Characteristics of steers of six beef breeds fattened from eight months of age and slaughtered at a target level of intramuscular fat II. Meat quality

Meat quality of Angus (AN), Simmental (SI), Charolais (CH), Limousin (LI), Blonde d'Aquitaine (BL), and Piedmontese (PI) steers (n=22 per breed group) was measured in the M. longissimus dorsi (M.l.d.) and the M. biceps femoris, regio glutea (M.b.f.). Animals were fattened in two subsequent series on a forage-based diet until a target level of 3.5% intramuscular fat (IMF) was reached aecording to real-time ultrasound assessments in the live animals or until 15 months of fattening had passed. Series 1 was performed in a tie-stall barn while a loosehousing system with straw bedding was applied in series 2. The actually measured IMF contents in M.l.d. were 3.35, 3.47, 3.49, 3.48, 2.34 and 2.40 % for AN, SI, CH, LI, BL and PI, respectively. Breed group differences in IMF content were mostly accompanied by a contrary Variation either in muscle water or protein content. Muscle cholesterol levels were similar for all breeds amounting to 47 and 51 mg/100 g on average in M.l.d. and M.b.f, respectively. Early and late postmortem muscle pH was relatively similar among breeds, but water-holding capacity, measured as losses due to drip, ageing, thawing and cooking, was unfavourably high in AN (drip loss excepted) in both muscles. Cooking loss tended to be lowest in PI, drip loss in SI. The AN showed the palest meat. In line with lightness, heme iron contents were clearly lowest in both muscles in the AN steers. There was no relationship found between IMF and shear force among breed groups. No significant differences between breed groups occurred in M.l.d. collagen solubility and shear force. Apart from breed differences, there were several differences noted between fattening series, namely clearly better water-holding capacity and lower shear force of the meat from series 2 (group housing) than from series 1 (tied System). The results indicate that in steers of similar IMF content and raised under the same feeding and management conditions, differences in most M.l.d. and M.b.f. quality traits were apparent, with the exception of shear force and M.l.d. collagen solubility.

Schlüsselwörter: Rindfleisch, Ochsen, Rasse, Fleischqualität, intramuskuläres Fett, Cholesterol 1 • Introduction Interest in label beef production is consistently increasing in Europe (NEUMANN and MARTIN, 1991;GERHARDY, 1994;ROCHE et al., 2000) in a Situation of declining consumption of red meat and a general decrease in acceptance of conventionally produced and marketed beef.Labels intend to provide a quality which is or can be especially oriented towards consumer demands including controlled beef quality and a greater consideration of the welfare of the animals (BOISSY et al., 2000).Although pricing Systems still heavily rely on carcass rather than on actual meat quality traits, labeis are expected to guarantee a certain level and, particularly, homogeneity of beef quality (BRANSCHEID and CLAUS, 1989) in order to fulftl the consumer's implicit expectations (TEMISAN, 1990).Current Swiss beef labeis encourage both the use of purebred beef breed cattle in order to profit from a higher carcass quality and, as category, steers to improve meat quality.This should provide products which can be distinguished from conventional Continental European beef which is based on calves largely representing a by-product of dairy husbandry.Visual perception of the products is often associated with beef quality and can affect purchase decision of the consumers.One main component of visual quality, together with colour, is marbling which expresses proportion and distribution of intramuscular fat (IMF) in the M. longissimus dorsi (M.l.d.).It is still unclear in how far known breed differences in meat quality traits such as water-holding capacity and tenderness (DUFEY, 1987(DUFEY, , 1988a;;TATUM et al., 1990;WHEELER et al., 1996) can be recovered under the condition of a similar extent of marbling.Furthermore, the repeatedly postulated low cholesterol content of Piedmontese beef (MONTANA RANGE, 2001) might get lost if really existing when the cattle is fed to high IMF contents.The objective ofthe present investigation was to compare the meat quality of six beef breeds fattened under the same conditions.The animals of this study were slaughtered at a target level of 3.5 % IMF.This level is higher than usually found in Switzeriand (DUFEY and CHAMBAZ, 1999).As reported in a first communication (CHAMBAZ et al., 2001), clear differences in growth and carcass quality developed between the breeds.Two breeds were unable to reach the target value in IMF, but were kept in the comparison in order to give information of their meat characteristics under the conditions applied.

2.
Materials and methods A total of 132 steers of six beef breeds, namely Angus (AN), Simmental (SI), Charolais (CH), Limousin (LI), Blonde d'Aquitaine (BL), and Piedmontese (PI), selected at 8 months of age from suckler beef rearing Systems, were fattened in two subsequent series (11 animals/breed/series).The first series was carried out in a tiestall barn and series 2 in a loose housing system with straw bedding.The animals had concentrate 220 g per kg of dry matter.Steers either were slaughtered when the b HTth " l^l 1 % re f ed 3 i % in the M -*" ' " » - (M' d (CHAMBA^etal 2001 Th"' ^ "f * , maXimUm °f 15 m °nths of fatte ™S (CHAMBAZ et al 2001).This approach resulted in large differences in slaughter age Members of the AN, SI, CH, LI, PI and BL groups finished the experiment at a^ average age of 381,509,529,610,683  SrpSr Ch ; T ^iSChe WCrkStätten GmbH ' D -Weilhenn) equipped with ä Sensor EB4 probe (Wtntion, CH-Gerzensee).Samples ofthe M.l.d. and M.b f were

ZZ
!?p Sn ?/"? 3SS SidC 48 h P>m -Meat C0l0Ur Was de termined with a Chroma-vfeWed L ' ^° <*> CH-Dietikon-Zürich) Wlying the light source D65, which fv.f K, • f 7 V Ä"* °-10 ° = ^-light), a* (red-green index) and b* (yellow-blue index when directly put onto fresh cuts of M.l.d. and M.b.f.48 h and 14 d p.m.These samples were sealed under vacuum as slices of 2 cm thickness.Separate samples of approximately 300 g were frozen at -30 °C until analysed for chemical composition heme iron and cholesterol content.Two slices, also vacuum-packed, were stored for 12 days at +2 °C for measurements in the aged meat.Weight los ;™ 8h ; S n P on ^f dete ™ ined / s W loss. Afterwards the samples were stored IITJ L iro bemg thaW6d at 2 ^ °C 0ver 24 h and then h ™^ for 5 min on a grill (type BP-50, Beergrill AG, CH-Zürieh) at 195 ± 5 °C by direct radiant heat with the samples repeatedly turned during heating.The apparatus was connected to an external electronical thermoregulator (Ematherm A, Trafag AG, CH-Männedorf) and a thermoprobe (Pt 100, Moser AG, CH-Hombrechtikon) to control temperature Aecording to prehminary assessments, this procedure resulted in a meat core temperature of approximately 68 °C.Within these procedures, losses due to thawing and cooking (directly after cooking) were recorded.In independent samples, drip losses were quant.f.ed as described by HONIKEL (1998) storing fresh 2 cm thick slices of each muscle for 48 h at 2 °C.Cooked samples cooled to ambient temperature were sheared by the original Warner-Bratzler device (model 3000, G-R Electric MFG Co Manhattan, Kansas, USA).Ten cores per sample of 1.27 cm diameter were obtained parallel to fibre orientation from the cooked slices aecording to KASTNER s^plts NRICKS °N °969) With an electrical dri » ^ a speed providing unifonr.
In the raw, homogenised muscle samples, heme iron contents were determined as Pigments aecording to BARTON (1967) by extracting pigments with acetone and spectral Photometrie determination at a wavelength of 640 nm on a Lambda 2 photometer (Perkin-Elmer, D-Überlingen).Heme iron was calculated assuming the conventiona ly applied value of 9.06 % heme iron in pigment.Cholesterol was enzymatically determined by a colorimetric method (BOEHRINGER MANNHEIM 1994) with the same spectrophotometer as described for heme iron analysis but at ä wavelength of 405 nm.As described in the guidelines (BOEHRINGER MANNHEIM, 1994), homogenized muscle samples were hot saponified during 30 min in advance of the cholesterol determination.Lyophilized samples of M.l.d. and M.b.f. were analysed for their contents of dry matter (3 h, 105 °C) as well as ash (total ash; 4 h, 550 °C; NAUMANN and BASSLER, 1997), fat (petrol ether extract;SLB, 1969) and protein (crude protein; KJELDAHL method; AOAC, 1995).Furthermore the lyophilized samples were analyzed for collagen content (hydroxyproline x 8) as described by ARNETH and HAMM (1971) adapted to the Technicon (Plainfield, New Jersey, USA) analyse chain.Collagen hydrothermal solubility at 90 °C was determined as outlined by KOPP et al. (1977).All chemical analyses were carried out in two replicates.
Data were statistically analysed with the NCSS program (version 1997, Hintze, Kaysville, Utah, USA).In a first evaluation, data of both series were included in order to be able to compare overall series differences, using a two-way ANOVA with breed and series as fixed effects and breed x series interactions.Because interactions between breed and series frequently occurred in meat quality traits, data were finally analysed separately for each series by one-way ANOVA with breed as fixed effect in the model.The Tukey test was used for multiple comparison among means regarding P < 0.05 as significant.

Results
On average of both experimental series the actually measured IMF contents in M.l.d.(means ± SD) were 3.35 ±1.12, 3.47 ± 0.93, 3.49 ±1.11, 3.48 ± 1.08, 2.34 ± 0.64 and 2.40 ± 0.63 % for AN, SI, CH, LI, BL and PI, respectively (Table 1).The corresponding levels found in M.b.f. were 3.35 ± 0.98, 3.92 ± 0.94, 3.02 ±1.14, 2.69 ± 0.93, 2.01 ± 0.75 and 1.71 ± 0.65 %.Therefore, at an overall slightly lower average IMF content of M.b.f. of 2.78 % compared to 3.09 % in M.l.d., the breed differences were roughly also reflected in M.b.f, particularly with again the lowest contents being found in BL and PI (partially significant against other breeds).Part of the within-breed Variation in IMF content resulted from significant series differences as the average IMF content achieved was lower in series 1 than in series 2. The M.l.d and M.b.f. of BL and PI had the highest protein content of all groups on average of both series, particularly when compared with AN, SI and CH, whereas LI held an intermediate position.Breed group differences in IMF content were mostly associated by contrary variations in muscle water and protein content.In line with protein and dry matter content there was a small but partially significant Variation in ash content of the two muscles.In contrast, no significant differences were observed between breed groups in cholesterol content of both muscles ranging at a similar level of around 47 and 51 mg/100 g in M.l.d. and M.b.f., respectively.The differences in early postmortem muscle pH were small and mostly not significant among groups for both muscles (Table 2).Moreover none of the animals presented a pH r value below 6.2, indicating that no PSE (pale, soft, exsudative) meat was found.Ultimate muscle pH, measured at 48 h p.m., was in the desired ränge with an average of 5.54 (max: 5.82) in the M.l.d. and of 5.50 (max: 5.79) in the M.b.f., when pooled over both series (slightly but significantly higher in series 2).None of the animals expressed an ultimate muscle pH above 6.0, regarded as the threshold level for DFD (dark, firm, dry) meat or DCB (dark cutting beef).In series 1, LI muscles had a significantly lower pH 48 in the M.l.d.compared to PI, with the other groups remaining intermediate.Table 3 describes traits of water-holding capacity of the meat.The LI showed the highest M.l.d.drip loss on average of both series.In M.b.f. the pattern was not the same, with the BL having the highest and AN and SI showing the lowest drip loss.Drip loss was higher in both muscles in series 2 especially in BL, whereas AN meat showed an inverse pattern.PI had the lowest ageing losses in both series for M.l.d. and M.b.f., whereas M.l.d. of BL and SI (only series 2) and M.b.f. of AN showed significantly higher ageing losses.On average of both series, the AN expressed the highest thawing and cooking losses for both muscles whereas the losses were lowest in BL and PI.This illustrates that, within the different losses measured for the same muscle, only thawing and cooking losses, making up the major proportion of total losses, expressed a similar pattern over breeds.Accordingly, the overall water-holding capacity was low in AN and high in PI.Muscles of animals in series 2 expressed a lower cooking loss than those in series 1. Water-holding capacity of M.b.f. was on average better than that of M.l.d. in all types of losses measured.SrS^nSSÄÄ^ """ S, 8n ' ti Can " y (P< °°5); Seri CS aVera6CS Wi l hi " ' he Same ^ wilh *-Meat-colour related traits are given in Table 4.The M.l.d. and M.b.f.ofthe AN and CH steers were lighter than those of the other groups, and PI steers had the darkest M.l.d. and one ofthe darkest M.b.f.In both muscles, BL and PI showed a trend to less reddtsh and yellowish meat than that of the other breed groups.As expected, during ageing for 12 days colour turned towards higher lightness and redness in both muscles (data of non-aged meat not shown).The Variation in M.l.d.lightness during ageing was significantly higher in AN than in SI with intermediate values in the other breed groups (data not shown).Heme iron content was lowest for AN and CH steers in both muscles and was highest for PI and SI in M.l.d. as well as for SI and LI in M.b.f.The M.b.f.heme iron content was higher by 44 % compared with that ofthe M.l.d.There was a significant negative correlation between heme iron content and L* in both muscles (r = -0.77 in M.l.d. and r = -0.65 in M.b.f, P < 0.001).Shear forces were not significantly different in M.l.d.among groups, but AN M.l.d.tended to have low values (Table 5).There was a significant decrease in shear values in the second series and especially in SI and PI.Unlike as in the M.l.d., there were significant differences among groups in shear force of M.b.f, but the trend remained similar.As expected, shear force and collagen contents were higher in the M.b.f.compared to M.l.d.although at similar collagen solubility.Muscle collagen content was higher in AN, SI and CH than in LI (only M.l.d.), BL and PI.There were no significant differences in M.l.d.collagen solubility among breed groups, but AN tended to have the highest solubility in both series, following the same trend as in shear force.Collagen solubility in M.l.d.increased in series 2 (+22 %) especially in the PI, LI and SI.

Discussion 4.1
Realized contents of intramuscular fat in steers fed on a forage-based diet Steers of four out of the six beef breeds evaluated were able to reach on average a target level of 3.5 % IMF in the M.l.d. which was desired in order to ensure the favourable impression of marbling (CHAMBAZ et al., 2001).However, no single PI and BL steer was able to reach this target level although age was almost twice as high as in AN when fattening was finally terminated (CHAMBAZ et al., 2001).This illustrates that some cattle breeds obviously do not have the inherent ability to deposit increasing amounts of IMF regardless of the length of the fattening period, although carcass fatness is still increasing in this period (SMITH, 1988).Most other breeds show a widely linear increase not only in adipose tissue but also in IMF up to high slaughter ages as described for instance by SZÜCS et al. (2001aSZÜCS et al. ( , 2001b) ) for German SI bulls.The high variability between animals of the same breed group in IMF content probably was mainly due to the restricted accuracy of the ultrasound method of IMF determination in live animals.However previous attempts analysing biopsies, as another potential way of estimation, proved to be far less precise because of the inhomogeneity ofthe distribution offat in the muscle.

4.2
Differences in meat quality of steers of different breed fattened to a target IMF level As expected there was an antiparallel relationship of the breed group differences in IMF and other compositional traits.The differences in chemical composition of muscle between BL and PI on one hand with low IMF and AN, SI, CH and LI on the other hand are in accordance with the results of BROWNING et al. (1990) who reported that muscles of leaner carcasses were higher in water and protein content.Nevertheless, the close inverse relationship between moisture and fat content described by these authors and others ( VAN KOEVERING et al., 1995) was less clear in the present study, especially in the M.b.f.where variations in IMF content were more associated with variations in protein content.This is probably due to the deliberately low differences in IMF.Accordingly, breed group differences in muscle protein and fat content were of very low importance in a dietetic sense.This is also true for muscle cholesterol content which was similar in all groups.Even muscles differed only slightly but equally by about 2-3 mg/100 g muscle tissue.This is in accordance with the results found by EICHORN et al. (1986), BAKER andLUNT (1990) as well as GARIEPY et al. (1999).RHEE et al. (1982) observed only a significant difference when comparing seven marbling-score categories when muscles "practically devoid" of marbling were compared with higher marbling scores which had higher cholesterol levels.However this contradicts findings of SLOVER et al. (1987) and VAN KOEVERING et al. (1995) reporting a positive relationship between fat and cholesterol content of raw beef.BROWNING et al. (1990) also noted a trend towards higher cholesterol levels from lean to fatter carcasses, but the difference of only 1.8 mg cholesterol/lOOg raw meat found is of little practical importance.Probably based on its low IMF content, there is a frequently cited opinion that meat of PI cattle has a considerably lower cholesterol content than meat of other breeds (MONTANA RANGE, 2001).However, neither the present data, where PI steers had been fattened to untypically high IMF contents for this breed, nor other scientific studies (BAKER and LUNT, 1990;GARIEPY et al., 1999;RULE et al., 1999) found any clear differences in muscle cholesterol between PI or PI crosses and other breeds.Although PSE meat is not a major problem in beef, DUFEY (1987, 1988b, 1989) occasionally recorded too low pHi-values in meat of bulls fattened in a tied stall, whereas no PSE meat was found here even in series 1 where the animals were kept under the same housing conditions as in the studies cited.The absence of DCB in this study is not surprising.DCB is a result of a reduced glycogen content in the muscle prior to slaughter and is often associated with stress caused for instance by mixing cattle or shipping fatigue.The duration of transport to the slaughter plant did not exceed 1 hour.Moreover the use of steers instead of bulls reduced the probability of DCB as steers have lower stress susceptibility (WARRISS, 1990).In normal beef, p.m. glycolysis reduces pH to 5.8 or lower within 48 h (KREIKEMEIER, 1998;IMMONEN and PUOLANNE, 2000).In an ultimate pH ränge of 5.8-6.0,defined as borderline DCB, meat already tends to have an abnormal colour and an increased risk of spoilage, particularly when vacuum-packed (WARRIS, 1990), although quality is still intermediate between true DCB and normal beef (VOISINET et al., 1997).This type of borderline dark cutters was found here in two out of 132 carcasses.Despite the low Variation in pH, breed groups differed to a certain degree in water-holding capacity, meat colour and tenderness related items.The ränge in drip loss found in the present study was mostly favourable as a level of up to 4.5 % is still considered as acceptable (ENDER and AUGUSTINI, 1998) whereas higher rates are undesired in retail packaging and so impair the appearance of the product at sale.Muscle tissue of M.l.d. but not of M.b.f. from LI steers showed drip losses on average just at the acceptable limit, and several members of this group had too high drip losses from M.l.d.aecording to the threshold level given above.Cooking losses were highest in both muscles in AN steers.Accordingly, differences between breeds were not related between drip loss and cooking loss, except in PI.This can be explained by the different compartments of bound water being stressed by the procedures.Drip losses are passive losses which strongly depend on loosely-bound water only affected by gravity, whereas meat was subjected to mechanical and thermal Stressors along with ageing (vaeuum), freezing/thawing and cooking.Accordingly, no correlation between drip loss and cooking loss was noted earlier (HONIKEL, 1986).CROSS et al. (1984) andCROUSE et al. (1985) did not find a difference in cooking losses between AN, SI, CH and Hereford as well as between AN and SI, respectively.The level of cooking losses were about two-fold lower in our study compared to those found by GERHARDY et al. (1995) with meat from 16 and 20 months old bulls and still lower than those found in young heifers by SCHEEDER et al. (1996) of 24 % and 25 % for M.l.d. and M.b.f, respectively.However, Variation in cooking methods might have played an important role in this respect.Apart from marbling, meat colour influences the visual appeal of meat to retail purchasers (SHORTHOSE and HARRIS, 1991) and thus is an important criterion for purchase decision of the consumer (FRENCH et al., 2000).Consumers appear to prefer beef which is neither extremely pale nor dark, with the ränge of L* values between 34 and 40 being considered as normal (ENDER and AUGUSTINI, 1998).On this basis, the meat of some AN and CH steers can be judged as to be too pale.An increasing IMF content tends to increase meat lightness (SHORTHOSE and HARRIS, 1991;FRENCIA and MONVOISIN, 1993), but the differences between BL and PI on one hand and the four other breeds on the other hand were not consistent.Obviously some independent breed differences in lightness exist.Similar to the present study, LIBORIUSSEN et al. (1977) found a higher heme iron content ofthe M.l.d. of SI sires compared with LI, CH and BL.However, the LI meat was lighter than that of CH which was not the case in our study.Accordingly, the heme iron content, the effective part of the pigments of bovine meat (RENERRE, 1982), similarly varied between breeds as the L* value.Heme iron is also of interest for human nutrition because of its high bioavailability (VARNAM and SUTHERLAND, 1995).In this view, AN and, partially, CH beef is inferior to beef ofthe other breeds.Overall, the animals of the present study provided a very tender M.l.d.applying the thresholds of 38 N in Warner-Bratzler shear force for very tender meat and of 45 N for tender meat as given by SHACKELFORD et al. (1991) and VAN KOEVERING et al. (1995) which were obtained under similar measurement conditions.In accordance with SCHEEDER et al. (1996), the differences in peak shear force found between M.l.d. and M.b.f. were low.Thus even the M.b.f., where shear forces were extremely low for PI and AN in series 2, could be used as steak instead of using it as lower priced tT tS ;i , K r l iCU i a u y ?en SiZC °f this CUt is not limitin g due to bi § sized carcasses.Only the M.b.f.ofthe SI in the first series had to be judged on average as too tough to be marketed as steak.However, SCHEEDER et al. (1996) noticed that peak shear force data provide little explanation for the collagen component of tenderness thus indicating that there might nevertheless exist clear differences in tenderness of both muscles.There was no clear relationship between IMF content and shear force among breed groups.The residual Variation in IMF content among the breeds after fattening to a ,^ , u l6Vel W3S C °ntraiy t0 the re P eated 'y ^ated inverse relationship between MF and shear force (AUGUSTINI and LÜDDEN, 1992;SHACKELFORD et al 1994aX For instance PI meat in series 2 had the lowest IMF content and the lowest J n e l; orc f m both muscles.This is in accordance with SHACKELFORD et al (1994b) who found that meat from PI crosses with AN or Hereford dams, despite havmg a lower IMF content as the steers in our study, showed the lowest shear force among eleven genetic groups including AN and CH.This finding is also well m P T P r^o ^J^T UM Ct aL (1 " 0) Wh ° COm P ared Ge,bvieh > Red A N and PI sires.BLUMER (1963) and DIKEMAN (1996) reported that IMF content is a poor predictor SiSS ™ CSS a w,.nly a °C0UntS f °r 5 t0 10 % of the variability in tenderness.WHEELER et al. (1996) also found an increase in tenderness at a common marbling end point when using PI crosses with various dams instead of twelve other breed sires KOCH et al. (1976) observed a difference of less than 3 N in shear force among AN and CH on one hand and LI and SI on the other hand when data were adjusted to a similar slaughter age.Though statistically significant, the differences were small and the values remained in the acceptable ränge of tenderness.DUFEY (1987) compared tenderness of M.l.d.samples from bulls of purebred SI, SI with 75 % Red Holstein blood, pure-bred Brown Swiss and Brown Swiss with 25 % Holstein blood The sensory analysis revealed that the SI bulls had the toughest meat apart from the Brown Swiss crosses, and an improvement in meat quality took place with a high Red Holstein blood proportion.In another study, DUFEY (1988a) compared SI and Brown steers, their crosses with AN, and Holstein in taste panel evaluations of tenderness for M.l.d. and M. semimembranosus samples with the result that meat from SI steers was rated low in tenderness.Crossbreeding with AN positively influenced tenderness In a third study, DUFEY (1989) compared tenderness of M.l.d.samples from Fl bulls sired by SI and BL with various dams (SI, Brown Swiss, Holstein).No significant differences were found in tenderness between the sire groups.BRANSCHEID and HERZOG (1996) noted that crossbreeding of German SI with CH and PI improved tenderness (shear force and panel tenderness).Aecording to this trend described in literature, the present data also showed a higher shear force of meat of SI steers compared to the other breed groups but this tendency was only apparent in the first series.

4.3
Fattening series differences in meat quality of steers (tied housing vs loose housing) Steers ofthe present study originated from two different fattening series, and housing system was changed from a tied system to group housing on straw beds without change in the diet.The Statistical evaluation of the data revealed several effects of fattening series on meat quality.In detail, the IMF content was higher, the cooking losses were clearly lower, collagen solubility in M.l.d. was higher and, finally, shear force was lower in series 2. From the design chosen, no clear attribution ofthe effects to housing system is possible and time-dependent effects as well as random trends in the animals selected from the breeds cannot be excluded.However, based on the results of other studies, some of the effects might nevertheless be explained by the Variation in the housing Situation.Using the tie-stall barn in series 1 resulted in a restricted movement and subsequent Joint and leg problems which reduced growth rate at the end of fattening for the oldest animals (CHAMBAZ et al., 2001).Management practices which alter growth and muscle accretion rates can have a profound effect on muscle proteinases and collagen characteristics.An improved growth rate may result in a decreased calpastatin activity at 24-31 h p.m. and consequently in an improved tenderness (SHACKELFORD et al., 1994b).THOMPSON et al. (1996) found that meat from steers growing fastest prior to slaughter was more tender and had higher ucalpain and calpastatin activities 2 h p.m.They concluded that a high u-calpain activity is likely to be related to a high rate of protein degradation and this should result in an increase ofthe myofibrillar fragmentation index and, therefore, tenderness.From their study they presumed that the level of u-calpain activity close to slaughter had been more important in determining tenderness than the calpastatin activity at 2 h p.m. Fast-growing cattle would also have a more intense protein tumover and therefore fewer heat-stable intermolecular collagen crosslinks (FISHELL et al., 1985;MCCORMICK, 1994).This could partly explain the higher collagen solubility and the lower shear force in series 2 in all breed groups.As collagen solubility was higher in M.l.d. in series 2, but not in M.b.f, certain additional series differences in the animals selected can be assumed.SI and PI had the highest series difference in slaughter age (63 days and 45 days younger in series 2 than in series 1, respectively; CHAMBAZ et al., 2001) and simultaneously showed the highest improvement of shear forces in series 2. Although the increase in the toughness of meat with animal age is well known (SHORTHOSE and HARRIS, 1990) it can be assumed that these effects were not determined by age alone.Firstly, all breed groups showed a decrease in shear force even those which were older in series 2 than in series 1 such as AN, CH and LI (25, 23 and 9 days older in series 2).Secondly, there were no significant correlations between age and shear forces within breed groups and series (data not shown).However from the age-related changes in collagen properties, nevertheless a certain decrease in tenderness with age can be expected.It is difficult to explain the better water-holding capacity found in series 2 by a single factor since animal genetics, age and slaughter conditions are known to significantly contribute to the expression of variations in these traits.

5.
Conclusions The overall goal of the present study was to compare a wide ränge in beef breeds in their differences in meat quality when a constant IMF content of 3.5 % is achieved.The target level in IMF, however, excluded two breeds, namely PI and BL, as not appropriate for this attempt.In the other four breeds, the target was reached on average but individual Variation in IMF was still high within breed due to the restricted accuracy of the in vivo ultrasound method applied for determination.From the present material, without further attempt to reduce variability in IMF content, certain breed differences in important quality traits such as colour, water-holding capacity and tenderness-related variables were found.Differences were particularly high between AN and the other suitable breeds (SI, CH, LI) with unfavourable properties of AN beef in total water-holding capacity and favourable estimates for tenderness-related data.However, differences in shear force among breed groups were small, with all breed groups being well above minimum levels of acceptance, even in a basically tougher muscle such as the M.b.f, except in SI.The significantly better growth characteristics obtained in the steers of the loose housing system compared with those in the tie-stall barn seemed to be the major cause of the improvement in meat quality in the second series.Apart from the inability ofthe BL and PI to reach a high IMF content under the present feeding and housing conditions, all breeds were suitable for premium beef quality marketing.
and 690 days, respectively.The coefficient S and 48 h post mortem (p.m.) in the M.l.d. at the 10 th rib and in the Musculus biceps femoris, regio glutea (M.b.f.) with a portable pH meter(WTW 197S

Table 1
Chemical composition of the M.l.d. and the M.b.f.(wet weight) of the steers originating from different beef breeds (n-11 per series) (Chemische Zusammensetzung des M.l.d.und des M.b.f.der Ochsen von verschiedenen Mastrassen; n=l 1 je VersuchsdurchgangV AN = Angus, SI = Simmental, CH = Charolais, LI = Limousin, BL = Blonde d'Aquitaine, PI = Piedmontese Means within one line without a common superScript differ significantly (P<0.05);series averages within the same variable with dilTerent superscnpts are significantly different (P<0.05).

Table 2
Development of pH ofthe M.l.d. and the M.b.f.ofthe steers originating from different beef breeds (n=ll per series) (Verlauf des pH-Wertes von M.l.d.und M.b.f.der Ochsen von verschiedenen Mastrassen; n=ll je Means within one line without a common superscript differ significantly (P<0.05);series averages within the same variable with different superscripts are significantly different (P<0.05).
Table3Water holding capacity ofthe M.l.d. and the M.b.f.ofthe steers originating from different beef breeds (n=l 1 per series) (Wasserbindungsvermögen des M.l.d.und des M.b.f.der Ochsen von verschiedenen Mastrassen; n=l 1