many appropriate panels or have performed many similar tests, and are very
familiar with product category characteristics and testing procedures.


Affective methods
Procedures that focus on how consumers (users of the products) react when given samples
to evaluate are called affective methods. The reactions that consumer panelists are asked to
convey are whether they like/dislike, prefer, or accept/reject samples. Consumer
responses may indirectly relate to the presence or absence of specific attributes. Do consumers
like the product? How well do they like it? Which sample is more spicy? Or more
tender? Do they prefer the product well enough to always purchase this brand over
another? Do they accept this product, even though they would prefer one less spicy? The
panelists used in these studies must be users of the product categories. Consumer panels
require larger numbers of people than do trained panels in order to sample or test
responses from a user population and then extrapolate the conclusions to a general population.
The consumer respondents are not trained or screened, except to determine demographic
profiles for relating to larger populations. The focus in consumer/affective testing
is the behavior of the panelist in relation to the product as the stimulus presented to the
consumer.


Determining which type of test
There are six fundamental questions to determine whether to use a difference/discriminative
test method or an affective test method.
1. Do the samples differ?
2. If so, on what sensory parameters do the samples differ?
3. Can the difference be quantified?
4. What is the direction of the difference? (i.e., more salty, less hard?)
5. How does this compare to similar products?
6. Does this have importance at the consumer level?
Generally, from the order of questions, that difference/discriminative or descriptive
tests come first, so that the characteristics of the product are known. Then consumers are
asked for acceptance, preference, or like/dislike in order to assess whether the known differences
are important to the consumer.
In product development and marketing research, another approach is emerging.
Consumer research determines how the concept of new products might be accepted and
determines what characteristics consumers want. Products are then designed with those
characteristics, using trained panels to screen and evaluate prototypes.
In any case, the purpose and function of the panel type remains the same. Consumer
panels are large to represent the feelings or purchase behavior of people toward the test
product. Trained panels are small numbers of people who have been screened to have good
acuity of the senses and whose task is to pinpoint discernible differences in samples.


Considerations in conducting sensory tests
This section will focus more on the smaller panels used for difference testing. In-house panels
can be made up of staff or students within the company or department. However,
screening and training are important and in order to screen and train panelists, they must
first be selected for their ability to detect small differences in aroma, taste, or texture.
Panelists must also be able to describe the characteristics. Although taste and smell are of
great importance, so is good health, a positive attitude, and motivation to perform the tests
without bias. Willingness and reliability to attend and participate in training and testing
sessions are equally important. A very important point to remember is that training panelists
involves more than explaining a scoresheet. Trained panelists function as sensitive
instruments, making responses to specific tasks that are totally separate from their personal
opinions of like/dislike.


Sample presentations and preparation
Samples presented for evaluation must come from a common and uniform source. This
is a difficult aspect of sensory testing when dealing with muscle food products, because
they are not as homogeneous as some other samples, such as grains or liquids. Choices
for sampling poultry meat depend on the test question, one being how many samples
are needed at one time. Another factor is how samples are to be cooked, sectioned,
and presented so that each panelist receives nearly identical samples under identical conditions.
The actual sample presented to the panelist should be uniform in size. Serving temperature
should be uniform throughout the sample piece. Appropriate implements should
be provided for evaluation (fork, toothpick). Filtered water should be provided between
samples for mouth cleansing to prevent taste carryover. Sometimes, unsalted crackers or
apples or other products are needed as well.
Methods of preparing the product are also determined by the test objective. Some studies
have been conducted where roasting in an oven was appropriate. Questions to be considered
were placement of pieces on the pan, placement of pan in oven, how to check
internal temperature without disrupting the cooking cycle, whether to roast covered or
uncovered, what oven temperature to use, and what internal temperature to use.
An example of a test preparation and sampling scheme used by Lyon and Lyon1
involved cooking broiler breasts in heat-and-seal bags immersed in water. This procedure
provided the best control for sample identification by labeling the bags and handling a
large number of samples during cooking. It was also appropriate to record individual
breast weights before and after cooking to determine cook yield and to conduct further
analysis on the cooked meat and fluid/solids liberated during heating.2–4 The effects of
cooking method on subsequent quality attributes of broiler breast meat have been
reported.5–7


Testing room
The area or room where panelists are presented samples and perform the tests requires specific
environmental controls, such as constant, comfortable temperature and humidity, and
freedom from extraneous odors, noise, and other distractions. This control is necessary
because human testers are designed to perceive and process many stimuli constantly and
unconsciously. In order that panelists might concentrate on smaller numbers of specific
stimuli (i.e., the test sample), they must be given an area that minimizes any stimuli other
than those of the test. In addition to the environmental controls, individual booths are
needed so that samples are presented to the panelist in isolation from other panelists to
avoid distraction and to avoid any collaboration on the part of panelists. Afloor plan for a
self-contained sensory laboratory is shown in Figure 7.2.
Lighting must also be controlled. If appearance of the sample is an important task of
the test, the lighting must not provide shadows and the spectrum of the light must be
appropriate for the use of the sample. On the other hand, if taste or mouth texture are key
aspects of the test, then special lighting might be needed to mask differences that the
panelists would use as cues to selecting different samples based on appearance rather than
the taste or texture under investigation. Some labs use red, green, or even blue lighting. A
monochromatic light often used is sodium vapor lighting that imparts an even spectrum of
orange, brown.
Test areas can range from portable partitions set up at a table to large testing facilities
housing a complete sensory evaluation laboratory including a waiting area, a training
room, testing booths with computerized data input systems, serving areas, and
kitchen/preparation areas. The key point is that the more control there is over the environment
where the test is performed, the more confidence the evaluator has that
the panelists are responding to stimuli in the product rather than stimuli to their surroundings


Specific sensory test formats
Difference/discriminative tests
Difference/discriminative tests are conducted under the premises that the panelist
evaluates a set of samples and determines whether any samples differ from another. If a
significant number of panelists detect a difference, then a true difference exists. The treatments
are known to the experimenter who scores the test responses as correct or not
correct and determines significance from tables based on the number of samples, number
of panelists, and the statistical probabilities of chance in selecting the correct sample.
Details on test features and the statistical tables for data interpretation can be found in several
popular textbooks.8–9
Difference tests include triangle, duo-trio, paired comparison, Anot A, two of five, and
three of five. These tests usually involve determining if two treatments differ. Multiples of
either are presented and the panelists must select one or two based on stated criteria.
Responses are recorded for whether the answer is correct or incorrect.


Triangle test
In the triangle test, panelists are presented with three coded samples, two the same, one different
(odd). Each panelist has a one in three (33.3%) chance of choosing the correct sample
by random selection. Therefore, the total responses must be higher than one third in order
to conclude that a true difference exists. The task for the panelist is to taste or smell, etc.,
the three coded samples in given order and to indicate which is different. Sample order is
randomized by the experimenter to avoid bias. Usually there is no qualifier to the test question
such as, “Which sample is different in sweetness?” Such a question tends to lead the
panelist to look only for sweetness when there may be other cues that determine the true
differences in the samples. An example of a triangle test scoresheet is shown in Figure 7.3.


Duo-trio test
In the duo-trio test, three samples are given. One is marked as “Reference” and the other
two samples are given codes. One of the coded samples is the same as the “Reference.” The
task for the panelist is to select the coded sample that is the same as the reference. The panelist
has a one chance out of two (50%) to select the correct sample by random selection.
Either of the two samples can be used as a reference throughout the whole test, or the selection
for reference can be alternated. The panelist is not given a specific characteristic to
focus on, but must decide which sample is the same as “Reference.”


Two-out-of-five test
In the two-out-of-five test, a panelist receives five coded samples. Two of the samples
belong to one set and the other three samples to another set. The task of the panelist is to
identify the set of two alike samples. The probability of guessing the right answer in this
test is 1 in 10, and is therefore considered more efficient than the triangle test. However, a
disadvantage is that sensory fatigue can be greater, especially if the test is used for taste or
oral texture. This test is used successfully with tasks involving visual, auditory, or tactile
senses.


Paired-comparison test
In a paired-comparison test, the respondent receives two coded samples (a pair) and is
asked to evaluate both, comparing the intensity of some specific characteristic. The specific
response is to record which of the two has the greater (or lesser) intensity of that attribute
being studied. In this test, a specific attribute may be given to the panelist to focus on in the
evaluation, such as which is sweeter.
Usually in difference tests, the task is to determine whether or not a difference exists.
If there is a difference, further tests might be presented to determine on what basis the samples
might differ or in what direction the samples might differ.

Ranking tests
Ranking tests are similar to directional-difference paired-comparison tests, except that
more than one sample is presented and panelists are asked to place samples in the set in
some sort of order. For example, rank from most tender to least tender. Rank samples from
most sweet to least sweet. These are examples of evaluating samples on a specific criteria,
i.e., tenderness or sweetness, and of indicating the direction of difference in that characteristic.
Aconsumer test (large number of untrained panelists) could also be a ranking test if
the task requested is that he/she place the samples in order based on least acceptable to
most acceptable or vice versa.


Category scaling
Difference tests can also involve category scales in which products are tested for specific
attributes and the panelists are asked to rate the amount that the characteristic is present.
Category scales can also be used for consumer testing in which the specific attribute rated
is degree of like/dislike, acceptance, or preference. The format of the scales can be numbers
(i.e., 1 to 5) anchored with a specific term, such as very tender, moderately tender, etc. The
scale can be unstructured, anchored only at ends and middle with either adjectives or faces
(i.e., frown to smile). The panelist marks on a line from left to right to indicate the point
their response to the product attribute is on the continuum. The response on the line is
measured with a ruler or automatically if a computerized system is used. The values of the
response, whether as the structured category scales or unstructured line scales, are analyzed
for their distribution variances by analysis of variance.


Descriptive analysis
Descriptive analysis is a form of sensory testing in which trained panelists determine the
perceptible attributes in a product set and score the intensity of the attributes that are
present. Flavor or texture may be profiled, or a profile can be developed for all the major
important attributes of a product from its initial appearance to the feeling left in the mouth
after the sample is swallowed.


Flavor profile
The first flavor profile method was introduced in 1949 by the Arthur D. Little Company.8
Flavor characteristics are described and quantified in a consensus manner by trained sensory
panelists. Much of the work of the panelists is done around a table where they first
analyze products individually and then discuss their responses as a group. The order that
aroma, flavor, and mouthfeeling characteristics appear is important. Asimplified intensity
scale is used to indicate where the sample is in an attribute range from detectable to very
strong. Because the final result is usually a group decision, statistics are not used to analyze
the data.


Texture profile
A method of evaluating sensory texture characteristics of products and relating these to
instrumental rheology principles was developed at General Foods Research in the early
1960s.10–12 Attributes were classified and defined to describe texture from the first bite to
after swallowing. Terminology and references were developed to illustrate various classifications
of characteristics. Mechanical characteristics dealt with resistance to breakdown
(hardness, cohesiveness, springiness). Geometrical characteristics dealt with the size,
shape, and orientation of the individual components or particles that form the structure
and how they behave when that structure is disturbed by force, such as chewing. Finally,
the moisture and fat properties were also considered to be part of the texture modality.
Evaluating these characteristics required trained panels. A scaling system that allowed
food references to be ranked or scored with the intensity of a predominant characteristic
that the food displayed was also developed. With numbers to indicate intensity, the results
of individual panelists could be statistically analyzed.


Other profiling methods
Building on the work of the original flavor profile and texture profile, variations of the
descriptive profiling methods have emerged, some now trademarked by their creators,
including Quantitative Descriptive Analysis (QDA)9 and Sensory Spectrum.8 Both of these
involve development of descriptive language by the panels and providing intensity values
that can be statistically analyzed. There are some differences in the way that terminology
is developed. Also, QDA uses an intensity scale that the panel selects based on the range of
products to be evaluated. Sensory Spectrum developed a universal scale to measure the
intensity of any identified character note in comparison to another. For example, a 0 to 15
scale could be used to rate sweetness of beverages comparing them to the sweetness of
sucrose solutions ranging from 2% (score of 2) to 10% (score of 10). Another example is the
grape character note scored as a 4 in grape Kool-Aid and a score of 12 for grape-note intensity
in Welch’s grape drink (Table 7.1). Against this background of intensive training by the
panel, the intensity of brothy notes in chicken soup or stewed chicken could be scored by
one panel and understood by another sensory panel trained in the same descriptive
method. Language or terms to describe the individual attributes are developed by the
panel members with a panel leader to guide them and provide references.
Variations of these methods have also been reported and used successfully. Free-choice
profiling lets panelists develop their own terms and score the intensities. Advanced statistical
procedures are needed to interpret the results.


Rating scales
The rating scales that are used with these methods take the form of a continuous line
that represents a low or no level of intensity to a very high level. Sometimes intensity terms
are presented as anchors along the line at distinct intervals or at the ends. When the
intervals are clearly marked, the scale is said to be structured. When there are no marked
points between the lines, the scale is unstructured and the panelist uses a mental cue for
intensity.


Instrumental methods of analysis
Texture is considered the most important characteristic of poultry meat and is the attribute
most affected by age of the bird and processing procedures. Because of the importance of
texture, a great deal of emphasis has been placed on instrumental procedures to evaluate
the structure of muscle fibers. Bourne13 noted several important truths about instrumental
texture measurements. Many tests are applicable to more than one type of food, so it is
more useful to classify the texture measurements by type of test rather than by commodity.
He noted that the basic process of chewing food to break down the food for swallowing
occurs, regardless of what kind of food is in the mouth. Another truth was that the fundamental
instrumental tests were developed by scientists and engineers interested in the
theory and practice of materials or construction to measure well-defined rheological properties.
Those theories may not be as useful in measuring what is happening in the mouth
during mastication. As a matter of fact, the expectations of the tests are opposite for the two
groups of scientists. The engineer wants to measure the strength of material in order to
design a structure that will withstand forces applied to it without breaking, while the food
scientist wants to measure the strength of food, and frequently weakens the structure so
that it will break easier. In this situation, food texture measurement might be considered
more of a study of the weakness of materials rather than strength of materials.
Instrumental procedures to estimate tenderness of meat have been studied and widely
accepted by researchers and quality control (QC) personnel since the 1950s. These procedures
offer repeatability to obtain numerical values that should relate to tenderness. The
danger of using instrumental procedures is putting too much value in the “number” without
understanding what it really represents. Texture has historically been viewed in an
overly simplistic manner, so the research was geared toward finding a single measurement
or number to encompass the entire mastication process and arrive at an either/or decision:
tender or tough. An accurate description of poultry meat tenderness involves more than a
single instrumental value, since most treatments alter postmortem biochemical events and
affect not only tenderness, but also moisture-binding characteristics such as juiciness and
moisture release.


Selected texture instrumental methods
Unless noted otherwise, the focus of this section will be on breast muscle/meat because
this economically significant part of the carcass has received the vast majority of the
research attention. The breast has received this attention because of its postmortem biochemistry
(see Chapter 4) and subsequent fiber characteristics that impact finished product
quality. It should be noted that any of these methods can be used to evaluate leg/thigh
meat and ground/comminuted products as well as intact meat products. One simply
needs to determine the objective of the analysis and choose the appropriate method.
Shearing may be most important for whole-muscle while compression may be best for
frankfurters or cohesiveness for restructured products like nuggets and patties.
The majority of the instrumental data used to determine tenderness in cooked poultry
meat have been generated on the Warner-Bratzler (W-B) or the Kramer Shear Press (KSP).
These procedures are designed to shear or cut through fibers of muscle. Another technique,
instrumental Texture Profile Analysis (TPA) data, has been used to generate texture information
for poultry meat products. An in-depth discussion of the concept and measurement
of food texture was published by Bourne13 and will only be briefly summarized here.


Shear test
Shear tests have been used for many years. Samples are positioned so that a single blade or
multiple blades cut perpendicular to the fibers. The basic principle of the test is that the
total force to cut through the sample is related to the tenderness/toughness of the cooked
sample. The force has historically been recorded in weight measurements (i.e., lb, kg), but
these can be converted to the force unit of Newtons, if appropriate.


Warner-Bratzler shear device. The W-B shear device has been used to shear or cut red
meat and poultry samples for the last 50_ years.14 The device is small and portable, consisting
of a rectangular blade with a triangular hole cut from the center. This blade is
attached to a circular fan scale. The sample of known dimensions, usually a circular core
for red meat or a rectangular strip for poultry, is placed in the triangular notch of the single
blade. Two bars are lowered by a hydraulic motor and the sample is pushed across the
apex of the triangular notch. As the bars are lowered across the sample, the peak force to
shear across the fibers is recorded in lb or kg on the circular fan scale. The benefits of this
device are its reliability, ruggedness, ease of use, portability, and low cost (less than $1200).
The device lends itself to on-site quality control work. The limiting factor is that only peak
load or peak shear force is generated during the test, so the researcher or QC personnel
must have sufficient background sensory panel data to add validity to the shear values
(Figure 7.4).


Kramer Shear Press (KSP). The other shear test that has been extensively used for red
meat and poultry texture research is performed with a shear cell based on the KSP.15 The
shear test cell is composed of two main parts, a metal box with slots which holds the sample
and a top part with 5 or 10 blades spaced to fit into the slots. This device is attached to
a system designed to move the multiple blades down and through a rectangular sample
placed in the cell. The multiple blades are lowered across the sample. They initially compress
and then shear across the fibers forcing the resulting strips out the bottom of the slotted
cell. Results are recorded as kg/g of sample weight. The KSP is rugged, but it is much
heavier, less portable, and more expensive. It has been modified from its original design to
predict quality of lima beans and used to measure textural properties of a variety foods
including fruits and other vegetables.
Both of the blade designs of the original W-B and KSP systems have been reproduced
on other instruments such as the Instron Universal Testing Machine™ (UTM); (Instron
Corp., Canton, MA) and the Texture Technologies Texture Analyzer™ (Texture
Technologies Corp., Scarsdale, NY). The multiple blade cell is also referred to as the Allo-
Kramer shear cell. The W-B blade and an Allo-Kramer shear cell are pictured in Figures 7.5,
7.6, respectively. The newer systems are accompanied by software to program the
machines and to record more dimensions of the force/distance or force/time curves.


Texture profile analysis
The instrumental TPAwas introduced as a way to generate multiple textural attributes for
food.10–12 The need for a multiple-point test was reinforced by Breene16 who noted that
texture is complex and multiple point procedures would be more useful than single point
procedures. The TPA was recently updated by Meullenet et al.17
A typical two-curve TPA for chicken meat is shown in Figure 7.7. The significant
attributes are noted and defined. Significant attributes such as hardness, springiness, cohesiveness,
and chewiness can be separated and analyzed. ATPAsample is usually a circular
core taken from the cooked meat. Adecision must be made by the researcher on percent of
compression during the test. In the literature, ranges reported for percent compression
range from 60 to 80% of the original height of the core. Compressing less than 60% usually
does not compress the sample enough to result in measurable changes, while compressing
more than 80% usually destroys the sample matrix so much during the first compression
that the second compression curve yields little or no information. The core to be evaluated
is placed on a flat metal plate, and the top metal plate attached to the load cell is positioned
to contact the sample (initial point). The percent compression is converted to cross head
travel from the initial point. After the first compression and return of the cross head to the
initial point, the cross head is immediately engaged for the second compression. The TPA
is more of a research tool than the shear tests. The TPAis more sensitive and versatile than
the W-B or KSP shears. However, the purchase and maintenance costs for instruments such
as the Instron UTM or the Texture Analyzer are much higher, and they are not as portable.


Sample considerations for shear or profile tests
Regardless of the type of instrumental test, sample dimensions play a major role in the
results and should always be described or referenced.18, 19 Physical characteristics of the
product as used by the consumer should be taken into account when evaluating the meat
sample. For example, if the treatment imposed has a direct effect on meat thickness due to
muscle contraction (postmortem/postchill deboning time), then the difference in thickness
should be part of the test. However, if the research goal is to evaluate the sensitivity of
instrumental procedures, then uniform sample dimensions (height and width) would be
required. Asampling scheme for both sensory and instrumental tests from a broiler breast
muscle used in this lab is illustrated in Figure 7.8.
Relationships between instrumental procedures and sensory panels for texture
As noted earlier, there is a danger in reducing the complex continuum of texture to a
single objective number. A number of studies have been conducted to help determine the
relationship between instrumental and sensory data related to texture.20–22

Intact muscle samples were used in a study20 to correlate sensory scores from a fivemember
untrained panel to KSP values. Processing treatments were used to simulate a
wide range of texture in the cooked meat. The authors noted that the KSP and sensory
scores were correlated, that KSP values greater than 8 kg/g of sample weight were tender
to very tender, and that due to a wide 95% confidence interval the five-member untrained
panel was too small to measure sensory reaction. Two studies by Lyon and Lyon21, 22
increased the number of untrained panelists to 24 to determine the texture relationship of
broiler breast meat to 4 instrumental tests. Then 4 breast muscle deboning times ranging
from after feather removal (0 hour postmortem) to after whole carcass aging for 24 hours
were used to provide the texture spectrum from tough to tender. The four instrumental
tests were the bench-top W-B, Allo-Kramer (i.e., KSP), W-B attached to an Instron (I-WB),
and a single blade version of the multi-bladed KSP (SB-AK). All shearing apparati, except
for the bench-top W-B were attached to an Instron UTM.
Results are summarized in tabular form in Table 7.2. The significance of the results is
that instead of a single number for each shear test, a range of values corresponding to the
sensory panel perception of tenderness was established for each test. The sensory scale is
not an either/or (tough or tender), but a gradation of values from very tough to very tender.
These data are used by quality control personnel to verify process control and ensure
optimum tenderness for customers.
Lyon and Lyon23 reported on the relationship between the TPA and a trained panel’s
response to intact broiler breast meat by using 4 postmortem deboning times (_5 min, 2, 6,
24 hours) and two cook methods (heat-seal bags in water and microwave) as variables. In
a series of sessions, the 8-member trained panel developed 17 attributes and rating scales
to evaluate texture (Table 7.3). The attributes developed by the panel to evaluate the samples
represented a 4-stage profile ranging from the first compression with molar teeth without
biting through the sample (stage 1) to impressions at the point of swallowing and the
“afterfeel” properties in the mouth (stage 4). Instrumental TPA attributes of hardness,
springiness, cohesiveness, and chewiness were calculated. Meat from muscles deboned 5
min and 2 hours postmortem was significantly different from those deboned 6 or 24 hours
postmortem for 16 of the 17 sensory attributes. No sensory differences were noted for meat
from muscles deboned 6 or 24 hours postmortem. Muscles removed 5 min postmortem had
significantly higher hardness and chewiness values than those deboned 2, 6, or 24 h. Within
deboning time, the panel scored meat cooked via microwaves as more juicy and wet and
as having less residual particles and toothpack compared to the meat cooked in water. By
TPA, the microwave cooked meat was more cohesive and chewy than the meat cooked in
water.
The panel results significantly correlated to the instrumental TPA. For example,
the muscles removed 5 min postmortem were more springy, cohesive, harder, produced
more saliva on chewing, had a larger bolus size, were harder to swallow, and had more
toothpack. Needless to say, this is significantly more than a single force shear value and
adds to the broad spectrum of attributes that we term “texture.” The impact and complexity
of juiciness are evident in the panel results.


Ground poultry meat texture studies
A series of studies published in the late 1970s and 1980s24–26 characterized the texture of
poultry products made from ground and comminuted meat with various ingredients. All
three studies utilized both sensory panel methods and instrumental texture measurements.
A wide range of quality attributes was evaluated (proximate composition, water-holding
capacity, color, rancidity, and cook loss). In one of the studies,24 use of mechanically
deboned poultry meat as the meat source (with and without skin) in conjunction with two
levels of structured protein fiber (15 and 25%) was evaluated by a 5-member trained panel
using the QDA technique. In addition, a scale to reflect overall impression of the products
was included.
In another study,25 six patty formulations containing different amounts of mechanically
deboned broiler meat (MDBM), hand deboned fowl meat (HDFM), and structured protein
fiber (SPF) were characterized for proximate composition, rancidity (measured as thiobarbituric
acid or TBA values), color (Hunter L, a, b values), force to shear (W-B), and sensory
properties. Sensory properties were evaluated using QDA. As the level of MDBM
decreased, moisture and protein contents, lightness (L values), and shear values increased
correspondingly; fat content, redness (a values), and TBA values decreased. Sensorially, as
the level of MDBM decreased, the products were perceived as being lighter, more chewy
and elastic, and less juicy. Based on the instrumental and sensory data, the authors noted
that interchangeable ratios of 40:60/60:40 MDBM and HDFM could be incorporated with
SPF to yield products of good quality. These multiple point results illustrate the benefits of
integrating instrumental and sensory analysis to arrive at decisions involving finished
product quality. Aspider-web diagram illustrating part of the results is shown in Figure 7.9.

Data points are placed on the various lines representing each attribute. The center represents
a value of “0” and the values increase away from the center point. The differences
in attributes such as outer appearance, chewiness, elasticity, particle size/shape, and overall
impression are easily noted. In this example, the combination of 40% MDBM: 60% HDFM
was superimposed on the 100% MDBM patty product for visual comparison of attributes.
In yet another study, Lyon et al.26 used TPA to determine differences between mixed
and flake-cut MDPM in patties containing either 15 or 25% SPF. The six-member trained
panel also evaluated juiciness using a seven-point intensity scale. Positive, significant correlation
coefficients between instrumental and sensory measures of hardness, springiness,
and chewiness indicated that the Instron and the panel were in good agreement.


Color
Color is very complex and is a major component of appearance in poultry meat or products.
Instrumental methods to measure color of an object are based on a light source and
a detector. Objects absorb and reflect light wavelengths that are detected by an instrument
or an observer. Results of instrumental detectors have little meaning unless validated by
the human observer. Therefore, numerical values provided by colorimeters are almost
always associated with a color/appearance term in order to understand the meaning. For
example, “lightness” is associated with “L values,” “redness” with “a values” and “yellowness”
with “b values” when an “Lab” color coordinate system is used. A typical colorimeter
used in research and quality assurance is shown in Figure 7.10.
Fletcher27 reviewed poultry meat color, color measurements, methods used to measure
color, and summarized color defects associated with poultry. The review of meat color
covered raw meat and many of the factors that affect meat color such as sex, age, strain,
processing procedures, cooking temperature, and freezing. Of particular significance at the
present time are the factors that influence “pinking” of breast meat. The significance from
both quality and safety standpoints is the assumption of insufficient cooking time/temperature.
This is a problem with immediate economic ramifications (returned shipments of
cooked product). Other specific color defects include the relationship between lightness
(paleness) and poor protein functionality,28 and the consumer objection to color variation
between meat pieces in a retail package.29


Flavor
Flavor analyses of poultry or poultry meat involves methods to extract compounds that are
assumed to contribute to aroma. Taste is usually associated with the basic solutions of salt,
sweet, sour, and bitter, while aroma is associated with stimulation of receptors in the nasal
cavity by volatiles released by foods. Instruments that separate compounds and indicate
their concentrations include gas chromatography (GC), high pressure liquid chromatography
(HPLC), and sensing devices referred to as “electronic noses.”
GC and HPLC are methods that separate extracts of the food into individual compounds.
Although individual compounds and classes of compounds have been identified,
they must be related to sensory response by descriptors. Sensory descriptors determined
by a trained panel for evaluation of cooked chicken were developed by Lyon30 (Table 7.4).
Farmer31 listed as many as 34 main compounds that are considered to be key components
of cooked chicken flavor. Taken alone, the individual compounds do not always exhibit the
aroma of individual perceived aromas from the samples. For example, 2-acetyl-pyrroline,
a key odor compound in cooked poultry meat, is described as “popcorn.” Re-combining
certain chemicals to create an aroma-specific character note is not always successful.
Sensory panels can tell the difference.
The electronic nose is a name given to instruments comprised of arrays of materials
(metal oxides, conducting polymers) that record an electrical charge or resistance response
when a stream of volatiles is passed over them. Several sensors of varying materials give
varying responses so that a pattern emerges for a given sample. Key to the instruments is
analysis of data by multivariate statistical programs that can detect pattern differences and
also develop algorithms that can later recognize this pattern as belonging to a certain sample.
Supposedly, the technique is based on how a stream of volatiles will pass across the
receptors in a human’s nose, detect the differences, and attach a recognition to the pattern
for future identification.



Factors that influence or contribute to meat quality
Many factors influence poultry meat quality. Some factors are more significant than others
(see Chapters 2–4). Rigor condition of the breast muscle at the time of removal from its
skeletal restraints (deboning time) significantly affects the texture of this economically
important part of the carcass (Chapter 4). Time of breast muscle removal involves postmortem
muscle biochemistry (pH decline, lactic acid increase, ATP depletion) as well as
physiology (gross and microscopic muscle fiber contraction, i.e., sarcomere lengths). Time
of breast muscle removal is really a “double-edged sword” depending on the condition of
the meat in the finished product. On one side, higher pH noted in prerigor muscle equates
to increased water-holding and emulsifying capacity which are important for ground and
comminuted products. On the other side, the same high pH equates to objectionable
toughness in intact cooked meat. Froning and Neelakantan32 reported that a pH of 5.9 or
higher could be used to indicate a prerigor condition in broiler and turkey breast meat, and
that the pH was below this value within 30 min of death. Lyon et al.33 reported prerigor pH
values of 6.1 and 6.3 for broilers and mature hens, respectively, within 20 min of death, but
values lower than 5.9 after 1.5 hours.
The relationship between postmortem time, muscle biochemistry, and ultimate texture
has been illustrated by many researchers. Lyon et al.34 noted that broiler breast muscles
deboned immediately postchill had significantly higher pH values, 6.22, and the cooked
meat required greater force to shear, 15.19 kg compared to muscles deboned at 1, 2, 4, 6, 8,
or 24 hours postchill. The most rapid pH decline was noted during the first hour postchill,
and no significant differences in pH or shear values were noted after 4 hours postchill time
prior to deboning. Sams and Janky35 evaluated the effects of water and brine chilling on
broiler breast meat pH and tenderness. They added a “hot boned” group of muscles which
were removed from the carcasses immediately after feather removal (picking). The other 2
treatments were breast removal after 1 hour chilling, and after 24 hours aging. For the
muscles chilled in water, the highest pH and KSP values were noted for the “hot boned”
group, 6.4 and 10.2 kg/g, respectively. The muscles removed after chilling were intermediate,
and the lowest pH and KSP values were noted for the 24-hour aged group. Dawson
et al.36 using similar conditions noted the same postmortem deboning time and shear force
(KSP, kg/g of sample weight) pattern for broiler breast meat, with highest KSP values
noted at 0.17 hours postmortem holding time and the lowest at 24.33 hours, 17.8 and 4.1
kg/g, respectively.
There are contradictions in the literature as to the effect of sex of the broiler on tenderness
of the cooked meat. Simpson and Goodwin 37 and Farr et al.38 reported that shear
values for male broilers were significantly lower than those for females. Other researchers,
including Goodwin et al.39 reported that sex of the bird did not influence tenderness. Lyon
et al.2 reported on the effects of postchill broiler breast muscle deboning time, fillet holding
time, and sex of the bird on tenderness. The two sexes of birds were raised under commercial
conditions, processed on separate days, and the data analyzed within each sex. Mean
raw breast weights were 163 and 122 grams for males and females, respectively.
Under the conditions of the Lyon et al.2 study in which weight of the breast samples
was not controlled or adjusted, it would appear that meat from female broilers was more
tender for all treatments. The larger size and weight of the male breasts probably contributed
to the increase in force to shear the samples. For the larger muscles removed from
the males immediately postchill (0 hour of aging), there would have been an accompanying
loss in area and an increase in thickness due to muscle shortening facilitated by
elevated ATP levels. This shortening pattern of the pectoralis muscle was reported by Papa
and Lyon.40 As noted earlier, since the condition of the sample is part of the variable being
studied, the height and/or width of the sample was not standardized.
To determine the practical importance of the numerical decrease in W-B values, Lyon
and Lyon21 superimposed the shear values on a tenderness scale established by a 24-
member untrained sensory panel. Frequency distribution of W-B values for female samples
which correspond to the panel’s perception of tenderness are illustrated in Figure 7.11.
Since fillet holding time prior to freezing was not significant, the data were combined into
a single value for postchill deboning time. Eighty-five percent of muscles removed from
female broilers immediately postchill (0 hour) would be classified as “moderately to very
tough” (categories 4 and 5). This percentage decreased to 43% if the muscles were left on
the skeleton for 1 hour. Muscles removed 24 hours postchill were all in the “moderately to
very tender” portion of the scale (categories 1 and 2) for both sexes. Without the sensory
panel perception data, the W-B shear values have less meaning and while the numbers can
be statistically analyzed, their practical importance would be limited.



Conclusions
Poultry meat quality is a complex issue which will become increasing important as more
new products are introduced to consumers. Students, researchers, quality control, and
management personnel must appreciate this complexity and all work together to provide
the appropriate and complete information. The “marriage” of sensory and instrumental
methodology is critical to providing the correct answers and making the best decisions
about product quality.