- Email: [email protected]
Variation in light scattering and water-holding capacity along the porcine Longissimus dorsi muscle
Meat Science 15 (1985) 203-214
Variation in Light Scattering and Water-Holding Capacity Along the Porcine Longissimus dorsi Muscle K. Lundstr6m & G. Malmfors Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences, S-750 07 Uppsala, Sweden (Received: 25 February, 1985)
SUMMARY The meat quality of M. longissimus muscle was evaluated in 851 pigs by using the Fibre Optic Probe (FOP) at three sites in the muscle and in both halves of the carcass. A systematic difference between sites was found, with the lowest light scattering (indicating the best meat quality) in the mid-part of the muscle and higher light scattering in the anterior and posterior parts. A non-systematic variation was also observed, with the opposite pattern in some animals, even though it was not frequent. A negative influence of the shackling wasfound, yielding higher FOP values in the shackled haiti Drip loss measurements in the Longissimus muscle, taken from another lOOpig carcasses, were evaluated using three methods. Drip loss, too, showed a considerable variation along the Longissimus muscle, with minimum losses around the last rib. Repeatability estimates, calculated from two non-consecutive pieces of the Longissimus dorsi muscle from each carcass, varied from 0"4 when keeping samples vacuum packed for 2 days, to 0"5 when the samples were either kept in plain plastic bags for 2 days or in a meat container with a squared inset ]br 1 or 2 days.
INTRODUCTION Interest in assessing meat quality in pig carcasses has increased during recent years. This is due to the very strict quality requirements when 203 Meat Science 0309-1740/85/$03.30 © Elsevier Applied Science Publishers Ltd, England, 1985. Printed in Great Britain
204
K. Lundstrdm, G. MalmJbrs
exporting pig meat, but also to attention the poor meat quality expressed by the consumer on the domestic market. Generally, a subjective method is used when the meat quality of the Longissimus muscle is evaluated commercially. As the muscle is treated as a primal cut, only cut surfaces at the shoulder and in the lumbar region are available for examination. If the muscle is not uniform in meat quality, there will be a certain degree of erroneous classification, as well as mistakes depending on the subjective evaluation. With new equipment now available, such as the Fibre Optic Probe, objective measurements of meat quality can be made on several muscles in a carcass and at several sites within a muscle. When pig meat quality is evaluated, some method of colour measurement is often used. Even more sophisticated methods, such as measuring the scattering coefficient, can only be regarded as indirect measures of meat quality. Consequently, there has been a wish to measure a more direct meat quality parameter such as water-holding capacity or drip loss of a muscle. Lack of reliable and speedy methods for the measurement of drip loss has prevented this variable from being included in meat quality assessment in the pig-progeny testing scheme. The twin purposes of these studies were (i) to examine the variation in light scattering along the Longissirnus muscle and (ii) to develop and test a method for measuring water-holding capacity or drip loss that would be simple, rapid and accurate.
MATERIALS A N D M E T H O D S
Variation in light scattering along the Longissimus muscle The animals used in the first study were ordinary pigs brought for slaughter at a slaughterhouse in the southeast of Sweden. Only a small sample (851) of the total number of pigs slaughtered in one week was used in this study. Several farms were represented and the mean carcass weight was 76kg (standard deviation, 4.9 kg). The animals were stunned with carbon dioxide and shackled by the left hind foot at the stunning. The carcasses remained shackled during the bleeding and scalding procedure (15-20 min in total). Evaluation of meat quality was made with the Fibre Optic Probe (TBL, Leeds, Great Britain) the day after slaughter. The Fibre Optic Probe
Variation in meat quality along the Longissimus muscle
205
determines the back-scatter of light transmitted into the meat. The recorded values are higher than normal for PSE muscles and lower for D F D (MacDougall & Jones, 1981). The recordings were made at three sites in the Longissimus muscle, on both left and right halves. The sites were: (i) between the 9th and 10th thoracic vertebrae (shoulder cut), (ii) at the tip of the last rib (mid-loin) and (iii) between the 5th and 6th lumbar vertebrae (lumbar cut). The probe was inserted between the spines into the middle of the muscle on the split carcass. The loins were divided into two quality groups--PSE and normal-depending on the Fibre Optic Probe value (FOP value). The muscles were classified as PSE when the FOP value equalled or exceeded 55 (G. Bjfirstorp, pers. comm., 1984).
Water-holding capacity The animals used in the second study were 100 Swedish Landrace × Yorkshire crossbred pigs. They came from one herd and had a mean carcass weight of 75 kg (standard deviation, 2.6 kg). The pigs were electrically stunned with a low voltage (85 V, 20 s) using a restrainer and shackled by the left leg only during bleeding (8 rain). The day after slaughter, the Longissimus dorsi muscle from the left side of the carcass was excised. Due to the method used to assess the meat content of the carcass, the Longissimus muscle was first cut at the last rib. The anterior part was divided into four similar pieces and the posterior part into two (Fig. 1). The six pieces from each carcass were weighed (mean value, 318 g, standard deviation, 56 g) and distributed among three different methods for evaluating drip loss. The two pieces used for each method were not consecutive. (1) Samples were placed with a cut surface facing down in a tray supplied with a net bottom and a squared inset (Fig. 2). The tray was then placed in a larger container, covered with an inverted container and kept at + 4 °C. The samples were reweighed after 1 and 2 days' storage. (2) Samples were kept in plain plastic bags at + 4 ° for 2 days and reweighed after dabbing with a paper napkin. (3) The samples were vacuum packed and then subjected to the same treatment as in (2) above. In addition, the Ultra-X capillary volumeter, as developed by
K. Lundstr6m, G. Malmfors
206
Drip toss, % 9 • Vacuumpacked,I 2 days ~
/
O Heatcontainer,Q 2 days ~
\
~
,
J / / ~
I
1
Jim
~
Plastic bag, ~ 2days ~ [] Meatcontainer,[] \ 1 day
I
~--
I
2
I
3
I
~
I
5
astlrib
I
6
Piece
I
"
Fig. 1. Drip loss measurements along M. longissimus dorsi. The six pieces are indicated on the Figure where the anterior cut is between the 4th and 5th thoracic vertebrae and the posterior cut is between the 5th and 6th lumbar vertebrae.
H o f m a n n (1975, 1977), was used on days 1 and 3 on the anterior muscle surface, cut at the last rib. Two recordings (30 s) were made on each muscle sample, using the same gypsum plate, one near the spine and one laterally.
Statistical methods Data were analysed with the Statistical Analysis System (SAS Institute Inc., 1982) using the General Linear Models procedure. For variation in light scattering the following model was used. Yijkl = l"t + ai + h~ + s k + (hs)j~ + eijkl
Variation in meat quality along the Longissimus muscle
207
Fig. 2. Equipment for measuring drip loss. Each side in the squared inset is 130 mm. The outside container is covered with an inserted container to prevent evaporation.
where Y~kl--- the ijklth observation; # = general mean; a i = effect of the ith animal (i = 1 , 2 , . . . ,i851);ihj = effect of t h e j t h half (j = 1,2); s k = effect of the kth site of measurement (k = 1, 2, 3); (hs)j~ = effect of the interaction between t h e j t h half and the kth site and euk t = residual r a n d o m term with variance tr e2 . The effect of animal was regarded as r a n d o m and the effects of half, site of measuring and the interaction between half and site as fixed. For water-hoMing capacity the following model was assumed to describe the data: Yo.kl = # + di + ao + Pk + b~'Cijkt + eijkt where Yukt= the ijklth observation; # = general mean; d i = effect of the ith day of slaughter (i = 1, 2 , . . . , 10); aij= effect of t h e j t h animal on the ith day of slaughter (j = 1,2, . . . , 100);pk = effect of the kth piece (k = 1, 2 . . . . ,6); b~ = linear regression on sample weight calculated within the kth piece; Xokt=weight of the ijklth sample (day 1) and e~k~= residual r a n d o m term with variance a 2e. All effects except the effect of animal were regarded as fixed. Repeatability estimates were calculated as intraclass correlations (t), with t calculated from the components of variance.
208
K. Lundstr6m, G. Malmfors
t--
O"a2 2 O"a +
2 O"e
where t7a2 is the component of variance for animal and t7e2 is the residual variance. For the various methods evaluating drip loss, the intraclass correlations were calculated between the two non-consecutive pieces of the Longissimus dorsi muscle taken from each carcass. For the capillary volumeter methods, the intraclass correlations were calculated between the two recorded values on the same muscle sample taken at the last rib. The standard errors of the intraclass correlations were calculated with the approximate formula described by Falconer (1981). RESULTS
Variation in light scattering Overall means for the F O P value from the right and left sides measured at the three sites can be seen in Table 1. The percentage of PSE musculature (FOP values > 55) is also shown. In accordance with the high mean values, the percentage of PSE was highest in the shoulder site at both halves. The lower meat quality in the left half at all sites (side of shackling) is of particular interest. Besides the significant effects of carcass half (P < 0.001) and site (P < 0.001) in the analysis of variance, the interaction between half and site TABLE 1 Overall Means of F O P Values for the Right and Left Sides of M. longissimus dorsi Measured at the Shoulder, Mid-loin and Ham Sites (n =851). The Percentage of PSE Muscles with F O P Values > 55 is Also Given Region Shoulder
Mid-loin
Ham
Right side Mean (SD) PSE (~o)
47.6 (16.6) 39.5 (14.5) 42.6 (14.1) 19.9 11.6 12.2
Left side ~ Mean (SD) PSE (~o)
50.6 (18.1) 40.4 (15-3) 43.8 (13.4) 23.6 12.1 13.2
a Side of shackling.
Variation in meat quality along the Longissimus muscle
209
TABLE 2 Uniformity of the Longissimus Muscle in the Left and Right Sides of the Carcass. The Sites are Shown in the Order Shoulder/Mid-loin/Ham a Left side
Right side
Per cent (Number) Per cent (Number) Normal/normal/normal PSE/normal/normal PSE/PSE/PSE ?SE/PSE/normal PSE/normal/PSE Normal/normal/PSE Normal/PSE/normal Normal/PSE/PSE
72.4 8.7 6.2 5.1 3.6 3.2 0.7 0.1
(616) (74) (53) (43) (31) (27) (6) (1)
74.8 7.2 5.4 4.6 2.7 3.6 1.2 0.5
(637) (61) (46) (39) (23) ( 31) (10) (4)
a The musculature was considered to be PSE when the FOP value was
>55. was also significant (P < 0.01). The shackled left half, with its higher average F O P values, showed a greater difference between F O P values at the three sites studied. The shackled half was thus less uniform in meat quality and this may explain the significant interaction between half and site. The differences in F O P values between the right and left halves were highly significant for the shoulder (P < 0.001) and lumbar sites (P < 0.01) and not far from significant for the mid-loin (P = 0.067). For further evaluation of the variation in light scattering in the Longissimus muscle, the degree of uniformity was examined. The threshold value of 55 was used to differentiate between normal and PSE musculature. As can be seen in Table 2, most Longissimus muscles are homogeneously normal in all sites (72.4 ~o in the left half and 74-8 ~ in the right half). There are, however, more loin muscles that are PSE only in the shoulder site than in all sites. The combinations PSE/PSE/normal, PSE/normal/PSE or normal/normal/PSE are approximately equally frequent, while the remaining combinations are very rare.
Water-holding capacity Mean values of drip loss of M. longissimus dorsi measured at different sites are presented in Fig. 1. Drip loss was greatest when the vacuum bag method was used (7.4 %) and least when keeping the samples in the meat
K. Lundstr6m, G. Malmfors
210
TABLE 3 Repeatedly of Water-Holding Capacity of M. longissimus dorsi Estimated as Intraclass Correlations ( + Standard Errors)
Method
Intraclass correlation
Level of significance Piece~ Regression on sample weight
Meat container 1 day 2 days Plastic bag, 2 days Vacuum bag, 2 days Capillary volumeter a Day 1 Day 3
0.50 0.51 0.52 0.39
+ + + +
0.08 0.08 0.08 0.09
*** *** *** ***
0.64 + 0.06 0.43 + 0-08
*** *
NS NS NS
w
a For the capillary volumeter, the difference between points of recording is tested (dorsally versus laterally). Levels of significance: NS = not significant (P > 0.05); * P < 0'05; • ** P_< 0.001.
container for one day (3.8 ~o)- A highly significant difference between the various sites in the Longissimus muscle was found for all methods (Table 3). This can also be seen in Fig. 1, where it is clearly evident that the drip loss is least in the middle of the muscle piece No. 3. The initial sample weight influenced the drip loss when evaluated by the plastic bag method. A significant interaction was found between sample site and the regression on sample weight, and the slope of the regression varied along the Longissimus muscle. The repeatability estimates, calculated as intraclass correlations, ranged from 0-39 for the vacuum bag method to 0.64 for the capillary volumeter method, one day after slaughter (Table 3). As the repeatabilities for drip loss and for the capillary volumeter method have different meanings, they cannot be compared (see 'Statistical methods' section). DISCUSSION The Longissimus dorsi muscle is often used as an indicator muscle for evaluating meat quality in pig carcasses. Despite the length of the Longissimus muscle, only one site of measurement is generally used. In
Variation in meat quality along the Longissimusmuscle
211
the Swedish pig-progeny testing scheme, meat colour is measured on a cross-section of M. longissimus dorsi cut at the tip of the last rib. The variation along the muscle is thus not taken into consideration. In both studies presented here, meat quality, expressed as light scattering or water-holding capacity, was, on average, best in the middle of the Longissimus muscle and poorer in both the anterior and the posterior parts. Assessing meat quality in the Longissimus muscle at the last rib thus underestimates the PSE problem in this muscle. Apart from the systematic variation along the Longissimus muscle, there was also a non-systematic variation, with the various loin muscles showing individual patterns. Due to the design of our materials, the interaction between animal and region could not be evaluated. In the literature, diverging pictures are given of the variation in meat quality along the Longissimus muscle. Similar to our results, Briskey (1969) mentioned that the Longissimus dorsi muscles could be PSE at both ends, with normal mid-portions--but also with normal ends and PSE-affected mid-portions. A maximum in drip loss around the 8th thoracic vertebra, with less drip loss at both the 5th and 1lth thoracic vertebrae and minimum drip loss at the last rib, was found by Grosse & Otto (1978). However, they also found the lowest water-holding capacity at the last rib when the filter press method was used. They explained the lack of consistency between methods by a significant interaction between muscle and site of measurement. Both Topel et al. (1966) and Linke & Heinz (1972) reported lighter colour in the anterior part of Longissimus compared with the posterior part. K. O. Honikel (pers. comm., 1984), also obtained a lower pH 1 value and higher drip loss (measured after 5 days) in the Longissimus dorsi muscle at the 3/4 last rib compared with measurements performed at the 13/14 rib. In earlier experiments, he also found a faster pH fall and higher drip loss in the lumbar region than in the centre part. In contrast to our results, Lawrie & Gatherum (1962) reported lighter colour and lower ultimate pH values in the middle of the Longissimus muscle, compared with the ends. Biedermann & Granz (1979) also found lower pill values in the mid-portion than at the ends and reported the converse pattern in only two out of seventy Longissimus muscles. Taylor & Dant (1971) recorded lower drip loss from the anterior part of the loin compared with the posterior part. A highly significant negative influence of side of shackling on M. semimembranosus was reported by Fischer & Augustini (1981). They measured pHi, rigor value, water-holding capacity, lactate, glycogen,
212
K. Lundstr6m, G. Malrnfors
ATP and R value, and found less favourable values for all parameters in the half by which the pigs had been shackled. The ultimate pH values in M. semimembranosus were not affected. No difference between the pH 1 values in the Longissimus muscle was found in their study. As a possible explanation for the faster glycolysis on the shackled side, the authors suggested the working load due to the weight of the hanging carcass. In our study, the difference in light scattering between halves was not uniform. The least difference was found at the last rib, a greater difference at the lumbar site and the greatest difference at the shoulder site. It is interesting that the greatest difference between halves was found where the highest FOP values were obtained. A possible explanation for our results might be a certain systematic variation in working load along the Longissimus muscle. Isometric, or even negative, contractions are thus performed by the Longissimus muscle, especially on the shackled side, due to the weight of the carcass. If the shackling immediately after slaughter could be modified; for instance, by unloading of the fore part, this might lead to better pig meat quality in general. Recording the water-holding capacity of pig muscle, for instance, in the pig-progeny testing scheme, is an old desideratum. Methods such as the filter press method (Grau & Hamm, 1952) or the capillary volumeter method (Hofmann, 1975; 1977; 1982) are theoretically possible, but judged to be too time consuming. Various measurements of the drip loss of the loin muscles have been performed, but usually without evaluation of the repeatability of the methods used. In this investigation, the repeatability studies included drip loss after keeping the samples in a meat container without individual cover, in a plain plastic bag and in a vacuum bag after vacuum packing. In addition, the capillary volumeter method was tried. This method was, however, considered to be too timeconsuming, and also slightly tricky due to air channels in some of the gypsum plates. The repeatability estimates for drip loss were of moderate size and also fairly equal around 0.5. The lowest estimate (0.4) was found for the vacuum bag method, in contrast to the high estimate of 0.72 found by Lundstr6m et al. (1977) after vacuum packing loin slices. In that study, however, consecutive slices were used for the evaluation. Even if an adjustment were made for the effect of site in the present study, this would deduct only that part of the variation which is systematic. The nonsystematic variation that might occur along Longissimus dorsi will thus lower the repeatability estimates when calculated as in the present investigation.
Variation in meat quality along the Longissimus muscle
213
In conclusion, we can recommend the method of keeping the samples in the meat container for one day. Despite the simplicity of this method, where temperature and humidity are standardized, the method is reasonably accurate, quick and causes no deterioration of the meat. In both the Swedish pig-progeny testing scheme and in the cutting procedure for the pig carcasses used at the University of Agricultural Sciences, the loin is cut both at the tip of the last rib and between the 5th and 6th lumbar vertebrae (Andersson, 1980) thus yielding a piece (about 650 g of pure muscle) that can be used for this drip loss measurement after defatting and deboning. The variation in meat quality along the Longissimus muscle and also between the sides should be taken into consideration when meat quality is evaluated. The negative effect of side of shackling will otherwise give a systematic error in meat quality assessment; for instance, in the progeny test. The difference between the sides due to shackling was, however, least at the tip of the last rib, where the meat quality measurements are carried out for Swedish pig-progeny testing. More research on the subject of variation within and between muscles, as well as the use of single indicator muscles, is certainly needed.
ACKN O W L E D G E M ENTS The Cooperative Slaughterhouse and Food Industry in Kristianstad is gratefully thanked for allowing us to further analyse their data. Valuable discussions with Dr Ingemar Hansson are greatly appreciated.
REFERENCES Andersson, K. (1980). Studies on crossbreeding and carcass evaluation in pigs. Dissertation, Sveriges lantbruksuniversitet, Institutionen f6r husdjursf6r/idling och sjukdomsgenetik, Rapport 46. Uppsala. Biedermann, G. & Granz, E. (1979). Zffchtungskunde, 51, 59. Briskey, E. J. (1969). In: Recent points of view on the condition and meat quality of pigsjor slaughter. (Sybesma, W., van der Wal, P. G. & Walstra, P. (Eds)). IVO, Zeist. Falconer, D. S. (1981). Introduction to quantitative genetics. Longman, London, New York. Fischer, K. & Augustini, Chr. (1981). Fleischwirts., 61, 1187.
214
K. Lundstr6m, G. Malmfors
Grau, R. & Hamm, R. (1952). Fleischwirts., 4, 295~. Grosse, F. & Otto, E. (1978). Arch. Tierzucht, Berlin, 21, 399. Hofmann, K. (1975). Fleischwirts., 55, 25. Hofmann, K. (1977). Fleischwirts., 57, 727. Hofmann, K. (1982). Fleischwirts., 62, 1604. Lawrie, R. A. & Gatherum, D. P. (1962). J. Agric. Sci., 58, 97. Linke, H. & Heinz, G. (1972). Fleischwirts., 52, 208. Lundstr6m, K., Nilsson, H. & Hansson, I. (1977). Swedish J. agric. Res., 7, 193. MacDougall, D. B. & Jones, S. J. (1981). In: Theproblem of dark cutting in beef. (Hood, D. E. &oTarrant, P. V. (Eds)), Topics in Veterinary Medicine and Animal Science 10, Martinus Nijhoff Publishers, The Hague, 328-39. SAS Institute Inc. (1982). SAS User's Guide: Statistics. SAS Institute Inc., Cary, NC. Taylor, A. A. & Dant, S. J. (1971). J. Fd Technol., 6, 131. Topel, D. G., Merkel, R. A., Mackintosh, D. L. & Hall, J. L. (1966). J. Anim. Sci., 25, 277.
Variation in Light Scattering and Water-Holding Capacity Along the Porcine Longissimus dorsi Muscle K. Lundstr6m & G. Malmfors Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences, S-750 07 Uppsala, Sweden (Received: 25 February, 1985)
SUMMARY The meat quality of M. longissimus muscle was evaluated in 851 pigs by using the Fibre Optic Probe (FOP) at three sites in the muscle and in both halves of the carcass. A systematic difference between sites was found, with the lowest light scattering (indicating the best meat quality) in the mid-part of the muscle and higher light scattering in the anterior and posterior parts. A non-systematic variation was also observed, with the opposite pattern in some animals, even though it was not frequent. A negative influence of the shackling wasfound, yielding higher FOP values in the shackled haiti Drip loss measurements in the Longissimus muscle, taken from another lOOpig carcasses, were evaluated using three methods. Drip loss, too, showed a considerable variation along the Longissimus muscle, with minimum losses around the last rib. Repeatability estimates, calculated from two non-consecutive pieces of the Longissimus dorsi muscle from each carcass, varied from 0"4 when keeping samples vacuum packed for 2 days, to 0"5 when the samples were either kept in plain plastic bags for 2 days or in a meat container with a squared inset ]br 1 or 2 days.
INTRODUCTION Interest in assessing meat quality in pig carcasses has increased during recent years. This is due to the very strict quality requirements when 203 Meat Science 0309-1740/85/$03.30 © Elsevier Applied Science Publishers Ltd, England, 1985. Printed in Great Britain
204
K. Lundstrdm, G. MalmJbrs
exporting pig meat, but also to attention the poor meat quality expressed by the consumer on the domestic market. Generally, a subjective method is used when the meat quality of the Longissimus muscle is evaluated commercially. As the muscle is treated as a primal cut, only cut surfaces at the shoulder and in the lumbar region are available for examination. If the muscle is not uniform in meat quality, there will be a certain degree of erroneous classification, as well as mistakes depending on the subjective evaluation. With new equipment now available, such as the Fibre Optic Probe, objective measurements of meat quality can be made on several muscles in a carcass and at several sites within a muscle. When pig meat quality is evaluated, some method of colour measurement is often used. Even more sophisticated methods, such as measuring the scattering coefficient, can only be regarded as indirect measures of meat quality. Consequently, there has been a wish to measure a more direct meat quality parameter such as water-holding capacity or drip loss of a muscle. Lack of reliable and speedy methods for the measurement of drip loss has prevented this variable from being included in meat quality assessment in the pig-progeny testing scheme. The twin purposes of these studies were (i) to examine the variation in light scattering along the Longissirnus muscle and (ii) to develop and test a method for measuring water-holding capacity or drip loss that would be simple, rapid and accurate.
MATERIALS A N D M E T H O D S
Variation in light scattering along the Longissimus muscle The animals used in the first study were ordinary pigs brought for slaughter at a slaughterhouse in the southeast of Sweden. Only a small sample (851) of the total number of pigs slaughtered in one week was used in this study. Several farms were represented and the mean carcass weight was 76kg (standard deviation, 4.9 kg). The animals were stunned with carbon dioxide and shackled by the left hind foot at the stunning. The carcasses remained shackled during the bleeding and scalding procedure (15-20 min in total). Evaluation of meat quality was made with the Fibre Optic Probe (TBL, Leeds, Great Britain) the day after slaughter. The Fibre Optic Probe
Variation in meat quality along the Longissimus muscle
205
determines the back-scatter of light transmitted into the meat. The recorded values are higher than normal for PSE muscles and lower for D F D (MacDougall & Jones, 1981). The recordings were made at three sites in the Longissimus muscle, on both left and right halves. The sites were: (i) between the 9th and 10th thoracic vertebrae (shoulder cut), (ii) at the tip of the last rib (mid-loin) and (iii) between the 5th and 6th lumbar vertebrae (lumbar cut). The probe was inserted between the spines into the middle of the muscle on the split carcass. The loins were divided into two quality groups--PSE and normal-depending on the Fibre Optic Probe value (FOP value). The muscles were classified as PSE when the FOP value equalled or exceeded 55 (G. Bjfirstorp, pers. comm., 1984).
Water-holding capacity The animals used in the second study were 100 Swedish Landrace × Yorkshire crossbred pigs. They came from one herd and had a mean carcass weight of 75 kg (standard deviation, 2.6 kg). The pigs were electrically stunned with a low voltage (85 V, 20 s) using a restrainer and shackled by the left leg only during bleeding (8 rain). The day after slaughter, the Longissimus dorsi muscle from the left side of the carcass was excised. Due to the method used to assess the meat content of the carcass, the Longissimus muscle was first cut at the last rib. The anterior part was divided into four similar pieces and the posterior part into two (Fig. 1). The six pieces from each carcass were weighed (mean value, 318 g, standard deviation, 56 g) and distributed among three different methods for evaluating drip loss. The two pieces used for each method were not consecutive. (1) Samples were placed with a cut surface facing down in a tray supplied with a net bottom and a squared inset (Fig. 2). The tray was then placed in a larger container, covered with an inverted container and kept at + 4 °C. The samples were reweighed after 1 and 2 days' storage. (2) Samples were kept in plain plastic bags at + 4 ° for 2 days and reweighed after dabbing with a paper napkin. (3) The samples were vacuum packed and then subjected to the same treatment as in (2) above. In addition, the Ultra-X capillary volumeter, as developed by
K. Lundstr6m, G. Malmfors
206
Drip toss, % 9 • Vacuumpacked,I 2 days ~
/
O Heatcontainer,Q 2 days ~
\
~
,
J / / ~
I
1
Jim
~
Plastic bag, ~ 2days ~ [] Meatcontainer,[] \ 1 day
I
~--
I
2
I
3
I
~
I
5
astlrib
I
6
Piece
I
"
Fig. 1. Drip loss measurements along M. longissimus dorsi. The six pieces are indicated on the Figure where the anterior cut is between the 4th and 5th thoracic vertebrae and the posterior cut is between the 5th and 6th lumbar vertebrae.
H o f m a n n (1975, 1977), was used on days 1 and 3 on the anterior muscle surface, cut at the last rib. Two recordings (30 s) were made on each muscle sample, using the same gypsum plate, one near the spine and one laterally.
Statistical methods Data were analysed with the Statistical Analysis System (SAS Institute Inc., 1982) using the General Linear Models procedure. For variation in light scattering the following model was used. Yijkl = l"t + ai + h~ + s k + (hs)j~ + eijkl
Variation in meat quality along the Longissimus muscle
207
Fig. 2. Equipment for measuring drip loss. Each side in the squared inset is 130 mm. The outside container is covered with an inserted container to prevent evaporation.
where Y~kl--- the ijklth observation; # = general mean; a i = effect of the ith animal (i = 1 , 2 , . . . ,i851);ihj = effect of t h e j t h half (j = 1,2); s k = effect of the kth site of measurement (k = 1, 2, 3); (hs)j~ = effect of the interaction between t h e j t h half and the kth site and euk t = residual r a n d o m term with variance tr e2 . The effect of animal was regarded as r a n d o m and the effects of half, site of measuring and the interaction between half and site as fixed. For water-hoMing capacity the following model was assumed to describe the data: Yo.kl = # + di + ao + Pk + b~'Cijkt + eijkt where Yukt= the ijklth observation; # = general mean; d i = effect of the ith day of slaughter (i = 1, 2 , . . . , 10); aij= effect of t h e j t h animal on the ith day of slaughter (j = 1,2, . . . , 100);pk = effect of the kth piece (k = 1, 2 . . . . ,6); b~ = linear regression on sample weight calculated within the kth piece; Xokt=weight of the ijklth sample (day 1) and e~k~= residual r a n d o m term with variance a 2e. All effects except the effect of animal were regarded as fixed. Repeatability estimates were calculated as intraclass correlations (t), with t calculated from the components of variance.
208
K. Lundstr6m, G. Malmfors
t--
O"a2 2 O"a +
2 O"e
where t7a2 is the component of variance for animal and t7e2 is the residual variance. For the various methods evaluating drip loss, the intraclass correlations were calculated between the two non-consecutive pieces of the Longissimus dorsi muscle taken from each carcass. For the capillary volumeter methods, the intraclass correlations were calculated between the two recorded values on the same muscle sample taken at the last rib. The standard errors of the intraclass correlations were calculated with the approximate formula described by Falconer (1981). RESULTS
Variation in light scattering Overall means for the F O P value from the right and left sides measured at the three sites can be seen in Table 1. The percentage of PSE musculature (FOP values > 55) is also shown. In accordance with the high mean values, the percentage of PSE was highest in the shoulder site at both halves. The lower meat quality in the left half at all sites (side of shackling) is of particular interest. Besides the significant effects of carcass half (P < 0.001) and site (P < 0.001) in the analysis of variance, the interaction between half and site TABLE 1 Overall Means of F O P Values for the Right and Left Sides of M. longissimus dorsi Measured at the Shoulder, Mid-loin and Ham Sites (n =851). The Percentage of PSE Muscles with F O P Values > 55 is Also Given Region Shoulder
Mid-loin
Ham
Right side Mean (SD) PSE (~o)
47.6 (16.6) 39.5 (14.5) 42.6 (14.1) 19.9 11.6 12.2
Left side ~ Mean (SD) PSE (~o)
50.6 (18.1) 40.4 (15-3) 43.8 (13.4) 23.6 12.1 13.2
a Side of shackling.
Variation in meat quality along the Longissimus muscle
209
TABLE 2 Uniformity of the Longissimus Muscle in the Left and Right Sides of the Carcass. The Sites are Shown in the Order Shoulder/Mid-loin/Ham a Left side
Right side
Per cent (Number) Per cent (Number) Normal/normal/normal PSE/normal/normal PSE/PSE/PSE ?SE/PSE/normal PSE/normal/PSE Normal/normal/PSE Normal/PSE/normal Normal/PSE/PSE
72.4 8.7 6.2 5.1 3.6 3.2 0.7 0.1
(616) (74) (53) (43) (31) (27) (6) (1)
74.8 7.2 5.4 4.6 2.7 3.6 1.2 0.5
(637) (61) (46) (39) (23) ( 31) (10) (4)
a The musculature was considered to be PSE when the FOP value was
>55. was also significant (P < 0.01). The shackled left half, with its higher average F O P values, showed a greater difference between F O P values at the three sites studied. The shackled half was thus less uniform in meat quality and this may explain the significant interaction between half and site. The differences in F O P values between the right and left halves were highly significant for the shoulder (P < 0.001) and lumbar sites (P < 0.01) and not far from significant for the mid-loin (P = 0.067). For further evaluation of the variation in light scattering in the Longissimus muscle, the degree of uniformity was examined. The threshold value of 55 was used to differentiate between normal and PSE musculature. As can be seen in Table 2, most Longissimus muscles are homogeneously normal in all sites (72.4 ~o in the left half and 74-8 ~ in the right half). There are, however, more loin muscles that are PSE only in the shoulder site than in all sites. The combinations PSE/PSE/normal, PSE/normal/PSE or normal/normal/PSE are approximately equally frequent, while the remaining combinations are very rare.
Water-holding capacity Mean values of drip loss of M. longissimus dorsi measured at different sites are presented in Fig. 1. Drip loss was greatest when the vacuum bag method was used (7.4 %) and least when keeping the samples in the meat
K. Lundstr6m, G. Malmfors
210
TABLE 3 Repeatedly of Water-Holding Capacity of M. longissimus dorsi Estimated as Intraclass Correlations ( + Standard Errors)
Method
Intraclass correlation
Level of significance Piece~ Regression on sample weight
Meat container 1 day 2 days Plastic bag, 2 days Vacuum bag, 2 days Capillary volumeter a Day 1 Day 3
0.50 0.51 0.52 0.39
+ + + +
0.08 0.08 0.08 0.09
*** *** *** ***
0.64 + 0.06 0.43 + 0-08
*** *
NS NS NS
w
a For the capillary volumeter, the difference between points of recording is tested (dorsally versus laterally). Levels of significance: NS = not significant (P > 0.05); * P < 0'05; • ** P_< 0.001.
container for one day (3.8 ~o)- A highly significant difference between the various sites in the Longissimus muscle was found for all methods (Table 3). This can also be seen in Fig. 1, where it is clearly evident that the drip loss is least in the middle of the muscle piece No. 3. The initial sample weight influenced the drip loss when evaluated by the plastic bag method. A significant interaction was found between sample site and the regression on sample weight, and the slope of the regression varied along the Longissimus muscle. The repeatability estimates, calculated as intraclass correlations, ranged from 0-39 for the vacuum bag method to 0.64 for the capillary volumeter method, one day after slaughter (Table 3). As the repeatabilities for drip loss and for the capillary volumeter method have different meanings, they cannot be compared (see 'Statistical methods' section). DISCUSSION The Longissimus dorsi muscle is often used as an indicator muscle for evaluating meat quality in pig carcasses. Despite the length of the Longissimus muscle, only one site of measurement is generally used. In
Variation in meat quality along the Longissimusmuscle
211
the Swedish pig-progeny testing scheme, meat colour is measured on a cross-section of M. longissimus dorsi cut at the tip of the last rib. The variation along the muscle is thus not taken into consideration. In both studies presented here, meat quality, expressed as light scattering or water-holding capacity, was, on average, best in the middle of the Longissimus muscle and poorer in both the anterior and the posterior parts. Assessing meat quality in the Longissimus muscle at the last rib thus underestimates the PSE problem in this muscle. Apart from the systematic variation along the Longissimus muscle, there was also a non-systematic variation, with the various loin muscles showing individual patterns. Due to the design of our materials, the interaction between animal and region could not be evaluated. In the literature, diverging pictures are given of the variation in meat quality along the Longissimus muscle. Similar to our results, Briskey (1969) mentioned that the Longissimus dorsi muscles could be PSE at both ends, with normal mid-portions--but also with normal ends and PSE-affected mid-portions. A maximum in drip loss around the 8th thoracic vertebra, with less drip loss at both the 5th and 1lth thoracic vertebrae and minimum drip loss at the last rib, was found by Grosse & Otto (1978). However, they also found the lowest water-holding capacity at the last rib when the filter press method was used. They explained the lack of consistency between methods by a significant interaction between muscle and site of measurement. Both Topel et al. (1966) and Linke & Heinz (1972) reported lighter colour in the anterior part of Longissimus compared with the posterior part. K. O. Honikel (pers. comm., 1984), also obtained a lower pH 1 value and higher drip loss (measured after 5 days) in the Longissimus dorsi muscle at the 3/4 last rib compared with measurements performed at the 13/14 rib. In earlier experiments, he also found a faster pH fall and higher drip loss in the lumbar region than in the centre part. In contrast to our results, Lawrie & Gatherum (1962) reported lighter colour and lower ultimate pH values in the middle of the Longissimus muscle, compared with the ends. Biedermann & Granz (1979) also found lower pill values in the mid-portion than at the ends and reported the converse pattern in only two out of seventy Longissimus muscles. Taylor & Dant (1971) recorded lower drip loss from the anterior part of the loin compared with the posterior part. A highly significant negative influence of side of shackling on M. semimembranosus was reported by Fischer & Augustini (1981). They measured pHi, rigor value, water-holding capacity, lactate, glycogen,
212
K. Lundstr6m, G. Malrnfors
ATP and R value, and found less favourable values for all parameters in the half by which the pigs had been shackled. The ultimate pH values in M. semimembranosus were not affected. No difference between the pH 1 values in the Longissimus muscle was found in their study. As a possible explanation for the faster glycolysis on the shackled side, the authors suggested the working load due to the weight of the hanging carcass. In our study, the difference in light scattering between halves was not uniform. The least difference was found at the last rib, a greater difference at the lumbar site and the greatest difference at the shoulder site. It is interesting that the greatest difference between halves was found where the highest FOP values were obtained. A possible explanation for our results might be a certain systematic variation in working load along the Longissimus muscle. Isometric, or even negative, contractions are thus performed by the Longissimus muscle, especially on the shackled side, due to the weight of the carcass. If the shackling immediately after slaughter could be modified; for instance, by unloading of the fore part, this might lead to better pig meat quality in general. Recording the water-holding capacity of pig muscle, for instance, in the pig-progeny testing scheme, is an old desideratum. Methods such as the filter press method (Grau & Hamm, 1952) or the capillary volumeter method (Hofmann, 1975; 1977; 1982) are theoretically possible, but judged to be too time consuming. Various measurements of the drip loss of the loin muscles have been performed, but usually without evaluation of the repeatability of the methods used. In this investigation, the repeatability studies included drip loss after keeping the samples in a meat container without individual cover, in a plain plastic bag and in a vacuum bag after vacuum packing. In addition, the capillary volumeter method was tried. This method was, however, considered to be too timeconsuming, and also slightly tricky due to air channels in some of the gypsum plates. The repeatability estimates for drip loss were of moderate size and also fairly equal around 0.5. The lowest estimate (0.4) was found for the vacuum bag method, in contrast to the high estimate of 0.72 found by Lundstr6m et al. (1977) after vacuum packing loin slices. In that study, however, consecutive slices were used for the evaluation. Even if an adjustment were made for the effect of site in the present study, this would deduct only that part of the variation which is systematic. The nonsystematic variation that might occur along Longissimus dorsi will thus lower the repeatability estimates when calculated as in the present investigation.
Variation in meat quality along the Longissimus muscle
213
In conclusion, we can recommend the method of keeping the samples in the meat container for one day. Despite the simplicity of this method, where temperature and humidity are standardized, the method is reasonably accurate, quick and causes no deterioration of the meat. In both the Swedish pig-progeny testing scheme and in the cutting procedure for the pig carcasses used at the University of Agricultural Sciences, the loin is cut both at the tip of the last rib and between the 5th and 6th lumbar vertebrae (Andersson, 1980) thus yielding a piece (about 650 g of pure muscle) that can be used for this drip loss measurement after defatting and deboning. The variation in meat quality along the Longissimus muscle and also between the sides should be taken into consideration when meat quality is evaluated. The negative effect of side of shackling will otherwise give a systematic error in meat quality assessment; for instance, in the progeny test. The difference between the sides due to shackling was, however, least at the tip of the last rib, where the meat quality measurements are carried out for Swedish pig-progeny testing. More research on the subject of variation within and between muscles, as well as the use of single indicator muscles, is certainly needed.
ACKN O W L E D G E M ENTS The Cooperative Slaughterhouse and Food Industry in Kristianstad is gratefully thanked for allowing us to further analyse their data. Valuable discussions with Dr Ingemar Hansson are greatly appreciated.
REFERENCES Andersson, K. (1980). Studies on crossbreeding and carcass evaluation in pigs. Dissertation, Sveriges lantbruksuniversitet, Institutionen f6r husdjursf6r/idling och sjukdomsgenetik, Rapport 46. Uppsala. Biedermann, G. & Granz, E. (1979). Zffchtungskunde, 51, 59. Briskey, E. J. (1969). In: Recent points of view on the condition and meat quality of pigsjor slaughter. (Sybesma, W., van der Wal, P. G. & Walstra, P. (Eds)). IVO, Zeist. Falconer, D. S. (1981). Introduction to quantitative genetics. Longman, London, New York. Fischer, K. & Augustini, Chr. (1981). Fleischwirts., 61, 1187.
214
K. Lundstr6m, G. Malmfors
Grau, R. & Hamm, R. (1952). Fleischwirts., 4, 295~. Grosse, F. & Otto, E. (1978). Arch. Tierzucht, Berlin, 21, 399. Hofmann, K. (1975). Fleischwirts., 55, 25. Hofmann, K. (1977). Fleischwirts., 57, 727. Hofmann, K. (1982). Fleischwirts., 62, 1604. Lawrie, R. A. & Gatherum, D. P. (1962). J. Agric. Sci., 58, 97. Linke, H. & Heinz, G. (1972). Fleischwirts., 52, 208. Lundstr6m, K., Nilsson, H. & Hansson, I. (1977). Swedish J. agric. Res., 7, 193. MacDougall, D. B. & Jones, S. J. (1981). In: Theproblem of dark cutting in beef. (Hood, D. E. &oTarrant, P. V. (Eds)), Topics in Veterinary Medicine and Animal Science 10, Martinus Nijhoff Publishers, The Hague, 328-39. SAS Institute Inc. (1982). SAS User's Guide: Statistics. SAS Institute Inc., Cary, NC. Taylor, A. A. & Dant, S. J. (1971). J. Fd Technol., 6, 131. Topel, D. G., Merkel, R. A., Mackintosh, D. L. & Hall, J. L. (1966). J. Anim. Sci., 25, 277.