- Email: [email protected]
Plasma and muscle cortisol measurements as indicators of meat quality and stress in pigs
Meat Sciem'e 39(1995} 237-246 1994 Elsevier Science Limited Printed in Great Bri,ain 0309- ! 740/'¢5 ' ~07.00
ELSEVIER
Plasma and Muscle Cortiso| Measm'ements as indicators ll" of Meat Quaaty and Stress in Pigs F. D. Shaw, G. R. Trout CSIRO Division of Food Science and Technology, Meat Research Laboratory, PO Box 12 Cannon Hill, Queensland 4170, Australia
& C. P. McPhee Queensland Department of Primary Industries, Animal Research Institute, Yeerongpilly, Queensland 4105, Australia (Received 4 June 1993; revised version received 2 February 1994; accepted 7 February 1994)
ABSTRACT Post-slaughter bhmd .~'amples and mu,~'ch' sample.~' were colh'cted li'om pigs slaughtered at the completion o.f a live-an#rod peiJbrmance trial. There were two Ihws of pigs hi which the halothane alh, le (n) was segregating. The lines were a lean line selected for rapM lean growth ami an tmselected fat line. There were homozygous normal (NN), homou,gous halothane positive (nn) and heterozygous (Nn) genoOTes in both lines. Cortisol was measured in the plasma of the blood samples and in muscle juice obtained by high-speed centrifugation. Meat quality was assessed using pH, colour, fibre-optic probe, drip loss and cure yieM measurenwnts. Plasma cortisoi concentrations in the .['at line were sign([icantly (P < 0.05) greater than those in the lean line but concentrations dM not differ significantly.[br the three halothane genoO,pes. Carcasses classO~'ed as dark, firm and dl3' (DFD) had significantly (P < 0.05) greater muscle cortisol concentrations than those class(lied as normal Plasma and muscle cortisol concentrations of carcasses classified as pale, soft and exudative ( PSE) did not dif[br sign(tic'an@from those classified as normal. Correlations between muscle cortisol and meat quality attributes were generally highly significant (r = 0.31 to r = 0.51, P < 0.001) There was a highly significant correlation (r = 0.73, P < 0.0001) between plasma and muscle cortisol concentrations. 237
238
F. D. Shaw, G. R. Trout, C P. McPhee
INTRODUCTION Stressful situations may lead to elevated plasma cortisoi values and it is common practice to measure plasma cortisol concentrations when assessing the welfare aspects of different transport, lairage and stunning treatments of pigs (Laursen, 1983; Forslid & Augustinsson, 1988; Nyberg et al., 1988; Warriss et al., 1992). As pale, soft and exudative (PSE) pork and dark, firm, and dry (DFD) pork are "stress-related" conditions, cortisol measurements might be expected to be of value in investigations into these conditions. However, in one trial in which plasma cortisol values indicated a greater than normal amount of ante-mortem stress, the combined incidence of PSE and DFD was less than 3% (Gregory et al., 1987). It might be expected that there would be a closer relationship between muscle cortisol values and muscle properties than between plasma cortisol values and muscle properties. Thus, the measurement of cortisol in a liquid derived from muscle may assist in meat-quality investigations. Additionally, there may be occasions when it is possible to obtain a muscle, but not a blood, sample. For example, investigations into meat-quality or animal-welfare problems may not be requested until after all animals involved have been slaughtered. Furthermore, the collection of blood samples at abattoirs during exsanguination can present practical problems. In many abattoirs, sticking takes place in confined areas and post-stunning convulsive movements by the pigs may make collection of a post-slaughter blood sample difficult. Additionally, there may be problems in subsequently relating a specific blood sample to a specific carcass. For these reasons, it was decided to measure cortisol levels in muscle juice obtained after high-speed centrifugation. MATERIALS AND METHODS Animals The blood and muscle samples were obtained from 165 pigs slaughtered at the completion of a live-animal performance trial. There were two lines of pigs in which the halothane allele (n) was segregating. These lines were a lean line selected for rapid lean growth and an unselected fat line. There were homozygous normal (NN), homozygous haiothane positive (nn) and heterozygous (Nn) genotypes in both lines (McPhee et al., 1992). Pigs were housed in separate pens and fed ad libitum from 25 kg to an average of 96 kg liveweight. They were intermingled for several hours weekly and prior to transport to the abattoir. This mixing familiarised pigs with each other so that agonistic encounters had fallen to a low level by the time they were consigned to the abattoir. On eight separate occasions, a group of 18-24 animals was transported overnight by road (600 km, 10-11 h) to an abattoir where they were slaughtered 5°-6 h after arrival. The animals were stunned electrically while in a V-restrainer and exsanguination commenced within 10 s of stunning. On all occasions animals were slaughtered between 0900 and 1100 h. Meteorological records were consulted to obtain estimates of representalive ambient temperatures at time of departure from the piggery and relevant
Cortisol measurements as indicators of meat quality in pigs
239
ambient temperatures and relative humidities at time of slaughter at the abattoir. Sample collection and meat-quality measurement
Blood samples (10 ml) were collected at exsanguination into tubes containing EDTA (12 mg). The tubes were placed on ice immediately after collection. Within 2 h of collection of the blood sample the plasma was obtained by centrifugation in a refrigerated centrifuge (5°C). Longissimus dorsi muscles were removed after overnight chilling of the carcasses. Weighed muscle samples, of 3-4 g were centrifuged at 100,000 g for 0.5 h in stainless-steel tubes (Bouton et ai., 1972). The volume of the centrifugally expressed muscle juice (predominantly sarcoplasmic fluid) was recorded. Both the expressed muscle juice and the blood plasma were frozen (-20°C) immediately after centrifugation and stored for subsequent cortisol analysis (within three months). The plasma and muscle-juice samples were assayed for total (free and protein bound) cortisol using a radio-immunoassay method (Incstar Corporation, Stillwater, Minnesota, USA). Both plasma and muscle juice cortisol concentrations were expressed as nmol/litre. Comparable results were obtained when muscle juice cortisol concentrations were calculated as nmol cortisol/g muscle tissue. Meat quality was assessed on a section of the Longissimus dorsi muscle, between the second-last rib and the 5th/6th lumbar vertebrae, which was removed from the carcasses 24 hours post mortem. The following quality attributes were measured: ultimate pH (pHu), visual colour score (1 = extremely pale, 8 = extremely dark), Minolta Chroma Meter L,a,b values (Illuminant D65, 2° observer angle, Model CR-200, Minolta Camera Co. Ltd, Osaka, Japan), FOP values (Fibre Optic Probe; Integrated Electro Optics, Barnsley, England), 48-hour drip loss and cured yield (Trout, 1992). Effect of time post mortem on muscle cortisol measurements
To determine if the storage time post mortem affected cortisol levels, on two occasions nine pork loins (three PSE, three normal and three DFD) were collected 24 h post mortem and analysed for muscle cortisol either immediately or after three or six days of vacuum-packaged storage at I°C. Statistical analyses
The data were analysed by an analysis of variance procedure for unbalanced design (Harvey, 1985). Fixed effects were slaughter day (1-8), line (lean and fat), halothane genotype (NN, Nn, nn), sex (entire male, female) and meat quality (DFD, PSE, normal). Error was composed of second and higher-order interactions and animal variation. The 165 animals used in this experiment were the progeny of I I sires and 27 dams in the lean line and 14 sires and 31 dams in the fat line. Since sires and dams were selected on the basis of their halothane genotypes their effects could not be considered as strictly random. The carcasses were subdivided into three groups; PSE (Minolta L values > 49.7 and pHu < 6.0), DFD (pHu > 6.0) and normal (all carcasses not in the previous two groups) (Trout, 1992).
240
F. D. Shaw, G. R. Trout, C P. McPhee
TABLE 1 Least Square Means (+SE) of Plasma and Muscle Cortisol Concentrations (nmol/litre) for Carcasses Classified as PSE, Normal or DFD
Plasma Muscle n
PSE
Normal
DFD
459 (+41) 81 (+13)"h 22
389 (+19) 69 (+6)b 85
420 (-+26) 100 (_+7)" 58
"'bRow means with the same, or no, superscript are not significantly different (P > 0.05). RESULTS The data in Table I are the least square means of plasma and muscle cortisol levels of carcasses classified as normal, PSE or D F D . There were no significant differences between the three categories for plasma cortisol (P > 0.05) but, for muscle cortisol, D F D carcasses had a higher cortisol level than normal carcasses (P < 0.05). The correlations between plasma and muscle cortisol concentrations and meat-quality attributes were quite variable (Table 2). Muscle cortisol levels were more highly correlated with meat-quality attributes than were plasma cortisol levels. In Table 3 the least square means of plasma cortisol levels for the three halothane genotypes in the two lines are given. Mean plasma cortisol levels were higher (P < 0.05) in the fat line than in the lean line but there were no statistically significant differences between the three genotype groups. For muscle cortisol, there were no statistically significant differences for either line or genotype. TABLE 2 Correlations between Plasma and Muscle Cortisol Concentrations and Meat Quality Attributes tbr (1) All Samples (2) Normal and PSE ~amples (3) Normal and DFD Samples Only
Measurement
Correlation co¢:~cient All samples Pktsma cortisoi
Plasma cortisoi Tristimulus L FOP pHu Colour score Drip loss Cure yield
--0.21'* -0.23** 0.25** 0.28*** -0.20* 0.05
Muscle cortisol
0.73*** -0.42*** -0.33*** 0.51"** 0.47*** -0.35*** 0.31"**
* Significance level P < 0.05. ** Significance level P < 0.01. *** Significance level P < 0.00. l
Normal & PSE only Plasma cortisol
-0.02 -0.08 0.07 0.22* -0.15 -0.11
Normal & DFD only
Muscle cortisol
Plasma cortisol
Muscle cortisol
0.66*** -0.08 0.00 0.12 0.30* -0.20* 0.01
-0.72*** -0.19' -0.47*** -0.21' -0.41'** 0.26** 0.53*** 0.28*** 0.49*** -0.16 -0.35*** 0.04 0.36***
241
Cortisol measurements as indicators o f meat quality in pigs
TABLE 3
Least Square Means (±SE) of Plasma Cortisol Concentrations (nmol/litre) for the Three Halothane Genotypes in the Two Lines NN
Nn
nn
Mean
Lean line
393 (±23)
350 (±24)
394 (+59)
Fat line
474 (±25)
433 (+22)
425 (+33)
379 (+22)" n=76 444 (±15)0 n=89
Mean
434 (±17)" n = 71
392 (±16)" n = 68
410 (±33)" n = 26
"'°Marginal means followed by difli~rent subscripts are significantly different (P < 0.05).
TABLE 4
Least Square Means (±SE) of Muscle Cortisol Concentrations (nmol/litre) for the Two Sexes in the Two Lines
Lean line Fat line Mean
Malt,
Female
Mean
88 (±9)''h 83 (±7)''h 86 (±5)
74 (±8)0 102 (+9)" 88 (+6)
81 (±6) 92 (±6)
Cell means followed by different subscripts are significantly different (P < 0.05). In Table 4 the least square means of muscle cortisol levels for the two sexes in the two lines are given. Muscle cortisol levels did not differ between the sexes and the apparently higher level in the fat line than in the lean line did not reach significance (P < 0.1). A significant line x sex interaction (P < 0.05) resulted from an unusually high level in females in the fat line. There was considerable variation between slaughter days in the levels of both plasma and muscle cortisol. The least square means of the cortisol concentrations for the different days are shown in Fig. 1. The reason for this large variation is ~ Muscle ~i
= 5OO
~
140
- ~2o
Plasma
O
100 '~
8O ~
u
= 300
0
60 u
~ 200 2O ~
~ 100 I
2
3
5
4
6
7
8
Day
Fig. 1. Least square means of muscle and plasma cortisol concentrations (nmol/litre) on different slaughter days.
242
F. D. Shaw, G. R. Trout, C P. McPhee
unclear. It was thought that the difference may have been due to the seasonal variation in temperature and humidity. The ambient temperature at the piggery at about the time of departure on the different days ranged from 18°C to 28°C while ambient temperature at the abattoir at the time of slaughter on the different days ranged from 17°C to 26°C. Relative humidity at the abattoir for the different slaughter days varied between 30% and 88%. When the data were analysed, however, there were no significant correlations between either of the temperature and humidity measurements and plasma or muscle cortisol concentrations. The time post mortem (up to six days) had no effect on the muscle cortisol. The average muscle cortisoi concentrations at 1, 3 and 6 days post mortem were 41.9, 39.1 and 39.0 nmol/litre respectively (standard error of determination of 28.3). As with the main experiment reported in this paper, muscle juice from DFD muscle had higher cortisol concentration (64.3 nmol/litre) than muscle juice from the PSE (30.6 nmol/litre) or normal (25.2 nmol/litre) muscle (P < 0.05). DISCUSSION Cortisol measurements
One advantage in measuring cortisol is that the act of stunning and sticking has little, if any, effect on this plasma constituent (Shaw & Tume, 1992) and thus blood samples collected at exsanguination correctly indicate the concentration immediately prior to slaughter. This is not the case with the cateeholamines as dramatic elevations (up to 60-fold) of these plasma constituents occur following electrical stunning (Troeger & Woltersdorf, 1988). One factor that may need to be taken it~to consideration when interpreting plasma cortisoi values in pigs is the circadian variation in cortisol secretion (Spencer, 1979). However, in these trials this factor should be of minimal importance as all animals were slaughtered at approximately the same time of day (0900-1100). There is evidence that changes in cortisol concentration due to circadian variation are minimal at these times (Spencer, 1979). Cortisol and meat quality
Published relationships between blood cortisol levels anti mcat quality are not conclusive. Warriss & Brown (1985) reported a relationship between plasma cortisoi level and DFD (ultimate P > 5.9). The mean cortisol values for three groups with DFD levels of 5, 23 and 45'7,, were 160, 248 and 450 nmol/litre, respectively. In the study reported in the present paper, elevated cortisol levels occurred in DFD carcasses (Table l). This difference was significant for muscle cortisol (P < 0.05) but not for plasma cortisol (P> 0.05). In contrast, Moss & Robb (1978) found that holding animals in lairage overnight increased the percentage of DFD yet resulted in a significant reduction in plasma cortisol level. However, in the study by Moss & Robb the mean cortisol values for the two groups, 181 nmol/litre for the normal slaughter group and 140 nmol/litre for the overnight lairage group, although significantly different, were both relatively low. in a study reported by Gregory et al. (1987), the mean plasma cortisol level was high (364 (_+ 24) nmol/litre) but the incidence of DFD was low (3"/0). On the
Cortisoi measurements as indicators o f meat quality in pigs
243
basis of the cortisol values, and values for other plasma constituents, these researchers concluded that there was a greater-than-normal amount of antemortem stress. In spite of this fact, only one out of 44 carcasses was DFD. However, a possible explanation for this contradiction was that these researchers defined D F D carcasses as those in which the ultimate pH of the longissimus dorsi muscle was greater than 6.2. This is a rather stricter definition of D F D than that commonly used. Conceivably, if the more commonly used cut-off point for D F D of 5.9 or 6.0 were used it is possible that more carcasses would have been classified as DFD. These authors did comment that, based on reflectance measurements, the colour of many of the samples was bordering on DFD. PSE is regarded as a 'stress-related' condition; however, in this study, plasma or muscle cortisol concentrations it: samples from PSE carcasses did not differ significantly from those in samples from normal carcasses t,'r,r,l,~ ,,,~,.,, 1). This poi~t is also illustrated in Table 2. When correlations between cortisol levels and meatquality attriLutes were carried out using all the data, the correlations were significant (P < 0.05) in most cases. If, however, only the normal and PSE samples were considered, most correlations were low and non-significant (P > 0.05). Other studies have shown that biochemical indicators of stress are related to the incidence of PSE. For example, Simmons (1989) used plasma creatine phosphokinase (CPK) as an indicator of stress and found a relationship between low, medium and high CPK levels and meat quality parameters. High levels of CPK were found in the animals which later exhibited PSE pork. Warriss et al. (1992), state that cortisoi mainly measures psychological stress while CPK measures physical stress. Combining the results of our study, in which cortisol concentrations (psychological stress) were not elevated in the PSE group, with those of Simmons (1989), in which CPK concentrations (physical stress) were elevated in the PSE group, could thus lead to the conclusion that PSE occurs in response to a physical stress rather than a psychological one. Cortisol and stress
The interpretation of plasma cortisol values is the subject of much debate. Herd (1989) states that one should avoid the tautological identification of stressors as being procedures or situations associated with an elevation in cortisol. R ushen (1986) acknowledges that plasma cortisol levels do rise when animals are subjetted to unpleasant treatments but questions whether the unpleasantness of the treatment per se is responsible for the rise and whether the extent of the rise can be used to index the unpleasantness of the experience. Animal-welfare and ethical considerations prevent the conduct of experiments in which one deliberately increases the magnitude of a stressor with a view to measuring the resultant increase in plasma cortisol concentration. However, there are numerous experiments which provide evidence that a pig's response to a stressor may include an increase in plasma cortisol concentration. Treatments that have produced significant increases (above baseline or control values) in plasma cortisol include haemorrhage (O'Benar et al., 1987); treadmill exercise (Heinze & Mitchell, 1989; Zhang et al., 1989); confinement (Becker et al., 1985) and heat stress (Klemcke et al., i989). Thus, there is evidence that a wide range of stressors produce significant elevations in cortisol. Becker et al. (1985) specifically state that their data substantiate the use of cortisol as an indicator of stress in swine.
244
F. D. Shaw, G. R. Trout. C P. McPhee
Reported values for plasma cortisol concentrations in blood samples collected at exsanguination vary widely. At the lower end are mean values of 127 (+ 27) (Forslid & Augustinsson, 1988) and 150 (+ 21) (Shaw & Tume, 1990) while at the upper end are values of 343 (+ 30) (Shaw & Tume, 1990) and 364 (+ 24) nmol/litre (Gregory et al., 1987). Even the latter values are considerably less than a mean value of 684 (+ 42) nmol/litre reported in blood samples from live pigs after an ACTH challenge (Zhang et al. 1991). The mean value for all pigs in the present experiment was 404 (+ 10) nmol/litre while the maximum mean value recorded on any one slaughter day during these trials was 576 (+ 34) nmoi/litre. Thus, the plasma cortisol concentrations are greater than most values reported in the literature. By far the dominant source of variation in cortisol levels in the present study was due to the variation between the different slaughter days. This was despite apparently generally uniform procedures for the handling of pigs during transport and at the abattoir. Ambient temperature, or humidity, did not appear to contribute significantly to this variation as there were no significant correlations between cortisol concentrations and temperature or humidity. Cortiso| and carcass composition
The finding of a greater mean cortisol value in the fat line, in comparison to that in the lean line, is not entirely unexpected. There is some direct evidence, and considerable indirect evidence, of relationships between fat content and plasma cortisol concentrations. For example, Wynn (1991) showed that immunisation of mildly stressed ewes against cortisol resulted in decreased fat content while in another experiment it was found that undisturbed animals suffered a decrease in carcass fatness, while animals subjected to low-level stress had more carcass fat. Moss (1992) ~tates that increased corticosteroid production can result in catabolism ot" muscle tissue and increased deposition of fat. In man and animals, obesity occurs in conjunctiorl with the hyperadrenal state of Cushing's syndrome. In cattle, bulls and .~teers generally have lower plasma cortisol concentrations, and less fat, than cows and heifers (Mitchell et ai., 1988). Cortisol and genotypes
in the present study, mean muscle and plasma cortisol levels for the three genotypes did not differ significantly. Nyberg et al. (1988) found that in some cases there was a ~;ignificant interaction between halothane genotype and treatment. For example, cortisol concentrations in transported nn animals were lower than in untransported nn animals; whereas transported NN animals had higher cortisol concentrations than untransported NN pigs. Consequently, Nyberg et al. (1988) concluded that there was no simple relationship between plasma cortisol and halothane genotype. Apparently conflicting results were reported by Mitchell & Heffron (19g.l) and Marp!e et al. (1974) with the former reporting stress-susceptible pigs (h~dotlaane-tested) having lower resting cortisol concentrations while the latter group reported higher cortisol concentrations in pigs from herds bred to be stress-susceptible. The time of collection of a blood sample in relation to the time of a specific stress may be an important factor in determining plasma cortisol concentration in stress-susceptible animals. Marple & Cassens (1973) reported that stress-susceptible swine had an increased metabolic
Cortisol measurements as indicators of meat quality in pigs
245
clearance of cortisol. Labelled cortisol was removed from the blood of stresssusceptible pigs at approximately five times the rate from that of normal pigs. Thus, immediately after a specific stress, stress-susceptible animals may have a marked elevation in plasma cortisol which returns to resting level within a comparatively short time. However, stress-susceptible pigs have an increased cortisol production rate in comparison to normal swine (Marple & Cassens, 1973).
Relationship between plasma and muscle cortisol The correlation coefficient of 0.727 (P <0.0001) indicates a close relationship between plasma and muscle cortisol concentrations. Thus, if practical considerations prevent the collection of a blood sample, the measurement of muscle cortisol can be used to provide a guide to the stress status, at time of slaughter, of a group of pigs.
ACKNOWLEDGEMENTS This work was supported by the Pig Research and Development Corporation. The competent technical assistance provided by Ms S. Dyson and Mr T. Larsen is gratefully acknowledged. The breeding and growing of the genetic lines was carried out by Mr L. Daniels at the Queensland Department of Primary Industries (QDPI) research piggery at Biloela. Mr H. Kramer of QDPI identified the halothane genotypes.
REFERENCES Becker, B. A., Nienaber, J. A., Christenson, R. K., Manak, R. C., DeShazer, J A. & Hahn, G. L. (1985). Am. J. Vet. Res., 46, 1034. Bouton, P. E., Harris, P. V. & Shorthose, W. R. (1972). d. Food Sci., 37, 351. Forslid, A. & Augustinsson, O. (1988). Acta. Physiol. Seand., 132, 223. Gregory, N. G., Moss, B. W. & Leeson, R. H. (1987). Vet. Rec., 121,517. Harvey, W. R. (1985). User's guide to L S M L / M W Mixed Model Least-Squares and Max#hum Likelihood Computer Program. Ohio State Univ., Colombus, OH, 46 pp. (Mimeo). Heinze, P. H. & Mitchell, G. (1989). Vet. Rec., 124, 163. Herd, R. M. (1989). Aust. Vet. J., 66, 341. Klemcke, H. G., Nienaber, J. A. & Hahn, G. L. (1989). Proc. Sot'. Exptl Biol. Med., 190, 42. Laursen, A. M. (1983). In Stunning of animals Jbr slaughter, ed. G. Eikelenboom. Martinus Nijhoff, Boston, p. 70. Marple, D. N. & Cassen, R. G. (1973). J. Anita. Sci., 36, 1139. Marple, D. N., Cassens, R. G., Topel, D. G. & Christian, L. L. (1974). J. Anita. St,i., 38, 1224. McPhee, C. P., Daniels, L. J., Kramer, H. L., Noble, J. W. & Trout, G. R. (1992). Proc. Aust. Assoc'. Anita. Breed. Genet., 10, 76. Mitchell, G. & Heffron, J. J. A. (1981). J. South Afr. Vet. Assoc., 52, 109. Mitchell, G., Hattingh, J. & Ganhao, M. (1988). Vet. Rec., 123, 201. Moss, B. W. (1992). In The Chemistry of Muscle Based Foods. Royal Society of Chem. istry, p. 69.
246
F. D. Shaw. G. R. Trout. C. P. McPhee
Moss, B. W. & Robb, J. D. (1978). J. Sci. Food Agric., 29, 689. Nyberg, L., Lundstrom, K., Edfors-Lilja, I. & Rundgren, M. (1988). J. Anim. Sci., 66, 1201. O'Benar, J. D., Hannon, J. P., Peterson, J. L. & Bossone, C. A. (1987). Am. J. Physiol., 252, R953. Rushen, J. (1986). Aust. Vet. J., 63, 359. Shaw, F. D. & Tume, R. K. (1990). Proc. llth Congr. Inter. Pig. Vet. Soc., Lausanne, p. 427. Shaw, F. D. & Tume, R. K. (1992). Meat Sci., 32, 311. Simmons, N. J. (1989). MSc Dissertation, University of Bristol. Spencer, G. S. G. (1979). Horm. Metab. Res., 11,586. Troeger, K. & Woltersdorf, W. (1988). Fieischwirtschaft, 68, 249. Trout, G. R. (1992). Proc. 38th. Inter. Congr. Meat Sci. Technol., France. p. 983. Warriss, P. D. & Brown, S. N. (1985). J. Sci. Food Agric., 36, 87. Warriss, P. D., Brown, S. N., Edwards, J. E., Anil, M. H. & Fordham, D. P. (1992). Vet. Rec.~ 131, 194. Wynn, P. (1991). In Project Guide 1991-92., Meat Research Corporation, Sydney, NSW, p. 120. Zhang, S. H., Hennessy, D. P. & Cranwell, P. D. (1989). In Manipulating Pig Production !!, Australasian Pig Science Association, Werribee, Australia, p. 212. Zhang, S. H., Hennessy, D. P., McCauley, I. & Cranweli, P. D. (1991). In Manipulating Pig Production IH, Australasian Pig Science Association, Attwood, Australia, p. 65.
ELSEVIER
Plasma and Muscle Cortiso| Measm'ements as indicators ll" of Meat Quaaty and Stress in Pigs F. D. Shaw, G. R. Trout CSIRO Division of Food Science and Technology, Meat Research Laboratory, PO Box 12 Cannon Hill, Queensland 4170, Australia
& C. P. McPhee Queensland Department of Primary Industries, Animal Research Institute, Yeerongpilly, Queensland 4105, Australia (Received 4 June 1993; revised version received 2 February 1994; accepted 7 February 1994)
ABSTRACT Post-slaughter bhmd .~'amples and mu,~'ch' sample.~' were colh'cted li'om pigs slaughtered at the completion o.f a live-an#rod peiJbrmance trial. There were two Ihws of pigs hi which the halothane alh, le (n) was segregating. The lines were a lean line selected for rapM lean growth ami an tmselected fat line. There were homozygous normal (NN), homou,gous halothane positive (nn) and heterozygous (Nn) genoOTes in both lines. Cortisol was measured in the plasma of the blood samples and in muscle juice obtained by high-speed centrifugation. Meat quality was assessed using pH, colour, fibre-optic probe, drip loss and cure yieM measurenwnts. Plasma cortisoi concentrations in the .['at line were sign([icantly (P < 0.05) greater than those in the lean line but concentrations dM not differ significantly.[br the three halothane genoO,pes. Carcasses classO~'ed as dark, firm and dl3' (DFD) had significantly (P < 0.05) greater muscle cortisol concentrations than those class(lied as normal Plasma and muscle cortisol concentrations of carcasses classified as pale, soft and exudative ( PSE) did not dif[br sign(tic'an@from those classified as normal. Correlations between muscle cortisol and meat quality attributes were generally highly significant (r = 0.31 to r = 0.51, P < 0.001) There was a highly significant correlation (r = 0.73, P < 0.0001) between plasma and muscle cortisol concentrations. 237
238
F. D. Shaw, G. R. Trout, C P. McPhee
INTRODUCTION Stressful situations may lead to elevated plasma cortisoi values and it is common practice to measure plasma cortisol concentrations when assessing the welfare aspects of different transport, lairage and stunning treatments of pigs (Laursen, 1983; Forslid & Augustinsson, 1988; Nyberg et al., 1988; Warriss et al., 1992). As pale, soft and exudative (PSE) pork and dark, firm, and dry (DFD) pork are "stress-related" conditions, cortisol measurements might be expected to be of value in investigations into these conditions. However, in one trial in which plasma cortisol values indicated a greater than normal amount of ante-mortem stress, the combined incidence of PSE and DFD was less than 3% (Gregory et al., 1987). It might be expected that there would be a closer relationship between muscle cortisol values and muscle properties than between plasma cortisol values and muscle properties. Thus, the measurement of cortisol in a liquid derived from muscle may assist in meat-quality investigations. Additionally, there may be occasions when it is possible to obtain a muscle, but not a blood, sample. For example, investigations into meat-quality or animal-welfare problems may not be requested until after all animals involved have been slaughtered. Furthermore, the collection of blood samples at abattoirs during exsanguination can present practical problems. In many abattoirs, sticking takes place in confined areas and post-stunning convulsive movements by the pigs may make collection of a post-slaughter blood sample difficult. Additionally, there may be problems in subsequently relating a specific blood sample to a specific carcass. For these reasons, it was decided to measure cortisol levels in muscle juice obtained after high-speed centrifugation. MATERIALS AND METHODS Animals The blood and muscle samples were obtained from 165 pigs slaughtered at the completion of a live-animal performance trial. There were two lines of pigs in which the halothane allele (n) was segregating. These lines were a lean line selected for rapid lean growth and an unselected fat line. There were homozygous normal (NN), homozygous haiothane positive (nn) and heterozygous (Nn) genotypes in both lines (McPhee et al., 1992). Pigs were housed in separate pens and fed ad libitum from 25 kg to an average of 96 kg liveweight. They were intermingled for several hours weekly and prior to transport to the abattoir. This mixing familiarised pigs with each other so that agonistic encounters had fallen to a low level by the time they were consigned to the abattoir. On eight separate occasions, a group of 18-24 animals was transported overnight by road (600 km, 10-11 h) to an abattoir where they were slaughtered 5°-6 h after arrival. The animals were stunned electrically while in a V-restrainer and exsanguination commenced within 10 s of stunning. On all occasions animals were slaughtered between 0900 and 1100 h. Meteorological records were consulted to obtain estimates of representalive ambient temperatures at time of departure from the piggery and relevant
Cortisol measurements as indicators of meat quality in pigs
239
ambient temperatures and relative humidities at time of slaughter at the abattoir. Sample collection and meat-quality measurement
Blood samples (10 ml) were collected at exsanguination into tubes containing EDTA (12 mg). The tubes were placed on ice immediately after collection. Within 2 h of collection of the blood sample the plasma was obtained by centrifugation in a refrigerated centrifuge (5°C). Longissimus dorsi muscles were removed after overnight chilling of the carcasses. Weighed muscle samples, of 3-4 g were centrifuged at 100,000 g for 0.5 h in stainless-steel tubes (Bouton et ai., 1972). The volume of the centrifugally expressed muscle juice (predominantly sarcoplasmic fluid) was recorded. Both the expressed muscle juice and the blood plasma were frozen (-20°C) immediately after centrifugation and stored for subsequent cortisol analysis (within three months). The plasma and muscle-juice samples were assayed for total (free and protein bound) cortisol using a radio-immunoassay method (Incstar Corporation, Stillwater, Minnesota, USA). Both plasma and muscle juice cortisol concentrations were expressed as nmol/litre. Comparable results were obtained when muscle juice cortisol concentrations were calculated as nmol cortisol/g muscle tissue. Meat quality was assessed on a section of the Longissimus dorsi muscle, between the second-last rib and the 5th/6th lumbar vertebrae, which was removed from the carcasses 24 hours post mortem. The following quality attributes were measured: ultimate pH (pHu), visual colour score (1 = extremely pale, 8 = extremely dark), Minolta Chroma Meter L,a,b values (Illuminant D65, 2° observer angle, Model CR-200, Minolta Camera Co. Ltd, Osaka, Japan), FOP values (Fibre Optic Probe; Integrated Electro Optics, Barnsley, England), 48-hour drip loss and cured yield (Trout, 1992). Effect of time post mortem on muscle cortisol measurements
To determine if the storage time post mortem affected cortisol levels, on two occasions nine pork loins (three PSE, three normal and three DFD) were collected 24 h post mortem and analysed for muscle cortisol either immediately or after three or six days of vacuum-packaged storage at I°C. Statistical analyses
The data were analysed by an analysis of variance procedure for unbalanced design (Harvey, 1985). Fixed effects were slaughter day (1-8), line (lean and fat), halothane genotype (NN, Nn, nn), sex (entire male, female) and meat quality (DFD, PSE, normal). Error was composed of second and higher-order interactions and animal variation. The 165 animals used in this experiment were the progeny of I I sires and 27 dams in the lean line and 14 sires and 31 dams in the fat line. Since sires and dams were selected on the basis of their halothane genotypes their effects could not be considered as strictly random. The carcasses were subdivided into three groups; PSE (Minolta L values > 49.7 and pHu < 6.0), DFD (pHu > 6.0) and normal (all carcasses not in the previous two groups) (Trout, 1992).
240
F. D. Shaw, G. R. Trout, C P. McPhee
TABLE 1 Least Square Means (+SE) of Plasma and Muscle Cortisol Concentrations (nmol/litre) for Carcasses Classified as PSE, Normal or DFD
Plasma Muscle n
PSE
Normal
DFD
459 (+41) 81 (+13)"h 22
389 (+19) 69 (+6)b 85
420 (-+26) 100 (_+7)" 58
"'bRow means with the same, or no, superscript are not significantly different (P > 0.05). RESULTS The data in Table I are the least square means of plasma and muscle cortisol levels of carcasses classified as normal, PSE or D F D . There were no significant differences between the three categories for plasma cortisol (P > 0.05) but, for muscle cortisol, D F D carcasses had a higher cortisol level than normal carcasses (P < 0.05). The correlations between plasma and muscle cortisol concentrations and meat-quality attributes were quite variable (Table 2). Muscle cortisol levels were more highly correlated with meat-quality attributes than were plasma cortisol levels. In Table 3 the least square means of plasma cortisol levels for the three halothane genotypes in the two lines are given. Mean plasma cortisol levels were higher (P < 0.05) in the fat line than in the lean line but there were no statistically significant differences between the three genotype groups. For muscle cortisol, there were no statistically significant differences for either line or genotype. TABLE 2 Correlations between Plasma and Muscle Cortisol Concentrations and Meat Quality Attributes tbr (1) All Samples (2) Normal and PSE ~amples (3) Normal and DFD Samples Only
Measurement
Correlation co¢:~cient All samples Pktsma cortisoi
Plasma cortisoi Tristimulus L FOP pHu Colour score Drip loss Cure yield
--0.21'* -0.23** 0.25** 0.28*** -0.20* 0.05
Muscle cortisol
0.73*** -0.42*** -0.33*** 0.51"** 0.47*** -0.35*** 0.31"**
* Significance level P < 0.05. ** Significance level P < 0.01. *** Significance level P < 0.00. l
Normal & PSE only Plasma cortisol
-0.02 -0.08 0.07 0.22* -0.15 -0.11
Normal & DFD only
Muscle cortisol
Plasma cortisol
Muscle cortisol
0.66*** -0.08 0.00 0.12 0.30* -0.20* 0.01
-0.72*** -0.19' -0.47*** -0.21' -0.41'** 0.26** 0.53*** 0.28*** 0.49*** -0.16 -0.35*** 0.04 0.36***
241
Cortisol measurements as indicators o f meat quality in pigs
TABLE 3
Least Square Means (±SE) of Plasma Cortisol Concentrations (nmol/litre) for the Three Halothane Genotypes in the Two Lines NN
Nn
nn
Mean
Lean line
393 (±23)
350 (±24)
394 (+59)
Fat line
474 (±25)
433 (+22)
425 (+33)
379 (+22)" n=76 444 (±15)0 n=89
Mean
434 (±17)" n = 71
392 (±16)" n = 68
410 (±33)" n = 26
"'°Marginal means followed by difli~rent subscripts are significantly different (P < 0.05).
TABLE 4
Least Square Means (±SE) of Muscle Cortisol Concentrations (nmol/litre) for the Two Sexes in the Two Lines
Lean line Fat line Mean
Malt,
Female
Mean
88 (±9)''h 83 (±7)''h 86 (±5)
74 (±8)0 102 (+9)" 88 (+6)
81 (±6) 92 (±6)
Cell means followed by different subscripts are significantly different (P < 0.05). In Table 4 the least square means of muscle cortisol levels for the two sexes in the two lines are given. Muscle cortisol levels did not differ between the sexes and the apparently higher level in the fat line than in the lean line did not reach significance (P < 0.1). A significant line x sex interaction (P < 0.05) resulted from an unusually high level in females in the fat line. There was considerable variation between slaughter days in the levels of both plasma and muscle cortisol. The least square means of the cortisol concentrations for the different days are shown in Fig. 1. The reason for this large variation is ~ Muscle ~i
= 5OO
~
140
- ~2o
Plasma
O
100 '~
8O ~
u
= 300
0
60 u
~ 200 2O ~
~ 100 I
2
3
5
4
6
7
8
Day
Fig. 1. Least square means of muscle and plasma cortisol concentrations (nmol/litre) on different slaughter days.
242
F. D. Shaw, G. R. Trout, C P. McPhee
unclear. It was thought that the difference may have been due to the seasonal variation in temperature and humidity. The ambient temperature at the piggery at about the time of departure on the different days ranged from 18°C to 28°C while ambient temperature at the abattoir at the time of slaughter on the different days ranged from 17°C to 26°C. Relative humidity at the abattoir for the different slaughter days varied between 30% and 88%. When the data were analysed, however, there were no significant correlations between either of the temperature and humidity measurements and plasma or muscle cortisol concentrations. The time post mortem (up to six days) had no effect on the muscle cortisol. The average muscle cortisoi concentrations at 1, 3 and 6 days post mortem were 41.9, 39.1 and 39.0 nmol/litre respectively (standard error of determination of 28.3). As with the main experiment reported in this paper, muscle juice from DFD muscle had higher cortisol concentration (64.3 nmol/litre) than muscle juice from the PSE (30.6 nmol/litre) or normal (25.2 nmol/litre) muscle (P < 0.05). DISCUSSION Cortisol measurements
One advantage in measuring cortisol is that the act of stunning and sticking has little, if any, effect on this plasma constituent (Shaw & Tume, 1992) and thus blood samples collected at exsanguination correctly indicate the concentration immediately prior to slaughter. This is not the case with the cateeholamines as dramatic elevations (up to 60-fold) of these plasma constituents occur following electrical stunning (Troeger & Woltersdorf, 1988). One factor that may need to be taken it~to consideration when interpreting plasma cortisoi values in pigs is the circadian variation in cortisol secretion (Spencer, 1979). However, in these trials this factor should be of minimal importance as all animals were slaughtered at approximately the same time of day (0900-1100). There is evidence that changes in cortisol concentration due to circadian variation are minimal at these times (Spencer, 1979). Cortisol and meat quality
Published relationships between blood cortisol levels anti mcat quality are not conclusive. Warriss & Brown (1985) reported a relationship between plasma cortisoi level and DFD (ultimate P > 5.9). The mean cortisol values for three groups with DFD levels of 5, 23 and 45'7,, were 160, 248 and 450 nmol/litre, respectively. In the study reported in the present paper, elevated cortisol levels occurred in DFD carcasses (Table l). This difference was significant for muscle cortisol (P < 0.05) but not for plasma cortisol (P> 0.05). In contrast, Moss & Robb (1978) found that holding animals in lairage overnight increased the percentage of DFD yet resulted in a significant reduction in plasma cortisol level. However, in the study by Moss & Robb the mean cortisol values for the two groups, 181 nmol/litre for the normal slaughter group and 140 nmol/litre for the overnight lairage group, although significantly different, were both relatively low. in a study reported by Gregory et al. (1987), the mean plasma cortisol level was high (364 (_+ 24) nmol/litre) but the incidence of DFD was low (3"/0). On the
Cortisoi measurements as indicators o f meat quality in pigs
243
basis of the cortisol values, and values for other plasma constituents, these researchers concluded that there was a greater-than-normal amount of antemortem stress. In spite of this fact, only one out of 44 carcasses was DFD. However, a possible explanation for this contradiction was that these researchers defined D F D carcasses as those in which the ultimate pH of the longissimus dorsi muscle was greater than 6.2. This is a rather stricter definition of D F D than that commonly used. Conceivably, if the more commonly used cut-off point for D F D of 5.9 or 6.0 were used it is possible that more carcasses would have been classified as DFD. These authors did comment that, based on reflectance measurements, the colour of many of the samples was bordering on DFD. PSE is regarded as a 'stress-related' condition; however, in this study, plasma or muscle cortisol concentrations it: samples from PSE carcasses did not differ significantly from those in samples from normal carcasses t,'r,r,l,~ ,,,~,.,, 1). This poi~t is also illustrated in Table 2. When correlations between cortisol levels and meatquality attriLutes were carried out using all the data, the correlations were significant (P < 0.05) in most cases. If, however, only the normal and PSE samples were considered, most correlations were low and non-significant (P > 0.05). Other studies have shown that biochemical indicators of stress are related to the incidence of PSE. For example, Simmons (1989) used plasma creatine phosphokinase (CPK) as an indicator of stress and found a relationship between low, medium and high CPK levels and meat quality parameters. High levels of CPK were found in the animals which later exhibited PSE pork. Warriss et al. (1992), state that cortisoi mainly measures psychological stress while CPK measures physical stress. Combining the results of our study, in which cortisol concentrations (psychological stress) were not elevated in the PSE group, with those of Simmons (1989), in which CPK concentrations (physical stress) were elevated in the PSE group, could thus lead to the conclusion that PSE occurs in response to a physical stress rather than a psychological one. Cortisol and stress
The interpretation of plasma cortisol values is the subject of much debate. Herd (1989) states that one should avoid the tautological identification of stressors as being procedures or situations associated with an elevation in cortisol. R ushen (1986) acknowledges that plasma cortisol levels do rise when animals are subjetted to unpleasant treatments but questions whether the unpleasantness of the treatment per se is responsible for the rise and whether the extent of the rise can be used to index the unpleasantness of the experience. Animal-welfare and ethical considerations prevent the conduct of experiments in which one deliberately increases the magnitude of a stressor with a view to measuring the resultant increase in plasma cortisol concentration. However, there are numerous experiments which provide evidence that a pig's response to a stressor may include an increase in plasma cortisol concentration. Treatments that have produced significant increases (above baseline or control values) in plasma cortisol include haemorrhage (O'Benar et al., 1987); treadmill exercise (Heinze & Mitchell, 1989; Zhang et al., 1989); confinement (Becker et al., 1985) and heat stress (Klemcke et al., i989). Thus, there is evidence that a wide range of stressors produce significant elevations in cortisol. Becker et al. (1985) specifically state that their data substantiate the use of cortisol as an indicator of stress in swine.
244
F. D. Shaw, G. R. Trout. C P. McPhee
Reported values for plasma cortisol concentrations in blood samples collected at exsanguination vary widely. At the lower end are mean values of 127 (+ 27) (Forslid & Augustinsson, 1988) and 150 (+ 21) (Shaw & Tume, 1990) while at the upper end are values of 343 (+ 30) (Shaw & Tume, 1990) and 364 (+ 24) nmol/litre (Gregory et al., 1987). Even the latter values are considerably less than a mean value of 684 (+ 42) nmol/litre reported in blood samples from live pigs after an ACTH challenge (Zhang et al. 1991). The mean value for all pigs in the present experiment was 404 (+ 10) nmol/litre while the maximum mean value recorded on any one slaughter day during these trials was 576 (+ 34) nmoi/litre. Thus, the plasma cortisol concentrations are greater than most values reported in the literature. By far the dominant source of variation in cortisol levels in the present study was due to the variation between the different slaughter days. This was despite apparently generally uniform procedures for the handling of pigs during transport and at the abattoir. Ambient temperature, or humidity, did not appear to contribute significantly to this variation as there were no significant correlations between cortisol concentrations and temperature or humidity. Cortiso| and carcass composition
The finding of a greater mean cortisol value in the fat line, in comparison to that in the lean line, is not entirely unexpected. There is some direct evidence, and considerable indirect evidence, of relationships between fat content and plasma cortisol concentrations. For example, Wynn (1991) showed that immunisation of mildly stressed ewes against cortisol resulted in decreased fat content while in another experiment it was found that undisturbed animals suffered a decrease in carcass fatness, while animals subjected to low-level stress had more carcass fat. Moss (1992) ~tates that increased corticosteroid production can result in catabolism ot" muscle tissue and increased deposition of fat. In man and animals, obesity occurs in conjunctiorl with the hyperadrenal state of Cushing's syndrome. In cattle, bulls and .~teers generally have lower plasma cortisol concentrations, and less fat, than cows and heifers (Mitchell et ai., 1988). Cortisol and genotypes
in the present study, mean muscle and plasma cortisol levels for the three genotypes did not differ significantly. Nyberg et al. (1988) found that in some cases there was a ~;ignificant interaction between halothane genotype and treatment. For example, cortisol concentrations in transported nn animals were lower than in untransported nn animals; whereas transported NN animals had higher cortisol concentrations than untransported NN pigs. Consequently, Nyberg et al. (1988) concluded that there was no simple relationship between plasma cortisol and halothane genotype. Apparently conflicting results were reported by Mitchell & Heffron (19g.l) and Marp!e et al. (1974) with the former reporting stress-susceptible pigs (h~dotlaane-tested) having lower resting cortisol concentrations while the latter group reported higher cortisol concentrations in pigs from herds bred to be stress-susceptible. The time of collection of a blood sample in relation to the time of a specific stress may be an important factor in determining plasma cortisol concentration in stress-susceptible animals. Marple & Cassens (1973) reported that stress-susceptible swine had an increased metabolic
Cortisol measurements as indicators of meat quality in pigs
245
clearance of cortisol. Labelled cortisol was removed from the blood of stresssusceptible pigs at approximately five times the rate from that of normal pigs. Thus, immediately after a specific stress, stress-susceptible animals may have a marked elevation in plasma cortisol which returns to resting level within a comparatively short time. However, stress-susceptible pigs have an increased cortisol production rate in comparison to normal swine (Marple & Cassens, 1973).
Relationship between plasma and muscle cortisol The correlation coefficient of 0.727 (P <0.0001) indicates a close relationship between plasma and muscle cortisol concentrations. Thus, if practical considerations prevent the collection of a blood sample, the measurement of muscle cortisol can be used to provide a guide to the stress status, at time of slaughter, of a group of pigs.
ACKNOWLEDGEMENTS This work was supported by the Pig Research and Development Corporation. The competent technical assistance provided by Ms S. Dyson and Mr T. Larsen is gratefully acknowledged. The breeding and growing of the genetic lines was carried out by Mr L. Daniels at the Queensland Department of Primary Industries (QDPI) research piggery at Biloela. Mr H. Kramer of QDPI identified the halothane genotypes.
REFERENCES Becker, B. A., Nienaber, J. A., Christenson, R. K., Manak, R. C., DeShazer, J A. & Hahn, G. L. (1985). Am. J. Vet. Res., 46, 1034. Bouton, P. E., Harris, P. V. & Shorthose, W. R. (1972). d. Food Sci., 37, 351. Forslid, A. & Augustinsson, O. (1988). Acta. Physiol. Seand., 132, 223. Gregory, N. G., Moss, B. W. & Leeson, R. H. (1987). Vet. Rec., 121,517. Harvey, W. R. (1985). User's guide to L S M L / M W Mixed Model Least-Squares and Max#hum Likelihood Computer Program. Ohio State Univ., Colombus, OH, 46 pp. (Mimeo). Heinze, P. H. & Mitchell, G. (1989). Vet. Rec., 124, 163. Herd, R. M. (1989). Aust. Vet. J., 66, 341. Klemcke, H. G., Nienaber, J. A. & Hahn, G. L. (1989). Proc. Sot'. Exptl Biol. Med., 190, 42. Laursen, A. M. (1983). In Stunning of animals Jbr slaughter, ed. G. Eikelenboom. Martinus Nijhoff, Boston, p. 70. Marple, D. N. & Cassen, R. G. (1973). J. Anita. Sci., 36, 1139. Marple, D. N., Cassens, R. G., Topel, D. G. & Christian, L. L. (1974). J. Anita. St,i., 38, 1224. McPhee, C. P., Daniels, L. J., Kramer, H. L., Noble, J. W. & Trout, G. R. (1992). Proc. Aust. Assoc'. Anita. Breed. Genet., 10, 76. Mitchell, G. & Heffron, J. J. A. (1981). J. South Afr. Vet. Assoc., 52, 109. Mitchell, G., Hattingh, J. & Ganhao, M. (1988). Vet. Rec., 123, 201. Moss, B. W. (1992). In The Chemistry of Muscle Based Foods. Royal Society of Chem. istry, p. 69.
246
F. D. Shaw. G. R. Trout. C. P. McPhee
Moss, B. W. & Robb, J. D. (1978). J. Sci. Food Agric., 29, 689. Nyberg, L., Lundstrom, K., Edfors-Lilja, I. & Rundgren, M. (1988). J. Anim. Sci., 66, 1201. O'Benar, J. D., Hannon, J. P., Peterson, J. L. & Bossone, C. A. (1987). Am. J. Physiol., 252, R953. Rushen, J. (1986). Aust. Vet. J., 63, 359. Shaw, F. D. & Tume, R. K. (1990). Proc. llth Congr. Inter. Pig. Vet. Soc., Lausanne, p. 427. Shaw, F. D. & Tume, R. K. (1992). Meat Sci., 32, 311. Simmons, N. J. (1989). MSc Dissertation, University of Bristol. Spencer, G. S. G. (1979). Horm. Metab. Res., 11,586. Troeger, K. & Woltersdorf, W. (1988). Fieischwirtschaft, 68, 249. Trout, G. R. (1992). Proc. 38th. Inter. Congr. Meat Sci. Technol., France. p. 983. Warriss, P. D. & Brown, S. N. (1985). J. Sci. Food Agric., 36, 87. Warriss, P. D., Brown, S. N., Edwards, J. E., Anil, M. H. & Fordham, D. P. (1992). Vet. Rec.~ 131, 194. Wynn, P. (1991). In Project Guide 1991-92., Meat Research Corporation, Sydney, NSW, p. 120. Zhang, S. H., Hennessy, D. P. & Cranwell, P. D. (1989). In Manipulating Pig Production !!, Australasian Pig Science Association, Werribee, Australia, p. 212. Zhang, S. H., Hennessy, D. P., McCauley, I. & Cranweli, P. D. (1991). In Manipulating Pig Production IH, Australasian Pig Science Association, Attwood, Australia, p. 65.