Feed Efficiency, Carcass Characteristics, and Sensory Quality of Lambs, With or Without Prolific Ancestry, Fed Diets with Different Protein Supplements' M. H. Fahmy2~3,J. M. Boucher4, L. M. Poste5, R. Gregoire4, G. Butlers, and J. E. Comeau7
Agriculture Canada and Quebec Ministry of Agriculture, Fishery, and Food
ABSTRACT: Data were collected on 130 intact male lambs fed diets based on roughages supplemented with fish meal, soybean meal, or corn gluten-blood meal and slaughtered at 43 kg live weight. A nonsupplemented group served as a control. The lambs represented Romanov (R), Finnsheep (F), a new breed developed in Canada (DLS), Coopworth (C1, and Suffolk (3, three first crosses of DLS with R, F, and Booroola Merino (B), and their backcrosses to DLS. Supplemented lambs consumed 16 to 22% less ( P < .05) silage than control lambs. Average daily gains of lambs fed fish meal (226 g) and corn gluten-blood meal (217 g) were higher and feed conversion ratio (4.99 and 5.11) lower than that of lambs fed soybean meal (189 g and 5.48) or control (186 g and 5.76) diets ( P < .05). The cost of feed per kilogram of gain or per kilogram of lean produced was similar in the four treatments but was between C$.23 and .65 cheaper in the protein-supplemented groups when the number of days to reach slaughter weight was considered. The effect of diet on
carcass traits and meat quality were minimal. Meat of lambs fed the three protein supplements was less juicy than that of control lambs. With a few exceptions, most of the significant differences among genetic groups in growth, carcass, and sensory traits were mainly between prolific (R and F) and meat-type breeds (C, S, and DLS). Gain in weight was highest in S lambs (199 g/d), but F and R first crosses were the youngest a t slaughter (196 and 198 d). The F lambs had higher dressing and kidney fat percentages than meat-type breeds. The DLS lambs had the largest longissimus muscle area (14.0 cm2), whereas C had the smallest (10.7 cm2).The B crosses had larger longissimus muscle area than R and F crosses. The R lambs had more lean and less fat in the 12th rib, whereas C lambs had the lowest lean and a high bone percentage. The toughest and the most tender roasts were those of R and B crossbred lambs, respectively. Roasts from F lambs had the most intense lamb flavor.
Key Words: Protein Supplements, Growth, Feed Conversion Efficiency, Carcass, Sensory Evaluation, Sheep J. h i m . Sci. 1992. 70:1365-1374
'Contribution No. 357 of Lennoxville Research Station, No. of Food Res. Centre, No. RO90 of Res. Program Service. The authors are grateful to Claire Corriveau, Ramez Fahmy, Christine Cellier, and Marjolaine St. Louis for carcass evaluation at Lennoxville, to J. Elsaesser and V. E. Agar for sensory evaluation at FRC, to Gilles Gagnon and his team for care of animals and data collection at L a Pocatiere, to Andre Belleau for data processing, to Louise Boisvert for preparing the manuscript, and to B. K. Thompson for helpful suggestions.
894
2To whom requests for reprints should be addressed. 3Agric. Canada, Lennoxville Res. Sta., Quebec, J1M 123. 4Quebec Ministry of Agric., Fishery, and Food, Deschambault Res. Sta., GOA 1SO. 5Agric. Canada, Food Res. Centre, Ottawa, K1A OC6. 6Agric. Canada, Res. Program Service, Ottawa, K1A OC6. 7Agric. Canada, La Pocatiere Exp. Farm, Quebec, GOR 1ZO. Received July 22. 1991. Accepted December 31, 1991.
-
1365
Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
1366
FAHMY ET AL.
Introduction
mented with up to 200 g per lamb daily of a grain mixture of different cereals. Before starting the feeding trials, the lambs were dewormed, treated The new trend toward using prolific ewes and against enterotoxemia and tetanus, and then their crosses in sheep farming to improve productivity has increased choices among combinations injected with selenium and vitamins A, D, and E. of prolific and prolific x native lambs for market. The lambs were fed from an initial weight of about Research information concerning growth, carcass, 23 kg to a slaughter weight of about 43 kg live and meat quality of these lambs is meager. weight. Average age of lambs at the beginning of Research on carcass and sensory quality of the feeding test varied between 105 d for DLS and Finnsheep and their crosses (Thomas et al., 1976; 117 d for R, whereas for S it was 83 d. Dahmen et al., 1979; Fahmy, 1979, 1985; Kemp et The feeding trials were carried out in a separate al., 1981; Magid et al., 19811 and on Romanov sheep barn with a concrete floor divided into 12 (Theriez and Tissier, 1975; Sanudo and Sierra, pens of equal size (16 m21.Each pen provided 1982; Fahmy, 1986; Pommier et al., 19901 has adequate feeding space and drinking facilities. indicated that these prolific breeds produce acStraw was used for bedding. ceptable market lambs. These studies also showed Management and Feeding Treatments. The feeding that excessive fat deposition occurs in the body study started on May 30, 1988. The animals were cavity, whereas fat cover on the carcass is often deprived of feed and water for 12 h then weighed. inadequate to attain higher grades. They were distributed among 12 pens to form Little is known about the most effective way of groups of equal initial weights as much as possifeeding the prolific breeds and their crosses. ble. Each pen contained 11 lambs, one of each of Although the excessive deposition of body cavity the 11 genetic groups. Only nine 1/2 B lambs were fat is believed to be genetic in origin, it is possible available; therefore, 1/4 B lambs were used to that feeding diets rich in protein may favor more complete three pens. One C and one DLS lamb rapid development of lean tissue and lower fat that started the study were omitted from the deposition. Hassan and Bryand (1986) showed that analysis for failure to grow. lambs given fish meal had greater gain than All the animals were fed a basal diet composed unsupplemented lambs. Other evaluations of proof 75% grass silage and 25% barley-corn concentein supplements include those of Ganev et al. trate, presented separately. This basal diet was (1979) and Silva and 0rskov (19881, who fed supplemented with one of the following protein soybean and fish meal, and of Stock et a1. (1983) sources: fish meal, soybean meal, or a mixture of and Nelson et al. (19851, who fed blood and corn gluten and blood meal. A fourth diet with no soybean meal. However, none of these studies protein supplement served as a control. Composiinvolved lambs of prolific breeds or studied sen- tion of the different diets and their nutritive values sory qualities of meat, and therefore the present are presented in Table 1. The barn was divided study was designed to provide such information. into three sections of four pens each. The four treatments were assigned within each section to ensure equal representation of the different treatMaterials and Methods ments among sections of the barn. The lambs were fed once a day at 0930. The Experimental Design amounts offered and refused were weighed and recorded daily. The amounts offered were adjusted Animals. This study was conducted at L a according to the amounts refused; if there was e Pocatiere Experimental Farm using 130 intact 5% (on a DM basis) left for three consecutive days, male lambs. The lambs represented two prolificthe amounts were increased by 5%. If however, the types, Romanov (R1 and Finnsheep (F),and three amounts refused exceeded 5% for 3 d, the level of meat-types, DLS (a new breed developed in Canfeeding was reduced by 5%. ada, Fahmy, 19881, Coopworth (C1, and Suffolk (SI. Measurements and Traits Studied. Data were In addition, six crosses were available, first collected on body weight at 14-d intervals (mainly crosses resulting from mating R, F, and Booroola to monitor rate of gain) and on feed consumption Merino (B1 rams to DLS ewes and backcrosses daily for 96 d from the beginning of the feeding resulting from mating rams from these first test. Because some lambs were heavier at the crosses to DLS ewes. All the lambs were born at L a Pocatiere between January 25 and March 1, beginning or gained faster than others on feed, they reached the assigned slaughter weight earli1988 except S, which were born between February er; accordingly, not all animals completed 96 d on 17 and April 4 and were purchased from a private test and number of animal-days was calculated breeder after weaning. The lambs were weaned at and used to obtain feed conversion ratio. Some 50 d of age and, until the start of the study, were animals did not reach slaughter weight a t the end allowed ad libitum access to grass silage suppleDownloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
1367
PROTEIN SUPPLEMENTATION FOR LAMBS
of the 96-d feeding test period. These lambs continued to receive the same diets but feed consumption was not monitored. The growth rate reported is that made during the 96-d feeding test period. Lambs were slaughtered at a local abattoir when they attained about 43 kg live weight. The chilled carcasses were weighed 3 d after slaughter and the ratio of carcass to slaughter weight was calculated (dressing percentage). The carcasses were then divided into three wholesale cuts, shoulders, loin-racks, and legs, and each was expressed as a percentage of chilled carcass weight. Kidney and pelvic fat was removed from the body cavity and weighted. The l l t h , 12th, and 13th ribs from the right side were separated from the rest of loin-rack area and transferred to the laboratory for further measurements. For sensory quality evaluation, a sample comprising the gth to 13th ribs from the left side was separated from the carcass and frozen until all the animals were slaughtered. The frozen samples packed with dry
ice were transported to the Food Research Centre in Ottawa for evaluation. On arrival at the Lennoxville Station laboratory, color readings of the longissimus muscle were taken on fresh cuts after the exposed surface of the muscle was removed to avoid any effect of oxidation. The color readings were taken with a LabScan Model LS 5100 (Hunter Laboratories, Reston, VAL The search head was equipped with a green filter (555 nm peak) and calibrated using a grey meat color standard. Three measurements were taken, Hunter L, a, and b (CIE, 19761, which measure, respectively, the color between black (0) and white (1001, red (+I and green (-1, and yellow (+I and blue (-1. The samples were then frozen. The other measurements taken on thawed samples of the 12th rib were backfat thickness over the longissimus muscle (average of three measurements a t the two extreme ends and halfway between them) and area of the longissimus muscle measured from a tracing on acetate paper using a planimeter (Planix, Tamaya, Tokyo, Japan).
Table 1. Composition and chemical analysis of the different diets fed to lambs (percentage of total diet) Diet supplement
Item Diet composition Grass silage Barley Corn Salt Ca, PO supplement Magnesium, 54% Mineral and vitamin Subtotal Protein supplement Fish meal (8O0/0 CPIa Soybean meal (48% CPI Blood meal (80% CP) corn gluten meal (60% CP) Total diet Chemical analysis of the combined diet and supplement CP TDN Calcium Phosphorus Magnesium Sulfur Cost of diet CC$/l,OOO kg DMI Concentrate Silage
Fish meal
Soybean meal
Corn gluten and blood meal
Control
75.0 16.0 7.6 .2 1.1 .04 .o 1 100.00
75.0 18.0 7.8 .2 1.1 .04 .01 100.00
75.0 16.0 7.6 .2 1.1 .04 .01 100.00
75.0 18.0 7.8 .2 1.1 .04 .01 100.00
-
Added relative to basal diet 6.9 1.9 2.5
105.1
108.9
O '/
5.1
OO /
104.4
100.00
of Total diet, DM basis
16.0 65.6 1.o .8 .2 .2
18.0 68.5 .7 .4 .2 .2
16.0 65.7 .7 .4 .2
13.4 65.0 .7 .4 .2 .2
194.3 80.0
185.4 80.0
181.6 80.0
139.2 80.0
*Percentage of protein in protein supplement on a n as-fed basis.
Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
.2
1368
FAHMY ET AL.
The 12th rib was separated from surrounding ribs, trimmed, weighed, and dissected into lean, fat, and bone, each expressed as a percentage of total 12th rib weight. Connective tissues surrounding the longissimus muscle were removed and the muscle was then ground. Dry matter content of the longissimus muscle was determined by freezedrying the entire sample. A subsample was used to determine intramuscular fat content using Soxhlet high temperature extraction units (Tecator, Hoganas, Sweden). Another subsample was used to determine total and soluble collagen applying the method of Hill (19661. Only samples of the pure breeds were analyzed for collagen content. Previous experience involving five of the crosses used in this study showed no significant differences among the crosses (M. H. Fahmy, unpublished data). Two lambs, one with abnormal lean, fat, and bone percentages and the other with excessively high collagen contents, were considered to be outliers, and the measurements from these animals were removed from the analyses.
Sensory and Instrumental Meat Quality Evaluation Sample Preparation. Lamb roasts were thawed a t 4°Cfor 24 h and oven-ready weights were record-
ed. The roasts were placed on metal racks in aluminum pans and cooked, uncovered, at 163" C in electric household ovens (Modern Maid) to an internal temperature of 75" C. Temperature was determined by a Type-T copper constantan thermocouple (Thermo-Electric, Montreal, Canada) inserted into the geometric center of the longissimus muscle of each roast. After a 10-minstanding time, final cooking weights were recorded. The longissimus muscle was excised from surrounding bone and fat then cut into 4-mm-thick slices with a Hobart electric slicer. The samples were kept warm (40 to 44" C) in a foil-lined, covered container placed in a 75°C oven for a maximum of 15 min. Each panelist received two slices of each sample from the same position ke., Slices 1 and 2 went to Panelist 1; Slices 3 and 4 went to Panelist' 2, etc.). The slices were served to panelists in random order. Training. Seven selected assessors with previous meat evaluation experience were trained (in four 30-min sessions) for this study. The scales and anchor points were derived through discussions. Tenderness, juiciness, and lamb flavor were selected to start training based on past experience on lamb projects. The panelists were encouraged to use other descriptors to describe intersample differences. The presence or absence of off-flavor was evaluated by use of a binomial response (yes or no). To familiarize the panelists with the intensity of lamb flavor, two roasts were prepared. Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
One was roasted to an internal temperature of 70°C and the other was placed in boiling water to an internal temperature of 75" C. At another sitting, three roasts were roasted to internal temperatures of 65, 70, and 80°C. The third and fourth sittings were run-through sessions, and each roast w a s roasted to a temperature of 75°C. Sensory Panels. The roasts were evaluated in 22 sittings, four roasts at a time. Each of seven selected trained assessors received a set of four samples, one from each roast. The first 14 sittings (56 roasts) were a balanced, incomplete block design (BIBD, Cochran and Cox, 1957) used to compare eight of the genetic groups (pure breeds and first crosses) wherein the panelists compared four roasts from different genetic group on the same diet. Two additional sittings (resulting in 16 sittings in total were held so that each diet could be evaluated in four sittings. A second design was employed to compare the sensory characteristics of the four diets; in this design three replicates of each of two backcrosses (1/4 F and 114 R) were presented in six sittings. At each sitting, roasts of lambs from one cross raised on each of the four diets were compared. Water a t room temperature, unsalted crackers, and two wedges of a Japanese pear were provided to remove residual flavors from the mouth. Assessors evaluated the perceived intensity of lamb flavor, tenderness, and juiciness using a 15-cm, unstructured line scale (Larmond, 1977). From left to right the descriptive anchor points (1.5 cm from each end) were as follows: tenderness, very tough to very tender; juiciness, very dry to very juicy, and lamb flavor, slight to intense. The presence or absence of off-flavor was also evaluated as indicated previously.
Statistical Analysis The data on growth and carcass were analyzed using the GLM procedure of SAS (19851. The mathematical model used was Yijk] = u + bi + f j + bfij + gk + fgik + eijk, where Y is the performance of a n individual lamb, u is the overall mean, b is the block effect, f is the feeding treatment effect, bf is the amongpen error used to test f, g is the genetic group effect, fg is the feeding treatment x genetic group interaction, and e is the among-animal error used to test g and fg. The feeding treatment effect was compared to the among-pen error because feeding treatments were applied to the whole pen. In no instance was among-pen error significantly greater than among-animal error. When treatment effect was significant compared to among-animal error (with greater power), as was the case for two traits, this result was reported in the text. The means presented are adjusted for the slight imbalance in the data. Traits not expressed as percent-
PROTEIN SUPPLEMENTATION FOR LAMBS
ages were adjusted to a common weight by including slaughter weight as a covariate in the analysis model. The covariate initial age (at the beginning of the test) was also considered and discussed in the text, but the figures in the tables are not adjusted for this effect. Six orthogonal contrasts were used to partition the effect due to genetic groups (10 dfl. These were, among the three genetic group types with 2 df (pure breeds vs first and back crosses), meat-type pure breeds (C, S, and DLS) vs prolific-type pure breeds (R and F), among meat type (2 dfl, R vs F, among first crosses, and among backcrosses b o t h with 2 do. Differences among the four diets were noted by examining Fisher’s protected lsd. The first 16 sessions of the meat quality data were examined for diet x genetic group interactions using SAS (1985).None was found significant, so the breed results were based on 14 sessions, because this design was balanced for genetic group. These were examined as a BIBD using GENSTAT (Alvey et al., 1983). The mathematical model used W a s Yijk = U + Si + pi + Spij + gk + sgik + eijk, where Y is the sensory measurement, u is the overall mean, s is the sitting or session effect, p is the panelist effect, sp is the session x panelist interaction, g is the effect of genetic group, sg is the session x genetic group interaction or the among-roast error used to test g, and e is the among-sample error. The sensory data from the six sessions conducted to test the diet effect were examined using SAS (1985) and applying the mathematical model Yijkl = U + gi + bi + gbii + Pk + gbpijk + fl + gfil + gbfijl + eijkl, where Y, u, g, p, and e are as defined in the previous model and b is the replicate effect, gb is the among-session error, gbp is the session x panelist interaction, f is the feeding treatment, gf is genetic group x feeding treatment interaction, and gbf is the among-roast error used to test f and gf. Again, contrasts and lsd were used to examine the effect of genetic groups and diets, respectively. Frequencies of off-flavors were also examined in contingency tables.
Results Effect o f Feeding Treatment. The total gain made by the lambs fed the different diets was similar because initial weights were balanced among treatments and slaughter weight was fixed at around 43 kg live weight. However, the lambs fed the control and soybean meal diets averaged 12 d longer time on feed ( P e .05) than lambs fed the other two diets. During the feeding test (96 dl, ADG of lambs fed fish meal (226 g) and corn glutenblood meal (217 g) were higher (P < .01) than those Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
1369
of lambs fed the soybean meal (189 g) and those fed the control diet (186 g). Feed conversion ratio was lowest in lambs fed fish meal (4.99) and corn gluten-blood meal (5.111, followed by those fed soybean meal (5.481, which in turn were superior to those fed the control diet (5.76, P < .01, Table 2). Lambs fed diets supplemented with protein consumed between 16 and 22% less silage ( P e .01) than those fed the nonsupplemented diet (Table 21. The consumption of concentrate mixture was similar in the four groups. Using the prevailing prices at the time this experiment was conducted, the cost of feed per kilogram of live weight, dressed weight, or per kilogram of lean meat produced (using the dissection of the 12th rib as a n indicator of carcass lean content) was similar in the control and the supplemented groups (Table 2). Extending the calculations further to take into account the rate of turnover of capital (multiplying by number of anirnal-days/lOO), lambs fed the control diet tended to cost more to produce 1 kg of lean than those fed corn gluten-blood meal and fish meal diets, although not significantly so (P < .10 > .05, Table 2). The effect of feeding treatment on carcass traits and composition was slight (Table 2). Compared to among-animal error, treatment effects on dressing and kidney fat percentages were significant at the 5 % level. The analysis of the 12th rib showed that lambs fed the control diet also had less lean and more fat content than lambs fed the supplemented diets. No other significant diet effects were observed in color, intramuscular fat, DM, or total collagen content. The interaction of diet x genetic group was significant only for kidney fat percentage. This resulted mainly from two high values ( > 5%)for F fed soybean meal. When these values were removed, the interaction became nonsignificant. Effect o f Genetic Group. The initial weight ranged between 20.5 kg for F and 29.1 kg for 1/2 F, but all lambs were slaughtered at about 43 kg live weight (Table 31. Average daily gain to slaughter was generally low for the different genetic groups, ranging from 141 g/d for R lambs to 199 g/d for S lambs. Finnsheep-DLS first-cross lambs were the youngest (196 d) at slaughter, whereas F lambs were the oldest (264 d). When the covariate initial age was added to the model to adjust for the younger age of S at the beginning of the test, the orthogonal contrast comparing pure breeds vs crosses was nonsignificant for ADG and age at slaughter. Significant differences observed among genetic groups in many of the carcass traits studied were mainly between prolific and meat-type lambs (Table 3). Finnsheep lambs had higher dressing percentage than meat-type lambs such as S and C.
1370
FAHMY ET AL.
The proportions of the three wholesale cuts varied within narrow ranges among the different genetic groups. The two prolific breeds, R and F, had slightly higher shoulder proportions than meattype animals (39.3 vs 37.2%),and the crosses were intermediate. As expected, F lambs had the highest kidney fat percentage (3.70/0),followed by R (3.4%) and R-crosses (2.9%). Meat-type breeds and B-crosses had the least kidney fat percentage. Conversely, the meat-type lambs had thicker backfat than the prolific lambs. The DLS lambs had the largest and C lambs the smallest longissimus muscle area (14.0vs 10.7 cm2). Among the
first crosses, the B-cross had larger longissimus muscle area then F- and R-crosses. Percentage of lean, fat, and bones in the 12th rib varied slightly between the different genetic groups. Romanov lambs had significantly more bone and less fat than F, whereas C lambs had the lowest lean and highest bone percentages among the other meattype breeds studied. Romanov lambs were among the lowest in intramuscular fat (8.9%), significantly lower than in F lambs (1 1.7%). Some differences in color of the longissimus muscle were significant, mainly between prolific (paler) and meat-types and between prolific and B first-cross (paler). Romanov
Table 2. Least squares means for average daily gain, feed conversion ratio, and carcass characteristics of lambs fed different protein supplements in diets Feeding treatment
Corn gluten Item No. of lambs Initial wt, kgb Initial age, d Slaughter wt, kg Feed consumption, kg of DMAamb SilageC ConcentrateC Total' Gain in wt, kg/lambc Feed conversion ratioC No. of animal-daysc ADG on test, g Cost per lambCd Cost per kg gainCd Cost per kg dressedCd Cost per kg leanCd Cost, adj. for dayCd Carcass percentages Dressing Leg Loin-rack Shoulder Kidney fat 12th Rib measurements Avg. fat thickness, mm Longissimus muscle area cm2 Muscle, % e Fat, 0% e Bone, Oh e Hunter color L Hunter color a Hunter color b Intramuscular fat, O/O Dry matter, % Total collagen content, % Soluble collaaen - content. %
Fish meal
Soybean meal
32 25.0 111 44.2'
33 24.3 110 42.58
56.7' 22.3 79.d 15.8 4.99f 70. If 226' 8.87 .56 1.37 3.29 2.3 1
and blood meal
Control
SEM'
33 25.3 109 43.1'8
32 24.6 108 43.3fg
.82 1.5 .32
55.2' 23.0 78.2' 14.2 5.488 76.0'8 1898 8.23 .58 1.39 3.30 2.51
52.4' 20.4 72.9' 14.3 5.1 1'8 66. 1' 217' 7.90 .55 1.36 3.29 2.18
67.58 22.0 89.58 15.6 5.76' 83.88 1868 8.47 .55 1.33 3.37 2.83
41.4 33.4 28.9 37.8 2.8
42.4 33.5 29.1 37.6 2.9
40.8 33.6 28.6 37.8 2.6
42.2 32.9 29.0 38.1 3.2
.53 .26 .22 .18 .13
3.9 12.6 42.0' 36.3' 20.6 26.1 5.1 5.7 9.6 24.5 2.05 .33
4.4 12.3 41.Ef 37.5' 19.5 25.5 5 .O 5.4 10.1 24.7 2.01 .29
4.2 11.9 41.6f 35.8' 21.2 26.1 5.3 5.7 9.6 24.6 2.08 .35
4.5
11.9 39.91 39.28 19.8 25.5 5.3 5.6 9.6 24.6 2.08 .32
.20 .26 .35 .53 .43 .54 .18 .15 .49 .15
NS NS
.08 .o1
NS NS
aMaximum standard error of the mean based on 6 df. bInitial weight for four animals was missing. CAdjusted means based on three pens. dAll prices are in Canadian dollars. e o n e outlier removed; see text. 'SgMeans followed by different letters are significantly different (P *,**P5 .05 and P 5 .01, respectively; NS, not significant.
Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
,051 except that of diet on juiciness (P c .05).
and F lambs had more (P < .05) total and soluble collagen in the longissimus muscle than the meattype breeds (Table 3). However, when initial age was included in the analysis model, the contrast of prolific vs meat-type in soluble collagen became nonsignificant . Sensory Evaluation. Diet x breed interactions were found nonsignificant, both within the 16 sittings to compare breed group and the six sittings to compare diets. No significant differences among diets were observed in any of the cooking traits (Table 4). Only juiciness showed a significant diet effect: meat of lambs fed the control diet was more juicy than that of those fed the three protein-supplemented diets. The comparison of the five pure breeds and the three first crosses indicated no significant effects observed for the traits of cooking rate, weight loss, and drip loss (Table 51. The sensory analysis results showed significant differences in tenderness among the crosses despite a good deal of variation among roasts for this trait. The 1/2 R and 1/2 F were tougher than
the 1/2 B. No significant effect was found for juiciness. Meat of F lambs exhibited the most intense flavor (10.51,significantly more than for R lambs or the three meat-type breeds. The amongcrosses contrast for flavor was barely nonsignificant ( P = .054). A significantly higher percentage of off-flavors was found in roasts from prolific breeds than in roasts from meat-type breeds (11.8 VS 4.8%).
Discussion Protein supplementation of the diet of sheep a t different stages of their life cycle has been studied for relation to growth, feed efficiency, milk and wool production, and, to a lesser extent, for carcass quality (Bowman and Paterson, 1988; Gunderson et al., 1988; Tayer and Bryant, 1988; Hovel1 and IZlrskov, 1989; Vipond et al., 1989). There is general agreement that protein supplementation has a beneficial effect on growth and carcass traits, manifested in a n increase in daily
Table 5. Cooking data and sensory meansa and their standard errors (SEM) Pure breedsb Meat-type Trait Cooking rate, min/100 g Weight loss, g Drip loss, O/O Tenderness Juiciness Flavor Off-flavor,%
Prolific-type
First crosses
F
S
C
DLS
R
F
11 14.9 4.1 9.3 7.7
13 16.3 4.9 7.1 8.4
10 15.8 3.7 6.8 7.8
15 18.5 3.9 6.8 7.7
11 15.5
11 16.0
14 16.0
11 13.2
6.9 1.05
3.8 7.5 8.7
3.9 5.9 7.6
3.5 6.3 8.4
3.8 9.1 7.7
.41 .88 .37
8.1 1.7
7.4 4.8
8.1 7.8
8.5 12.1
10.5 11.5
7.9 7.3
8.4 10.0
9.3 3.5
.40 3.31
1/2
R
1/2
1/2
B
SEM
SOCc 5
2,4 2
&Means were based on at least 54 observations and the standard error on 35 df. = Coopworth; DLS = a new breed developed in Canada; R = Romanov; F = Finnsheep; B = Booroola Merino. 'SOC = significant orthogonal comparisons, codes as follows: 1, purebreeds vs first crosses; 2, meat-type vs prolific-type, 3, among meat-types; 4, R vs F; 5 , among first crosses.
bS = Suffolk; C
Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
PROTEIN SUPPLEMENTATION FOR LAMBS
intakes and improved daily gains, carcass weights, and carcass characteristics. However, in many cases the improvements failed to reach significant levels. Many different sources of protein supplementation were compared (Loerch and Berger, 1981; Umunna et al., 1982; Stock et al., 1983; Yilala and Bryant, 1985; Butler and Laycock, 1987; Duff et al., 1988).For the traits examined, these studies as well as the present study have shown small differences from the source of supplementary protein. When soybean meal was compared with other sources of protein supplement, lambs fed the soybean meal showed slightly inferior gains and feed conversion efficiency (Stock et al., 1983). It seems, however, that protein supplementation may have a n effect in reducing fat deposition in the body, as evidenced from the present data. Similar findings were reported by Vipond et al. (1989) using fish meal supplementation. The taste panel did not detect significant differences among meat samples of lambs fed the different supplemented diets, indicating that these diets may have little effect on meat tenderness, juiciness, and flavor, although there are some indications that lambs fed the soybean meal supplement were slightly more tender and had less off-flavor than those fed fish meal and corn gluten-blood meal or no supplement. Similar results were reported by Shqueir et al. (19841, who found no difference among diets (one of which included liquefied fish) on lamb flavor. The absence of a diet x genetic group interaction in all the traits studied (except for kidney fat percentage) indicates that prolific and nonprolific sheep and purebred and crossbred lambs respond similarly to differences in the diets offered. Comparative studies involving many genetic groups of sheep are scarce. Kempster and Cuthbertson (19771 surveyed carcass characteristics of the main types of British sheep and showed that many differences exist between breeds. Growth, carcass characteristics, and meat quality of lambs from breeds used in the present study were reported in numerous publications (Theriez and Tissier, 1975; Thomas et al., 1976; Dahmen et al., 1979; Fahmy, 1979, 1985, 1986; Kemp et al., 1981; Pommier et al., 1990). However, few comparative studies involving prolific breeds and their crosses are found in the literature (Jakubec and Krizek, 1988; Young and Dickerson, 19881. Considering age a t slaughter as a n indicator of overall growth rate, 1/2 F and 1/2 R lambs were youngest, followed by S lambs, whereas F and 112 B lambs ranked last. The two prolific breeds had the highest percentage of kidney fat on all diets; therefore, it does not seem that the protein supplementation applied in this study changed this tendency. In contrast, the Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
1373
three nonprolific breeds had a thicker fat cover on the body. Both prolific breeds had a higher percentage of off-flavor than did a standard meat-type breed such a s the S breed. In conclusion, the protein supplementation applied in this study had some beneficial effect on the traits studied, although the differences in many cases failed to reach statistical significance. The different genetic groups responded in a similar manner to these treatments. There were many differences among the genetic groups, mainly between prolific and nonprolific breeds. The first and back crosses with the DLS were generally similar in performance.
Implications Supplementing the diets of growing lambs with soybean meal or industry byproducts to increase protein content may seem at first to be an expensive proposal; however, when one considers the better feed conversion and the shorter period on feed, protein supplementation may prove to be advantageous. Products such as fish meal, which have been reported to affect flavor of meat in chicken, had no such effect on sheep meat, and therefore can be recommended in lamb diets. Based on the results of this study, wide-scale use of prolific breeds and their crosses to increase productivity of sheep will not have an adverse effect on quality of the carcasses or the meat available for consumers.
Literature Cited Alvey, N. G., C. F. Banfield, R. I. Baxter, J. C. Gower, W. J. Krzanowski, P. W. Lane, P. W. Leech, J. A. Nelder, R. W. Payne, K. M. Phelps, C. E. Rogers, J.G.S. Ross, H. R. Simpson, A. D. Todd, G. Tunnicliffe-Wilson, R.W.M. Wedderburn, R.P. White, and G. N. Wilkinson. 1983. GENSTAT. A general statistical program. Numerical Algorithms Group Ltd., Oxford, UK. Bowman, J.G.P., and J. A. Paterson, 1988. Evaluation of corn gluten feed in high-energy diets for sheep and cattle. J. Anim. Sci. 66:2057. Butler, G., and K. A. Laycock. 1987. Supplementation of silage based diets for lambs. Occasional Symposium, Br. Gassl. SOC. 21:113. CIE. 1976. Commission Internationale de 1’Eclairage. 18th Session, Sept. 1975. CIE Publication No. 36. London. Cochran, W. G., and G. M. Cox. 1957. Experimental Designs (2nd Ed.). John Wiley and Sons, New York. Dahmen, J. J., D. D. Hinman, J. A. Jacobs, and D. 0. Everson. 1979. Performance and carcass characteristics of Suffolk sired lambs from Panama and Finn x Panama dams. J. Anim. Sci. 49:55. Duff, G. C., A. L. Goetsch, K. M. Landis, A. C. Hardin, S. R. Stokes, Z. B. Johnson, and K. L. Hall. 1988. Mixing or alternating dietary crude protein sources and performance of wether lambs. Can. J. Anim. Sci. 68:569. Fahmy, M. H. 1979. Body and carcass measurements of DLS
1374
FAHMY ET AL.
and Finn x DLS lambs slaughtered at three light body weights. World Rev. Anim. Prod. XV(4):ll. Fahmy, M. H. 1985. The accumulative effect of Finnsheep breeding in crossbreeding schemes: Growth and carcass traits. Can. J. h i m . Sci. 65:811. Fahmy, M. H. 1986. Preliminary results on fertility, prolificacy, lamb production and carcass traits of Romanov sheep in Canada. Proc. 3rd World Congr. Genet. Appl. Livest. Prod. IX:559. Lincoln, NE. Fahmy, M. H. 1988. Development of DLS sheep: lamb production of the pure breeds, initial crosses and first generation DLS. World Rev. Anim. Rod. 24(3):77. Ganev, G., E. R. Brskov, and R. Smart. 1979. The effect of roughage on concentrate feeding and rumen retention time on total degradation of protein in the rumen. J. Agnc. Sci. (Camb.1 93651. Gunderson, S. L., A. A. Aguilar, D. E. Johnson, and J. D. Olson. 1988. Nutritional value of wet corn gluten feed for sheep and lactating dairy cows. J. Dairy Sci. 71:1204. Hassan, S. A,, and M. J. Bryand. 1986. The response of store lambs to dietary supplements of fish meal. 2. Effects of level of feeding. Anim. Prod. 42:233. Hill, F. 1966. The solubility of intramuscular collagen in meat animals of various ages. J. Food Sci. 31:151. Hovel], F.D.D., and E. R.Brskov. 1989. The role of fish meal in rations for sheep. Tech. Bull. Int. Assoc. Fish Meal Manufacturing No. 23. Herts, UK. Jakubec. V., and J. Krizek. 1988. Breeding and exploitation of prolific breeds in Czechoslovakia. J. Agnc. Sci. Finl. 60:505. Kemp, J. D., D. G. Ely, J. D. Fox, and W. G. Moody. 1981. Carcass and meat characteristics of crossbred lambs with and without Finnish Landrace breeding. J. h i m . Sci. 52: 1026.
Kempster, A. J., and A. Cuthbertson. 1977. A survey of the carcass characteristics of the main types of British lamb. Anim. Prod. 25:165. Lamond, E. 1977. Laboratory methods of sensory evaluation of food. Publ. 1637. Communication Branch, Agriculture Canada, Ottawa. Loerch, S. C., and L. L. Berger. 1981. Feedlot performance of steers and lambs fed blood meal, meat and bone meal, dehydrated alfalfa and soybean meal as supplemental protein sources. J. h i m . Sci. 531198. Magid, A. F., V. B. Swanson, J. S. Brinks, G. E. Dickerson, J. A. Crouse, and G. M. Smith. 1981. Border Leicester and Finnsheep crosses. 111. Market lamb production from crossbred ewes. J. h i m . Sci. 52:1272. Nelson, M. L., I. G. Rush, and T. J. Klopfenstein. 1985. Protein
Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
supplementation of ammoniated roughages. 11. Wheat straw supplemented with alfalfa, blood meal or soybean meal fed to wintering steers. J. Anim. Sci. 61:245. Pommier, S. A,, M. H. Fahmy, L. M. Poste, and G. Butler. 1990. Effect of sex, electrical stimulation and conditioning time on carcass and meat characteristics of Romanov lambs. Food Qual. Preference 1:127. Sanudo, C., and I. Sierra. 1982. A study on the carcass and meat quality in Romanov x Aragon crossbreds. I. A description and comparison of two types of lambs. Anal. Fac. Vet. univ. Zaragoza 16/17, p 285. SAS. 1985. SAS User’s Guide: Statistics. (5th Ed.]. SAS Inst. Inc., Cary, NC. Shqueir, A. A,, D. C. Church, and R. 0. Kellems. 1984. Evaluation of liquefied fEh in digestibility and feedlot performance studies with sheep. Can. J. h i m . Sci. 64:889. Silva, A. T., and E. R. Brskov. 1988. The effect of five different supplements on the degradation of straw in sheep given untreated barley straw. Anim. Feed. Sci. Technol. 19289. Stock, R., T. Klopfenstein, D. Brink, S. Lowry, D. Rock, and S. Abrams. 1983. Impact of weighing procedures and variation in protein degradation rate on measured performance of growing lambs and cattle. J. h i m . Sci. 571276. Tayer, S. R., and M. J. Bryant. 1988. The response of store lambs to dietary supplements of fish meal. 3. Effects of the preceding pattern of growth, Anim. h o d . 47:393. Theriez, M., and M. Tissier. 1975. L’utilisation des races prolifiques. Valeur d‘blevage des anima- c r o i a s et qualit& des carcasses. 1 Journ. Rech. Ovine Caprine. SPEOC,11:64. Thomas, D. L., J. V. Whiteman, and L. E. Walters. 1976. Carcass traits of lambs produced by crossbred dams of Finnsheep, Dorset and Rambouillet breeding and slaughtered at two weights. J. Anim. Sci. 43:373. Umunna, N. N., T. J. Klopfenstein, S. Hasimoglu, and W. R. Woods. 1982. Evaluation of corn gluten meal with urea as a source of supplementary nitrogen for growing calves and lambs. Anim. Feed Sci. Technol. 7:375. Vipond, J. E., M. E. King, E. R. Brskov, and G. Z. Metherill. 1989. Effect of fish meal supplementation on performance of overfat lambs fed on barley straw to reduce carcass fatness. Anim. h o d . 48:131. Yilala, K., and M. J. Bryant. 1985. The effects upon the intake and performance of store lambs of supplementing grass silage with barley fish meal and rapeseed meal. Anim. h o d . 40:lll. Young, L. D., and G. E. Dickerson. 1988. Performance of Booroola Merino and Finnsheep crossbred lambs and ewes. J. Agric. Sci. Finl. 60:492.
Agriculture Canada and Quebec Ministry of Agriculture, Fishery, and Food
ABSTRACT: Data were collected on 130 intact male lambs fed diets based on roughages supplemented with fish meal, soybean meal, or corn gluten-blood meal and slaughtered at 43 kg live weight. A nonsupplemented group served as a control. The lambs represented Romanov (R), Finnsheep (F), a new breed developed in Canada (DLS), Coopworth (C1, and Suffolk (3, three first crosses of DLS with R, F, and Booroola Merino (B), and their backcrosses to DLS. Supplemented lambs consumed 16 to 22% less ( P < .05) silage than control lambs. Average daily gains of lambs fed fish meal (226 g) and corn gluten-blood meal (217 g) were higher and feed conversion ratio (4.99 and 5.11) lower than that of lambs fed soybean meal (189 g and 5.48) or control (186 g and 5.76) diets ( P < .05). The cost of feed per kilogram of gain or per kilogram of lean produced was similar in the four treatments but was between C$.23 and .65 cheaper in the protein-supplemented groups when the number of days to reach slaughter weight was considered. The effect of diet on
carcass traits and meat quality were minimal. Meat of lambs fed the three protein supplements was less juicy than that of control lambs. With a few exceptions, most of the significant differences among genetic groups in growth, carcass, and sensory traits were mainly between prolific (R and F) and meat-type breeds (C, S, and DLS). Gain in weight was highest in S lambs (199 g/d), but F and R first crosses were the youngest a t slaughter (196 and 198 d). The F lambs had higher dressing and kidney fat percentages than meat-type breeds. The DLS lambs had the largest longissimus muscle area (14.0 cm2), whereas C had the smallest (10.7 cm2).The B crosses had larger longissimus muscle area than R and F crosses. The R lambs had more lean and less fat in the 12th rib, whereas C lambs had the lowest lean and a high bone percentage. The toughest and the most tender roasts were those of R and B crossbred lambs, respectively. Roasts from F lambs had the most intense lamb flavor.
Key Words: Protein Supplements, Growth, Feed Conversion Efficiency, Carcass, Sensory Evaluation, Sheep J. h i m . Sci. 1992. 70:1365-1374
'Contribution No. 357 of Lennoxville Research Station, No. of Food Res. Centre, No. RO90 of Res. Program Service. The authors are grateful to Claire Corriveau, Ramez Fahmy, Christine Cellier, and Marjolaine St. Louis for carcass evaluation at Lennoxville, to J. Elsaesser and V. E. Agar for sensory evaluation at FRC, to Gilles Gagnon and his team for care of animals and data collection at L a Pocatiere, to Andre Belleau for data processing, to Louise Boisvert for preparing the manuscript, and to B. K. Thompson for helpful suggestions.
894
2To whom requests for reprints should be addressed. 3Agric. Canada, Lennoxville Res. Sta., Quebec, J1M 123. 4Quebec Ministry of Agric., Fishery, and Food, Deschambault Res. Sta., GOA 1SO. 5Agric. Canada, Food Res. Centre, Ottawa, K1A OC6. 6Agric. Canada, Res. Program Service, Ottawa, K1A OC6. 7Agric. Canada, La Pocatiere Exp. Farm, Quebec, GOR 1ZO. Received July 22. 1991. Accepted December 31, 1991.
-
1365
Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
1366
FAHMY ET AL.
Introduction
mented with up to 200 g per lamb daily of a grain mixture of different cereals. Before starting the feeding trials, the lambs were dewormed, treated The new trend toward using prolific ewes and against enterotoxemia and tetanus, and then their crosses in sheep farming to improve productivity has increased choices among combinations injected with selenium and vitamins A, D, and E. of prolific and prolific x native lambs for market. The lambs were fed from an initial weight of about Research information concerning growth, carcass, 23 kg to a slaughter weight of about 43 kg live and meat quality of these lambs is meager. weight. Average age of lambs at the beginning of Research on carcass and sensory quality of the feeding test varied between 105 d for DLS and Finnsheep and their crosses (Thomas et al., 1976; 117 d for R, whereas for S it was 83 d. Dahmen et al., 1979; Fahmy, 1979, 1985; Kemp et The feeding trials were carried out in a separate al., 1981; Magid et al., 19811 and on Romanov sheep barn with a concrete floor divided into 12 (Theriez and Tissier, 1975; Sanudo and Sierra, pens of equal size (16 m21.Each pen provided 1982; Fahmy, 1986; Pommier et al., 19901 has adequate feeding space and drinking facilities. indicated that these prolific breeds produce acStraw was used for bedding. ceptable market lambs. These studies also showed Management and Feeding Treatments. The feeding that excessive fat deposition occurs in the body study started on May 30, 1988. The animals were cavity, whereas fat cover on the carcass is often deprived of feed and water for 12 h then weighed. inadequate to attain higher grades. They were distributed among 12 pens to form Little is known about the most effective way of groups of equal initial weights as much as possifeeding the prolific breeds and their crosses. ble. Each pen contained 11 lambs, one of each of Although the excessive deposition of body cavity the 11 genetic groups. Only nine 1/2 B lambs were fat is believed to be genetic in origin, it is possible available; therefore, 1/4 B lambs were used to that feeding diets rich in protein may favor more complete three pens. One C and one DLS lamb rapid development of lean tissue and lower fat that started the study were omitted from the deposition. Hassan and Bryand (1986) showed that analysis for failure to grow. lambs given fish meal had greater gain than All the animals were fed a basal diet composed unsupplemented lambs. Other evaluations of proof 75% grass silage and 25% barley-corn concentein supplements include those of Ganev et al. trate, presented separately. This basal diet was (1979) and Silva and 0rskov (19881, who fed supplemented with one of the following protein soybean and fish meal, and of Stock et a1. (1983) sources: fish meal, soybean meal, or a mixture of and Nelson et al. (19851, who fed blood and corn gluten and blood meal. A fourth diet with no soybean meal. However, none of these studies protein supplement served as a control. Composiinvolved lambs of prolific breeds or studied sen- tion of the different diets and their nutritive values sory qualities of meat, and therefore the present are presented in Table 1. The barn was divided study was designed to provide such information. into three sections of four pens each. The four treatments were assigned within each section to ensure equal representation of the different treatMaterials and Methods ments among sections of the barn. The lambs were fed once a day at 0930. The Experimental Design amounts offered and refused were weighed and recorded daily. The amounts offered were adjusted Animals. This study was conducted at L a according to the amounts refused; if there was e Pocatiere Experimental Farm using 130 intact 5% (on a DM basis) left for three consecutive days, male lambs. The lambs represented two prolificthe amounts were increased by 5%. If however, the types, Romanov (R1 and Finnsheep (F),and three amounts refused exceeded 5% for 3 d, the level of meat-types, DLS (a new breed developed in Canfeeding was reduced by 5%. ada, Fahmy, 19881, Coopworth (C1, and Suffolk (SI. Measurements and Traits Studied. Data were In addition, six crosses were available, first collected on body weight at 14-d intervals (mainly crosses resulting from mating R, F, and Booroola to monitor rate of gain) and on feed consumption Merino (B1 rams to DLS ewes and backcrosses daily for 96 d from the beginning of the feeding resulting from mating rams from these first test. Because some lambs were heavier at the crosses to DLS ewes. All the lambs were born at L a Pocatiere between January 25 and March 1, beginning or gained faster than others on feed, they reached the assigned slaughter weight earli1988 except S, which were born between February er; accordingly, not all animals completed 96 d on 17 and April 4 and were purchased from a private test and number of animal-days was calculated breeder after weaning. The lambs were weaned at and used to obtain feed conversion ratio. Some 50 d of age and, until the start of the study, were animals did not reach slaughter weight a t the end allowed ad libitum access to grass silage suppleDownloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
1367
PROTEIN SUPPLEMENTATION FOR LAMBS
of the 96-d feeding test period. These lambs continued to receive the same diets but feed consumption was not monitored. The growth rate reported is that made during the 96-d feeding test period. Lambs were slaughtered at a local abattoir when they attained about 43 kg live weight. The chilled carcasses were weighed 3 d after slaughter and the ratio of carcass to slaughter weight was calculated (dressing percentage). The carcasses were then divided into three wholesale cuts, shoulders, loin-racks, and legs, and each was expressed as a percentage of chilled carcass weight. Kidney and pelvic fat was removed from the body cavity and weighted. The l l t h , 12th, and 13th ribs from the right side were separated from the rest of loin-rack area and transferred to the laboratory for further measurements. For sensory quality evaluation, a sample comprising the gth to 13th ribs from the left side was separated from the carcass and frozen until all the animals were slaughtered. The frozen samples packed with dry
ice were transported to the Food Research Centre in Ottawa for evaluation. On arrival at the Lennoxville Station laboratory, color readings of the longissimus muscle were taken on fresh cuts after the exposed surface of the muscle was removed to avoid any effect of oxidation. The color readings were taken with a LabScan Model LS 5100 (Hunter Laboratories, Reston, VAL The search head was equipped with a green filter (555 nm peak) and calibrated using a grey meat color standard. Three measurements were taken, Hunter L, a, and b (CIE, 19761, which measure, respectively, the color between black (0) and white (1001, red (+I and green (-1, and yellow (+I and blue (-1. The samples were then frozen. The other measurements taken on thawed samples of the 12th rib were backfat thickness over the longissimus muscle (average of three measurements a t the two extreme ends and halfway between them) and area of the longissimus muscle measured from a tracing on acetate paper using a planimeter (Planix, Tamaya, Tokyo, Japan).
Table 1. Composition and chemical analysis of the different diets fed to lambs (percentage of total diet) Diet supplement
Item Diet composition Grass silage Barley Corn Salt Ca, PO supplement Magnesium, 54% Mineral and vitamin Subtotal Protein supplement Fish meal (8O0/0 CPIa Soybean meal (48% CPI Blood meal (80% CP) corn gluten meal (60% CP) Total diet Chemical analysis of the combined diet and supplement CP TDN Calcium Phosphorus Magnesium Sulfur Cost of diet CC$/l,OOO kg DMI Concentrate Silage
Fish meal
Soybean meal
Corn gluten and blood meal
Control
75.0 16.0 7.6 .2 1.1 .04 .o 1 100.00
75.0 18.0 7.8 .2 1.1 .04 .01 100.00
75.0 16.0 7.6 .2 1.1 .04 .01 100.00
75.0 18.0 7.8 .2 1.1 .04 .01 100.00
-
Added relative to basal diet 6.9 1.9 2.5
105.1
108.9
O '/
5.1
OO /
104.4
100.00
of Total diet, DM basis
16.0 65.6 1.o .8 .2 .2
18.0 68.5 .7 .4 .2 .2
16.0 65.7 .7 .4 .2
13.4 65.0 .7 .4 .2 .2
194.3 80.0
185.4 80.0
181.6 80.0
139.2 80.0
*Percentage of protein in protein supplement on a n as-fed basis.
Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
.2
1368
FAHMY ET AL.
The 12th rib was separated from surrounding ribs, trimmed, weighed, and dissected into lean, fat, and bone, each expressed as a percentage of total 12th rib weight. Connective tissues surrounding the longissimus muscle were removed and the muscle was then ground. Dry matter content of the longissimus muscle was determined by freezedrying the entire sample. A subsample was used to determine intramuscular fat content using Soxhlet high temperature extraction units (Tecator, Hoganas, Sweden). Another subsample was used to determine total and soluble collagen applying the method of Hill (19661. Only samples of the pure breeds were analyzed for collagen content. Previous experience involving five of the crosses used in this study showed no significant differences among the crosses (M. H. Fahmy, unpublished data). Two lambs, one with abnormal lean, fat, and bone percentages and the other with excessively high collagen contents, were considered to be outliers, and the measurements from these animals were removed from the analyses.
Sensory and Instrumental Meat Quality Evaluation Sample Preparation. Lamb roasts were thawed a t 4°Cfor 24 h and oven-ready weights were record-
ed. The roasts were placed on metal racks in aluminum pans and cooked, uncovered, at 163" C in electric household ovens (Modern Maid) to an internal temperature of 75" C. Temperature was determined by a Type-T copper constantan thermocouple (Thermo-Electric, Montreal, Canada) inserted into the geometric center of the longissimus muscle of each roast. After a 10-minstanding time, final cooking weights were recorded. The longissimus muscle was excised from surrounding bone and fat then cut into 4-mm-thick slices with a Hobart electric slicer. The samples were kept warm (40 to 44" C) in a foil-lined, covered container placed in a 75°C oven for a maximum of 15 min. Each panelist received two slices of each sample from the same position ke., Slices 1 and 2 went to Panelist 1; Slices 3 and 4 went to Panelist' 2, etc.). The slices were served to panelists in random order. Training. Seven selected assessors with previous meat evaluation experience were trained (in four 30-min sessions) for this study. The scales and anchor points were derived through discussions. Tenderness, juiciness, and lamb flavor were selected to start training based on past experience on lamb projects. The panelists were encouraged to use other descriptors to describe intersample differences. The presence or absence of off-flavor was evaluated by use of a binomial response (yes or no). To familiarize the panelists with the intensity of lamb flavor, two roasts were prepared. Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
One was roasted to an internal temperature of 70°C and the other was placed in boiling water to an internal temperature of 75" C. At another sitting, three roasts were roasted to internal temperatures of 65, 70, and 80°C. The third and fourth sittings were run-through sessions, and each roast w a s roasted to a temperature of 75°C. Sensory Panels. The roasts were evaluated in 22 sittings, four roasts at a time. Each of seven selected trained assessors received a set of four samples, one from each roast. The first 14 sittings (56 roasts) were a balanced, incomplete block design (BIBD, Cochran and Cox, 1957) used to compare eight of the genetic groups (pure breeds and first crosses) wherein the panelists compared four roasts from different genetic group on the same diet. Two additional sittings (resulting in 16 sittings in total were held so that each diet could be evaluated in four sittings. A second design was employed to compare the sensory characteristics of the four diets; in this design three replicates of each of two backcrosses (1/4 F and 114 R) were presented in six sittings. At each sitting, roasts of lambs from one cross raised on each of the four diets were compared. Water a t room temperature, unsalted crackers, and two wedges of a Japanese pear were provided to remove residual flavors from the mouth. Assessors evaluated the perceived intensity of lamb flavor, tenderness, and juiciness using a 15-cm, unstructured line scale (Larmond, 1977). From left to right the descriptive anchor points (1.5 cm from each end) were as follows: tenderness, very tough to very tender; juiciness, very dry to very juicy, and lamb flavor, slight to intense. The presence or absence of off-flavor was also evaluated as indicated previously.
Statistical Analysis The data on growth and carcass were analyzed using the GLM procedure of SAS (19851. The mathematical model used was Yijk] = u + bi + f j + bfij + gk + fgik + eijk, where Y is the performance of a n individual lamb, u is the overall mean, b is the block effect, f is the feeding treatment effect, bf is the amongpen error used to test f, g is the genetic group effect, fg is the feeding treatment x genetic group interaction, and e is the among-animal error used to test g and fg. The feeding treatment effect was compared to the among-pen error because feeding treatments were applied to the whole pen. In no instance was among-pen error significantly greater than among-animal error. When treatment effect was significant compared to among-animal error (with greater power), as was the case for two traits, this result was reported in the text. The means presented are adjusted for the slight imbalance in the data. Traits not expressed as percent-
PROTEIN SUPPLEMENTATION FOR LAMBS
ages were adjusted to a common weight by including slaughter weight as a covariate in the analysis model. The covariate initial age (at the beginning of the test) was also considered and discussed in the text, but the figures in the tables are not adjusted for this effect. Six orthogonal contrasts were used to partition the effect due to genetic groups (10 dfl. These were, among the three genetic group types with 2 df (pure breeds vs first and back crosses), meat-type pure breeds (C, S, and DLS) vs prolific-type pure breeds (R and F), among meat type (2 dfl, R vs F, among first crosses, and among backcrosses b o t h with 2 do. Differences among the four diets were noted by examining Fisher’s protected lsd. The first 16 sessions of the meat quality data were examined for diet x genetic group interactions using SAS (1985).None was found significant, so the breed results were based on 14 sessions, because this design was balanced for genetic group. These were examined as a BIBD using GENSTAT (Alvey et al., 1983). The mathematical model used W a s Yijk = U + Si + pi + Spij + gk + sgik + eijk, where Y is the sensory measurement, u is the overall mean, s is the sitting or session effect, p is the panelist effect, sp is the session x panelist interaction, g is the effect of genetic group, sg is the session x genetic group interaction or the among-roast error used to test g, and e is the among-sample error. The sensory data from the six sessions conducted to test the diet effect were examined using SAS (1985) and applying the mathematical model Yijkl = U + gi + bi + gbii + Pk + gbpijk + fl + gfil + gbfijl + eijkl, where Y, u, g, p, and e are as defined in the previous model and b is the replicate effect, gb is the among-session error, gbp is the session x panelist interaction, f is the feeding treatment, gf is genetic group x feeding treatment interaction, and gbf is the among-roast error used to test f and gf. Again, contrasts and lsd were used to examine the effect of genetic groups and diets, respectively. Frequencies of off-flavors were also examined in contingency tables.
Results Effect o f Feeding Treatment. The total gain made by the lambs fed the different diets was similar because initial weights were balanced among treatments and slaughter weight was fixed at around 43 kg live weight. However, the lambs fed the control and soybean meal diets averaged 12 d longer time on feed ( P e .05) than lambs fed the other two diets. During the feeding test (96 dl, ADG of lambs fed fish meal (226 g) and corn glutenblood meal (217 g) were higher (P < .01) than those Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
1369
of lambs fed the soybean meal (189 g) and those fed the control diet (186 g). Feed conversion ratio was lowest in lambs fed fish meal (4.99) and corn gluten-blood meal (5.111, followed by those fed soybean meal (5.481, which in turn were superior to those fed the control diet (5.76, P < .01, Table 2). Lambs fed diets supplemented with protein consumed between 16 and 22% less silage ( P e .01) than those fed the nonsupplemented diet (Table 21. The consumption of concentrate mixture was similar in the four groups. Using the prevailing prices at the time this experiment was conducted, the cost of feed per kilogram of live weight, dressed weight, or per kilogram of lean meat produced (using the dissection of the 12th rib as a n indicator of carcass lean content) was similar in the control and the supplemented groups (Table 2). Extending the calculations further to take into account the rate of turnover of capital (multiplying by number of anirnal-days/lOO), lambs fed the control diet tended to cost more to produce 1 kg of lean than those fed corn gluten-blood meal and fish meal diets, although not significantly so (P < .10 > .05, Table 2). The effect of feeding treatment on carcass traits and composition was slight (Table 2). Compared to among-animal error, treatment effects on dressing and kidney fat percentages were significant at the 5 % level. The analysis of the 12th rib showed that lambs fed the control diet also had less lean and more fat content than lambs fed the supplemented diets. No other significant diet effects were observed in color, intramuscular fat, DM, or total collagen content. The interaction of diet x genetic group was significant only for kidney fat percentage. This resulted mainly from two high values ( > 5%)for F fed soybean meal. When these values were removed, the interaction became nonsignificant. Effect o f Genetic Group. The initial weight ranged between 20.5 kg for F and 29.1 kg for 1/2 F, but all lambs were slaughtered at about 43 kg live weight (Table 31. Average daily gain to slaughter was generally low for the different genetic groups, ranging from 141 g/d for R lambs to 199 g/d for S lambs. Finnsheep-DLS first-cross lambs were the youngest (196 d) at slaughter, whereas F lambs were the oldest (264 d). When the covariate initial age was added to the model to adjust for the younger age of S at the beginning of the test, the orthogonal contrast comparing pure breeds vs crosses was nonsignificant for ADG and age at slaughter. Significant differences observed among genetic groups in many of the carcass traits studied were mainly between prolific and meat-type lambs (Table 3). Finnsheep lambs had higher dressing percentage than meat-type lambs such as S and C.
1370
FAHMY ET AL.
The proportions of the three wholesale cuts varied within narrow ranges among the different genetic groups. The two prolific breeds, R and F, had slightly higher shoulder proportions than meattype animals (39.3 vs 37.2%),and the crosses were intermediate. As expected, F lambs had the highest kidney fat percentage (3.70/0),followed by R (3.4%) and R-crosses (2.9%). Meat-type breeds and B-crosses had the least kidney fat percentage. Conversely, the meat-type lambs had thicker backfat than the prolific lambs. The DLS lambs had the largest and C lambs the smallest longissimus muscle area (14.0vs 10.7 cm2). Among the
first crosses, the B-cross had larger longissimus muscle area then F- and R-crosses. Percentage of lean, fat, and bones in the 12th rib varied slightly between the different genetic groups. Romanov lambs had significantly more bone and less fat than F, whereas C lambs had the lowest lean and highest bone percentages among the other meattype breeds studied. Romanov lambs were among the lowest in intramuscular fat (8.9%), significantly lower than in F lambs (1 1.7%). Some differences in color of the longissimus muscle were significant, mainly between prolific (paler) and meat-types and between prolific and B first-cross (paler). Romanov
Table 2. Least squares means for average daily gain, feed conversion ratio, and carcass characteristics of lambs fed different protein supplements in diets Feeding treatment
Corn gluten Item No. of lambs Initial wt, kgb Initial age, d Slaughter wt, kg Feed consumption, kg of DMAamb SilageC ConcentrateC Total' Gain in wt, kg/lambc Feed conversion ratioC No. of animal-daysc ADG on test, g Cost per lambCd Cost per kg gainCd Cost per kg dressedCd Cost per kg leanCd Cost, adj. for dayCd Carcass percentages Dressing Leg Loin-rack Shoulder Kidney fat 12th Rib measurements Avg. fat thickness, mm Longissimus muscle area cm2 Muscle, % e Fat, 0% e Bone, Oh e Hunter color L Hunter color a Hunter color b Intramuscular fat, O/O Dry matter, % Total collagen content, % Soluble collaaen - content. %
Fish meal
Soybean meal
32 25.0 111 44.2'
33 24.3 110 42.58
56.7' 22.3 79.d 15.8 4.99f 70. If 226' 8.87 .56 1.37 3.29 2.3 1
and blood meal
Control
SEM'
33 25.3 109 43.1'8
32 24.6 108 43.3fg
.82 1.5 .32
55.2' 23.0 78.2' 14.2 5.488 76.0'8 1898 8.23 .58 1.39 3.30 2.51
52.4' 20.4 72.9' 14.3 5.1 1'8 66. 1' 217' 7.90 .55 1.36 3.29 2.18
67.58 22.0 89.58 15.6 5.76' 83.88 1868 8.47 .55 1.33 3.37 2.83
41.4 33.4 28.9 37.8 2.8
42.4 33.5 29.1 37.6 2.9
40.8 33.6 28.6 37.8 2.6
42.2 32.9 29.0 38.1 3.2
.53 .26 .22 .18 .13
3.9 12.6 42.0' 36.3' 20.6 26.1 5.1 5.7 9.6 24.5 2.05 .33
4.4 12.3 41.Ef 37.5' 19.5 25.5 5 .O 5.4 10.1 24.7 2.01 .29
4.2 11.9 41.6f 35.8' 21.2 26.1 5.3 5.7 9.6 24.6 2.08 .35
4.5
11.9 39.91 39.28 19.8 25.5 5.3 5.6 9.6 24.6 2.08 .32
.20 .26 .35 .53 .43 .54 .18 .15 .49 .15
NS NS
.08 .o1
NS NS
aMaximum standard error of the mean based on 6 df. bInitial weight for four animals was missing. CAdjusted means based on three pens. dAll prices are in Canadian dollars. e o n e outlier removed; see text. 'SgMeans followed by different letters are significantly different (P *,**P5 .05 and P 5 .01, respectively; NS, not significant.
Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
,051 except that of diet on juiciness (P c .05).
and F lambs had more (P < .05) total and soluble collagen in the longissimus muscle than the meattype breeds (Table 3). However, when initial age was included in the analysis model, the contrast of prolific vs meat-type in soluble collagen became nonsignificant . Sensory Evaluation. Diet x breed interactions were found nonsignificant, both within the 16 sittings to compare breed group and the six sittings to compare diets. No significant differences among diets were observed in any of the cooking traits (Table 4). Only juiciness showed a significant diet effect: meat of lambs fed the control diet was more juicy than that of those fed the three protein-supplemented diets. The comparison of the five pure breeds and the three first crosses indicated no significant effects observed for the traits of cooking rate, weight loss, and drip loss (Table 51. The sensory analysis results showed significant differences in tenderness among the crosses despite a good deal of variation among roasts for this trait. The 1/2 R and 1/2 F were tougher than
the 1/2 B. No significant effect was found for juiciness. Meat of F lambs exhibited the most intense flavor (10.51,significantly more than for R lambs or the three meat-type breeds. The amongcrosses contrast for flavor was barely nonsignificant ( P = .054). A significantly higher percentage of off-flavors was found in roasts from prolific breeds than in roasts from meat-type breeds (11.8 VS 4.8%).
Discussion Protein supplementation of the diet of sheep a t different stages of their life cycle has been studied for relation to growth, feed efficiency, milk and wool production, and, to a lesser extent, for carcass quality (Bowman and Paterson, 1988; Gunderson et al., 1988; Tayer and Bryant, 1988; Hovel1 and IZlrskov, 1989; Vipond et al., 1989). There is general agreement that protein supplementation has a beneficial effect on growth and carcass traits, manifested in a n increase in daily
Table 5. Cooking data and sensory meansa and their standard errors (SEM) Pure breedsb Meat-type Trait Cooking rate, min/100 g Weight loss, g Drip loss, O/O Tenderness Juiciness Flavor Off-flavor,%
Prolific-type
First crosses
F
S
C
DLS
R
F
11 14.9 4.1 9.3 7.7
13 16.3 4.9 7.1 8.4
10 15.8 3.7 6.8 7.8
15 18.5 3.9 6.8 7.7
11 15.5
11 16.0
14 16.0
11 13.2
6.9 1.05
3.8 7.5 8.7
3.9 5.9 7.6
3.5 6.3 8.4
3.8 9.1 7.7
.41 .88 .37
8.1 1.7
7.4 4.8
8.1 7.8
8.5 12.1
10.5 11.5
7.9 7.3
8.4 10.0
9.3 3.5
.40 3.31
1/2
R
1/2
1/2
B
SEM
SOCc 5
2,4 2
&Means were based on at least 54 observations and the standard error on 35 df. = Coopworth; DLS = a new breed developed in Canada; R = Romanov; F = Finnsheep; B = Booroola Merino. 'SOC = significant orthogonal comparisons, codes as follows: 1, purebreeds vs first crosses; 2, meat-type vs prolific-type, 3, among meat-types; 4, R vs F; 5 , among first crosses.
bS = Suffolk; C
Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
PROTEIN SUPPLEMENTATION FOR LAMBS
intakes and improved daily gains, carcass weights, and carcass characteristics. However, in many cases the improvements failed to reach significant levels. Many different sources of protein supplementation were compared (Loerch and Berger, 1981; Umunna et al., 1982; Stock et al., 1983; Yilala and Bryant, 1985; Butler and Laycock, 1987; Duff et al., 1988).For the traits examined, these studies as well as the present study have shown small differences from the source of supplementary protein. When soybean meal was compared with other sources of protein supplement, lambs fed the soybean meal showed slightly inferior gains and feed conversion efficiency (Stock et al., 1983). It seems, however, that protein supplementation may have a n effect in reducing fat deposition in the body, as evidenced from the present data. Similar findings were reported by Vipond et al. (1989) using fish meal supplementation. The taste panel did not detect significant differences among meat samples of lambs fed the different supplemented diets, indicating that these diets may have little effect on meat tenderness, juiciness, and flavor, although there are some indications that lambs fed the soybean meal supplement were slightly more tender and had less off-flavor than those fed fish meal and corn gluten-blood meal or no supplement. Similar results were reported by Shqueir et al. (19841, who found no difference among diets (one of which included liquefied fish) on lamb flavor. The absence of a diet x genetic group interaction in all the traits studied (except for kidney fat percentage) indicates that prolific and nonprolific sheep and purebred and crossbred lambs respond similarly to differences in the diets offered. Comparative studies involving many genetic groups of sheep are scarce. Kempster and Cuthbertson (19771 surveyed carcass characteristics of the main types of British sheep and showed that many differences exist between breeds. Growth, carcass characteristics, and meat quality of lambs from breeds used in the present study were reported in numerous publications (Theriez and Tissier, 1975; Thomas et al., 1976; Dahmen et al., 1979; Fahmy, 1979, 1985, 1986; Kemp et al., 1981; Pommier et al., 1990). However, few comparative studies involving prolific breeds and their crosses are found in the literature (Jakubec and Krizek, 1988; Young and Dickerson, 19881. Considering age a t slaughter as a n indicator of overall growth rate, 1/2 F and 1/2 R lambs were youngest, followed by S lambs, whereas F and 112 B lambs ranked last. The two prolific breeds had the highest percentage of kidney fat on all diets; therefore, it does not seem that the protein supplementation applied in this study changed this tendency. In contrast, the Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
1373
three nonprolific breeds had a thicker fat cover on the body. Both prolific breeds had a higher percentage of off-flavor than did a standard meat-type breed such a s the S breed. In conclusion, the protein supplementation applied in this study had some beneficial effect on the traits studied, although the differences in many cases failed to reach statistical significance. The different genetic groups responded in a similar manner to these treatments. There were many differences among the genetic groups, mainly between prolific and nonprolific breeds. The first and back crosses with the DLS were generally similar in performance.
Implications Supplementing the diets of growing lambs with soybean meal or industry byproducts to increase protein content may seem at first to be an expensive proposal; however, when one considers the better feed conversion and the shorter period on feed, protein supplementation may prove to be advantageous. Products such as fish meal, which have been reported to affect flavor of meat in chicken, had no such effect on sheep meat, and therefore can be recommended in lamb diets. Based on the results of this study, wide-scale use of prolific breeds and their crosses to increase productivity of sheep will not have an adverse effect on quality of the carcasses or the meat available for consumers.
Literature Cited Alvey, N. G., C. F. Banfield, R. I. Baxter, J. C. Gower, W. J. Krzanowski, P. W. Lane, P. W. Leech, J. A. Nelder, R. W. Payne, K. M. Phelps, C. E. Rogers, J.G.S. Ross, H. R. Simpson, A. D. Todd, G. Tunnicliffe-Wilson, R.W.M. Wedderburn, R.P. White, and G. N. Wilkinson. 1983. GENSTAT. A general statistical program. Numerical Algorithms Group Ltd., Oxford, UK. Bowman, J.G.P., and J. A. Paterson, 1988. Evaluation of corn gluten feed in high-energy diets for sheep and cattle. J. Anim. Sci. 66:2057. Butler, G., and K. A. Laycock. 1987. Supplementation of silage based diets for lambs. Occasional Symposium, Br. Gassl. SOC. 21:113. CIE. 1976. Commission Internationale de 1’Eclairage. 18th Session, Sept. 1975. CIE Publication No. 36. London. Cochran, W. G., and G. M. Cox. 1957. Experimental Designs (2nd Ed.). John Wiley and Sons, New York. Dahmen, J. J., D. D. Hinman, J. A. Jacobs, and D. 0. Everson. 1979. Performance and carcass characteristics of Suffolk sired lambs from Panama and Finn x Panama dams. J. Anim. Sci. 49:55. Duff, G. C., A. L. Goetsch, K. M. Landis, A. C. Hardin, S. R. Stokes, Z. B. Johnson, and K. L. Hall. 1988. Mixing or alternating dietary crude protein sources and performance of wether lambs. Can. J. Anim. Sci. 68:569. Fahmy, M. H. 1979. Body and carcass measurements of DLS
1374
FAHMY ET AL.
and Finn x DLS lambs slaughtered at three light body weights. World Rev. Anim. Prod. XV(4):ll. Fahmy, M. H. 1985. The accumulative effect of Finnsheep breeding in crossbreeding schemes: Growth and carcass traits. Can. J. h i m . Sci. 65:811. Fahmy, M. H. 1986. Preliminary results on fertility, prolificacy, lamb production and carcass traits of Romanov sheep in Canada. Proc. 3rd World Congr. Genet. Appl. Livest. Prod. IX:559. Lincoln, NE. Fahmy, M. H. 1988. Development of DLS sheep: lamb production of the pure breeds, initial crosses and first generation DLS. World Rev. Anim. Rod. 24(3):77. Ganev, G., E. R. Brskov, and R. Smart. 1979. The effect of roughage on concentrate feeding and rumen retention time on total degradation of protein in the rumen. J. Agnc. Sci. (Camb.1 93651. Gunderson, S. L., A. A. Aguilar, D. E. Johnson, and J. D. Olson. 1988. Nutritional value of wet corn gluten feed for sheep and lactating dairy cows. J. Dairy Sci. 71:1204. Hassan, S. A,, and M. J. Bryand. 1986. The response of store lambs to dietary supplements of fish meal. 2. Effects of level of feeding. Anim. Prod. 42:233. Hill, F. 1966. The solubility of intramuscular collagen in meat animals of various ages. J. Food Sci. 31:151. Hovel], F.D.D., and E. R.Brskov. 1989. The role of fish meal in rations for sheep. Tech. Bull. Int. Assoc. Fish Meal Manufacturing No. 23. Herts, UK. Jakubec. V., and J. Krizek. 1988. Breeding and exploitation of prolific breeds in Czechoslovakia. J. Agnc. Sci. Finl. 60:505. Kemp, J. D., D. G. Ely, J. D. Fox, and W. G. Moody. 1981. Carcass and meat characteristics of crossbred lambs with and without Finnish Landrace breeding. J. h i m . Sci. 52: 1026.
Kempster, A. J., and A. Cuthbertson. 1977. A survey of the carcass characteristics of the main types of British lamb. Anim. Prod. 25:165. Lamond, E. 1977. Laboratory methods of sensory evaluation of food. Publ. 1637. Communication Branch, Agriculture Canada, Ottawa. Loerch, S. C., and L. L. Berger. 1981. Feedlot performance of steers and lambs fed blood meal, meat and bone meal, dehydrated alfalfa and soybean meal as supplemental protein sources. J. h i m . Sci. 531198. Magid, A. F., V. B. Swanson, J. S. Brinks, G. E. Dickerson, J. A. Crouse, and G. M. Smith. 1981. Border Leicester and Finnsheep crosses. 111. Market lamb production from crossbred ewes. J. h i m . Sci. 52:1272. Nelson, M. L., I. G. Rush, and T. J. Klopfenstein. 1985. Protein
Downloaded from https://academic.oup.com/jas/article-abstract/70/5/1365/4705276 by guest on 11 February 2018
supplementation of ammoniated roughages. 11. Wheat straw supplemented with alfalfa, blood meal or soybean meal fed to wintering steers. J. Anim. Sci. 61:245. Pommier, S. A,, M. H. Fahmy, L. M. Poste, and G. Butler. 1990. Effect of sex, electrical stimulation and conditioning time on carcass and meat characteristics of Romanov lambs. Food Qual. Preference 1:127. Sanudo, C., and I. Sierra. 1982. A study on the carcass and meat quality in Romanov x Aragon crossbreds. I. A description and comparison of two types of lambs. Anal. Fac. Vet. univ. Zaragoza 16/17, p 285. SAS. 1985. SAS User’s Guide: Statistics. (5th Ed.]. SAS Inst. Inc., Cary, NC. Shqueir, A. A,, D. C. Church, and R. 0. Kellems. 1984. Evaluation of liquefied fEh in digestibility and feedlot performance studies with sheep. Can. J. h i m . Sci. 64:889. Silva, A. T., and E. R. Brskov. 1988. The effect of five different supplements on the degradation of straw in sheep given untreated barley straw. Anim. Feed. Sci. Technol. 19289. Stock, R., T. Klopfenstein, D. Brink, S. Lowry, D. Rock, and S. Abrams. 1983. Impact of weighing procedures and variation in protein degradation rate on measured performance of growing lambs and cattle. J. h i m . Sci. 571276. Tayer, S. R., and M. J. Bryant. 1988. The response of store lambs to dietary supplements of fish meal. 3. Effects of the preceding pattern of growth, Anim. h o d . 47:393. Theriez, M., and M. Tissier. 1975. L’utilisation des races prolifiques. Valeur d‘blevage des anima- c r o i a s et qualit& des carcasses. 1 Journ. Rech. Ovine Caprine. SPEOC,11:64. Thomas, D. L., J. V. Whiteman, and L. E. Walters. 1976. Carcass traits of lambs produced by crossbred dams of Finnsheep, Dorset and Rambouillet breeding and slaughtered at two weights. J. Anim. Sci. 43:373. Umunna, N. N., T. J. Klopfenstein, S. Hasimoglu, and W. R. Woods. 1982. Evaluation of corn gluten meal with urea as a source of supplementary nitrogen for growing calves and lambs. Anim. Feed Sci. Technol. 7:375. Vipond, J. E., M. E. King, E. R. Brskov, and G. Z. Metherill. 1989. Effect of fish meal supplementation on performance of overfat lambs fed on barley straw to reduce carcass fatness. Anim. h o d . 48:131. Yilala, K., and M. J. Bryant. 1985. The effects upon the intake and performance of store lambs of supplementing grass silage with barley fish meal and rapeseed meal. Anim. h o d . 40:lll. Young, L. D., and G. E. Dickerson. 1988. Performance of Booroola Merino and Finnsheep crossbred lambs and ewes. J. Agric. Sci. Finl. 60:492.
