Effects of Size and Rate of Maturing on Carcass Composition of Pasture-or Feedlot-Developed Steers *

Steers (n = 335) of known genetic backgrounds from four fundamentally different growth types were subjected to two production systems to study the main effects and possible interactive effects on carcass composition. Growth types were animals with genetic potential for large mature weight (LL), intermediate mature weight-late maturing (IL), intermediate mature weight-early maturing (IE), and small mature weight-early maturing (SE). Each year, in a nine year study, calves of each growth type were weaned and five steers of each growth type were developed on pasture or feedlot and harvested at approximately 20 and 14 mo of age, respectively. Data recorded were chilled carcass weight and percentages of forequarter, foreshank, chuck, rib, plate, brisket, hindquarter, round, rump, shortloin, sirloin, flank, lean, fat, bone, and retail cuts. The growth type×production system interaction was an important source of variation in chilled carcass weight (p = 0.0395) and percentage retail cuts (p = 0.001), lean (p = 0.001), fat (p = 0.001), rump (p = 0.0454), shortloin (p = 0.0487), and flank (p = 0.001). The ranking of the growth type x production system means for percentage lean was LL-pasture>IL-pasture = IE-pasture = SE-pasture>LL-feedlot, IL-feedlot>IE-feedlot = SE-feedlot. The growth type× production system interaction was non-significant (p>0.05) for forequarter, foreshank, chuck, rib, plate, brisket, hindquarter, round and bone. Growth types of IE and SE yielded greater (p<0.05) mean forequarter than did growth types of IL and LL (51.6±0.3 and 51.5±0.3 vs. 51.1±0.3 and 50.8±0.3%). Mean bone was highest (p<0.05) for the LL growth type and lowest (p<0.05) for the SE growth type (19.5±0.5 vs. 16.8±0.5%). Mean bone was greater (p<0.05) for the pastured steers than for the feedlot steers (21.8±0.8 vs. 14.5±0.6%). These data indicate that growth type responded differently in the two production systems and that these results should be helpful in the match of genetics to production resources. (Asian-Aust. J. Anim. Sci. 2006. Vol 19, No. 5 : 661-671)


INTRODUCTION
Consumers are demanding leaner, palatable beef (Savell et al., 1987), and the cattle industry should establish methods of targeting market end points with reduced carcass fat and maximum carcass value (May et al., 1992).Forage-fed beef carcasses have less fat than grain-finished carcasses (Robinson et al., 2001), thus forage-fed beef provide consumers with leaner meat than grain-finished beef.Pasture-based beef production systems offer economic benefits that surpass feedlot finishing because forage traditionally is less expensive than feed grain per unit of energy and protein (Dixon and Stockdale, 1999).More research is needed on all-forage development systems (NRC, 1976), as pasture-based systems are a viable alternative energy source for beef production (Allen et al., 1996).
Growth type has the potential to influence beef carcass composition (Priyanto et al., 1999) by altering the relative growth patterns of muscle, bone, and fat (Berg and Butterfield, 1968).Carcass composition varies among cattle growth types (Koch et al., 1976(Koch et al., , 1979(Koch et al., , 1982;;Stiffler et al., 1985: Griffin et al., 1992) and it is important to understand how various cattle types can optimally produce lean, highquality beef (May et al., 1992).For example, European breeds developed on a high plane of nutrition, typically have higher yields of lean edible product than English breeds (Koch and Dikeman, 1977;Koch et al., 1979Koch et al., , 1982)).In the past, differences in beef growth types were reflected in differences among breeds; however, increased emphasis on selection for size over the last two decades has resulted in a broad array of growth types within most breeds (Brown et al., 2005).
As a result, it is often difficult to isolate the true effects of forage and grain feeding on carcass composition because the cattle involved are harvested at carcass weights which vary widely within and between studies (Muir et al., 1998).Therefore, feed-type differences have been confounded with plane-of-nutrition effects, such that cattle fed the higher energy feedlot diet have been heavier and fatter in comparison with those fed the forage-based diet (Williams et al., 1983;Bidner et al., 1986;Muir et al., 1998).Consequently, it is difficult to ascertain the extent to which

Effects of Size and Rate of Maturing on Carcass Composition of
Pasture-or Feedlot-Developed Steers* the lighter weight and lower fat content has been due to lower growth rate or to forage rather than grain in the diet (Davies, 1977;Steen et al., 2003).
When divergent growth types are finished on highenergy diets to exploit lean tissue accretion, considerable variation in carcass fatness usually results (Brown et al., 2005).For comparisons of carcass composition, studies have used various standardization techniques.These standardization techniques have been reviewed and discussed by Muir et al. (1998).The correct match of nutrient requirement of growth type to feed resources in various production systems could provide for more acceptable lean to fat composition and eliminate the need for these adjustments.
Nutrient requirements among growth types and possible alternative feeding systems indicate more data are needed on how carcass composition and feed interact (Flora, 2001).Variation in carcass composition among growth types and production systems suggest a correct match of growth type and nutritional requirements to production system would aid in eliminating excess fatness and provide for lean tissue accretion.Indeed, Robinson et al. (2001) indicated an important factor in producing carcass beef is to correctly match beef growth type to available inputs within a production system.Therefore, the objective of this study was to determine the percentage yield of primal and subprimal cuts, carcass lean, fat and bone of four fundamentally different growth types of steers developed in either pasture or feedlot production systems.Divergent growth types included in the study were chosen to increase the possibility of determining differences in carcass composition among growth types and production systems.

Animals
Three hundred thirty-five steers representing four genetically different beef growth types were developed on pasture or in a feedlot and harvested to study the interaction of growth type and production system on carcass traits.Five calves from each beef growth type were assigned to each production system (pasture vs. feedlot) in each year of a nine-year study.Eighteen steers were removed from the study because of chronic bovine respiratory disease or injury.An additional 7 steers were removed because some of their carcass traits were outliers (n = 335).It was by random chance that a few more steers of two of the four biological types were removed.The smallest growth type× production system subclass contained 39 steers; therefore, removal of the steers should not have been an important source of bias in these data.
Beef growth types were determined by growth curve parameters of mature weight and rate-of-maturing of the cattle herds represented.Growth types included genetic potential for large mature weight-late maturing (LL, n = 79), intermediate mature weight-late maturing (IL, n = 88), intermediate mature weight-early maturing (IE, n = 87), and small mature weight-early maturing (SE, n = 81).The LL steers were Chianina, Charolais, or crosses between these breeds.The IL steers were either Red Poll or Hereford, the IE steers were current-pedigree Angus, and the SE steers represented a sample of small Angus cattle that were like those popular in the U.S. in the 1950's.The beef growth types were selected due to their broad variation in available growth curves and maturity patterns and their combined impact on carcass traits.Growth types were established using the three-parameter growth curve model described by Brody (1945).With the exception of the Chianina cowherd, composite growth curves of these herds were presented and discussed by Johnson et al. (1990).Mean estimated mature weight and maturing rate in the Chianina cow herd were 636 kg and 0.041%/mo, respectively, (unpublished data).Brown et al. (1991) also characterized size and maturing rate differences between these beef growth types.

Production system
Steers used in this study were born in the spring, received no creep feed, and were weaned at approximately 7 mo of age.Each year, after weaning, one half of the steers of each beef growth type (5 of each growth type) were allocated to a pasture production system.Pasture-developed steers grazed in the cool seasons on tall fescue (Festuca arundinacea Schreb.) that was overseeded with rye, ryegrass, and red clover (Secole cereale, Lolium multiforum, and Trifolium pretense, respectively).Warm season grazing consisted of tall fescue and bermudagrass (Cynodon dactylon) overseeded with sudan (Sorghum vulgare) in addition to some millet (Pennisetum glaucum).Forage availability was appraised weekly by experienced personnel and was found to be adequate for steer growth above maintenance (unpublished data) except in the second year where steers received supplemental prairie hay due to drought conditions.Steers in the pasture production system grazed unimproved pasture until overseeded pasture was available about December 1 st of each year.Then, steers were allowed to graze pastures for 330 d and slaughtered at approximately 20 mo of age.
Upon weaning, the other half of the steers of each beef growth type (5 of each growth type) were allocated to a feedlot production system and fed a ration that contained 33% cotton seed hulls, 43% cracked corn, 9.5% crimped oats, 14.5% soybean meal and 2.2% calcium carbonate.Also 2,200 IU of vitamin A were added per kilogram of feed.As formulated (NRC, 1976), the diet contained 1.6 Mcal Ne m and .9Mcal Ne g /kg DM and 12% CP (Brown et al., 1991).Feedlot steers were given ad libitum access to feed for 210 d and slaughtered at 14 mo of age.
In both production systems, steers had free access to fresh water and a commercial mineral mixture that contained 12.5 to 15% calcium and 12% phosphorus.A detailed description of the management of steers in each production system in this study is given by Camfield et al. (1999).All steers within a given production system were harvested at a similar age and all beef growth types had similar opportunity for development.Throughout the study, husbandry was in accordance with guidelines recommended by the Consortium (1998).

Slaughter and fabrication
Body weights were recorded for both pasture-and feedlot-developed steers at the University of Arkansas, Savoy Unit, before shipping study animals 21 km to the University of Arkansas Red Meat Abattoir in Fayetteville, AR, where feed and water were withheld overnight.Preslaughter shrunk body weights were taken prior to stunning.After dressing, splitting, determining hot carcass weight, and dressing percentage, carcasses were chilled and stored in a cooler for 96 h at 2°C.Upon completion of chilling, carcasses were weighed (CCW) and ribbed between the 12 th and 13 th ribs and carcass measurements taken by trained personnel.Carcass measurements were obtained 96 h postmortem in order to more efficiently utilize labor and processing facilities.Main effect means for carcass traits of steers in the study, as influenced by beef growth type within pasture-or feedlot production systems, have been summarized by Camfield et al. (1999).The interaction effect means for beef growth type x production system for carcass traits of these steers have been reported by Brown et al. (2005).
The left sides were then fabricated into primal/ subprimal and retail cuts, lean trim, fat and bone following procedures outlined by the Institutional Meat Purchase Specifications for Fresh Beef (IMPS; USDA, 1988).The forequarter of the left side was fabricated into the rib, square cut chuck, arm section, wholesale plate, brisket and foreshank.Fat from the cod, pelvic and kidney regions, hanging tender and kidney were removed from the hindquarter.The flank was removed from the loin and separated into lean trim, fat, and flank.The round was separated into the knuckle, rump, hindshank, lean trim, fat, and bone.The wholesale loin was fabricated to yield the sirloin and shortloin, both of which were eventually fabricated into retail cuts.Each primal/subprimal cut was further processed into steaks, roasts, lean trim, and bone.Finally, all subcutaneous and accessible intermuscular fat was removed to produce cuts free of trimmable fat.Weights for each cut were recorded at each stage of fabrication.Right sides were quartered, then fabricated in the same manner as the left side.
Percentage of total lean was calculated by combining the weights of the lean from the right fore-and hindquarter and dividing by the chilled weight of the right side.Percentage of total fat was calculated as the sum of kidney, pelvic, heart and subcutaneous, and accessible intermuscular fat from all retail cuts from the right side divided by the chilled carcass weight of the right side.The combined weight of bone removed from the right forequarter and hindquarter was divided by the chilled weight of the right side to calculate the percentage of total bone.Total retail cut yield was the total weight of retail cuts from both sides expressed as a percentage of the chilled carcass weight.

Statistical analysis
Data were analyzed according to methods of least squares analysis of variance with unequal subclass numbers.Sources of variation in CCW and composition traits were partitioned in a mathematical model that included terms for an overall mean, year, beef growth type, production system, year×beef growth type, year×production system, beef growth type×production system, year×beef growth type× production system, age within production system, and residual error.The 6-mo difference in mean harvest age of steers between production systems was considered to be part of the variance partitioned by the production-system effect.These data were analyzed as such because pasturedeveloped steers require additional time and input to approach a more suitable harvest weight and composition compared to that of the feedlot-developed steers.Composition trait data were not adjusted to a constant endpoint basis (i.e. 12 th and 13 th rib fat thickness) because variation of interest in the stated objective would be reduced or eliminated by this adjustment.All analyses were performed using the general linear models (GLM) procedure of SAS (SAS Inst.Inc., Cary, NC).

RESULTS AND DISCUSSION
Year was an important source of variation (p<0.001) for all traits studied except 102-forequarter and 155-hindquarter (data not shown).Interactions involving year with growth type were significant (p<0.05) for all traits except 102forequarter, 103-rib, 118-brisket, 155-hindquarter, and 173shortloin.Interactions involving year with production system were significant for all traits except 102-forequarter.The year×growth type×production system interaction was non-significant (p>0.05) for all traits studied except 103-rib.Significant interactions involving year were expected and likely resulted from temporary environmental effects on pasture that made it impossible to exactly duplicate pastures from year to year (Vallentine, 1990).Also, as year was included in the statistical model, observation for CCW and carcass composition traits were adjusted to a mean year effect.Therefore, main effects means of year and interactions effect means involving years are not presented.
Within production system steer age at harvest accounted for substantial variation in all traits (p<0.001)except 103rib, 173-shortloin, and 113-chuck (p>0.05).Across production system, pastured developed steers were 6 mo older at harvest than the feedlot finished steers.Steers developed on pasture required additional time and inputs to approach a more suitable harvest weight and composition compared to the feedlot steers.Consequently, feed type differences are confounded with age of steer.This likely resulted because of the difficulties in achieving sustained high rates of gain in the pastured steers as the pasture in this study, even though of excellent quality, contained less energy than the feedlot diet.In addition, even at similar growth rates, steers consuming a feedlot diet deposit fat at a higher rate than pasture-fed steers (Tudor, 1992;Sainz et al., 1995).These results are in agreement with those of previous studies (Schaake et al., 1993;Camfield et al., 1999;Brown et al., 2005) where forage-fed steers were older at harvest than grain-finished steers when each group was fed to a set harvest weight endpoint or harvest composition endpoint.
The beef growth type×production system interaction was not significant (p>0.05) for 102-forequarter, 117foreshank, 113-chuck, 103-rib, 121-plate, 118-brisket, 155hindquarter, 164-round, and bone.Therefore, the main effect means for beef growth type and for production system are presented for these traits.
Least squares means and standard errors for the effects of beef growth type on percentages of carcass fore-and hindquarters, percentages of primals and subprimals, and percentage of bone in pasture-or feedlot developed steers are presented in Table 1.Growth types of IE and SE yielded greater (p<0.05)mean percentage of 102-forequarter than did growth types of IL and LL (51.6±0.3 and 51.5±0.3 vs. 51.1±0.3 and 50.8±0.3).The LL growth type had less (p<0.05)mean percentage of 102-forequarter than did the IL growth type (50.8±0.3 vs. 51.1±0.3).These results are not in agreement with those of other scientists (May et al., 1992;Arthur et al., 1995;Miller et al., 1996) who reported no difference in 102-forequarter among different biological types of cattle.
The LL growth type had greater (p<0.05)mean 117foreshank than the other three growth types (4.4±0.1 vs. 3.9±0.1,3.9±0.1 and 3.8±0.1%)while the SE growth type had less (p<0.05)mean 117-foreshank than the other three growth types.The mean percentage of 117-foreshank of the IL and IE growth types were similar (p>0.05) and intermediate to the LL and SE growth types.These results are supported by data of Miller et al. (1996) and Arthur et al. (1995) who reported greater 117-foreshank in the large-late maturing Brown Swiss and Charolais breeds respectively, when these breeds were compared to breeds of medium size.Results of this study are not in agreement with those of other studies (Belk et al., 1991;May et al., 1992;Priyanto et al., 1999).May et al. (1992) found no difference in mean percentage of 117-foreshank among large, medium and small steers of three muscle thicknesses (thick, average, and thin).Additionally, Priyanto et al. (1999) found no difference in mean 117-foreshank among Hereford, Brahman, and Hereford x Brahman steers at 100 and 165 kg of carcass side weight.Conversely, Belk et al. (1991) reported that small-framed steers produced the highestyielding 117-foreshank when compared to other frame sizes.
Mean  May et al. (1992) and Miller et al. (1996).May et al. (1992) found similar mean percentage 113-chuck among large, medium and small steers with each size having three muscle thickness levels and 0.75 mm at 12 th rib fat thickness.Miller et al. (1996) reported no difference in mean percentage 113-chuck among Brown Swiss, Okie #1, Mexican #1, and Mexican #2 steers.
There was no difference (p>0.05) in mean percentage of 103-rib among the LL, IL, and IE growth types (7.5±0.1, 7.5±0.1,7.4±0.1,respectively), while the SE growth type had less (p<0.05)mean 103-rib (7.3±0.1%)than the other three growth types.These results are not in agreement with those of May et al. (1992) who found that small-framed steers had greater mean percentage of 103-rib than mediumor large-framed steers.However, Priyanto et al. (1999) found less 103-rib in Hereford and Hereford×Brahman steers compared to Brahman steers.Miller et al. (1996) also found that Brown Swiss and Okie steers yielded more 103rib than Mexican #2 steers and that the higher yields of rib was likely due to less muscling in the Mexican steers.
Growth types of IE and SE had greater (p<0.05)mean 121-plate than did growth types of LL and IL (8.3±0.2 and 8.4±0.2 vs. 7.1±0.2and 7.8±0.2%,respectively).The IL growth type had greater (p<0.05)yield of mean 121-plate when compared to the LL growth type (7.8±0.2 vs. 7.1±0.2%).These results are not in agreement with those of Belk et al. (1991) who reported that large framed steers had the highest-yielding plate when compared to other frame sizes and the work of Miller et al. (1996) who reported that Brown Swiss and Okie steers yielded more 121-plate than Mexican #2 steers.However, May et al. (1992) found similar numerical mean percentages of 121-plate among large-, medium-, and small-framed steers with 1.5 cm fat thickness and thick, average, or thin muscling.
There was no difference (p>0.05) in mean percentage of 118-brisket among the IL, IE, and SE growth types (4.0±0.1, 4.1±0.1 and 4.1±0.1,respectively).The LL growth type had less (p<0.05)mean 118-brisket (3.8±0.1%) when compared to the other three growth types.These results are not in agreement with those of May et al. (1992), Belk et al. (1991) and Miller et al. (1996) who reported no difference in mean percentage 118-brisket among different biological types of cattle.May et al. (1992) reported similar mean percentages of 118-brisket among steers of three frame sizes (large, medium and small) and three fat thickness levels (.75 cm, 1.5 cm and 2.25 cm).Also, Belk et al. (1991) found similar values for mean percentage 118brisket among large, medium, and small framed steers with three levels of muscle thickness (No. 1, No. 2, No. 3).Miller et al. (1996) found similar mean percentage of 118brisket among Brown Swiss, Okie #1, and Mexican #1 and #2 steers.
The LL growth type yielded greater (p<0.05)mean percentage of 115-hindquarter when compared to the other three growth types.Growth types of IE and SE were similar (p>0.05) for yield of mean percentage of 115-hindquarter.The IL growth type had less (p<0.05)mean 115-hindquarter than did the LL growth type (48.1±0.3 vs. 48.5±0.3%).These results are in agreement with those of Price (1976) who reported a 1.5% greater (p<0.05)mean 115-hindquarter for Charolais×Hereford compared to Hereford steers.These results are not in agreement with those of Garcia- de-Siles et al. (1977) who reported similar percentage of hindquarter for Holstein and Hereford steers.
The mean yield of 164-round was greater (p<0.05) for the LL growth type when compared to the other three growth types (20.3±0.4 vs. 18.2±0.4,17.9±0.4,and 17.5±0.4%).There was no difference (p>0.05) in mean percentage of 164-round between the IE and SE growth types.The IL growth type was intermediate to the LL and IE growth types for mean percentage of 164-round.The results of this study are supported by those of Miller et al. (1996), Belk et al. (1991) and May et al. (1992).Miller et al. (1996) found significant difference in mean percentage yield of 164-round among four biological types of steers.Mean percentage yield of 164-round for the four biological types of steers in this study was less than mean percentage of yield of 164-round found by Miller et al. (1996) for Brown Swiss (23.50),Okie #1 (23.44),Mexican #1 (23.81), and Mexican #2 (25.08) steers.The dairy×Brahman crossbred Mexican #2 steers had about 1.6% more wholesale round than Brown Swiss or Okie steers (Miller et al., 1996).In addition, Reiling et al. (1992) and Williams et al. (1989) reported that 164-round yields differed (p<0.05) between large-and small-framed steers.May et al. (1992) stated that regardless of frame size, and fat thickness the percentage yield of round tended to decrease as muscle score changed from thick to thin.The results of this study are not in agreement with those of Price (1976) who found a nonsignificant 0.3% difference in mean percentage round among Charolais x Hereford and Hereford steers.
Mean percentage of bone was highest (p<0.05) for the LL growth type and lowest (p<0.05) for the SE growth type (19.5±0.5 vs. 16.8±0.5,respectively).There was only a 2.7 kg difference (p<0.05) in mean bone weight between the two extreme growth types LL vs. SE (data not shown).The IL and IE growth types were intermediate to the LL and SE growth types and differed (p<0.05) for mean bone (18.5±0.5 vs. 17.8±0.5%).Mean percentage of bone for the four biological types of steers in this study, generally were higher than those reported by Miller et al. (1992) who reported 16.55, 14.46, 15.22, and 16.06 for Brown Swiss, Okie #1, Mexican #1, and Mexican #2 steers, respectively, but similar to those reported by Taylor (1982) who found percentage of bone for moderate growth rate Friesian (18.99 ±0.24) and Hereford (14.5±0.46) and for high growth rate Friesian (15.87±0.32)and Hereford (13.25±0.38).In the study by Priyanto et al. (1999), at 100 kg of side weight, Hereford steers had less (p<0.05)mean bone than Brahman and Brahman×Hereford steers.Additionally, Arthur et al. (1995) reported that mean percentage bone was 21.5 and 19.5 for yearling domestic and heavy export steers, respectively.These results are supported by those of Koch et al. (1979) and Koch et al. (1976).Koch et al. (1979) showed a small but significant and economically important difference in bone percentage between large-and mediumframed steers.They also found that large-framed Chianina crosses had the highest bone percentage and medium framed Angus crosses had the lowest bone percentage of all biological types studied.Koch et al. (1976) stated that large-framed Simmental and Charolais crosses had more bone than medium-framed Hereford, Angus, Limousin, and South Devon crosses.However, Koch et al. (1976) acknowledged that the differences in percentage bone were small on a weight-constant basis.However, Gregory et al. (1994) stated that total bone percentage for larger-framed steers (Charolais and Limousin) was lower than that of medium-framed steers (Angus, Red Poll and Hereford) when compared on an age-constant basis.
Least squares means and standard errors for the effects of production system on percentage of carcass fore-and hindquarters, percentage of primals and subprimals, and percentage bone in pasture-or feedlot-developed steers are presented in Table 2. Feedlot-developed steers yielded higher (p<0.05)mean 102-forequarter than did pasturedeveloped steers (52.7±0.4 vs. 49.8±0.5%).These results concur with those of Norton (1993) who also found a higher (p<0.05)mean percentage of forequarter in feedlotdeveloped steers compared to pasture-developed steers (51.9±0.1 vs. 51.6±0.1).These results are not in agreement with those of Murray et al. (1974) who compared carcass composition of steers in three production systems, two of which involved high growth (0.8 kg/d) and low growth (0.4 kg/d).The low growth steers yielded higher arithmetic mean percentage of forequarter than did the high growth rate steers (49.0 vs. 48.6).
There was no difference (p>0.05) in mean 113-chuck among the pasture-or feedlot-developed steers (26.5±0.5 and 26.2±0.4%).This result is not in agreement with that of Norton (1993) who found a greater mean percentage of 113chuck for pasture-developed steers when compared to mean percentage of 113-rib for feedlot-developed steers (27.1±0.1 vs. 26.2±0.1,respectively).Likewise there was no difference (p>0.05) in mean percentage of 103-rib between the pasture-or feedlot-developed steers.This result is not in agreement with that of Norton (1993) who reported a higher mean percentage of 103-rib for pasture-developed steers compared to feedlot-developed steers (7.5±0.03 vs. 7.3± 0.03).
Mean percentage of 115-hindquarter was greater (p<0.05) for pasture-developed than for feedlot-developed steers (48.6±0.5 vs. 47.3±0.4).These results are similar to those reported by Norton (1993) who reported 48.3±0.1 and 48.1±0.1% for pasture-and feedlot-developed steers, respectively.These results are not in agreement with those reported by Murray et al. (1974) who found a similar arithmetic mean percentage of hindquarter (51.0) for steers in low growth rate and high growth rate production system.Mean 164-round was greater for pasture-developed than for feedlot-developed steers (21.0±0.6 vs. 16.0±0.5%).Norton (1993) also reported a greater mean percentage of 164round for pasture-developed steers.In the Norton (1993) study, mean percentage of 164-round was 19.9±0.1 and 17.7 ±0.1 for pasture-and feedlot-developed steers, respectively.
The mean percentage of bone was greater (p<0.05) for the pastured steers than for the feedlot steers (21.8±0.8 vs. 14.5±0.6).This 7.3% difference in mean bone percentage among pasture or feedlot developed steers equated to only 1.6 kg difference in mean bone weight among the two production systems (data not shown).These results support those of Schaake et al. (1993), Sapp et al. (1996), and Sully and Morgan (1982) who found a higher mean percentage of bone for cattle in production systems where pasture was all or part of the developmental diet compared to cattle in production systems involving finishing on grain diets.Mean percentages of bone found by Schaake et al. (1993) by production system were: 21.2 off of fescue-clover pasture in the spring; 21.1 off of millet, bermudagrass, or sudan grass in summer; 20.7 off summer pasture and 75 d in feedlot; and 18.3 (SE = 0.41) after weaning, conditioning, and 170 d in feedlot.Sapp et al. (1996) found no difference in mean percentage bone for steers developed on pasture and steers developed on pasture plus approximately 180 d in feedlot (17.66 vs. 17.76) but mean percentage of bone in both production systems differed from mean percentage bone in feedlot-finished steers (16.67).In addition, mean bone for pasture (21.8±0.8%) and feedlot (14.6±0.6%)steers in this study was smaller than the 23.1 and 20.6% for pasture and feedlot steers reported by Sully and Morgan (1982).Murry et al. (1974) suggested that pasture steers have higher mean percentage of bone because they are older at harvest.The results in this study are not in agreement with those of Priyanto et al. (1999) and those of Murry et al. (1974).In the study by Priyanto et al. (1999) there was no difference in mean bone between pasture-and pen-fed steers.Murry et al. (1974) found that there was no difference in mean bone percentage among steers managed for three rates of daily gain.
The beef growth type×production system interaction was an important source of variation in CCW (p = 0.0395), retail cuts (p = 0.001), lean (p = 0.001), fat (p = 0.0101), rump (p = 0.0454), shortloin (p = 0.0482), and flank (p = 0.001) (mean squares shown).Although many studies (Bowling et al., 1978;Harrison et al., 1978;Williams et al., 1989;Sapp et al., 1996;Steen et al., 2003) have been conducted using different biological types (animal types) and production systems (management groups) no test of interaction of these effects were conducted except in the study by Smith et al. (1977) who found no significant animal type×management group interaction for any of the traits in the current study.
Table 3. Least squares means and standard errors for the beef growth type×production system interaction effect on chilled carcass weight (kg), primals and subprimals (%), lean and fat trim (%)and retail cuts (%) in pasture-or feedlot-developed steers Growth type 12 The beef growth type×production system interaction for CCW resulted in a change in ranking of mean CCW among the growth types in the two production systems (rank order interaction) (Table 3).The ranking of the beef growth type× production combination means for CCW was LL-feedlot>IL-feedlot = IE-feedlot>SE-feedlot>LL-pasture> IL-pasture>IE-pasture>SE-pasture.There was no difference (p>0.05) between the IE and IL steers for mean CCW in the feedlot-production system, whereas the IL steers had greater (p<0.05)mean CCW than the IE steers in the pasture production system.Given ample pasture resources for cattle development, IL steers will produce heavier CCW than IE steers.In contrast, should a feedlot-production system be chosen, the IL and IE steers should produce similar CCW.The LL steers had greater (p<0.05)mean CCW in both production systems, but the difference between mean CCW of the LL-feedlot combinations was greater than the difference between mean CCW of the LL-pasture and the IL-pasture combinations (33.4 vs. 30.9 kg, respectively).The difference between IL-pasture and IE-pasture combinations for mean CCW was 11.5 kg while the difference between the IL-feedlot and IE-feedlot combination for mean CCW was 6.4 kg, illustrating the difference in performance among these growth types in the two production systems.The SE steers had lesser (p<0.05)mean CCW when compared to the other beef growth types in both production systems, but there was greater (p<0.05)no difference between the IE-feedlot and SE-feedlot combinations for mean CCW (27.2 kg) than those was between the IE-pasture and SE-pasture combinations for mean CCW (17.2 kg).Chilled carcass weight was lower (p<0.05)for all pastured steers compared to the feedlotdeveloped steers, even after an additional 6-mo age advantage for the pasture-developed steers.Therefore, it would seem likely that the pasture production system did not meet nutrient requirements for maximizing growth and tissue accretion.
The interaction of beef growth type×production system for rump resulted from differences between mean percentages of rump of the beef growth types across the two production systems (an interaction of magnitude).The ranking of the beef growth type×production system combination means for percentage rump was LL-pasture> IL-pasture = IE-pasture = SE-pasture = LL-feedlot>IL-feedlot>IE-feedlot = SE-feedlot.The combinations of ILpasture, IE-pasture, and SE-pasture were similar (p>0.05) for mean rump (5.9±0.3,5.7±0.3 and 5.8±0.3%,respectively) and were different (p<0.05) from the LLpasture combination for mean percentage of rump (6.8±0.3),whereas the IE-feedlot and SE-feedlot combinations were similar (p>0.05) for yield of mean rump (4.2±0.2 and 4.1± 0.2%) and differed (p<0.05) from the yield of mean rump of the IL-feedlot combination (4.6±0.2%).The LL-feedlot combination yielded greater (p<0.05)mean rump than did the IL-feedlot combination (5.5±02 vs. 4.6±0.2%).Previous research has shown that large-framed steers produce higheryielding rump than either medium-or small-framed steers (May et al., 1992;Reiling et al., 1992).
The beef growth type×production system interaction for flank resulted from differences in mean percentage flank between the IE and SE growth types in the two production systems.The IE-feedlot and SE-feedlot combination means were similar (p>0.05) and different (p<0.05) from the ILfeedlot and LL-feedlot combination means for flank (7.9±0.2 vs. 7.4±0.3and 5.9±0.2%,respectively).The SEpasture and IE-pasture combinations were similar (p>0.05) for mean flank (4.7±0.3 vs. 4.6±0.2%).The IL-pasture combination for mean percentage flank was greater (p<0.05)than mean percentage flank of the LL-pasture combination.The LL-pasture combination had the lowest mean percentage flank of all combinations.The ranking of the beef growth type x production system interaction means for percentage flank was SE-feedlot = IE-feedlot>IL-feedlot>LL-feedlot>SE-pasture = IE-pasture>IL-pasture> LL-pasture.Previous research has shown that large-framed steers produce lower-yielding flank than either medium-or small-framed steers (May et al., 1992;Reiling et al., 1992).
The beef growth type×production system interaction for mean lean resulted from differences in the ranking of the IL, IE and SE growth types in the two production systems.The LL-pasture combination yielded greater (p<0.05)mean percentage lean than did the combinations of IL-pasture, IEpasture, and SE-pasture (70.5±1.4 vs. 65.5±1.5, 65.1±1.4,and 65.0±1.5%).Likewise, the LL-feedlot combination was greater (p<0.05) for mean percentage of lean than the combinations of IL-feedlot, IE-feedlot, and SE feedlot (62.0±1.2 vs. 55.1±1.1, 53.6±1.2, and 53.5±1.2).There was no difference in mean percentage lean among the IL-pasture, IE-pasture, and SE-pasture combinations (65.5±1.5, 65.1± 1.4, and 65.0±1.5%).Conversely, the IL-feedlot combination was greater (p<0.05) for mean percentage lean than the IE-feedlot and SE-feedlot combinations (55.1±1.1 vs. 53.6±1.2 and 53.5±1.2%).The ranking of the beef growth type×production system interaction means for percentage lean was LL-pasture>IL-pasture = IE pasture = SE-pasture>LL-feedlot>IL-feedlot>IE-feedlot = SE-feedlot.These data agree with Koch et al. (1976Koch et al. ( , 1979Koch et al. ( , 1982) ) and Dolezal et al. (1993), who found that larger-framed cattle produce leaner carcasses than small-framed cattle when compared on a weight constant or a constant time-on-feed basis.Tatum et al. (1986) reported that larger-framed cattle are leaner than small-framed cattle when compared at similar slaughter weights, because larger-framed cattle tend to fatten at relatively heavy weights.
In general, the ranking of beef growth type x production system interaction means for percentage fat was inversely proportional to the mean percentage of lean.This interaction resulted from differences (p<0.05) in the IL-and IE-feedlot combinations and the similarity (p>0.05) of the IL-and IE-pasture combinations.The combinations of IEfeedlot and SE-feedlot were similar (p>0.05) for mean percentage fat (32.1±1.5 and 32.7±1.4).The IL-feedlot and the LL-feedlot combinations were different (p<0.05) for mean fat (29.8±1.4 vs. 27.1±1.5%)and both combinations were different (p<0.05) from the IE-feedlot and SE-feedlot combinations.The IE-and SE-pasture combinations were similar (p>0.05) for mean fat (12.3±1.8 and 13.6±1.9%).When compared, IL-pasture and IE-pasture combinations were similar (p>0.05) for mean percentage fat (11.5±1.9 and 12.3±1.8).The LL-pasture combination had less mean fat than the IL-pasture, 12.3±1.8,and 13.6±1.9% respectively).The ranking of the beef growth type x production system interaction means for fat was SEfeedlot = IE feedlot>IL-feedlot>LL-feedlot>SE-pasture = IE-pasture = IL-pasture>LL-pasture.Numerous reports have indicated that cattle developed on a forage diet have a smaller proportion of fat than comparable animals finished on grain diets (Wanderstock and Miller, 1948;Dinius et al., 1975;Bidner et al., 1986).Stonaker et al. (1952), Koch et al. (1976Koch et al. ( , 1979Koch et al. ( , 1982)), and Butts et al. (1980aButts et al. ( , 1980b) ) reported that differences between large-and small-framed carcasses tended to be reduced when compared at similar levels of fatness, but larger cattle tend to be older and heavier because of their slower maturity rates and tendency to fatten at heavier weights.
The beef growth type×production system interaction for retail cuts resulted from differences in mean percentage retail cuts for some of the growth types in the two production systems (an interaction of magnitude).There was a 4.9% difference (p<0.05) between the LL-pasture and IL-pasture steers for mean retail cuts.Whereas, the difference (p<0.05) in mean retail cuts between the LLfeedlot and the IL-feedlot was 6.6%.The LL-growth type yielded greater (p<0.05)mean percentage of retail cuts in both production systems.There were no differences (p>0.05) in mean percentage retail cuts between the IE and SE growth types in both production systems, but the pasture system yielded greater (p<0.05)mean percentage retail cuts than the feedlot system (72.5±1.4 and 71.9±1.4 vs. 60.9±1.2 and 60.5±1.1,respectively).The ranking of the beef growth types×production system means for retail cuts was LL-pasture>IL-pasture = IE-pasture = SE-pasture>LL-feedlot> IL-feedlot>IE-feedlot = SE-feedlot.Gregory et al. (1994) showed differences in retail product between two trim levels (8 mm and 0 mm of fat trim).These differences tended to be less in breeds that are larger-framed with lower fat content than in smaller-framed breeds with more fat content.Koch et al. (1979) reported that larger-framed Chianina crosses had more retail product than either medium-framed Angus or Red Poll crosses when compared at a constant carcass weight.Koch et al. (1976) stated that large-framed Charolais, Limousin, and Simmental crosses had higher (p<0.05)percentages of retail product than either medium framed Hereford-Angus, Jersey, or South Devon crosses.

IMPLICATIONS
Differences in mean CCW between IL-and IE-pasture combinations and IL-and IE-feedlot combinations illustrate performance differences among beef growth types in two production systems.Given ample pasture resources, IL steers will produce heavier CCW than IE steers.In contrast, IL and IE steers should produce similar CCW under a feedlot production system.Pasture system yielded greater mean percentage retail cuts and bone than the feedlot system.The CCW was lower for all pastured steers compared to feedlot-developed steers indicating the pasture system did not meet nutrient requirements for maximum growth and tissue accretion.Beef growth type x production system interaction for mean percentage fat was inversely proportional to mean percentage lean and further indicated that cattle developed on forage have less fat than comparable animals finished on grain.While genetic potential may be more effectively exploited in a feedlot situation, a need exists within the cattle industry to correctly match nutrient requirement of animal growth type to resources regardless of production system.

Table 1 .
113-chuck differed (p<0.05)between the LL and IL growth types (26.9±0.3 vs. 26.3±0.3%)and these growth types had greater (p<0.05)yield of mean percentage of 113 -chuck than IE and SE growth types (26.1±0.3).There was no difference (p>0.05) in mean 113-chuck of the IE and SE growth types (26.1±0.3 and 26.1±0.3%).Our results are in Least squares means and standard errors for the effects of beef growth type on carcass forequarter and hindquarter (%), primals and subprimals (%) and bone (%) in pasture-or feedlot-developed steers

Table 2 .
Least squares means and standard errors for the effects of production system on carcass forequarter and hindquarter (%), primals and subprimals (%), and bone (%) in pasture-or feedlot- c, d Means in the same row with different superscripts differ (p<0.05).